-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

MWJ 2002.10.05 (October 28, 2002)
=================================

   Copyright 2002, GCSF Incorporated.  All rights reserved.


Top of the Week
---------------

**The big story**

  The Weekly Attitudinal, our obnoxious right-by-definition
  opinion feature, saw a little news item about Adobe and Monotype
  headed to court over font rights. From that, the _Attitudinal_
  has a comprehensive look at the bizarre world of font embedding
  and copyrights. Learn which major foundries only allow embedding
  if you don't need it; look at how those foundries are wielding
  the DMCA in wildly inappropriate ways, and remember how Adobe is
  getting no sympathy due to its own ham-handed invocation of DMCA
  on other PDF matters. It's an underreported, triple-size story
  that only the _Attitudinal_ has attempted, and we think it's
  worth the wait. We hope you'll agree.

**In other news**

  The theme of This Week's News is "technical information,"
  including nearly four dozen KnowledgeBase articles and two long
  looks at Apple's fear of Mac OS technology. We have the
  long-awaited "Power Macintosh G4 (Mirrored Drive Doors)
  Developer Note" and an exclusive look at why Java 1.4 has been
  missing from Mac OS X for the past year and a half - and it may
  be mostly Cocoa politics. There are new "Switch" commercials,
  updates to .Mac sign-ups and education pricing, fences mending
  between Apple and Sorenson Media, Steve Jobs quitting at the
  Gap, O'Reilly Conference coverage, and a look at _eWEEK_'s
  attempts to turn fantasy into fact. All that plus over eighty
  briefs from the week that was.

**Stuff to buy**

  Product Showcase spends most of its time with iSync, and we're
  not really sure why: it's marginally useful for people using Mac
  OS X 10.2's Address Book and iCal as primary personal
  information tools, but we suspect there aren't many of those,
  and that synchronization must not be critical to them or they
  wouldn't be doing that. We have full-size looks at that plus
  Eudora 5.2b10, REALbasic 5, StreamRipperX 1.0.1, Audio Hijack
  1.0, OnMyCommandCM 1.3.16, and Freeverse's just-ending Card Game
  Week, plus more in brief you may have missed.


This Week's News
----------------

**follow-ups**

* As Apple Computer had fervently hoped, .Mac sign-ups have
  accelerated past the 1,000-per-day figure cited in a 2002.09.17
  press release [1]. By the older numbers, Apple should have
  around 114,000 .Mac users by now, but a Tuesday press release
  [2] confirms that the total is now over 180,000. If it reaches
  220,000, Apple will have the 10% conversion rate that Phil
  Schiller said in July he considered "typical" for free-to-fee
  transitions. This time, however, Apple resisted saying how many
  new subscribers are coming on board each day, probably because
  everyone knows the big iTools conversion push is just about over -
  even if Apple tries extending the deadline again.

  [1] <http://www.apple.com/pr/library/2002/sep/17dotmac.html>
  [2] <http://www.apple.com/pr/library/2002/oct/01dotmac.html>

  To further the push, Apple unveiled .Mac for Education [3], a
  limited version of the service that goes for US$59 per account
  with a ten-account minimum, something schools can order via
  purchase order. The 40% discount reflects in capability:
  education accounts get 10MB of E-mail storage (instead of 15MB),
  and a 50MB iDisk instead of 100MB. Meanwhile, major news site
  such as MacNN continue to report [4] on .Mac glitches and
  downright outages, concerns that have plagued the service for
  years, and for which Apple still has no real answers. (MWJ_
  2002.09.28)

  [3] <http://www.apple.com/education/eddotmac/>
  [4] <http://www.macnn.com/news.php?id=16714>

* Without providing any details, Apple Computer on Wednesday
  announced it has settled its lawsuit with Sorenson Media. The
  announcement [5], in full (aside from boilerplate company
  promotion): "Apple and Sorenson Media today announced the
  settlement of a pending federal court lawsuit that had been
  brought by Apple in April, 2002. The settlement provides for the
  dismissal of all claims and counterclaims." Apple had sued [6]
  Sorenson for providing video compression in Flash MX products
  from Macromedia; Apple asserted that it had an exclusive license
  to all of Sorenson's video compression for QuickTime, as
  Sorenson's compression had been QuickTime's standard from
  QuickTime 3.0 through QuickTime 5, supplanted only this year by
  MPEG-4 video. Sorenson Media, in turn, filed counterclaims [7]
  against Apple in August, quoting a Steve Jobs E-mail saying
  Apple didn't have rights to all future Sorenson video
  compression products, just to the one in QuickTime. There's no
  word on what the settlement is, but it appears that both
  companies have declared, "Never mind." (MWJ_ 2002.05.04,
  2002.08.06)

  [5] <http://www.apple.com/pr/library/2002/oct/02settlement.html>
  [6] <http://www.sorenson.com/content.php?pageID=35&id=21&nav=7>
  [7] <http://www.sorenson.com/content.php?pageID=35&id=22&nav=7>

* Having noted several times that Apple had not released a
  developer note for the Power Macintosh G4 (Mirrored Drive Doors)
  computers, it seems only fair to note the 2002.09.27 release of
  the Power Macintosh G4 (Mirrored Drive Doors) Developer Note
  [8]. (We know it's dated 2002.09.17, but it was released on
  2002.09.27.) The note answers many questions about the machines,
  though not all: for example, it doesn't detail the physical
  dimensions of optical drives you can put in the computer, and
  even though a KnowledgeBase article [9] does (148 mm wide, 42 mm
  tall, 198 mm long), it still doesn't list how wide the drive
  tray can be and still fit through the door in the enclosure.
  It's the kind of information a Developer Note should provide,
  and now that it doesn't, we're kind of bummed that we looked so
  hard for it to come out.

  [8] <http://developer.apple.com/techpubs/hardware/Developer_Notes/Macintosh_CPUs-G4/PowerMacG4/index.html>
  [9] <http://docs.info.apple.com/article.html?artnum=75370>

  It's still a useful reference in case you need to deal with this
  current Power Macintosh model. One part refers you to a
  KnowledgeBase article [10] about the eMac and iMac power button
  because, like those consumer computers, the Power Macintosh G4
  (Mirrored Drive Doors) includes no Programmer's Switch or NMI
  button, even though another part of the note - cut and pasted
  from the Power Macintosh G4 (QuickSilver 2002) version, no doubt -
  says they're still there. We learn that the machine does not
  come with a 4-pin to 6-pin FireWire cable (unlike the consumer
  models), that the internal modem appears as a USB device on bus
  2 (the same one shared with the ADC connector), and that the
  built-in sound hardware applies signal processing to the sound
  to make it sound good on Apple Pro Speakers (though not to sound
  going through headphones or the "line out" port).

  [10] <http://docs.info.apple.com/article.html?artnum=88330>

  The note does clear up a few mysteries, documenting that the
  Level 3 cache speed ratio is 4:1 for the 867MHz and 1GHz models,
  but 5:1 for the 1.25GHz model. In other words, even though the
  high-end model has 25% faster processors, the cache RAM still
  runs at the same 250MHz speed, as the dual-1GHz model. This may
  explain why Bare Feats's early tests found the dual-1.25GHz
  model to run only 20% faster than the dual-1GHz mid-range Power
  Macintosh G4 (Mirrored Drive Doors) computer. We also learn that
  older AGP graphics cards won't work in the new machine, and that
  since the fans are connected to the heat sink, they disconnect
  when you open the case and therefore you should _not_ operate a
  Power Macintosh G4 (Mirrored Drive Doors) machine with the case
  open. There's more information about the U2 microcontroller that
  runs the system, the RAM specifications (very useful for RAM
  vendors or resellers), and lots more. Grab the note's PDF
  version if you want to know more about the hardware. (MWJ_
  2002.08.18, 2002.08.25, 2002.08.31)

**business news**

* After a false start or two with some Internet postings, Apple
  Computer has released a new set of five "Switch" television
  commercials that started airing in the US on Wednesday. Gianni
  Jacklone [11] is a young IT director who likes Mac OS X even
  though he's worked with Windows for 10 years (and Unix for 7)
  and has "hated Macs [his] entire life," proof that Apple is
  definitely not marketing to its installed base. Richard Ziskin
  [12] runs the American Umbrella Company, the "only umbrella
  manufacturer left in the United States" and apparently the only
  US company without a Web site. It's now run entirely on iMacs.
  "I can't believe I didn't switch from the PCs to the iMacs
  sooner," Ziskin says, before adding the memorable tag line, "My
  name is Richard Ziskin, and my dream is that it rains every
  single day!"

  [11] <http://www.apple.com/switch/ads/giannijacklone.html>
  [12] <http://www.apple.com/switch/ads/richardziskin.html>

  The other three new ads are Apple's first real "celebrity" ads
  of the campaign (a funny one featuring "Saturday Night Live"
  alumnus Will Ferrell was shown at the Macworld Expo keynote in
  New York but never aired or put on the Web). Professional surfer
  Kelly Slater [13] explains that with iMovie, he learned in about
  an hour how to edit his own surfing movies together, and the
  commercial even shows excerpts (the first of the campaign to cut
  away from the person telling the story). Professional
  skateboarder and video game namesake Tony Hawk [14] tells a
  similar story, but he uses Final Cut Pro to make both
  professional skate videos and home movies of his oldest son
  skating, again with excerpts of the videos. Finally, DJ Qbert
  [15], a "scratch DJ" (who creates new rhythms and sounds largely
  by scratching vinyl records) describes how he made his first
  album on a Mac, which is like a "visual tape recorder" he can
  use to see everything and arrange it how he wants. The five new
  ads join the ten already in circulation (including Ellen Feiss's
  ad, which we have actually seen on television now) for a total
  of fifteen in the current campaign. It remains to be seen how
  threatened PC users or IT directors will try to minimize these
  stories.

  [13] <http://www.apple.com/switch/ads/kellyslater.html>
  [14] <http://www.apple.com/switch/ads/tonyhawk.html>
  [15] <http://www.apple.com/switch/ads/djqbert.html>

* Steve Jobs resigned [16] this week from the board of directors
  of Gap, Inc., where he had served for just over three years
  (MWJ_ 1999.09.11). Most stories, including the one at CNet News
  [17], noted the "interlocking membership" between Gap and Apple:
  Jobs joined Gap's board just four months after Gap CEO Mickey
  Drexler joined Apple's board in May 1999 (MWJ_ 1999.05.29).
  Apple and Gap were targeting the same markets, and more than one
  observer has commented that Apple's US retail stores have
  Gap-like sensibilities. That's no accident, either: the
  _Business_Journal_ reported [18] over a year ago how Apple hired
  Fisher Development to build its retail stores. That's "Fisher
  Development" as in "Robert Fisher," brother of Gap founder and
  chairman Donald Fisher.

  [16] <http://vocuspr.vocus.com/vocuspr30/xsl/gappr/Profile.asp?Entity=PRAsset&EntityID=25659&XSL=PressRelease&PublishType=Press+Release>
  [17] <http://news.com.com/2100-1040-960735.html>
  [18] <http://sanjose.bizjournals.com/sanjose/stories/2001/06/04/focus1.html>

  There are multiple reasonable interpretations: you can conclude
  that the "good old boys" network was at work so that Jobs and
  Drexler and Fisher all rewarded each other, or you can conclude
  that Apple decided it wanted Gap-like stores and did whatever it
  had to do to get Gap's cooperation, including putting Jobs on
  Gap's board. The relationship, however, seems tied to Drexler.
  Not enough stories make it explicit: Drexler announced in May
  that he would retire as soon as Gap named a successor as CEO.
  Gap did just that on 2002.09.26, appointing [19] 15-year Walt
  Disney Company theme park leader Paul Pressler as the new CEO.
  One week later, Jobs resigned and Gap appointed former Coca-Cola
  Europe executive Penny Hughes to replace him. Drexler, however,
  remains on Apple's board, and no longer has daily
  responsibilities at the Gap.

  [19] <http://vocuspr.vocus.com/vocuspr30/xsl/gappr/Profile.asp?Entity=PRAsset&EntityID=25355&XSL=PressRelease&PublishType=Press+Release>

**business news in brief**

* Apple's next US retail store is to open today, 2002.10.05, in
  the Towson (MD) Town Center, north of Baltimore
  <http://www.apple.com/retail/towson/>

* After that, Apple opens a US retail store at Somerset center in
  Troy, MI on 2002.10.12
  <http://www.apple.com/retail/somerset/>

* Dell Computer on Tuesday raised its revenue guidance for the
  quarter ending 2002.11.01, saying the firm will now book US$9.1
  billion in revenue instead of previous guidance of US$8.9
  billion, with profits of US$0.21 per share (at the high end of
  previous guidance); note that either figure is more revenue than
  Apple Computer books in five quarters
  <http://news.com.com/2100-1001-960359.html>

* Although Intuit has yet to announce any definitive features for
  QuickBooks Pro 5 for Mac OS X (other than Mac OS X operation and
  better reporting), MYOB continues to chip away at QuickBooks's
  massively-larger feature set, announcing integration with
  SurePayroll, an online service that processes payroll and all
  taxes (with direct deposit) for per-period fees, directly
  integrating with MYOB AccountEdge
  <http://www.myob.com/us/egateway/msp/index.htm>

* ATI Technologies lost US$32.2 million on revenues of US$239.5
  million in its fourth fiscal quarter ended 2002.08.31 (using
  Generally Accepted Accounting Principles), compared to a US$11.6
  million loss on US$33.8 million in revenue a year ago; gross
  margins, however, are up from 23.2% to 32.8% as ATI shifts from
  a board-based business model to a chip-based model
  <http://mirror.ati.com/companyinfo/ir/2002/atiq402.pdf>

* Banc of America Securities initiated analyst coverage of Apple
  Computer this week, rating the stock "Market Perform" (somewhat
  equivalent to "neutral")
  <http://biz.yahoo.com/c/20021001/i.html?aapl>

* _Macworld_UK_ reports on important price changes from Apple
  Europe: the company has slashed basic margins to razor-thin
  margins, but will lower the prices (and increase the margins) on
  a dealer-by-dealer basis to reward resellers that provide a
  strong sales environment, after-sales support, and other
  customer-friendly amenities
  <http://macworld.co.uk/news/main_news.cfm?NewsID=5340>

* A Reuters report criticizes troubled Tyco Corporation for
  putting Jerome York on its board of directors because York is on
  Apple's board and also CEO of MicroWarehouse, a leading Apple
  reseller; once again, though, the story misstates the facts
  (Jobs was awarded a US$45 million jet and Apple took another
  US$45 million write-off to pay for options that are now
  underwater; Jobs did not get US$90 million worth of anything)
  <http://biz.yahoo.com/rc/021004/manufacturing_tyco_board_1.html>

* John Santoro, a 12-year veteran of Apple Computer (who started
  out in Apple II education marketing; we know these things) and
  was most recently iPhoto product manager, leaves the company to
  join the new PreClick Corporation, makers of "world's fastest
  and easiest photo organizing software," intending to surpass
  iPhoto's ease and capabilities on Windows
  <http://www.preclick.com/pr_20021001.html>

* US Assistant Attorney General for antitrust matters Charles
  James, who brokered the US-Microsoft antitrust settlement that's
  still controversial (and hasn't been approved by the court yet),
  is leaving the Justice Department to become general counsel for
  the massive oil company ChevronTexaco (which itself barely
  cleared antitrust merger hurdles during the Clinton
  administration)
  <http://zdnet.com.com/2100-1106-960649.html>

**press watch**

* Thirty years after Ron Zeigler and the Nixon White House,
  perhaps it's time to end the adage that repeating a lie five
  times makes it truth. In August, Daniel Drew Turner and Matthew
  Rothenberg of _eWEEK_ reported [20] that Apple would "tweak" new
  Macintosh computers to "prevent booting into Mac OS 9," and that
  it would "probably be enforced via a software feature in Pinot,
  the next major Mac OS X update after Jaguar," which had not been
  released at the article's press time. "Apple will most likely
  make the move by January's Macworld Expo/San Francisco."

  [20] <http://www.eweek.com/article2/0,3959,431382,00.asp>

  Of course, Apple did announce it would make new machines that
  don't boot into Mac OS 9, but eWEEK's details appear
  substantially wrong. The story describes machines that could
  boot into Mac OS 9 but are prevented from doing so via a "tweak"
  or software update, and there's no available evidence to support
  that. As MWJ_ pointed out at the time, a far more plausible
  explanation is that Apple will finally advance the motherboard
  to include new technologies that Apple doesn't want to support
  in native Mac OS 9, such as faster FireWire, USB, AGP, or even
  HyperTransport (MWJ_ 2002.09.18). Turner and Rothenberg's own
  article goes to show why disabling Mac OS 9 booting for
  non-technical reasons is ridiculous: large customers are already
  saying they won't purchase those machines. Apple is not in a
  position to turn away eager Power Macintosh customers in a slow
  technology economy.

  However, since the article, _eWEEK_ has published at least three
  more crowing about how Apple's announcement "confirmed a recent
  _eWEEK_ report." The original article was modified to link to a
  new story [21] "in which Apple confirms its plans," and that
  story spends its first two paragraphs talking about how Apple
  was just confirming what _eWEEK__had already reported - even
  though Apple did no such thing. A 2002.09.16 Rothenberg column
  boldly crows about the "scoop", saying, "I hate to say it, but
  we told you so," and "I'm delighted to note that Daniel Drew
  Turner and I were on the money back at the beginning of August,
  when we first flagged Apple's next major step to close the book
  on Mac OS 9."

  [21] <http://www.eweek.com/article2/0,3959,525179,00.asp>

  Rothenberg happily linked multiple times to the original article
  to drive home the point, even though the original did not
  mention advance warning of an announcement, nor does any
  available evidence support the original article's theory of why
  or how this is happening. In fact, had Apple not made the
  announcement so close to the _eWEEK_ article, it would likely be
  like any number of Rothenberg's other "scoops," such as his 1999
  insistence that Apple was about to replace keyboards with
  handwriting recognition, his 2000 assertion that you'd be using
  post-PowerPC machines by now, or his more recent escalation of
  Apple's cross-platform efforts into a "backup plan" for Mac OS
  X. Rothenberg, who formerly wrote MacWEEK.com's "Mac the Knife"
  column, tends to get onto these rumors and hold them for years,
  including the "Apple's about to dump the PowerPC" and
  "handwriting recognition is the wave of the future" memes.

  To an extent, this is harmless braggadocio, but _eWEEK_ is
  starting to change that. The site, and more specifically,
  Rothenberg, are linking to the "confirmed" report about booting
  into Mac OS 9 at every opportunity to lend credibility to other
  rumors that even the Knife likely wouldn't have gotten into
  print. In his most recent article [22], Rothenberg writes that
  "a few [of his scoops] (like the news that January's Macs will
  boot in Mac OS X only) have subsequently been confirmed by
  Apple. Others (such as our report [23] that Apple is working
  closely with IBM to hone the latter company's 64-bit GPUL
  processor for future Mac desktops and servers) should rise
  closer to the surface in the coming months."

  [22] <http://www.eweek.com/article2/1,3959,570988,00.asp>
  [23] <http://www.eweek.com/article2/0,3959,543317,00.asp>

  Should they? Actually, none of Rothenberg's "scoops" have been
  confirmed, unless you stretch "confirmed" on the Mac OS 9
  booting, and if you count InkWell in Mac OS X 10.2 as
  "confirmation" of Rothenberg's years-old reports of imminent
  handwriting recognition in the OS. But by saying "a few" have
  been confirmed and that the rest "should" be, Rothenberg makes
  it sound like he has a strong factual background for a "roadmap"
  he draws based entirely on fancy. He says the decision to
  eliminate Mac OS 9 booting was made by marketing instead of
  engineering, a silly concept because if the machines _could_ do
  it, someone would patch Mac OS 9 to make it work within a few
  weeks. He says that even the Classic environment might be
  discontinued by the end of 2003, but there's no evidence for
  that, either: everyone using Mac OS X picks native applications
  over classic ones, and there's absolutely no reason to cut off
  compatibility to millions of older programs that will never be
  updated.

  Of course, Rothenberg is entitled to his opinion, even if it's
  based on pure imagination (ideas come from _somewhere_, you
  know), but _eWEEK_ is lending gravitas to this house of cards by
  continually referring to how one part of it is "confirmed" by
  Apple. It's no such thing, and the entire "roadmap" is
  speculation that seems to us somewhat ungrounded in reality.
  Repeat it as often as you want - it won't become true.

* Scorn and fie to InfoWorld for muckraking. The mainstream press
  has completely ignored the O'Reilly Mac OS X Conference held
  this week, though sites like MacCentral have covered sessions
  for Mac-oriented readers (such as Stuart Cheshire's session [24]
  on Rendezvous, covering much of the same ground in MWJ_'s recent
  MacCyclopedia series on the topic, MWJ_ 2002.09.09, 2002.09.18).
  Yet when _InfoWorld_'s Paul Krill thought he could create
  controversy, suddenly the X-only conference was newsworthy.
  Krill's article [25] says that former Darwin lead engineer Fred
  Sanchez-Vega "criticized" the Darwin project, but the only
  issues Krill identifies are weak.

  [24] <http://maccentral.macworld.com/news/0210/03.rendezvous.php>
  [25] <http://www.infoworld.com/articles/hn/xml/02/10/03/021003hndarwin.xml>

  He quotes Sanchez-Vega as saying that Darwin is only interesting
  "because it's part of Mac OS X," but that's obvious to a novice
  programmer: Mac OS X is the only commercial operating system
  with a fully open-source backbone that anyone can modify at
  will. Sanchez-Vega also pointed out issues with fixing "easy"
  bugs in Darwin, but he did the same when he was still at Apple.
  It's a necessary side-effect of Apple controlling the major
  releases of Darwin to coincide with major Mac OS X releases.
  That, in turn, is necessary so that Apple's engineers can keep
  working on new features without releasing the source code before
  it's ready. There's nothing new here or even that controversial,
  and it's the only part of the entire conference _InfoWorld_
  chose to cover, looking for discord where there's only agreement
  that the process should improve. Scorn and fie, we say.

**press watch in brief**

* In MacCentral's coverage of Jordan Hubbard's O'Reilly Conference
  keynote speech, note how the most prominent audience questions
  are from people who fundamentally misunderstand Apple Computer:
  they want Mac OS X free, want it to run on any hardware they can
  build themselves, and want Apple's other programs (like iPhoto)
  to be open-source as well; apparently none of the Unix advocates
  actually bother to consider how a hardware company makes money
  on such efforts
  <http://maccentral.macworld.com/news/0210/02.hubbard.php>

* Sigh - Bob "Dr. Mac" LeVitus's mention of how Mac OS X 10.2 is
  winning over converts has the obligatory Ellen Feiss mention (at
  least the commercial has now aired on television in the US)
  <http://www.chron.com/cs/CDA/story.hts/tech/weekly/1592826>

* Even though Apple Computer delivered a major upgrade to Logic
  Platinum for Windows after purchasing Emagic in late June, has
  continued to sell the Windows software until this week, offers
  technical support for another full year, and has offered up to
  US$1300 in incentives to any Logic Windows user that wants to
  switch to Mac OS X - but, unsurprisingly, _Wired_ News's only
  spin is that Apple is "snubbing" PC users who are "furious" that
  running Windows doesn't insulate them from development concerns
  (MWJ_ 2002.07.06)
  <http://www.wired.com/news/mac/0,2125,55333,00.html>

* Once again, _BusinessWeek_ ignores Charles Haddad, its own
  weekly online Macintosh columnist, handing its in-print review
  of Mac OS X 10.2 to Stephen Wildstrom, a reporter with a spotty
  Apple track record who praises the software but mentions
  _nothing_ other than its ability to network with his beloved
  Windows computer, and that Apple "has released so little
  technical information" that Rendezvous is hard to figure out, in
  a column that came out after Apple released the entire
  Rendezvous source code
  <http://www.businessweek.com/magazine/content/02_40/b3802029.htm>

* In his look at Red Hat Linux 8, Hiawatha Bray of the
  _Boston_Globe_ says that Mac OS X is more of a threat to Linux
  than it is to Windows, and quotes a previously-unpublicized Giga
  Information Group study that says 36% of corporate software
  buyers are so fed up with Microsoft's new software licensing
  scheme that they're considering switching to another supplier
  <http://www.boston.com/dailyglobe2/273/business/Mac_poses_as_much_of_challenge_to_Linux_as_to_Windows+.shtml>

* _The_Manila_Times_ conducts an extensive interview with outgoing
  Apple Philippines manager Charlie Tang just before Tang quit,
  leaving the country for Canada
  <http://www.manilatimes.net/national/2002/sept/30/business/20020930bus13.html>

* MacCentral updates you on the future of IEEE wireless networking
  via an article from Stephen Lawson of the site's sister IDG News
  Service
  <http://maccentral.macworld.com/news/0209/30.wlan.php>

* _eWEEK_ says the Xserve is impressive and wins "Rookie of the
  Year" honors, but is not an "MVP" because it uses the PowerPC
  G4, even though one paragraph later, reviewer Henry Baltazar
  admits he has no hard numbers showing the G4 is better or worse
  at server tasks than Intel chips - but it _might_ be (the review
  also criticizes the box for its lack of hot-swappable power
  supplies and fans)
  <http://www.eweek.com/article2/0,3959,562239,00.asp>

* _Forbes_ columnist Stephen Manes, who has written favorably
  about Macintosh products and unfavorably about Windows in recent
  months (MWJ_ 2001.09.26, 2002.08.06) weighs in on "sleek"
  computers, finding the iMac to be the sleekest but also
  significantly slower than the Gateway Profile 4 in tasks
  performed in Photoshop Elements, the first we've heard of this
  potentially-useful benchmark for consumer computers
  <http://www.forbes.com/forbes/2002/1014/162.html>

* Dan Gillmor of the San Jose Mercury News, writing on the same
  topic on which he spoke this week at the O'Reilly Mac OS X
  Conference, says that Apple's refusal to implement "digital
  rights management" in the operating system to lock you out of
  media you may own is "whether by design or by accident, a
  Digital Rights Management operating system where the rights in
  question are the user's rights - and they are expansive."
  <http://www.siliconvalley.com/mld/siliconvalley/business/columnists/4193839.htm>

* Charles Haddad's latest entry for _BusinessWeek_ Online explores
  third-party Office v.X manuals, finding the best to be
  O'Reilly's "Office X for Macintosh: The Missing Manual [26]",
  including documentation of several Office features Haddad
  considers to be "hidden" even though they're in the menu bar
  (like the Document Map)
  <http://www.siliconvalley.com/mld/siliconvalley/4195675.htm>
  [26] <http://www.oreilly.com/catalog/offxtmm/>

* Leander Kahney's latest for _Wired_ News notes that Mac OS X is
  harder to "customize" than Mac OS 9 was, meaning that it's not
  as easy to replace standard interface elements with system-wide
  alternates; Kahney pays lip service to system stability but
  leaves the impression that Apple is just being mean and stingy
  <http://www.wired.com/news/technology/0,1282,55395,00.html>

* MacCentral's write up of its "2002 Standard of Excellence
  WebAward" somehow manages to leave out that the awards come from
  the Web _Marketing_ Association
  <http://maccentral.macworld.com/news/0210/02.webaward.php>

* David Coursey is writing a book "for would-be [Macintosh]
  switcher, intended to help them make the right choice," an
  outcome almost as likely as Microsoft helping you choose the
  right open-source software
  <http://www.zdnet.com/anchordesk/stories/story/0,10738,2882357,00.html>

* David Zeiler's new weekly _Baltimore_Sun_ column on Macintosh
  issues explores iCal, saying that because it's free, its
  "distinguishing features outweigh its flaws," despite "sluggish"
  performance and "many problems," concluding that "for those
  still pondering the move to Jaguar, iCal gives you one more
  reason [to] make the move"
  <http://www.sunspot.net/technology/pluggedin/bal-mac100302,0,5614276.column>

* The Mac Observer takes on a _Portsmouth_Herald_ article that
  incorrectly calls the iPod "one of the many ways to download
  music from the Internet" (that would apparently be through the
  iPod's non-existent Ethernet jack), though the article has since
  been corrected
  <http://www.macobserver.com/article/2002/10/03.11.shtml>

**internet news in brief**

* IDC says that over 31,000,000,000 E-mail messages move across
  the Internet every day, with "about two-thirds" being
  person-to-person communications; by 2006, it will increase to
  60,000,000,000 messages per day with "a little over half" being
  person-to-person and the rest being automated, including both
  legitimate notifications (like stock alerts) and the dreaded
  spam
  <http://maccentral.macworld.com/news/0209/27.mail.php>

* As noted by MacMinute and other sites, MacMusic, the site for
  music and audio creation on Macintosh computers, recently
  celebrated its fifth anniversary (we believe that's the
  "subwoofer anniversary" in this context)
  <http://www.macmusic.org/>

* On the 2002.10.02 edition of "Your Mac Life with Shawn King":
  Toast product manager Adam Fingerman, Tom Negrino (author of the
  shiny and toothsome "Microsoft Office v.X Inside Out"), and
  Peter Cohen
  <http://www.yourmaclife.com/>

* GoBlox Technologies opens "My iCal.com," an Xserve-hosted site
  that provides free WebDAV access for publishing your iCal
  calendars without requiring .Mac services
  <http://www.myical.com/>

* The "MacRaffle" site is offline, at least temporarily, while the
  Ohio Attorney General's office investigates whether the site was
  truly offering raffles of chance, which are illegal in Ohio (and
  most states) without special dispensation
  <http://www.macraffle.com/>

* The Wireless Ethernet Compatibility Alliance renames itself the
  "Wi-Fi Alliance" and opens a new Web site at a trendier URL
  <http://www.wi-fi.org/>

* The October 2002 issue of _Web_Page_Design_for_Designers_
  explores the useful CSS "box" element, with looks at Mozilla's
  powerful built-in CSS editor and a mini-review of Netscape
  Composer
  <http://www.wpdfd.com/editorial/wpd1002.htm>

* Mac-related Internet audio gets a bit more competitive again as
  MacRadio premieres "Andy Ihnatko's Weekly Write-Off", a new
  half-hour Internet radio show featuring America's 42nd Most
  Beloved Industry Figure
  <http://www.macradio.com/thursday/>

* Former MacOpinion and AppleLinks columnist John Martellaro, who
  stopped those gigs when he became Apple's senior marketing
  manager for science and technology on 2000.08.21 (MWJ_
  2000.08.19), is no longer with Apple and is now writing for
  OSXFAQ.com
  <http://www.osxfaq.com/Editorial/sci_tech/>

**apple's knowledgebase**

* Although Apple Computer has gotten better over the years at
  blending the differences between NextStep and classic Mac OS
  programming models, there are still areas where it's glaringly
  obvious that Apple's current NeXT-led programmers reject
  Apple-invented technologies for no particular reason. Such has
  it been with Internet Config [27], the public-domain system for
  storing your Internet-related configuration information.

  [27] <http://www.quinn.echidna.id.au/Quinn/Config/>

  Peter Lewis and Quinn "The Eskimo!" invented Internet Config in
  the mid-1990s because dozens of Net-related applications had to
  ask you for the same information: your E-mail address, your
  preferred home page, your signature, your mail server, and more.
  Most conveniently, Internet Config records your choice of
  protocol helpers, so every program knows what E-mail program,
  Web browser, FTP client, or even telnet software you want to
  use. It also records what applications map to MIME types and
  file name extensions so files can seamlessly move between the
  HFS-style file type and creator type metadata to the Unix world.

  With some evangelism by long-time Macintosh stalwarts, Internet
  Config really caught on, in no small part because Microsoft
  chose it to manage most preferences in Internet Explorer. It
  became part of the system with Mac OS 8.5 as the "Internet"
  control panel, and because so many Carbon applications require
  it, it's part of Mac OS X as well and always has been. And yet,
  time and time again, Apple refuses to use Internet Config in its
  Mac OS X code and barely acknowledges that it exists.

  Internet Config already contained a complete and useful mapping
  of file name extensions to creator types, completely capable of
  mapping almost any file you'd need to send to a Windows machine,
  and yet Apple's management wouldn't even consider using it
  instead of forcing file name extension on every program, no
  matter how grossly inappropriate they may be. You do not need
  Omni Outliner documents to end in ".ooutline" because there is
  no Windows version to exchange files with. You definitely don't
  need your AirPort Admin settings saved in ".basestation" files.
  Amazingly, Apple recommends [28] with a straight corporate face
  that you should search for AppleWorks files by searching for
  files with the extension ".cwk", as in "ClarisWorks," saying
  that file type "is not an accurate way to search for files,"
  even though that's only true because Apple told programmers to
  make it that way. It's a crock. Internet Config knows this. Avie
  Tevanian is clueless.

  [28] <http://docs.info.apple.com/article.html?artnum=107005>

  Or maybe not: since Internet Config so clearly disproves the
  NeXT notion that file system metadata must be eliminated, Apple
  may be deliberately deemphasizing the technology. Even though
  Internet Config used an open and understandable way to map file
  name extensions to launching applications, Apple invented a new,
  secret, proprietary way in "Launch Services," the undocumented
  method the Finder uses that you can only repair by deleting all
  its files. According to OmniWeb release notes [29], the system
  isn't even interested anymore in making sure Internet Config
  works the right way: "We now ask Launch Services which
  application handles a particular URL scheme before falling back
  on asking Internet Config. (This means that on 10.2 we now
  correctly determine the preferred Mail application and also
  correctly pass webcal URLs off to iCal, since apparently 10.2 no
  longer tries to keep Launch Services and Internet Config
  synchronized.)"

  [29] <http://www.omnigroup.com/applications/omniweb/releasenotes/>

  If you're waiting for Apple to make Internet Config do the right
  thing, don't hold your breath. Fortunately, you're not dependent
  on Apple: Internet Config is in the public domain, and anyone
  can improve it or even reimplement it. That's not necessary at
  the moment: the Mac OS X version contains [30] all of the
  "relevant" thread-safe Internet Config APIs, except those that
  support multiple preference files. But no one has stepped up to
  port the configuration program to Mac OS X, either, and Apple
  sure doesn't seem interested in letting you control Internet
  Config. Then you might want to _use_ it, and that would ruin
  everything. Look at how all the traditional Mac OS developers
  refusing to abandon file types in their Carbon applications has
  kept your disks from being overrun with idiocy so far. Apple
  sure doesn't want to repeat that mistake.

  [30] <http://developer.apple.com/techpubs/macosx/Carbon/networkcomm/InternetConfig/Internet_Config/Miscellaneous/CarbonCompatibility.html>

  Yet someone snuck something useful into Mac OS X 10.2, according
  to this KnowledgeBase article [31]: the Internet control panel
  in Classic, and therefore the Internet Config application,
  changes your Mac OS X settings. In other words, the Classic
  environment uses your native Internet Config settings, and
  therefore changes in Classic percolate back up to the native
  environment, so you finally have a way to change them. That's
  why trying to create a new preference set gives you an error
  -677: the native implementation supports only one preference
  file, so creating a new one is kind of off limits. When you boot
  into Mac OS 9, you'll still get the same, separate Internet
  Config sets you would have before - otherwise, Mac OS 9 would
  try to launch Mac OS X native browsers like OmniWeb or Chimera
  if they were your Mac OS X choices - but while you're in
  Classic, you get one unified settings architecture you can query
  or change from either Classic or native applications. Now if
  Launch Services would look to Internet Config first, the world
  would be a happier place.

  [31] <http://docs.info.apple.com/article.html?artnum=107093>

* The outcome isn't always great: sometimes, when Apple's
  wholesale abandonment of its own technology causes both
  incompatibility and embarrassment, the company apparently
  chooses to lie about it. KnowledgeBase article #107092 says [32]
  that there are two choices for saving compiled AppleScripts in
  the Mac OS X 10.2 Script Editor: "compiled script" and "Compiled
  Script". Apple says that these choices "produce identical
  scripts. It does not matter which format you choose; they only
  differ in the capitalization." This is very much not how it
  works. The lowercase "compiled script" option saves the script
  data in the resource fork of the saved file, just as Script
  Editor (and, of course, ever other scripting editor) has always
  done. The capitalized "Compiled Script" option saves the script
  data in the saved file's data fork, something not done before
  Mac OS X 10.1. Most applications that load scripts for their own
  use can't read the data-fork format, and so it's important for
  now to use the lowercase "compiled script" option in Script
  Editor, or to use a third-party program like Script Debugger
  that is more concerned with accuracy than politics. The issue
  has come up, among other places, on the Mailsmith-Talk mailing
  list, and despite several loud protests, Apple has not corrected
  the misinformation in the two-week-old KnowledgeBase article.
  Don't be fooled by it.

  [32] <http://docs.info.apple.com/article.html?artnum=107092>

**apple's knowledgebase in brief**

* As many Power Macintosh G4 (Mirrored Drive Doors) owners have
  discovered, the "Software Restore" discs with the new computers
  are substantially different than those with earlier models, and
  are the only way to restore Mac OS 9.2.2 or iDVD onto the
  computer - but they require the computer booted into Mac OS X
  10.2 before you can proceed
  <http://docs.info.apple.com/article.html?artnum=42929>

* Mac OS X 10.2 limits the amount of RAM shared among all Classic
  applications to 128MB or less
  <http://docs.info.apple.com/article.html?artnum=61528>

* You need DVD Studio Pro 1.2.1 or 1.5.1 updates to properly
  encode Macrovision on discs; 1.2 or 1.5 won't do it correctly
  <http://docs.info.apple.com/article.html?artnum=61634>

* If you don't want to use the iSync beta but want iCal's
  calendars on your iPod, Apple explains how to copy the files to
  the right iPod location yourself
  <http://docs.info.apple.com/article.html?artnum=61708>

* iCal saves your calendar files in "~/Library/Calendars"
  <http://docs.info.apple.com/article.html?artnum=61721>

* If you're using Final Cut Pro 3 with an ATI Rage 128 Pro or
  Radeon video card in thousands of collars, you may see green or
  blue tints on video material rendered in real time; to fix it,
  set your display to millions of colors
  <http://docs.info.apple.com/article.html?artnum=61772>

* More fun with units: Apple lists the minimum and maximum power
  consumption figures for single-processor and dual-processor
  Xserve units in Watts, and both idle and maximum thermal output
  in "BTU/hr", a unit that's the same thing as - Watts
  <http://docs.info.apple.com/article.html?artnum=75383>

* The Power Macintosh G4 (Mirrored Drive Doors) is not designed
  for professional audio equipment on its input and output jacks,
  so you'll hear a buzzing or humming sound unless you run
  XLR-connected equipment through a stereo isolation transformer
  <http://docs.info.apple.com/article.html?artnum=75408>

* The Backup program available to .Mac members "has only been
  tested to work with internal CD-R and DVD-R (SuperDrive) drives
  that come with Apple computers ... Backup may work with many of
  the same devices that work with iTunes, but they have not been
  tested and thus do not receive technical support"
  <http://macworld.co.uk/news/main_news.cfm?NewsID=5340>

* If pressing the power button on your eMac makes a popping sound
  and generates flashes both inside the case and on the CRT, and
  internal video then stops working, Apple says to take the
  computer to an authorized service provider (you'd think that
  would be advice for _any_ kind of "popping sound" associated
  with flashes, wouldn't you?)
  <http://docs.info.apple.com/article.html?artnum=95169>

* If setting up login items in Mac OS X - items the system
  launches for you when you log in - confuses you, Apple's set of
  six simple steps for creating login items with six paragraphs of
  footnotes is a good indication of why
  <http://docs.info.apple.com/article.html?artnum=106146>

* Apple shows what kernel panics look like both before and after
  Mac OS X 10.2, including how to find the log of a 10.2-or-later
  kernel panic in "/Library/Logs/"
  <http://docs.info.apple.com/article.html?artnum=106227>

* Similarly, Apple updates instructions on how to log and submit
  kernel panic information to accommodate changes in Mac OS X 10.2
  <http://docs.info.apple.com/article.html?artnum=106228>

* Mac OS X 10.0 through 10.1.5 change home folders of deleted
  users to belong to the "root" user in case a system
  administrator wants to look at them; Mac OS X 10.2 instead makes
  a disk image of the user's folder and deletes the actual folders
  so they don't hang around the "/Users" folder
  <http://docs.info.apple.com/article.html?artnum=106256>

* If connecting to SMB (Windows file sharing) servers in Mac OS X
  confuses you, Apple's set of four simple steps for doing so with
  twelve paragraphs of footnotes is a good indication of why
  <http://docs.info.apple.com/article.html?artnum=106471>

* Mac OS X has always required that you install the "optional"
  (but highly-recommended) BSD Subsystem if you want to use FTP or
  Remote Login (SSH or telnet), but in Mac OS X 10.2, it's also
  required if you want to use Internet Sharing
  <http://docs.info.apple.com/article.html?artnum=106472>

* The hidden Print Center option to delete all printers and print
  jobs (by holding down Option as you quit the program) is removed
  in Mac OS X 10.2, but you can still remove multiple printers or
  jobs by Command-clicking on each one in the "Printer List"
  window
  <http://docs.info.apple.com/article.html?artnum=106611>

* If trying to change a Mac OS X user's "short user name" (the one
  that his "home" folder is named after) confuses you, Apple's set
  of _thirty-nine_ simple steps for doing so with three footnotes,
  three notes, and one warning should be enough to convince anyone
  that sometimes it's "The power of Unix, the ease of Macintosh"
  and sometimes it's just the other way around
  <http://docs.info.apple.com/article.html?artnum=106824>

* If a disk doesn't appear as a choice in the Mac OS X 10.2
  Installer, Apple recommends repairing it with Disk Utility, then
  restarting, then disconnecting all third-party devices, then
  restoring any jumpers on the hard drive to their original
  settings, and finally erasing the drive (though you can stop
  after any step that makes the drive show up in the Installer)
  <http://docs.info.apple.com/article.html?artnum=106960>

* If Help Viewer crashes shortly after launching in Mac OS 10.2,
  upgrade to Mac OS X 10.2.1
  <http://docs.info.apple.com/article.html?artnum=107013>

* If you try to copy a file to a Windows file server using Mac OS
  X 10.2, you'll get error -43 if the file you're copying has any
  of the "?", "[", "]", "/", "\", "=", "+", "<", ">", ";", ":",
  "|", or "*" characters in the name (also the double hash mark
  ["] or comma [,] characters), as HFS Plus or UFS allows all of
  these characters in file names but Windows does not
  <http://docs.info.apple.com/article.html?artnum=107032>

* Don't install Apple Display Software 2.1.1 or earlier on top of
  Mac OS X 10.2; that OS already includes equivalent software, and
  the duplication makes the computer restart to a blue or gray
  screen (perhaps depending on your side of the Mason-Dixon line
  in the US)
  <http://docs.info.apple.com/article.html?artnum=107041>

* If AppleScript isn't your cup of tea, Apple recommends using
  drag-and-drop from Microsoft Entourage to import contacts into
  Mac OS X 10.2's Address Book
  <http://docs.info.apple.com/article.html?artnum=107051>

* Sherlock 3 lets you disable cookies, but in another stroke of
  "left hand, meet the right hand," doing so prevents the program
  from retrieving any AppleCare documents (all the _third-party_
  channels work fine, of course...)
  <http://docs.info.apple.com/article.html?artnum=107056>

* Mac OS X 10.2 does not include built-in drivers for PCMCIA (PC
  Card) Bluetooth adapter cards; contact your card's manufacturer
  for drivers if they didn't come with the card
  <http://docs.info.apple.com/article.html?artnum=107055>

* Mac OS 9 clients can't connect to Mac OS X 10.2 printer sharing,
  but you can run Mac OS 9's printer sharing service at the same
  time (although trying to share the same printer simultaneously
  through both protocols is asking for trouble)
  <http://docs.info.apple.com/article.html?artnum=107060>

* In Mac OS X 10.2, Preview properly asks you for a password when
  viewing a password-protected PDF file, closing a noted document
  security hole in earlier versions of Mac OS X (though it's not
  yet 100% clear to us that _all_ PDF-viewing APIs in Mac OS X
  respect the password or if it's just a Preview thing)
  <http://docs.info.apple.com/article.html?artnum=107064>

* If you're one of the many Mac OS X 10.2 users who finds Help
  Viewer unbearably slow to launch, Apple suggests just sticking
  with it and launching it a few times, allowing it to download
  new help content from the Internet; Apple also suggests turning
  off all network ports that aren't your primary Internet
  connection to help it along
  <http://docs.info.apple.com/article.html?artnum=107070>

* Ah, the real story: the dialog box in Mac OS X 10.2 that warns
  you certain kernel extensions "may reduce the security of your
  computer" appears when a kernel extension (a package, remember)
  has privileges that let anyone other than the "root" user change
  any of its files; "Fix and Use" changes the permissions and
  loads the kernel extension; it has nothing to do with hardware
  access or unsupported APIs
  <http://docs.info.apple.com/article.html?artnum=107072>

* If your local network supports reverse DNS, Mac OS X 10.2's
  "Connect to Server" dialog box uses reverse DNS lookup names for
  Windows file servers, reverting to the user-defined SMB
  workgroup name only if reverse DNS fails
  <http://docs.info.apple.com/article.html?artnum=107085>

* If turning on Windows file sharing on your Mac OS X 10.2
  computer (so Windows clients can connect to it) confuses you,
  Apple provides 20 simple steps in three categories with seven
  paragraphs of footnotes to make it easier
  <http://docs.info.apple.com/article.html?artnum=107085>

* Ah, the art of error messages: when Mac OS X 10.2 tells you "No
  file services are available at the URL" of a Windows file
  server, it means to say, "be sure you typed your password with
  correct uppercase and lowercase letters"
  <http://docs.info.apple.com/article.html?artnum=107084>

* The Mac OS X 10.2 Finder doesn't automatically refresh Windows
  file sharing windows, so if you want to see files added or
  deleted, force an update by adding or removing a file yourself
  (or by logging out of and into the server again)
  <http://docs.info.apple.com/article.html?artnum=107087>

* If you get the specific error "-50" when connecting to a Windows
  file sharing server, Apple says it's probably because whoever
  shared the volume on Windows didn't do it correctly, so Apple
  explains how to share a disk under Windows (the world is indeed
  becoming a strange place when the Tech Info Library is not only
  renamed "KnowledgeBase," like Microsoft's library, but starts
  solving Windows problems)
  <http://docs.info.apple.com/article.html?artnum=107095>

* Now that Mac OS X 10.2 repositions desktop icons when you change
  screen resolutions, exiting a full-screen game may leave Dock
  and desktop icons in the wrong place - if the game restores the
  original resolution at all (use System Preferences or the
  Display menu extra if it doesn't)
  <http://docs.info.apple.com/article.html?artnum=107094>

* If Mac OS X 10.2's "Internet" pane loses your .Mac password,
  open Keychain Access and delete the "iTools User" password item
  (if it exists), then create a new password item with your .Mac
  account information
  <http://docs.info.apple.com/article.html?artnum=107097>

* Mac OS X 10.1.5 and earlier used a customized set of tcsh
  settings in Terminal by default, derived by long NeXT tradition;
  Mac OS X 10.2 uses standard settings instead, but you can see
  the customized set at "/usr/share/tcsh/examples/README"
  <http://docs.info.apple.com/article.html?artnum=107106>

* Either updating to Mac OS X 10.2.1 or optimizing an HFS (not HFS
  Plus) volume may panic the kernel next time you restart the
  computer; if so, start up cleanly and delete the file
  "/var/db/BootCache.playlist" (this is easier to do in single
  user mode, but props to Apple for not requiring a command-line
  solution)
  <http://docs.info.apple.com/article.html?artnum=107111>

* Browsing Windows file servers (SMB browsing) in Mac OS X 10.2 is
  limited to your local subnet; you'll need to use URLs or IP
  addresses to connect to any Windows machines past the local
  router (though Windows itself doesn't have this problem)
  <http://docs.info.apple.com/article.html?artnum=107117>

* You can't install Mac OS X Server 10.2 on top of Mac OS X 10.2.1
  because the Installer objects at finding newer 10.2.1 versions
  of some Mac OS X Server 10.2 components; Apple explains how to
  fix it by moving the receipts for the French, German, Japanese,
  and BSD installation packages to a different folder that won't
  make the Installer choke
  <http://docs.info.apple.com/article.html?artnum=107124>

* Mac OS X Server 10.2 comes with an "Admin Tools" installer that,
  when run on regular Mac OS X 10.2, installs copies of QuickTime
  Streaming server, Workgroup Manager, Server Status, Server
  Settings, Server Monitor, Macintosh Manager, Network Image
  Utility, Open Directory Assistant, and Server Assistant
  <http://docs.info.apple.com/article.html?artnum=107125>

* If you run Mac OS X Server 10.2's Open Directory Assistant on a
  machine with Japanese as the primary language, it can only
  connect to a Mac OS X Server installation that also has Japanese
  as the primary language; Apple offers tips on how to change the
  server's primary language locally or remotely (believe it or
  not, there's now a command-line tool for this)
  <http://docs.info.apple.com/article.html?artnum=107126>

* Mac OS X 10.2's Windows file sharing implementation uses NetBIOS
  over TCP on ports 137, 138, and 139, as Apple explains with
  pointers to more well-known ports
  <http://docs.info.apple.com/article.html?artnum=107127>

* Mac OS X 10.2 will no longer guess which Mac OS 9 System Folder
  you want to use for the Classic environment: the first time it
  launches, you'll have to pick one if multiple valid bootable
  folders are available
  <http://docs.info.apple.com/article.html?artnum=107128>

* Ken Bereskin's Weblog says [33] that iChat peer-to-peer file
  transfer works when you "just drag-and-drop any document from
  the Finder to an online user on your buddy list," and that's how
  you have to do it: clicking the paper clip icon to send a file
  in iChat doesn't work until you've sent a text message to that
  buddy first
  <http://docs.info.apple.com/article.html?artnum=107129>
  [33] <http://radio.weblogs.com/0100676/2002/08/26.html>

* Error "1282" in Print Center in Mac OS X 10.2 means that, due to
  confluence of the stars and planets, your upgrade to 10.2 from
  10.1 did not create a "/var/spool/cups" directory belonging to
  the user daemon and the group admin with the right privileges;
  creating it yourself (with supplied instructions) stops the
  error and magically makes existing print queues reappear
  <http://docs.info.apple.com/article.html?artnum=107054>

**what a deal!**

* Apple Computer and Microsoft are addressing concerns about the
  price of Office by, well, cutting it drastically. The new
  "Office Party" promotion [34] lets you purchase Microsoft Office
  v.X for only US$199 as long as you purchase it at the same time
  as any new Macintosh (on the same invoice, from the same Apple
  authorized reseller in the United States). That's a savings of
  up to US$300, though only for people who are buying new
  computers (one copy of Office per qualifying computer), but
  that's not so bad: most people who are going to use Mac OS X on
  existing hardware have already upgraded in the year and a half
  since the OS was released. Most new Mac OS X users get that way
  when they buy powerful new hardware, and that's a good segment
  for Apple and Microsoft to target. Microsoft has its own page
  [35] listing several dealers you can contact, and also quoting
  Apple's Phil Schiller, something Apple's page does not do. Then
  again, Apple's "Office Party" has better typography.

  [34] <http://www.apple.com/promo/office/>
  [35] <http://www.microsoft.com/mac/officex/prodinfo/OfficePartyPromo.asp>

**what a deal in brief!**

* Pixels Digital offers steep discounts on pre-orders for the Mac
  OS X native Pixels 3D 4.1, to be released at the end of this
  month; new copies are available for US$399 (instead of US$799);
  upgrades from version 3.7 go for US$199 instead of US$399
  <http://store.pixels.net:2001/cgi-bin/WebObjects/Store.woa>

* Buy ProSQL Personal or Enterprise edition through 2002.10.07 and
  get a free copy of REALbasic 4.5, upgradable to version 5.0 (see
  "Product Showcase," this issue)
  <http://www.intellisw.com/prosql/fall2002promo.html>

* We told you that huge discounts on StuffIt Deluxe 7 would soon
  appear, and here's the first, offering the US$80 program for
  US$30 (the advertisement we received quoted our own "Delivery
  Dock" feature on the program, so we think quoting the offer is
  fair)
  <http://www.digitalriver.com/aladdinsys/62476>

* Discreet is bundling its Cleaner media compression program and
  Combustion rotoscoping tool for US$1497 through 2002.12.31,
  savings of over US$500
  <http://www.discreet.com/resellers/bundle/>

* Internet service provider Speakeasy is seeking new
  business-class broadband customers, offering a free 5GB iPod to
  new customers who sign up between 2002.10.03 and 2002.10.31
  <http://www.speakeasy.net/ipod/>

**other macintosh news**

* The Macintosh world has been in mystery and suspense for months,
  wondering which unusual and key luminary will be selected to
  deliver the keynote speech at January's Macworld Expo in San
  Francisco. After all, stretching back to July 1997's show in New
  York, the keynote spot has been occupied by a variety of
  fascinating and useful speakers promoting a variety of...you
  know we're full of it. Every US Macworld Expo since then, and
  most of the Tokyo expos, have featured keynote speeches by Steve
  Jobs, so absolutely no one is surprised at this week's
  announcement [36] that Jobs is keeping the job. The speech
  starts at 9:00 AM on 2003.01.07 (1800 GMT). The rumor mill is
  absolutely certain that Jobs will announce some kind of new
  hardware, or maybe some software.

  [36] <http://www.macworldexpo.com/macworld2003/V33/press.cvn?id=11&p_id=4>

**other macintosh news in brief**

* Apple France removes the iPod from store shelves after learning
  that its maximum volume, over 100dB, is higher than allowed by
  French law; the company has updated and replaced unsold units;
  they'll go back on sale next week, and Apple will release a
  software patch for existing iPods on 2002.10.08
  <http://news.com.com/2100-1040-960211.html>

* StuffIt 6.5.2 and earlier products that decompress Zip files may
  get a buffer overflow when unzipping "maliciously crafted"
  archives with very long file names inside; Aladdin recommends
  the free upgrade to StuffIt Expander 7, which does not have the
  newly-discovered issue
  <http://www.stuffit.com/expander/cert.html>

**developer news**

* One of Apple's dirty little technology secrets leaked out this
  week as the company invited developers to attend the J2SE 1.4.1
  on Mac OS X Kitchen [37], to be held 2002.10.29-2002.10.31 in
  Cupertino. That's "J2SE" as in "Java 2 Standard Edition," the
  current user-level version of Java (the beefier one is "J2EE"
  for "Java 2 Enterprise Edition." Sun first announced [38] Java
  1.4 in May 2001, with the first versions shipping late that same
  year. Java developers were hot for it because it includes XML in
  the core foundation for better work with Web services,
  accelerated Java2D graphics, more security, new classes, and all
  manner of other good stuff.

  [37] <http://adckitchen.apple.com/?Event=Java20021029>
  [38] <http://java.sun.com/pr/2001/05/pr010529-01.html>

  And yet Apple, despite flogging the "we have the most advanced
  Java of any operating system" horse since mid-2000, was silent.
  Developers asked about Java 1.4 implementation, and all Avie
  Tevanian would say is that Apple would have something to say
  about that someday. It even came up at the Xserve introduction
  event for the press in May.

  Some people knew the truth, but everyone abided by their
  non-disclosure agreements even though it was incredibly
  disappointing: it's taken Apple almost an extra year to make
  Java 1.4.1 available because, due to executive fiat, Apple's
  limited number of Java engineers have had to spend the past year
  rewriting the entire Java VM so that it calls Cocoa instead of
  Carbon.

  No, we're not kidding.

  Java lives on top of an operating system. When Java displays a
  menu, it calls through to the Mac OS, or Windows, or GNOME, to
  display the menu in the computer's native appearance. When it
  wants to read or write files, it calls through to the underlying
  operating system. Apple's implementation of Java has always
  called through to the underlying Mac OS, and the Mac OS X
  version has always called the Carbon APIs since it was an easy
  port from Mac OS 9 Java. (After all, that's supposedly why
  Carbon is there, remember?)

  So now, developers have been kept away from Java 1.4 for about a
  year because Apple management considers it important that Java
  routines to open a file call the Cocoa APIs instead of the
  Carbon File Manager. It's somehow vital that Java programs
  displaying windows have to open new NSView objects instead of
  allowing Java to call the Carbon Window Manager, even though -
  and this is the fun part - Cocoa calls through to the same
  Window Manager itself. Yes, that's right, Apple management has
  delayed a vital standards-based piece of software for many
  months to add another level of indirection to it. The magic of
  Cocoa!

  In the outline for the kitchen (a kind of developer workshop),
  Apple says it plans to explain technical reasons for the
  transition and future benefits, and there may be some. Both Java
  and Cocoa are object-oriented systems with much in common. Not
  many people are writing Cocoa applications in Java because it's
  something of a poor-performing hack that even Cocoa experts
  recommend against, and making Java more Cocoa-friendly could
  solve that, making it easier for Java programmers to write true
  Cocoa applications and, therefore, better and faster Mac OS X
  software.

  On the other hand, there are a ton of "transition issues" for
  existing Java developers, for all Carbon applications calling
  Java, and more. Apple has some sales to do on this issue,
  though. Most Carbon code rewritten in Cocoa is small code
  without lots of dependencies (that is, no external APIs) that's
  shifted because it's easier to maintain. Apple says J2SE 1.4.1
  will be easier to maintain, but that's about the only way it
  qualifies for the conversion test. If Apple can't sell huge
  benefits for this delay, including substantial performance
  increases for all applications (not just Cocoa Java programs),
  it has some explaining to do.

**developer news in brief**

* Derrick Story of the O'Reilly Network interviews Dan Wood of
  Karelia Software, author of Watson, with plenty of focus over
  the Sherlock 3 fracas; Wood says that while he's gotten good
  help from individuals or small groups at Apple, he questions
  Apple's habit of poaching small utilities for the OS, and
  wonders if ADC management has interest in any but the largest
  developers (the answer to that, of course, is "no": Apple's 1998
  developer reorganization, still in effect, was specifically to
  focus on the top 100 developers)
  <http://www.macdevcenter.com/pub/a/mac/2002/09/26/dev_osx.html>

* Apple's OpenGL documentation page now includes specifications
  for 33 OpenGL extensions implemented in Mac OS X 10.2
  <http://developer.apple.com/opengl/extensions.html>

* Continuing an unfortunate trend of keynote speakers who are at
  best tangentially related to Macintosh development, the 2002
  MacHack conference keynote speaker is Ken Arnold, an architect
  of Sun's JINI platform (that doesn't run on the Macintosh), lead
  engineer on JavaSpaces (that doesn't run on the Macintosh), and
  a BSD expert (and hence the tangent)
  <http://www.machack.com/pr/10_1_02.html>


Product Showcase: This Week's Top 25
------------------------------------

**1. iSync Beta**

  [X] It's been a week since Apple released the iSync beta version
  [39] for testing, and we're still not really sure it's worth
  playing with. First announced at Macworld Expo in New York back
  in July, iSync is supposed to be the personal information center
  of the digital hub, synchronizing your Address Book and iCal
  information with _all_ of your digital devices: your iPod, your
  Palm OS organizer, your Bluetooth-enabled cell phone, and even
  between work and home computers using a .Mac account. You'll
  keep all your personal information on one computer, and merely
  by launching iSync, you'll keep it up-to-date with all of your
  digital devices and remote computers.

  [39] <http://www.apple.com/isync/>

  That's the theory. The reality is more sobering. Although iSync
  installs three private frameworks ("SyncConduit," "SyncEngine,"
  and "SyncServices"), five kinds of synchronizers in
  "/System/Library/SyncServices" (meaning they belong to Mac OS X
  itself), and a synchronizing server application at
  "/System/Library/CoreServices/SyncServer.app" in addition to the
  iSync application itself, the program is remarkably inflexible.
  It can only synchronize Address Book and iCal settings, with no
  options for third-party PIMs at this time (no Palm Desktop, no
  Entourage, no Now Up-to-Date & Contact, no Chronos Organizer).
  It can't initiate a HotSync for a Palm OS device, requiring you
  to press the button on the HotSync cradle, just as Palm Desktop
  4.0 requires. In fact, iSync requires Palm Desktop 4 if you want
  to synchronize with a Palm OS device, although it then requires
  that you disable any actual synchronization [40] with any Palm
  Desktop conduits, including "notes" and "memos" that iSync
  doesn't support.

  [40] <http://docs.info.apple.com/article.html?artnum=61634>

  That all-or-nothing attitude pretty much pervades the beta
  iSync. You must have both Mac OS X 10.2.1 and iCal installed,
  and iCal _must_ be in the "/Applications" folder or iSync's
  installer goes meshugga. This is because Apple again relies on
  the pathname-dependent NeXT Installer framework that is perhaps
  50% as functional as the Apple Installer for Mac OS 9 was five
  years ago (Apple's upper engineering management has yet to learn
  that it's no plus to rely on a framework to do the work for you
  when the framework does it _badly_). The release notes [41] show
  lots of warnings that the program may duplicate or corrupt data,
  lots of holes where it doesn't yet work, and Apple has a
  separate list [42] of almost two dozen more warnings about ways
  iSync beta doesn't work as it should.

  [41] <http://docs.info.apple.com/article.html?artnum=61747>
  [42] <http://www.info.apple.com/usen/isync/>

  Of course, it's beta software, even though Steve Jobs said in
  July that iSync would be "released" in September (the beta
  version was released on 2002.09.28). It's bound not to work in
  some instances, and it may not work in others. However, iSync
  will only truly become useful to everyone when Apple fulfills
  more of the potential that Andy Ihnatko wrung out of product
  managers at the show. When there's a software development kit,
  then companies like Microsoft and Now Software and Chronos can
  start making their programs work with iSync. So can companies
  that want to synchronize Pocket PCs, or other kinds of cell
  phones, or whatever you may have.

  That program would be the center of a digital information hub.
  This program copies your Address Book and iCal information to
  several devices, and if you're lucky, can copy them back after
  you change them without creating duplicates or trashing the
  database (iSync provides several backup options for your data in
  case a synchronization someday eats your data). This beta
  version is useful only to people using Mac OS X 10.2.1 who keep
  personal information in Address Book and iCal, and who also want
  it copied to one of the four supported Ericsson or Sony Ericsson
  phone, an iPod, a Palm OS device, or another computer that can
  also use a .Mac account. If that's not you, don't feel like
  you're missing anything. If it is you, use iSync's backup
  options to protect yourself against beta problems.

  And be prepared for bumps in the road; Apple doesn't have
  everything completely thought through. We swear that this is a
  verbatim quote from the iSync 1.0 Beta Release Notes file: "This
  software is beta software. You should back up important data
  before you install and use it. To back up your data, you can use
  the menu command in iSync." Yes, that's right: Apple is telling
  you to use an iSync command to back up data
  _before_you_install_or_use_ that same iSync. The final version
  of this could take a while.

**2. AirPort 2.1.1**

  [X] Update to AirPort software only for Mac OS X 10.2 and later
  includes the latest v.4.0.7AirPort Base Station firmware for the
  "snow" (Dual Ethernet) model, works better with "some" ISPs
  (like Verizon) can disable configuration over the Ethernet (WAN)
  port, allows quoted or hexadecimal-escaped passwords to enable
  Internet Sharing, adds 128-bit encrypted computer-to-computer
  network setup in the AirPort menu extra, plus other bug fixes (a
  separate AirPort 2.0.5 release [43] for Mac OS X 10.1.5 and
  earlier provides the updated firmware and improved PPPoE status,
  plus the Verizon fix)

  [43] <http://docs.info.apple.com/article.html?artnum=120154>
  <http://docs.info.apple.com/article.html?artnum=120121>

**3. AOL Instant Messenger 4.5.863**

  [9][X] America Online's original instant message client gets
  various bug fixes, moves your personal files from the
  "/Users/Shared" folder to your own user folder under Mac OS X,
  and now supports long filenames
  <http://www.aim.com/get_aim/mac/latest_macosx.adp>

**4. Eudora 5.2b10**

  [9][X] Our last sighting of a new Eudora 5 release (with more
  than just localization changes) was over four months ago, with
  the long-delayed final release of Eudora 5.1.1 (MWJ_
  2002.06.01). That version ended a 14-month beta cycle by
  disabling some features Qualcomm couldn't get working natively
  in Mac OS X, such as URL Access (still broken in the OS) and SSL
  (broken on Qualcomm's end). Now Eudora is back in the
  pre-release cycle with Eudora 5.2b10 [44] for Mac OS 8.1 through
  9.2.2, and separately for Mac OS X. The release notes include a
  long list of changes, including fixing SSL on Mac OS X (10.2 and
  later), fixing a lot of toolbar button problems, more settings,
  options to not display PDF files inline (on Mac OS X), and lots
  more. They're all minor changes, but they'll add up for intense
  Eudora Eusers. Although the beta page says commercial release is
  expected in "mid-October," Qualcomm doesn't say of what year,
  and the beta has no expiration date built into it. Even so, the
  company recommends not purchasing a one-year license for the
  beta version, but instead urges you to wait for 5.2 final.
  Eudora remains free in "light" mode, as well as in the
  full-featured but ad-driven "Sponsored" mode. The full feature
  set without ads costs US$40 per year.

  [44] <http://www.eudora.com/betas/>

**5. OmniWeb 4.1.1**

  [X] The Omni Group's Cocoa-based US$30 Web browser is now final
  with RealOne player support, improved HTTPS and SSL support,
  Swedish and Finnish localizations, and bug fixes related to Mac
  OS X 10.2
  <http://www.omnigroup.com/applications/omniweb/>

**6. Microsoft Outlook Express 5.0.6**

  [9] Free E-mail client includes all recent bug fixes as of
  version 5.0.5 (MWJ_ 2002.07.13), including support for MSN Web
  mail accounts, and now including a fix for the recent
  man-in-the-middle security attacks [45] (MWJ_ 2002.09.09)

  [45] <http://www.microsoft.com/technet/treeview/?url=/technet/security/bulletin/MS02-050.asp>
  <http://www.microsoft.com/mac/DOWNLOAD/OE/oe5.asp>

**7. Toast Titanium 5.2**

  [9][X] US$90 CD and DVD optical media recording tool adds Mac OS
  X 10.2.x support, supports up to 52X CD-RW burners, reads and
  writes CD TEXT format tracks found on some new MP3 discs, and
  improves encoding performance for VideoCD (if you downloaded the
  abortive release of this last week, you may have to reinstall an
  earlier version from CD to successfully apply the update)
  <http://www.roxio.com/toastosx/>

**8. REALbasic 5 Announced**

  [9][X] REAL Software this week announced REALbasic 5.0 [46],
  coming in the first quarter of 2003. Last year's four-part
  series on REALbasic and its critics showed that REALbasic users
  have a habit of holding REAL Software's feet to the fire: if the
  company says it wants to put a certain feature in an upcoming
  release, every release after that is judged in part on whether
  or not it contains the feature. REAL has therefore stopped
  talking specifically about features it doesn't yet have
  implemented, saying only that that version 5 "includes a long
  list of Mac OS X 10.2 Jaguar-only features."

  [46] <http://www.realsoftware.com/corporate/pressreleases/RbWinAnnounce.html>

  The big news, however, is that this version of REALbasic works
  on Windows as well. Previous versions of REALbasic have all run
  on the Macintosh, but the professional versions have included a
  Windows cross-compiler so the same code could be built for
  Windows machines as well. However, REALbasic's development
  environment did not run on Windows, so you could not write
  REALbasic on Windows, or edit it, or even debug the code. If the
  Windows version of a REALbasic program has bugs, the programmer
  has had to solve it without stepping through the program or
  using other modern debugging tools.

  The new version 5 changes that, making the Windows and Macintosh
  versions equal in every significant way. That means not only
  that REALbasic for Windows can build and debug Windows programs,
  but also that they can build Mac OS and Mac OS X programs,
  though they'll need either the Professional edition or both
  Macintosh and Windows standard editions. The REALbasic 5 project
  and plug-in formats are cross-platform as well, and since
  REALbasic is already based on Microsoft's Visual BASIC syntax,
  the change could help more Windows-style BASIC programs move to
  the Macintosh.

  REAL Software is conducting an open pre-release cycle [47] for
  owners of REALbasic 4.5 who want to test the alpha versions for
  Macintosh or Windows, but note that they are alpha, and there's
  lots of work yet to be done (at least two or three months before
  golden master). Lorin Rivers of REAL Software said that the work
  done so far on the Windows version has exposed and corrected
  several problems in the system that just hadn't shown up on the
  more forgiving Mac OS, but that should make all code more robust
  in the long run. Pricing for the new releases has not been
  announced, but REAL's FAQ says the company does not plan to
  change "from our current pricing."

  [47] <http://www.realsoftware.com/realbasic/prerelease/>

**9. BootCD 0.5**

  [X] Charles Srstka's free-in-beta utility creates a working,
  bootable Mac OS X CD-ROM with the Finder, utilities of your
  choice, and a Dock that contains all those utilities; new
  version works with and requires Mac OS X 10.2, adds German and
  French localizations, lets you change the size of the disk image
  and the RAM disk, works with System Preferences and Console
  applications (and they're automatically copied to the boot CD),
  works with Drive10 but not Norton Utilities yet, and eliminates
  many but not all causes of the "hanging bug," which should now
  display an error window if it crops up
  <http://www.charlessoft.com/>

**10. StreamRipperX 1.0.1**

  [X] Broadcasters love streaming media because it's easier to
  control: you can download and copy an MP3 file or a QuickTime
  movie, but you can't save a stream. Once the stream is removed
  from the Internet, you can't watch or listen to it anymore,
  unless you hooked up audio equipment to your computer and
  recorded the stream as you would a radio broadcast.

  Of course, it doesn't have to be that way - the RTSP and RTP
  streaming formats that QuickTime and RealPlayer use are well
  documented. In theory, a program could capture the stream and
  reassemble the data portions into a more known file format, like
  a movie or MP3 file. Even though this seems like a natural tool
  (if one that would provoke RIAA ire), there haven't been many
  Macintosh tools to do it. In fact, at least one maker of
  anti-piracy tools hasn't bothered with Macintosh versions
  because it believes there are no Macintosh streaming capture
  tools. There's an open-source Unix command line tool,
  streamripper [48], for capturing streaming MP3 audio, but it's a
  bit obscure.

  [48] <http://streamripper.sourceforge.net/>

  Enter StreamRipperX 1.0.1 [49], an open-source Cocoa port of
  streamripper. The free application provides a full Mac OS X
  interface to capturing MP3 streams, including those from Icecast
  or SHOUTcast. The new version isn't really new - it debuted on
  2002.09.05, though we just learned of it this week - can capture
  streams you drag-and-drop from iTunes 3.0's "Radio" list, as
  well as Shoutcast-style ".pls" files you drag-and-drop from the
  Finder. It has tool tips, user interface improvements, a
  Japanese localization, and more. We wouldn't begin to cover
  legal and licensing issues with capturing copyrighted broadcasts
  for redistribution, and we note that StreamRipperX does not
  capture QuickTime video or audio streams, but it's there if you
  need it for your own purposes.

  [49] <http://streamripperx.sourceforge.net/>

**11. iPod 1.2.1**

  [9][X] iPod firmware update (available in separate Mac OS 9 and
  Mac OS X installers, though you must use iTunes 3 under Mac OS X
  10.1 or later to get access to all iPod 1.2 features) makes sure
  the battery icon correctly indicates a full charge
  <http://www.apple.com/ipod/download/>

**12. Audio Hijack 1.0**

  [X] Rogue Amoeba software's Audio Hijack 1.0 [50] isn't exactly
  a stream-ripping program like StreamRipperX, but it can do the
  same thing via a different method. Where StreamRipperX is
  designed to grab data from the Internet and save it to disk,
  Audio Hijack grabs whatever data you're hearing (through your
  current sound output device) and saves it, or filters it. That's
  the real point: since it captures data just before it hits the
  speakers, Audio Hijack adds a 10-band equalizer to _all_ audio
  piped through your system, allowing you to apply digital signal
  processing effects to it, even VST effects using the include VST
  hub. This can put audio out of sync slightly, but you can tell
  Audio Hijack to use a smaller buffer to keep it closer to real
  time. It's pretty cool: enhance audio from RealPlayer, capture
  Flash audio, capture Windows Media audio - anything you can hear
  (with the bizarre exception of Propellerhead's Reason
  synthesizer or sound from Classic applications), Audio Hijack
  can capture and process. For this power, you pay the grand total
  of sixteen American dollars (US$16), and a 15-day free trial is
  at your disposal.

  [50] <http://www.rogueamoeba.com/audiohijack/>

**13. Zingg 1.2**

  [X] Rainer Brockerhoff (of XRay fame) offers this free
  contextual menu to open any selected file with any eligible
  application; Mac OS X 10.2-only update lets you configure
  applications to always appear in the menu, can now use Japanese
  or other high-ASCII characters for the "Open With" string, and
  fixes a bug with Internet Explorer in Mac OS X 10.2
  <http://www.brockerhoff.net/zingg/>

**14. Goliath 1.0**

  [9][X] Free tool from the WebDAV Organization implements
  browsing of WebDAV servers much like you'd see in an FTP client,
  a great boon for Mac OS 9 where you can use a .Mac iDisk over
  fast WebDAV instead of slower AppleShare (Mac OS X 10.1 and
  later include a WebDAV file system that treats the servers like
  any other disk, but the program is still useful); version 1.0
  adds "Edit in External Application" features, enhanced SSL
  support (Mac OS X's WebDAV file system doesn't support WebDAV
  over SSL to our knowledge), better handling of long or Unicode
  file names, a reorganized menu bar, and much more.
  <http://www.webdav.org/goliath/>

**15. OnMyCommandCM 1.3.1a6**

  [X] We've mentioned Contextual Menu Workshop [51] in these pages
  multiple times, even though it's a developer-level framework, in
  the hopes it would encourage more programmers to write Mac OS X
  contextual menu plug-ins. They're still somewhat scarce compared
  to Mac OS 9 versions, but they're fully supported and an easy
  way to add functionality to the system, so the more the better.
  After all, you don't have to install every one you see.

  [51] <http://free.abracode.com/cmworkshop/>

  Abracode, the Contextual Menu Workshop folks, make several
  contextual menu items based on the Workshop code (including the
  useful FileUtilsCM [52], with options to set a file's metadata
  based on its filename extension and to refresh the Finder's view
  of an object). The latest is OnMyCommandCM 1.3.1a6 [53], a
  "command-line executor" that looks really nifty. Although
  there's no utility to edit the program's configuration files at
  the moment, so you have to use Property List Editor or edit an
  XML file, you can use those to add any Unix command you want to
  your contextual menu. The command can optionally work on the
  file or files you have selected by passing pathnames to the Unix
  command.

  [52] <http://free.abracode.com/cmworkshop/file_utils.html>
  [53] <http://free.abracode.com/cmworkshop/on_my_command.html>

  So what kinds of things [54] can you do with it? How about
  executing selected text as a Unix command? Compressing or
  decompressing a selected file with gzip (although we still
  prefer StuffIt Deluxe's more expansive contextual menu module)?
  Showing the man page for selected text, even in a Terminal
  window? Displaying your current screen saver as a desktop
  background (under Mac OS X 10.2 only)? If you repeatedly use
  command-line utilities by dragging-and-dropping files from the
  Finder to a Terminal (or other shell) window so you don't have
  to type pathnames, OnMyCommandCM is a good candidate to make
  those tasks part of the Aqua interface. This is the kind of
  cooperation we expected between Mac OS X's Unix and Macintosh
  sides, even if it is in the alpha phase (due to no command
  editor). The new version gets support for processing multiple
  selected objects, adds a preferences installer, extends
  activation checking, and works properly with Terminal in Mac OS
  10.1 through 10.1.5.

  [54] <http://free.abracode.com/cmworkshop/commands/commands.php>

**16. Eggplant for Mac OS X**

  [X] Interesting new cross-platform tool tests graphic
  applications on any platform that supports Virtual Network
  Computer (VNC) by recording your actions and capturing bitmaps
  around clicked objects to identify them later; one Mac OS X
  system running Eggplant can test any number of user-driven
  systems, with pricing based on number of concurrent uses;
  SenseTalk scripting language and interactive script development
  automates test scripts for repeated use on the same platform
  (pricing not available without a sales call)
  <http://www.redstonesoftware.com/products/index.html>

**17. Summary 2.3**

  [9][X] US$59-and-up Web server log analysis and traffic
  monitoring tool adds new "Country by GeoIP" and "Region by
  GeoIP" reports; adds "Interesting Referring Domains" report to
  detect patterns of hits that are, well, interesting; adds
  support for regular expression searches in a report, and makes
  available a Mac OS X command-line version
  <http://summary.net/summary.html>

**18. DiskSpy 1.1**

  [X] This free utility is a blast from the Norton Utilities past,
  blinking an indicator in the Mac OS X menu bar when the system
  reads from or writes to the hard disk (but without implementing
  a driver or patching frameworks, according to the author); new
  version takes less CPU time, doesn't flicker the disk reading
  icon, and gets new icons instead of arrows
  <http://dreamless.home.attbi.com/diskspy.html>

**19. Keyboard Maestro 1.2**

  [X] US$20 lightweight automation utility with program switcher,
  clipboard switcher, and hot keys now adds a window switcher for
  closing, minimizing and switching windows with a single
  keystroke; also adds a French localization, makes sure its
  inactive hot keys don't interfere with other programs' hot keys,
  and adds a new manual to the "Help" menu
  <http://www.keyboardmaestro.com/>

**20. Digital Performer 3.1**

  [9] US$795 digital audio and MIDI sequencing tool gets an
  updater with a DP 3.1 patch, FreeMIDI files, MAS update, and a
  REX shared Library - but no release notes
  <http://www.motu.com/english/software/dp/body.html>

**21. Reunion 8.0**

  [9][X] Best-selling Macintosh US$90 genealogy tool is now native
  for Mac OS X compatibility with new charting features, more
  accurate relationship descriptions, an expanded color palette,
  better numbering systems, LDS support for the massive LDS
  genealogy records, a "never married" designation, unlimited
  child statuses, multiple found lists, and a "sensitive"
  designation incase you want to exclude a particular child (don't
  think our parents haven't thought about this, particular for our
  callous use of double negatives), plus much more for the family
  tracker fan in everyone
  <http://www.leisterpro.com/doc/Version8/version8.html>

**22. Freeverse Card Game Week**

  [9][X] If you're like us, you have an extended staff member who
  got a Freeverse card game for Christmas two or three years ago,
  plays it obsessively throughout the week, and keeps Classic open
  in Mac OS X now solely to keep a card game running. Maybe that's
  just us, but maybe not: Freeverse's card games are enormously
  popular thanks to their Internet-capable play, their
  customizable avatars and spoken messages, and the fast game play
  that's both easy to use and complex to master. This week was
  Freeverse Card Game Week [55], in which the company finally
  updated six of its card games to run natively under Mac OS X,
  with compatible Mac OS 9 versions and fixes specific to each
  game. The updated games are Classic Cribbage 2.0 [56], 3D Hearts
  Deluxe 7.2 [57], 3D Euchre Deluxe 2.2 [58], 3D Spades Deluxe 3.2
  [59], 3D Pitch Deluxe 2.2 [60] (also known as "Setback"), and
  the most important one for our staff, still not quite out at
  press time, 3D Bridge Deluxe [61]. All of these games are US$20,
  except for Bridge, which is US$30. Freeverse hasn't updated
  Classic Gin Rummy [62] as part of Card Game Week, but once
  Bridge is out, Gin Rummy will be the only card game in the
  Uni-freeverse that's not Mac OS X native. Bring on the no
  trumps!

  [55] <http://dev.freeverse.com/phpbb2/viewtopic.php?t=129&sid=d4a1d2fb0c28350fc951979b098b5ef4>
  [56] <http://www.freeverse.com/cribbage.mgi>
  [57] <http://www.freeverse.com/hearts.mgi>
  [58] <http://www.freeverse.com/euchre.mgi>
  [59] <http://www.freeverse.com/spades.mgi>
  [60] <http://www.freeverse.com/pitch.mgi>
  [61] <http://www.freeverse.com/bridge.mgi>
  [62] <http://www.freeverse.com/ginrummy.mgi>

**23. "Mac OS X for Unix Geeks"**


  [X] O'Reilly & Associate's new US$25 (C$39) book [63] is a guide
  to figuring out the BSD Unix system and specific components and
  features a quick overview of the Terminal, linking and porting
  Unix software, and a description of building the Darwin kernel;
  if you have any interest in NetInfo, go _now_ and download and
  read the sample chapter [64] all about Mac OS X directory
  services, and then go buy the rest of the book
  [63] <http://www.oreilly.com/catalog/mosxgeeks/>
  [64] <http://www.oreilly.com/catalog/mosxgeeks/chapter/index.html>

**24. SkinnyEPS 2.0**

  [9][X] US$195 utility reduces EPS image file size down to as
  little as 7% of original size without "compromising" printed
  quality for page layout by eliminating massive amounts of image
  data that won't be used in most halftone or screen printing
  processes; new version 2.0 works natively in Mac OS X with
  Photoshop 7 as well as in Photoshop 6 and later on Mac OS 9
  <http://www.ultimate-tech.com/cgi-bin/store/shopdisplayproducts.asp?Action=Software&id=62&cat=SkinnyEPS>

**25. EyeTV 1.0.3**

  [X] Software for US$199 digital video recorder with program
  guide and scheduling now lets you create repeating programs for
  weekdays or a custom schedule, adds a slider for recording
  duration, can save non-standard "high quality" recordings to
  special CDs with Toast Titanium 5.2, can keep the entire live TV
  buffer in RAM, lets you resize the player window, and gets
  cosmetic fixes to remote control, plus bug fixes
  <http://www.elgato.com/eyeTV/index.html>


The Weekly Attitudinal[tm]: Type rights and wrongs
--------------------------------------------------

**Agfa-Adobe squabble hints at type casting**

  Oh, the commentariat chuckled last month when Adobe announced it
  could be going to court defending itself against violations of
  the Digital Millennium Copyright Act - the same law the company
  used a year ago to have an international programmer arrested
  solely because it could. "Hee hee," they said, "this is
  delicious." "Serves them right," said the online civil
  libertarians, and they were correct, as the _Attitudinal_ will
  explain somewhere around the third trimester of this tome.

  There's only one small problem: this time, Adobe's right.

  While the punditocracy was giggling with irony, the
  _Attitudinal_ smelled something more serious going on here, and
  started reading some license agreements. It turns out that one
  of the largest type foundries in the world has quietly started
  asserting rights to its fonts that it probably doesn't have. You
  could be violating font licenses every time you make a PDF file,
  at least if you do something unreasonable like distribute it to
  anyone. Just because it's the "portable" document format doesn't
  mean you can pass that file around, buster!

  Just as the advent of new media and a greater understanding of
  typography should make it a grand time for fonts, some of the
  biggest names in type are mired in the 1980s, wanting to
  restrict you from using their fonts for anything other than
  printing or making bitmap images. Using resolution-independent
  fonts to make resolution-independent documents is right out, and
  you didn't even know it. The _Attitudinal_, however, found out.

**The tip of the iceberg**

  Most of these outline antics have escaped public scrutiny, but a
  hint of the underlying drama poked through the surface in
  September. Adobe Systems announced that it had gone to court
  seeking declarative judgment in a font licensing dispute with
  Agfa-Gevaert, the German imaging conglomerate. Talk about
  yawn-inspiring! No one could possibly care about this, except
  maybe _you_.

  Over the past fifteen years, Agfa has bought its way to becoming
  a major player in typography. In 1987, Agfa-Gevaert purchased
  long-time digital typographical firm Compugraphic Corporation,
  acquiring thousands of fonts in the process. The two companies
  also merged with medial imaging firm Matrix to form Agfa
  Corporation. Miles acquired Agfa in 1992 as its imaging
  division, staying as such when Miles renamed itself Bayer in
  1995 after its most popular product, Bayer Aspirin. In 1997,
  Agfa acquired the century-old Monotype foundry, picking up
  thousands of the most used and desired fonts on the market.

  In 1999, Agfa Corporation broke off from Bayer to become a part
  of the Agfa-Gevaert group in Germany. Agfa's entire typographic
  asserts were merged into a new wholly-owned subsidiary, the Agfa
  Monotype Corporation [65]. In 2000, Agfa Monotype purchased the
  International Typeface Corporation (ITC) [66], acquiring
  thousands more fonts. Today, Agfa Monotype is the 500-pound
  gorilla in the typographic market, controlling Monotype, ITC,
  www.fonts.com [67], and lots more. Almost every commercial font
  today comes from Agfa Monotype, Linotype, or Adobe, with
  Bitstream and other foundries bringing up the rear. Agfa
  Monotype licenses hundreds of the fonts that other foundries
  sell, and creates the core Windows TrueType fonts that are
  hinted to within an inch of their lives for outstanding results
  on screen without using bitmap versions. Agfa Monotype developed
  Microsoft's Web fonts [68], too, and may be part of the reason
  why those fonts are no longer available for free download.

  [65] <http://www.monotype.com/>
  [66] <http://www.itcfonts.com/>
  [67] <http://www.fonts.com/>
  [68] <http://www.microsoft.com/typography/fonts/default.asp>

  Ah, yes, the dispute. In separate proceedings in San Jose and
  London, Adobe is seeking "declaratory relief" in a contract
  dispute with Agfa over font rights. ITC is a US corporation and
  is the subject of the San Jose action; Monotype was founded in
  London and Adobe is challenging it in English courts. In
  essence, Adobe is arguing with Agfa and asking a court to decide
  somewhat preemptively that Adobe is in the right, before one
  party alleges that the other has actually breached a contract. A
  declaratory judgment asks a court to declare (hence the term, of
  course) that one interpretation of the contract is correct.

  The dispute is hard to pinpoint since Adobe and Agfa aren't
  making the contracts in question public, but according to
  Adobe's announcement [69], it's all about font embedding. Adobe
  wants the US and British courts to declare that its contracts
  with ITC and Monotype give Adobe "the right to permit its
  customers to embed ITC [and Monotype] fonts in electronic
  documents." An Adobe vice-president said, "Many years ago Adobe
  anticipated the shift to electronic documents. At that time, we
  obtained the embedding rights from our font partners necessary
  to permit the creation of electronic documents. We are now
  defending the rights we obtained for our customers to continue
  to conduct business in the electronic age."

  [69] <http://www.adobe.com/aboutadobe/pressroom/pressreleases/200209/200209itc.html>

**Learning to embed** -- On the surface, this is almost
  certainly true. Adobe started work on what became Acrobat in the
  early 1990s, long before anyone anticipated a ubiquitous
  Internet or the Web's somewhat interesting idea of typography.
  Digital documents were the company's focus for years, arriving
  one small engineering milestone at a time. The first step was
  the Multiple Master font, a PostScript technology for rendering
  fonts between two hand-drawn outlines. Multiple Master fonts
  aren't hard to figure out: if you have a very skinny font and a
  very fat font, the control points for its curves (like you would
  see in Illustrator or FreeHand) are probably the same. If there
  are eight control points on a skinny "S", there are probably
  eight on a fat "S". They're just in different places - the
  skinny ones close to the center of the stroke, and the fat ones
  farther away.

  When you demand a Multiple Master font instance halfway between
  the "skinny" and "fat" reference points, the PostScript software
  moves each control point on the "S" to the midpoint of a line
  segment between the skinny reference point and the fat reference
  point. (Don't panic - it's not trigonometry, very much. Get some
  graph paper and sketch it out if you want, or just take the
  _Attitudinal's_ word for it. Or read about it in the technical
  specifications [70].) Once each control point is moved in such a
  fashion, the resulting glyph is "halfway" between skinny and
  fat. The same method applies for every axis in the font -
  height, weight, optical size, slant, retail price, whatever.
  Although incredibly useful in design, it was also the first step
  to digital documents - the first time PostScript could make
  good-looking fonts to fit an exact width or height.

  [70] <http://partners.adobe.com/asn/developer/type/designmm.html>

  Next was an important but long-forgotten step: SuperATM [71].
  Suppose you have a word processing document that uses Times New
  Roman for its body text, and you want to send it to a colleague
  who has the same word processor but not the same font. SuperATM
  took Adobe's new Multiple Master technology one step further by
  including a database of hundreds of Adobe and third-party font
  metrics - the exact sizes of each glyph in popular typefaces. If
  your colleague had SuperATM installed, any time a program like a
  word processor needs a font you don't have, the program would
  synthesize a substitution font from two generic Adobe Multiple
  Master fonts (Adobe Sans MM and Adobe Serif MM), using the exact
  metrics of the missing font (Times New Roman in this case). The
  resulting view would not look like the original document, since
  the substitution font does not look like Times New Roman, but it
  would fit perfectly. Every glyph would be the same size, and no
  text would rewrap to the next line.

  [71] <http://typedriverssupport.adobe.com/adobeknowbase/root/public/tm3522.htm?DREID=8564>

  That created semi-portable documents - to view a document
  accurately, the recipient needed SuperATM and the software that
  created the document, but not the fonts it used. (It was also
  behind Adobe's desperate plea to developers to ask the system
  for every font they wanted even if they knew it wasn't installed -
  otherwise, SuperATM didn't get a chance to synthesize a
  substitute.) Meanwhile, back at the printer, PostScript itself
  could faithfully reproduce a document's graphics and exact glyph
  position, but needed fonts with the right metrics to reproduce
  the document itself. Combining SuperATM's technology with
  PostScript's document description produced the framework Adobe's
  electronic document grail, Acrobat. SuperATM and other early
  digital document initiatives all came out of the same "Carousel"
  project at Adobe, and that's why the Acrobat creator type was
  and is 'CARO'. Isn't that fascinating? Kind of?

  These early electronic documents weren't very typographically
  refined. Unless the viewing computer had the right fonts
  installed, all the fonts in the electronic document were the
  substitute fonts from SuperATM technology - serviceable but not
  interesting, and not sufficient for companies like Apple or
  Adobe itself that have a strong typographical identity. Adobe
  knew from the start that Acrobat would need to embed real fonts,
  just as PostScript job files could embed font streams for
  faithful reproduction. Adobe added technology to include
  PostScript and TrueType glyphs in a PDF file, filling out the
  feature set for Acrobat 1.0.

  This made font foundries extremely nervous, including Adobe's
  own type division. If a PDF file could include an entire
  high-quality Type 1 or TrueType font, it would be easy enough
  for enterprising hackers (or PostScript experts) to extract the
  font from the document and have a free copy of it. Since users
  already viewed fonts as less than software and passed them
  around too freely for the foundries' taste, the font makers
  wanted assurances Acrobat wouldn't lead to massive font piracy.

  Adobe wasn't the only one working on digital documents, though.
  Apple was (in QuickDraw GX), as was IBM in OS/2. Through the
  front door or the back, a set of embedding bits evolved in the
  TrueType format, located in the 'OS/2' table inside the font.
  Various bits in that table control whether any glyphs of the
  font can be embedded in digital documents, or if glyphs can be
  embedded for viewing but not editing, or if the whole font can
  be included so that the recipient can edit the digital document
  (perhaps adding text that wasn't in the original). This is only
  tangentially related to subsetting a font, by the way. If you
  make a PDF document containing only the word "Apple" and embed
  the font in the PDF file, the only glyphs included are "A", "p",
  "l", and "e." Even if the font was embedded with editing
  privileges, you couldn't change it to read "Compaq" because the
  right glyphs aren't included. That'll teach you.

  The _Attitudinal's_ point is that embedding rights are different
  than the practical issue of using embedded fonts. Embedding a
  font doesn't mean you can use it to edit a document. Editing may
  work better if the font isn't embedded, and just because you
  managed to embed a font doesn't mean you had permission to do
  so.

**Adding your outlines** -- Oh, yes! Just as you may have a copy
  of a font that you didn't purchase, you may have purchased a
  font that allows embedding without having purchased the _rights_
  to embed that font in any digital documents. And you may not
  even know it! Embedding bits only tell applications whether to
  allow or deny embedding on a broad basis, not whether you have a
  legal copy of a font. Acrobat Distiller happily embeds fonts in
  PDF files if the embedding bits are set, even if you obtained
  those fonts by robbing the Linotype Factory Outlet at gunpoint
  with a CD burner.

  It's the licensing agreement that allegedly controls what rights
  you do and do not have for fonts you've purchased. Adobe, for
  its part, wants you to feel free to embed any of its fonts in
  any document you want - Adobe knows that the #1 destination for
  embedded fonts is PDF files, even though Microsoft Office allows
  it as well, and embedding good fonts makes PDF files, and
  therefore Acrobat, more attractive. Adobe therefore gives you
  almost everything. The standard shrink-wrap software license
  [72] for Adobe's fonts explains that you can use the fonts on up
  to five computers, back them up, install them on at least one
  printer, with the rest of the standard legal rigmarole. In
  section 2.6.5, however, Adobe explicitly allows embedding:

  [72] <http://www.adobe.com/type/browser/legal/adobe5eula.html>


    You may embed the font software, or outlines of the font
    software, into your electronic documents to the extent that
    the font vendor copyright owner allows for such embedding.
    The fonts contained in this package may contain both Adobe
    and non-Adobe owned fonts. You may fully embed any font
    owned by Adobe.


  Adobe only hints at the present dispute in its announcement, but
  the hints seem clear: several years ago, Adobe believes it
  negotiated the rights to allow full embedding of ITC and
  Monotype fonts in PDF documents, but Agfa now says this is not
  the case. It's not clear if Agfa expects more money from Adobe
  or from its customers to allow embedding its fonts. It is clear
  that Agfa is happy to sell you digital fonts as long as you
  don't actually try to use them in digital situations.

**Agfa's font dreams**

  Once you read Adobe's license agreement, provided you don't need
  a power nap, you may wonder how it differs from Agfa's standard
  end-user license agreement for its font products. You can find
  the agreements on either the ITC or Agfa Monotype sites. Both of
  them lead to special pages all about font embedding, and even
  though the ITC page [73] and the Monotype page [74] (at
  fonts.com) have different addresses and layouts, they contain
  the same text, adjusted only for the names "ITC" or "Agfa
  Monotype." Here's text from the ITC version of the page:

  [73] <http://www.itcfonts.com/about/embedding.asp>
  [74] <http://www.fonts.com/fontservices/services_home.asp?con=embedding>


    ITC, like most other font foundries, licenses fonts for use
    on a set number of workstations. The typical font can be
    used on five (5) workstations, while a font library provides
    for use on up to twenty (20) workstations. Our End User
    license [75] does not permit redistribution of font data
    beyond the set number of workstations without an additional
    license. Because of the introduction of font embedding, font
    streaming and font server technologies, and the use of
    embedded fonts into commercial products, ITC is introducing
    reasonably-priced font licenses for designers and
    publishers.


  The page goes on to describe four new font licenses. The
  "editable embedding" license is for fonts you want to embed in
  documents that are editable by the recipient, something that
  applies more to Microsoft Office documents than PDF files. The
  "Commercial Product Font Embedding License" for embedding fonts
  in any document that is "sold to end users," with given examples
  of "newsletters, reports, eBooks, eMagazines, etc. distributed
  commercially on CD-ROM or on the Internet," but that would seem
  to include PDF product manuals in for-sale software and
  shareware as well, since you're "selling" the document as part
  of the product.

  [75] <https://www.itcfonts.com/order/eula-print.asp>

  One new Internet license provides rights to use an Agfa font
  with technologies that download the font to a client computer
  and allow editing, like Microsoft's WEFT [76], Bitstream's
  TrueDoc [77], or Flash applications. A second Internet license
  allows keeping fonts on a server where multiple clients can use
  them to "create new documents or images, or edit content with
  font data that resides on the server." An easy example is
  Apple's iCards application [78], where fonts like Copperplate
  Gothic and Comic Sans live on the iCards server, and a
  WebObjects application uses them to add text to cards, making a
  "new image" that's mailed to your card's recipient.

  [76] <http://www.microsoft.com/typography/web/embedding/weft3/>
  [77] <http://www.bitstream.com/categories/developer/truedoc/index.html>
  [78] <http://www.mac.com/WebObjects/iCards.woa>

  The _Attitudinal_ encourages you to stop here for a moment and
  consider the absolute idiocy of the text on Agfa's page: "Our
  End User license does not permit redistribution of font data
  beyond the set number of workstations without an additional
  license." Since most font licenses are good for five
  workstations, Agfa is saying that if you embed a font into a
  digital document of any kind, you're only legally allowed to
  send that document to the five workstations licensed for the
  font - the exact same workstations that don't need you to embed
  the font since they're already licensed to use the font anyway.

**Bit by embedded bit** -- This is not Agfa's first brush this
  year with embedding restrictions. The subject came up earlier
  this year on Declan McCullagh's Politech mailing list, again via
  an interesting claim on Agfa's part. That discussion [79]
  included contributions by the _Attitudinal's_ publisher (he
  thinks he's _sooooo_ clever, particularly here [80] and here
  [81]), which the _Attitudinal_ happily appropriates.

  [79] <http://www.politechbot.com/cgi-bin/politech.cgi?name=agfa>
  [80] <http://www.politechbot.com/p-03506.html>
  [81] <http://www.politechbot.com/p-03529.html>

  The problem came about when Agfa used the big stick of the
  Digital Millennium Copyright Act (DMCA) on Tom Murphy VII, who
  goes by "Tom 7," a freeware font developer [82]. Murphy
  developed his fonts with Macromedia Fontographer, an editor
  released before the embedding bits of TrueType fonts were
  defined as they are today. Fontographer provides font developers
  with a choice of embedding permissions, but it does not set the
  bits properly in the resulting TrueType font file. Instead,
  Fontographer usually creates bits that are completely invalid as
  font embedding permissions, and never sets the bits the way you
  think it should. Fontographer provides _no_ way to change these
  bits to valid settings. None. It can't be done with
  Fontographer. You must remember this point: _every_ TrueType
  font built with Fontographer has problems, regardless of the
  font developer's intentions, because programs like Acrobat
  Distiller conservatively treat invalid embedding rights as no
  embedding rights [83]. Now, in some cases, Fontographer does
  generate bits that allow embedding, but it's purely by accident.

  [82] <http://fonts.tom7.com/>
  [83] <http://www.pdfzone.com/rich/fonts1.html>

  Rather than try to edit his fonts to include the least
  restrictive embedding rights - the setting he intended all along -
  Murphy wrote embed [84], a Windows program that finds the
  appropriate bits in the "OS/2" table of a TrueType font and sets
  them to zero. He placed both the program and its source in the
  public domain, and specifically noted on the distribution page,
  "Changing the embedding value does not give you license to
  distribute the fonts. You should only change this setting if you
  are the font creator, or something like that. Use at your own
  risk." Remember, the tools Murphy used to make his fonts would
  not allow him - the font creator - to set the embedding bits as
  he chose, so he wrote his own tool and distributed it so other
  font creators could have the same capability.

  [84] <http://www.andrew.cmu.edu/~twm/embed/>

  Five years later, an Agfa Monotype lawyer wrote to Murphy [85]
  (insert your own "Murphy's Law" joke here) and demanded that he
  withdraw the program from distribution, apparently under the
  belief that knowledge placed in the public domain can somehow be
  withdrawn. ("Linda? It's Monica. Can you call the press and tell
  them to drop this story?") Attorney Paul F. Stack wrote, "The
  distribution of this program, whether for free or for a fee,
  infringes my client's federal copyrights in their TrueType
  programs." Murphy wrote back to note that his program and page
  included absolutely _no_ Agfa copyrighted material, rejecting
  the demand to withdraw the programs.

  [85] <http://www.andrew.cmu.edu/~twm/embed/dmca.html>

  Stack didn't write back for two and a half months, but when he
  did, he again demanded that Murphy withdraw his program because
  offering it was "unlawful conduct." That precipitated more
  E-mail, articles on Slashdot and CNet News (in addition to the
  aforementioned Politech thread), and a lot of legal maneuvering
  that seems designed more to confuse than to clarify. The
  _Attitudinal_, of course, cannot stand for that, offering only
  the caveat that the _Attitudinal_ is neither a lawyer nor a
  legal entity.

  DMCA took copyright law further by outlawing technology or
  devices designed to circumvent measures intended to preserve
  intellectual property rights. Try saying that three times fast.
  For example, if you write software and protect it with a
  copy-protection scheme, DMCA makes it illegal to develop a
  program to crack that copy-protection scheme. The scheme is a
  "technological measure that effectively controls access to a
  work" protected by copyright law, and DMCA bars interfering with
  such schemes. Agfa's position is that changing the embedding
  bits on a font is "circumventing a technological measure that
  effectively controls access" to the font, and therefore any tool
  that changes these bits violates DMCA.

  This is a lot like arguing that the Finder should be illegal
  because it lets you clear a program's "locked" checkbox. In a
  long legal memo sent to Murphy, Stack argues that DMCA fully
  prohibits changing software switches that work as technological
  measures to control work, but his example is ridiculously off
  base. Stack points out that RealMedia files include a "copy
  switch," set by the content creator, that determines whether
  anyone viewing RealAudio or RealVideo files can save them to
  disk or merely view them as streams. Of course, as you're well
  aware, re-viewing a stream requires reconnecting to the Internet
  and hoping that the stream is still available; saving a copy
  makes it available at your leisure for all time and allows you
  to make copies to send to your friends, even though you may not
  have rights to do so.

  The problem, however, is that the "copy switch" really
  _does__"effectively control access" to the media in question.
  Standard RealMedia files (the ".ra" or ".ram" files you
  download) are text files that contain an RTSP URL for finding
  the stream. There is no way to download the stored media via
  HTTP, FTP, or other method - you can only view the stream in
  RealPlayer, and then you can save a copy only if the stream sent
  from the RealServer software tells your copy of RealPlayer that
  it's acceptable.

  The embedding bits in a TrueType or OpenType font, however, are
  in no way a "technological measure that effectively control
  access to a work." No value for embedding permissions prevents
  anyone from copying or pirating the font file that you install
  on your system to make that font available. Software is not
  bound by law to respect the embedding bits, either, and Acrobat
  Distiller v3 and earlier did not - those versions would embed
  any font you asked them to embed, regardless of the value or
  even the existence of embedding permissions.

  The "copy switch" in RealMedia files actually keeps you from
  accessing the stored media on disk or saving a copy. Embedding
  bits in a font on your hard drive do not stop any program from
  copying that font file. They are not copy protection, they're
  not encryption, they are nothing more than the font designer's
  intent for embedding. Agfa's position is even weaker by the
  realization that most of its fonts have the embedding bits set
  to _allow_ embedding. As with all forms of piracy, you can embed
  Agfa fonts in digital documents, but you're not supposed to
  unless you own the rights to do so.

  The only legitimate fear Agfa could possibly have is that
  customers are using tools like "embed" to change the embedding
  bits on fonts without purchasing an embedding license. In other
  words, even if someone did pirate the font file, they still
  wouldn't be able to embed the font in digital documents without
  something like "embed." This, too, is bilge. The exact
  specification of embedding bits [86] is open for anyone to see,
  and any number of font creation tools can change it, including
  tools that Agfa Monotype developed for its in-house use. In the
  name of this alleged embedding piracy, Agfa wants to restrict
  full manipulation of font embedding bits to itself and keep it
  out of the hands of developers like Murphy who use a tool that
  doesn't work properly. Macromedia is now a Web design company
  and Fontographer is in mothballs - it's never been updated to
  fix the embedding problem and probably never will be.

  [86] <http://partners.adobe.com/asn/developer/opentype/os2.html#fst>

  Agfa's demand that no tools change embedding bits speaks more of
  anticompetitive behavior than of a desire to protect copyrights,
  in the _Attitudinal's_ sage opinion. Were this argument to
  stand, most font editors on the planet would also be illegal,
  for they allow changing embedding permissions. So do Apple's
  font development tools [87], and even the venerable ResEdit and
  Resorcerer [88]. There is no attempt to hide the embedding
  information or encrypt it - nothing that common sense or DMCA
  would rule as trying to "limit" distribution of the font, just
  how it's embedded. Sure, the company says its embedding licenses
  are "reasonably-priced," but after six years of publication, the
  _Attitudinal_ has learned one rule well: if a company is afraid
  to post its prices, it knows it's charging too much.

  [87] <http://developer.apple.com/fonts/Tools/>
  [88] <http://www.mathemaesthetics.com/ResorcererInfo.html>

**Licensing font data** -- On its face, Agfa's license seems to
  be completely inappropriate, at least in the _Attitudinal's_
  homeland. When the company says, "Our End User license does not
  permit redistribution of font data beyond the set number of
  workstations without an additional license," it's standing on
  shaky ground: US law does not allow protecting "font data" via
  copyright.

  It's amazing but true. As the comp.fonts FAQ [89] describes, the
  Copyright Revision Act of 1976 explicitly removed typefaces as
  "artistic works" that may be protected by copyright. The US
  House of Representatives committee that reported the bill said,
  "The Committee does not regard the design of typeface, as thus
  defined, to be a copyrightable 'pictorial, graphic, or
  sculptural work' within the meaning of this bill and the
  application of the dividing line in section 101." In 1988, the
  US Copyright Office decided that this prohibited protecting
  electronic fonts, too: "data that merely represents an
  electronic depiction of a particular typeface or individual
  letterform [that is, a bitmapped font] is also not
  registerable." (57 FR 6201, for you fans of the
  _Federal_Register_.)

  [89] <http://nwalsh.com/comp.fonts/FAQ/cf_13.htm>

  The US law is definitely in the minority: just about every other
  country with copyright laws allows protecting font designs via
  copyright. The US also protects fonts designed outside the US by
  copyright - it has to, as a signatory to the Berne Convention
  that enforces international copyright laws. If it's protected by
  copyright in the country of design, it's protected here. It's
  merely US designers - and companies - that are out of luck. And
  despite Agfa's and Monotype's European heritage, Agfa Monotype
  and ITC are both wholly or partially US corporations. Remember,
  the _Attitudinal_ is not a lawyer or even particularly clever;
  the larger point is that this lack of copyright affects all US
  companies that sell fonts.

  The situation isn't totally bleak for font designers, though:
  the US Copyright Office does register _outline_ fonts for
  copyright protection. Why the split? The Copyright Office has
  ruled that outline fonts aren't really typefaces - they're
  computer programs. "The creation of scalable font output
  programs to produce harmonious fonts consisting of hundreds of
  characters typically involves many decisions in drafting the
  instructions that drive the printer. The expression of these
  decisions is neither limited by the unprotectable shape of the
  letters nor functionally mandated. This expression, assuming it
  meets the usual standard of authorship, is thus registerable as
  a computer program." (57 FR 6202)

  Therefore, because of bizarre US laws, bitmap fonts aren't
  protected. The raw outlines of fonts as you might find them in
  Illustrator or FreeHand after converting text to outlines aren't
  protected either, because that's merely a depiction of an
  unprotectable typeface. That's why all those knock-off foundries
  that offer "four billion outline fonts on one CD" can retrace
  the outlines of popular fonts: the typefaces themselves can't be
  protected. It's only the actual Type 1 font program or TrueType
  font that can be protected.

  Type 1 fonts contain hinting instructions that move their
  control points around depending on how the font fits the
  resolution of the output device. TrueType fonts also contain
  hints in a more powerful but lower-level programming language.
  Now, theoretically, the hinting may be protectable by copyright
  but the glyph data itself not subject to the same protections,
  but the Copyright Office or the US Courts have never been asked
  to split that hair, at least to the _Attitudinal's_ knowledge.
  You could possibly get into an argument that the curves or
  bitmaps are necessary data for the copyrighted program to run,
  but it's a dispute that won't be resolved here.

  So, for now, outline font "programs" can be protected by
  copyright. This is why Agfa and other foundries take such great
  pains to refer to their font "software" and not just to "fonts."
  It's also why the embedding page has such a major mistake: in
  the US, Agfa can't license "font data" because non-program
  descriptions of fonts are explicitly ineligible for copyright
  protection. You therefore can't need a license, since Agfa can't
  protect the bitmaps or the curves anyway. Oops.

  The whole mess suggests that embedding would be a non-issue if
  PDF and other formats embedded only the outlines of glyphs and
  not the hints that make them look good at low resolutions, but
  that's not happening, at least not in PDF. Adobe is very proud
  of the high quality of PDF output, even on relatively
  low-resolution devices like inkjet printers or typical desktop
  displays. You may not think of a 2880 DPI inkjet printer as "low
  resolution," but most people don't print at that resolution all
  the time, and it takes high-quality paper and ink to approach
  the same output you'd get on a printing press with similar
  resolution. Displays that barely reach 100 DPI are downright
  primitive compared to imagesetters, but PDF looks good on them
  anyway. Adobe's not about to give up the hints.

  It's a shame, because it might cut off Agfa's oxygen supply in
  the company's stunning attempt to demand more money from fonts
  you've already purchased. Despite setting embedding permissions
  on the fonts, the company now says you can't give documents with
  its embedded fonts to anyone who isn't already licensed to have
  the fonts - and therefore doesn't need them embedded in the
  document. It wants to force public domain tools for setting
  embedding bits in fonts you design to be taken off the market
  while keeping its own private tools for the same purpose. Agfa
  doesn't even want you to use its fonts in server applications
  that automatically draw text on demand, like iCards or that
  Amazon.com system that generates a graphic with your name in it.

**Shooting itself in the foot** -- Agfa's behavior is a really
  good example of why foundries may not be ready for US copyright
  protection on typefaces themselves. Just think: any time you use
  a bitmap font in any kind of image or computer document, you are
  "embedding" the bitmap data. Since Agfa is a German company, it
  doesn't have to report results, so you don't know how well or
  how poorly the font business is performing, but Agfa gives the
  appearance of a company trying hard to squeeze new sources of
  revenue out of a maturing digital typeface business. Agfa might
  want to charge you for "embedding" a bitmap font in an image on
  your Web site, in a QuickTime movie, or even in an icon.

  In other words, Agfa is willing to license its font programs to
  you, but only if all you ever publish is the output. No no, stay
  with the _Attitudinal_ here. If you purchase a Photoshop filter
  to manipulate images, you can't embed that filter into something
  you distribute, like shareware or a Photoshop action, even if it
  gives the output you want. Agfa seems to believe that embedding
  its font programs into PDF files is the same thing: you can use
  the font programs on your systems to produce whatever output you
  want, but you can't embed the program elsewhere. That's giving
  away Agfa's font program to run on someone else's computer when
  the PDF file is viewed.

  Unfortunately, the company also wants extra payments if you use
  the font on a server-based application. That's like saying you
  can't rig up a Web server to call Photoshop to create an image
  on the fly. Of course you can, as long as the server computer
  has a fully licensed copy of Photoshop. The license doesn't care
  if it's running on a server or on your desktop, but Agfa's
  license does. If Agfa could also protect bitmap images of fonts
  output by its outline font programs, the rights grab could be a
  lot worse.

  The industry sponsors a site called "TypeRight" that advocates a
  change in US law so typefaces are again protected by copyright.
  TypeRight seeks to reassure jittery type users that copyrighted
  typefaces won't unreasonably restrict their use, such as a
  foundry refusing you permission to use a typeface because it
  disagrees with your content. The best answer [90] the site has
  for this concern? "It's very unlikely this would happen." Isn't
  that reassuring?

  [90] <http://www.typeright.org/feature3.html>

  On the same page, the site tries to assuage fears that
  publishers will have to pay royalties every time they use a font
  like this: "When you purchase a font now, you pay a one-time fee
  to license certain rights of use. Although these rights may vary
  from foundry to foundry, nobody asks you to pay a fee every time
  you use that font. Copyright will not change this. What it will
  do is protect designers and foundries from those who take our
  work, rename it, and sell it as their own."

  The latter goal is admirable. Designing a font takes real
  artistic talent and the patience to spend dozens of hours
  fine-tuning outlines, two traits the _Attitudinal_ does not
  possess, on top of a vision for a consistent and expressible
  design throughout an entire glyph set whose baseline seems to
  grow each month thanks to increasing demand for Unicode
  characters. But to Agfa, that "one-time fee" to license "certain
  rights" already excludes using that font digitally. Printing?
  Sure, go ahead. Embedding in images? OK for now, since the font
  bitmaps can't be copyrighted. Embedding in digital documents?
  No, absolutely not, at least not without a more expensive
  license.

  If Agfa wants to restrict its typefaces so severely under the
  already-limited US copyright law, it's painful to imagine what
  kind of royalties or licenses it would demand if it had full
  protection for them. Fortunately, the marketplace is already
  providing a solution: foundries like Adobe and Bitstream that
  not only allow but encourage full embedding of their fonts in
  digital documents. These two companies correctly see that good
  typography in a PDF file makes you more likely to want a font
  for yourself, but no more likely to pirate it than if you'd seen
  it on a billboard or in a magazine. Typography fans purchase
  fonts to express their visions and communicate clearly. They
  want tools to assist, not obstruct their choice of media. "This
  is the perfect design, but we can only use it in print unless we
  pay more to license it for the online user manual." Pshaw.

  That's why the _Attitudinal's_ influence on MWJ_'s redesign,
  limited strictly to the points where someone finally had to
  _decide_ something, is mainly felt in the typography of the PDF
  versions. The fonts are Adobe's Minion Pro (for body text) and
  Myriad Pro Semi-Condensed (for headlines). The _Attitudinal_
  screamed for these not only because they are perhaps the
  best-drawn fonts of the past decade, nor only because they are
  OpenType fonts with vast glyph selections offering lots of
  growth opportunities for great typography. They're also Adobe
  Originals, meaning neither Agfa nor any other foundry can later
  come and demand that Adobe stop selling versions that allow
  embedding.

  The only thing these fonts are missing are the Apple Advanced
  Typography (AAT) tables, the post-GX descendants that let many
  Mac OS X applications automatically get beautiful typography and
  glyph selection with almost no work on your part. Adobe has
  never supported Apple's typography initiatives - other than to
  steal the ideas for OpenType but implement them in a way that
  only benefits Adobe's and Microsoft's programs - but everything
  else about them is nearly perfect, including embedding rights.
  You see Minion and Myriad in diverse uses these days, and it's
  not because they're cheap. It's because they work how you want
  them to work, visually and electronically. What more could you
  want?

**The DMCA stick**

  And so it comes full circle. According to Adobe's announcement,
  Agfa is threatening to file a DMCA complaint against Acrobat,
  presumably for embedding fonts without Agfa's permission. "Adobe
  believes these claims are being made to gain ITC and Agfa
  leverage in the contractual disputes. Adobe strongly disputes
  this claim and is asking the court to rule that there is no
  violation of the DMCA." The scarier thought is that Agfa really
  does believe that Acrobat violates the DMCA, just as the company
  has tried to bully freeware font developers into withdrawing
  tools that no pirate in his right mind would ever bother to use.
  If you've got enough font savvy to know how to change embedding
  bits, you've got more powerful tools than a Windows command-line
  tool in your arsenal.

  Yet it's somewhat satisfying to see Adobe threatened on DMCA
  grounds, or hoist with its own petard [91]. Over a year ago,
  Adobe initiated the only criminal DMCA prosecution to date,
  against Russian programmer Dmitry Sklyarov, the author of the
  Advanced eBook Processor. The program, which sold for US$99 from
  Russian software developer Elcomsoft, could take any encrypted
  PDF-based eBook file and break the pathetically weak encryption,
  allowing you to save it as an unprotected PDF file. This, of
  course, cheesed Adobe off no end, especially since the company
  and the programmer were in Russia.

  [91] <http://www.creators.com/lifestyle_show.cfm?next=2&ColumnsName=rky>

  It goes without saying that it is not a crime to break US laws
  outside the US, just as it's not a crime for people in the US to
  behave in ways forbidden by Russian law. Were it not so, there'd
  be no fine fraternity tradition of trundling to Tijuana. This
  grated Adobe's cheese even further, until the company learned
  that Sklyarov would be speaking at a hacker convention in Las
  Vegas in August 2001. Adobe has already tried falsehoods in
  legal documents to try to stop the distribution of Advanced
  eBook Processor, such as telling ISPs and payment processors
  that it contained copyrighted Adobe software, and that bullying
  tactic hadn't worked.

  So in July 2001, Adobe called the FBI. The company's employees -
  specifically eBooks product manager Kevin Nathanson,
  investigator Daryl Spano, and senior eBook engineering manager
  Tom Diaz - told FBI special agent Daniel J. O'Connell not only
  about Advanced eBook Processor, but also t