Comments

• Diamond Wakizashi

  2008-12-21 18:35:22

  How can a format that requires a Microsoft patent license be considered a standard? The ISO is as corrupt as it is useless.

• AlexH

  2008-12-21 19:04:44

  Someone, quick, tell Sun!

• Roy Schestowitz

  2008-12-21 20:32:06

  Sun supports Free software; Microsoft does not. It attacks it.

• Jo Shields

  2008-12-21 20:41:11

  Yet Sun develop OOXML support for OOo3? What's up with THAT?

• Roy Bixler

  2008-12-21 20:54:06

  The following patent agreement between Sun and Microsoft may be interesting:
  
  LIMITED PATENT COVENANT AND STAND-STILL AGREEMENT http://www.sec.gov/Archives/edgar/data/709519/000119312504155723/dex10109.htm
  
  As I understand it, it came about as part of a settlement of the private anti-trust case which Sun files against Microsoft. It's interesting to note that there is a section which specifically relates to Open Office and that it is treated differently than Star Office. Predictably, only Sun gets any protection and Microsoft reserves the right to sue sub-licencees.

• Roy Schestowitz

  2008-12-21 20:58:23

  If it's about patents, then there's lots of prior art to invalidate them. Microsoft did not invent office productivity software.

• Jo Shields

  2008-12-21 21:17:26

  As I understand it, it came about as part of a settlement of the private anti-trust case which Sun files against Microsoft. It’s interesting to note that there is a section which specifically relates to Open Office and that it is treated differently than Star Office. Predictably, only Sun gets any protection and Microsoft reserves the right to sue sub-licencees.

  
  
  Don't be SILLY. Fluffy bunnies and general huggable people at Sun, entering into an exclusionary patent deal which protects only its customers? If that had happened, mighty crusaders for Freedom would have had them shut down years ago!

• Roy Schestowitz

  2008-12-21 21:23:56

  See this for details.

• Jo Shields

  2008-12-21 21:28:58

  See this for details.

  
  
  What does SCO have to do with anything? We're talking about Sun's deliberate insertion of OOXML into OOo3 - and their (still active) exclusionary patent deal with Microsoft. If they were so changed, would they not have cancelled that arrangement?

• Dan O'Brian

  2008-12-21 21:41:46

  Roy Schestowitz said, December 21, 2008 at 3:58 pm
  
  If it’s about patents, then there’s lots of prior art to invalidate them. Microsoft did not invent office productivity software.

  
  
  Careful, Roy, your hypocrisy is showing there.
  
  You're saying not to worry because those patents can be invalidated via prior art. Yet, the same is true for Mono but you argue that you need to be afraid of Mono.
  
  Double standards, anyone?

• Roy Schestowitz

  2008-12-21 21:52:16

  Do you want to play with the patent maze or without it?

• Dan O'Brian

  2008-12-21 22:01:27

  What does that have to do with your double standards?

• Roy Schestowitz

  2008-12-21 22:09:50

  Sun wants to elevate OOo and ODF with the help of GNU/Linux; Microsoft denies GNU/Linux the access to Microsoft Office.
  
  Sun offers Java to GNU/Linux where it performs better than on Windows; Microsoft does not offer .NET or VS to GNU/Linux users.
  
  Corollary: Sun can live with GNU/Linux; Microsoft tries to destroy GNU/Linux.

• Jo Shields

  2008-12-21 22:17:32

  I'll just go run my JavaF... oh, wait. I'll just use DTrac... damn. Perhaps I can find some files on my ZF... balls
  
  Remind me again why OOXML from Sun is a shiny happy move of Free Software love, again?

• Roy Schestowitz

  2008-12-21 22:26:48

  JavaF* initially target wide reach, the rest aren't for Windows either.

• Dan O'Brian

  2008-12-21 22:32:37

  Roy: We're not talking about ODF here, we're talking about OOXML as implemented by Sun within OOo. You are admitting that Microsoft's patents can be invalidated by prior art, but this contradicts your opposition to Mono. Like with OOXML in OOo, Mono can protect itself from patent attacks by invalidating said patents by pointing to prior art.
  
  WRT Java on Linux faster than Java on Windows:
  
  I'm assuming you are referring the the Phoronix benchmark of Java 1.6_7 on Windows Vista vs Java 1.6_10 on Linux.
  
  I'm not sure if Java is faster on Linux or Windows, but the benchmark is extremely suspect considering they used different versions of the Java runtime. Even more so since Java 1.6_10 got a lot of performance optimizations (according to the Release Notes) since previous revisions. It was also pointed out that the Windows benchmarks used the Client VM whereas the Linux benchmarks used the Server VM. In other words, the benchmark failed at comparing apples to apples.
  
  Thus, I would not refer to that benchmark as empirical evidence that Java on Linux is faster than on Windows.
  
  Again, I'm not saying that Java is /not/ faster on Linux than Windows, I'm simply stating that the benchmark you are using as your empirical proof is bunk.

• Dan O'Brian

  2008-12-21 22:34:54

  JavaF* initially target wide reach, the rest aren’t for Windows either.

  
  
  The problem is that they aren't available for Linux.

• Roy Schestowitz

  2008-12-21 22:44:18

  It's funny -- though not surprising -- that you jump to Windows' defence.

• Jo Shields

  2008-12-21 22:50:02

  It's funny - though not surprising - that you think a flawed benchmark should be used regardless to help support your conclusions
  
  It's funny - though not surprising - that you ignore TWICE the statement that whilst the supposition may still be true (Java faster on Linux) that using a broken benchmark to do it is braindead & should not be condoned
  
  It's funny - though not surprising - that you continue to change the subject RE Sun's OOXML support
  
  It's funny - though not surprising - that you think double standards are fine

• Dan O'Brian

  2008-12-21 22:52:05

  I'm not jumping to anyone's defense, I'm merely stating that biased benchmarks do not make good sources of proof.
  
  I'd like to see the benchmarks run again using the same version of Java on both systems, using the same VM (client or server), on the same architecture (64bit vs 64bit or 32bit vs 32bit, rather than 32bit vs 64bit).
  
  Biased benchmarking does no one any good. The thing you want to walk away with after a benchmark vs a competitor is to figure out where you need improvement. If you cheated, then you don't know where you need improvement and so in the end will lose out.

• Roy Schestowitz

  2008-12-21 22:52:56

  Sun deserves the benefit of the doubt because it does not attack GNU/Linux like Microsoft.
  
  You're treating Microsoft like it's just another company (get clue).

• Roy Schestowitz

  2008-12-21 22:54:30

  I’m not jumping to anyone’s defense, I’m merely stating that biased benchmarks do not make good sources of proof.

  
  
  It sounds more like you're taking the role of "Microsoft spokesman". Spin, spin.

• Dan O'Brian

  2008-12-21 22:56:15

  We're not talking about giving Sun the benefit of the doubt regarding the benchmarks, we're talking about the benchmarks being biased.
  
  If you meant OOXML support in OOo, what doubt are you giving benefit?

• Roy Schestowitz

  2008-12-21 22:58:06

  Patents and all, since you are the one who brought it up (or your ilk).

• Dan O'Brian

  2008-12-21 22:59:29

  Who's spinning? Phoronix is certainly spinning something, if they really took benchmarking seriously, they would have at least used the same version of Java on both systems.
  
  The whole point of the benchmark was to show which platform Java ran faster on - so why would anyone run an old (pre-optimized) version of Java on one system and a recent (post-optimization) version on another?
  
  It just doesn't make sense.

• Dan O'Brian

  2008-12-21 23:02:46

  Patents and all, since you are the one who brought it up (or your ilk).

  
  
  Could you do me a favor and explain what benefit of what doubt you are giving to Sun regarding "patents and all"? I'm not sure I know what you mean.

• Roy Schestowitz

  2008-12-21 23:03:36

  Rant to Michael about it then. Reporters took his work at face value.

• Roy Schestowitz

  2008-12-21 23:05:03

  Re: Sun and patents; you don't believe they would pull a SCO, do you? Unlike Microsoft, for example...

• Jo Shields

  2008-12-21 23:10:31

  Reporters took his work at face value.

  
  
  Taking those results at face value would qualify you as an idiot. Whichever reporters did so would meet the requirements.
  
  

  Re: Sun and patents; you don’t believe they would pull a SCO, do you? Unlike Microsoft, for example…

  
  
  Not unless they had good reason to
  
  But as you say with irritating regularity, "the patenter matters" - why is Sun's work on OOXML above criticism, again? Since that would be Microsoft patents (should any be applicable)

• Roy Schestowitz

  2008-12-21 23:12:46

  Sun fought against OOXML. Now it's being thrown around in the wild.

• Jo Shields

  2008-12-21 23:14:27

  Oh yes, their so-called "fighting" against OOXML by devoting engineering resources to its implementation. I remember now.

• Dan O'Brian

  2008-12-21 23:14:33

  Well, I'm not sure if Sun would or not. Sun did, after all, fund SCO at the time of their copyright attacks on Linux. http://www.linux-watch.com/news/NS6700440248.html
  
  (had you asked me if I thought SCO would have sued before they did, I likely would have thought "no")

• Roy Schestowitz

  2008-12-21 23:19:25

  I didn't forget that and I took it into consideration.

• Jo Shields

  2008-12-21 23:21:23

  Sun fought against OOXML. Now it’s being thrown around in the wild.

  
  
  http://blogs.sun.com/GullFOSS/entry/office_open_xml_ooxml_filters
  
  If only we could all fight OOXML as hard as Sun

• Roy Schestowitz

  2008-12-21 23:25:44

  I've already remarked on this in that Web site at the time.

• Jo Shields

  2008-12-21 23:30:49

  Yet you count the above article from Sun engineering, saying "it is only natural that we care about OOXML now, too" and talking at length about Sun's plans to work on OOXML, as "fighting" it?

• Roy Schestowitz

  2008-12-21 23:32:58

  Why don't you also link to their posts that are against OOXML?

• Jo Shields

  2008-12-21 23:37:21

  Because I don't need to - I'm not saying they're "pro OOXML", and the link I gave doesn't say that either
  
  I'm saying they've spent significant resources on developing support for it and adding it to OOo
  
  And, as per usual where Sun are involved, you've completely ignored any of your accusations against Novell which are 100% equally applicable to Sun.
  
  Your subjectivity is disconcerting.

• Roy Schestowitz

  2008-12-21 23:42:46

  Sun did this much, much later.

• Dan O'Brian

  2008-12-21 23:43:35

  It's especially disconcerting since Sun not only has a similar patent deal with Microsoft (as Novell) and Sun is suspected of having funded SCO's attacks against Linux.

• Roy Schestowitz

  2008-12-21 23:45:20

  It was a settlement.

• Dan O'Brian

  2008-12-21 23:51:41

  Where is it stated that this was a settlement?

• Roy Schestowitz

  2008-12-21 23:52:54

  Here.

• aeshna23

  2008-12-22 03:26:19

  Some unfair attacks on Roy's logic are being launched here. There are massive differences between Microsoft and Sun in terms of both interest and culture. Microsoft's interest in defeating FOSS are obvious, but given Sun's past actions (e.g. Java) it's almost incomprehensible how it would now be in Sun's interest to defeat FOSS. Furthermore, Sun past actions--despite whatever earlier sins Sun may or may not committed--has created a pro-FOSS culture inside Sun. Microsoft blatantly has a culture antagonistic towards FOSS--baring some small areas of the companies which are probably there for cosmetic reasons.
  
  It is these differences between Sun and Microsoft that causes mono to be much more of a threat than OOXML in OOo--if the latter a threat at all. No double stand at all on Roy's behalf.
  
  Ironically, the Greek word for the type of judgment required to think about these issues is "Phronesis" often translated as "practical wisdom".

• Dan O'Brian

  2008-12-22 03:33:27

  but given Sun’s past actions (e.g. Java) it’s almost incomprehensible how it would now be in Sun’s interest to defeat FOSS.

  
  
  What about their funding of SCO to attack Linux?
  
  As far as OOXML, wasn't it just a few days ago that Roy published an article claiming that OOXML was a patent trap? Now it's all hunky-dory because he says it's possible that these patents can be invalidated via prior art.
  
  Why isn't it possible for patents concerning Mono to be invalidated via prior art? It's even easier to use prior art in the case of Mono (Java, scheme, smalltalk, etc).
  
  Besides, according to Roy - it's the patent holder that matters, not the patent. Microsoft is the one that would own patents on OOXML, so it is irrelevant if you believe Sun is a saint or not, according to Roy.

• aeshna23

  2008-12-22 04:03:53

  "What about their funding of SCO to attack Linux?"
  
  That was an earlier decision, before they committed to their present course. Firms can change their business strategy, but it is costly to do so. No doubt when Sun decided to free Java, the firm lost some value (stock price went down), but they hoped to regain that value by good will from their commitment to FOSS. A costly commitment to a strategy are usually hard to change for psychological as well as practical reasons.
  
  Dan, I have a question for you. I know Roy is strident in ways, but he is rational. Why are you so irrational? It's really quite unseemly. To be clear, by "irrational", I mean equivalencing thing that clearly aren't the same to make a rhetorical point.

• Roy Schestowitz

  2008-12-22 09:24:18

  Sun is not the bad guy here. It makes OOo, for starters.

• Jo Shields

  2008-12-22 09:37:08

  Sun is not the bad guy here. It makes OOo, for starters.

  
  
  And spends significant resources on adding OOXML support to it
  
  Either Sun are deliberately putting OOo users at risk for reasons unknown, or OOXML isn't the danger you're claiming it to be.
  
  Pick one.

• Roy Schestowitz

  2008-12-22 09:40:46

  Microsoft's ally (Novell) was already pushing it into its fork of OOo, which put pressure on Sun. Unlike Novell, Sun fought against OOXML.

• Jo Shields

  2008-12-22 09:42:17

  Yes, "fought", what with getting paid engineers to work on it and all
  
  And if you had the faintest idea what Go-OO was, perhaps you might understand WHY it got that feature sooner

• Victor Soliz

  2008-12-22 17:27:24

  The novell fans are just trying to say "Sun does evil too" Novell is ok. If Sun does evil too, let them fuck off from Free software as well? Is it really that hard? Mono and OOXML in OOo remain threats to free software, of course, we are now screwed up thanks to ISO and ECMA giving the title of 'open standard' to all sorts of lock-in that require licensing or patent deals, so really, it is a no brainner.
  
  It is irrelevant what Sun does or does not. Novell still made that one deal, Mono is still a danger, OOXML still terrible as an open standard. Dear Dan, diverting attention will not work forever. Thanks.
  
  It would be fun if this stayed on topic... Does anyone of the OOXML advocates have something to say about the noooxml link Roy has posted? Or should we all just ignore the actual article because Sun is evil too?

• AlexH

  2008-12-22 19:37:30

  @Victor: I think you're missing the point.
  
  Saying that there is a potential threat from patents in a product one company is pushing is one thing. Saying that there is a threat in multiple products from different companies and distributed widely is quite another. It's not just asking us to believe that Novell are ignorant of the threat; it's asking us to believe that Sun are ignorant, anyone distributing OOo is ignorant, etc. etc.
  
  It's too much. These businesses and organisations are aware of the problems with software patents. Asking us to believe that they are naive / unaware / whatever is too much.

• Roy Bixler

  2008-12-22 19:55:52

  I doubt many here would disagree that software patents in general are bad things which should never be granted. The point is that use of OOXML requires a patent licence from Microsoft. Microsoft has apparently refused to grant a licence to someone who has asked, by not responding to them in a reasonable period of time. Even if they did grant such a licence, it would most likely be done in such a way as to make it incompatible with the GPL, since they would certainly not make their grant available to sub-licencees. Put this together with Microsoft's rhetoric that "our intellectual property needs to respected", "Linux violates 235 of our patents", "Linux users have an undisclosed balance liability sheet". Given this background, it's not unreasonable to be paranoid about the use that Microsoft might make of their software patents.

• Roy Schestowitz

  2008-12-22 20:00:11

  They also use OOXML to legitimise software patents in places where these are not legal.

• AlexH

  2008-12-22 20:01:28

  @Roy B: it's not clear that they made the request along the right channels; and MS already make available to the general public a patent license that's royalty free, which is what OOo relies on.
  
  It's fine to be paranoid about software patents.

• Roy Schestowitz

  2008-12-22 20:05:09

  MS already make available to the general public a patent license that’s royalty free, which is what OOo relies on.

  
  
  But is this permanent? The thing about the OSP is that it's not a binding contract, AFAIK.
  
  http://brendanscott.wordpress.com/2008/01/18/more-on-the-osp/

• AlexH

  2008-12-22 20:09:21

  @Roy: it doesn't have to be a binding contract, because it's offered generally. It's a bit like the GPL in that regard - the GPL is also not a contract.
  
  What's important is that it is irrevocable: they cannot take it back. It's unilateral, it isn't limited: they can only suspend it if someone sues them first.

• Roy Schestowitz

  2008-12-22 20:15:40

  It's quite unsatisfactory as it is.
  
  http://brendanscott.wordpress.com/2008/08/04/the-updated-osp-and-free-software-interoperability/
  
  IANAL, but this^^ guy is.

• AlexH

  2008-12-23 07:49:02

  @Roy: the argument isn't that the OSP is the best thing since sliced bread; it clearly isn't. But it is good enough to implement OOXML without worry from patents; the link you've posted agrees and so do the people implementing OOXML.