Microsoft: Still evil after all these years, yea…
(Apologies to Paul Simon…)
So by now everyone has probably heard about the latest round of FUD emanating from Redmond, claiming that Linux et al violate 235 Microsoft patents. The most amusing yet direct response is undoubtedly that of Sun’s Tim Bray:
Litigate or shut up.
Speaking only for myself and not for Adobe (as always, but in this case I want to emphasize it particularly), I think this latest move by Microsoft is proof positive that for all of their claimed moves towards more openness, Microsoft is still just as evil as they ever were. Does it really matter if they submit OOXML and XPS to standards bodies if they attack the open source operating systems and applications that implement those standards? Does having evangelists who promote interoperability and open source mean anything when they turn around and attack the vendors they are supposed to interoperate with? This is a sad day for the industry. I guess the big news here is that at least Microsoft is attacking more directly now instead of hiding behind SCO. All in all, a sad day.
Of course, this also makes me even happier that I escaped the Redmond orbit and found my way to Adobe. Here’s a summary of recent moves on the openness front by the two companies, which to me speak volumes:
| Adobe | Microsoft |
|---|---|
| Turned PDF Specification over to ISO via AIIM | Promises to turn XPS over to ECMA or equivalent for rubber-stamping someday soon. |
| Donated Flash Player’s Tamarin JavaScript engine to Mozilla | Open sourced their thin “DLR” that sits on top of their .Net runtime, but the license isn’t GPL compatible or compatible with OSI’s Open Source Definition. |
| Open sourced our Flex runtime and our Javascript & MXML compilers. | C# and XAML compilers for Silverlight are closed source, as are the .Net runtime libraries. |
| Flash Player runs on Linux and Solaris. | Silverlight doesn’t and won’t anytime soon. Novell’s MONO folks are developing their own version, but what Microsoft patents are they likely to run afoul of? We know Microsoft has warned other developers off of building XAML-based engines previously (c.f. xamlon). |
| Apollo developers are doing extensive work on WebKit, beefing up its Windows support in particular. | IE still sucks. |
Here’s some other interesting reading:
- O’Reilly Network’s take on the announcement: “However, today’s news brings the major reason you should run away from depending on Microsoft technology like it had a case of Ebola.”
- Joe Wilcox of Microsoft Watch says: “Microsoft should carefully consider its allegations. SCO hasn’t done as well as hoped with its Linux patent infringement claims, and the company came out swinging with a lineage of recognizable Unix patents. Meanwhile, the U.S. Supreme Court is pushing back against patent lawsuits.”
- Simon Phipps of Sun says: “Very sad that they can’t work with the emerging paradigm like the rest of us instead of fighting it. No wonder so many developers hate them.”
- Mike Gunderloy,
formerly oflong-time Microsoft developer, posted an article before the news hit talking about software patents where he said “Microsoft itself is built on open intellectual property from the first three or four decades of computer science.” How apt and ironic. - Mary Jo Foley says “What’s got Microsoft so spooked? As the Fortune article noted, the GPL v3’s provisions regarding the Microsoft-Novell deal suggest that Microsoft itself could be considered a ‘Linux distributor,’ and thus beholden to the GPL v3 terms. And even if that doesn’t happen, under the GPL v3, other Linux distributors would be barred from doing deals like the one struck by Novell and Microsoft.”
- [Update 5:03pm] Via Stefan Tilkov, Charles Nutter says: “I’d hate to be an OSS developer or apologist at Microsoft today. If Sun did something like this I’d resign.”
Shebanator 
“Apollo developers are doing extensive work on WebKit, beefing up its Windows support in particular.”
Wow, they are? Funny how they’ve contributed exactly zero back to the WebKit project in the way of Windows improvements.
The Microsoft Permissive License is an open source license and I do not see how it is incompatible with the GPL. It does not have the OSI rubber-stamp to call it open source, but that has nothing to do with it being compatible or not. An organization needs to bring the license up with OSI to get it approved, and Microsoft for historical reasons chose not to (at one point they felt uneasy because the OSI site had a whole section on the “Halloween Documents”).
Maybe you could point us to the analysis that you did on it, my reading -and our lawyers reading- is that this is an open source compatible license. We distribute IronPython and we will distribute the DLR as part of our 100% open source distribution OpenSUSE and IronPython is also being distributed as part of Debian and they have historically been the stricter distribution when it comes to accepting what qualifies as open source and what does not qualify as open source.
Miguel
Mono Project, Novell.
Umm, not quite zero. Here are two I found quickly:
http://bugs.webkit.org/show_bug.cgi?id=13212
http://bugs.webkit.org/show_bug.cgi?id=6286
There still is a lot more coming. The bottom line is that the commitment to work together is there, even if the logistics haven’t been worked out, and Apollo engineers are commiters on WebKit.
Nice troll attempt, though.
Miguel,
Always a pleasure to here from you. I haven’t done any analysis of the legal terms myself. IANAL. However, I did go by the links I referred to in the article, and others that came up in the search on google. Here are a few more:
http://www.regdeveloper.co.uk/2005/10/19/shared_source_licences/
http://weblogs.mozillazine.org/gerv/archives/2005/10/microsoft_annou.html
http://blogs.msdn.com/jasonmatusow/archive/2005/10/19/482562.aspx
http://www.lessig.org/blog/archives/003188.shtml
http://lwn.net/Articles/156310/
The first has a quote from Jason Matusow of Microsoft: “It should be compatible with most open source licences, but I think it would still conflict with the GPL.”
Also, I wanted to comment on a few things:
* Tamarin source code size (minus PCRE): 98,000 lines of C++ and header files.
* DLR (as shipped in IronPython 2-Alpha): 94,000 lines of C# code.
Tamarin is great, and the DLR is still in the early betas, but am not sure why you felt compelled to call their DLR contribution “thin”.
Miguel.
So the first link from LWN effectively says that it abides by the four freedoms that the FSF expects from any license, and even that it follows the GPL model (The quotes from George Greve).
And Matusow basically was speculating.
So this is hardly proves your point, and as I said, Debian and our lawyers consider the MsPL an open source license that fits the OSI definition of free software.
Miguel.
It seems clear from the google searching i’ve done that many people think there may be issues with GPL compatibility. On the other hand, I couldn’t find anything on the internet from any open source organization who said “we’ve reviewed the terms with our lawyers and believe they are in fact GPL compatible”. But if your lawyers say you are OK then I certainly am not going to try and argue with them – its ultimately a judgment call on risk that every organization needs to make for themselves.
Also, Matusow’s comment may not be an official Microsoft legal interpretation, but given his job it s would be surprised if he didn’t at least get chance to review the terms before they were announced. Thus it seems likely his opinion meshes pretty well with that of the people who framed the license terms in the first place, and thus goes to Microsoft’s intentions.
On the thin thing: probably not the best choice of words. I wasn’t trying to imply anything about the size in LOC, just that its functionality is only a small layer on top of the larger CLR which remains proprietary. By contrast, the Tamarin code IS more or less the entire chunk of runtime VM that sits in Flash Player. Is it a bad thing ours is so much tighter than theirs? 🙂
Neither of those WebKit bugs you linked to has *anything* to do with Windows. One of the issues addresses a crash in Safari, and the other provides an interface around a concept that has yet to be defined.
The point I am making here is rather than showering Adobe with praise for their excellent open source work in relation to WebKit it would be worth understanding exactly what Adobe has contributed to date: two cross-platform bug fixes, and a bunch of build system files that are useless to anyone outside of Adobe. Perhaps in the future they will be more involved in the WebKit open source community but, please, save the congratulations until their contributions are more than fantasy.
Dear Anonymous Coward,
I’m not going to get into a debate with you over which WebKit bug fixes are worthy or not worthy, or about how many fixes have actually been contributed by Adobe engineers (my two examples took me all of 30 seconds to find). It sounds to me like you have an ax to grind here that has little to do with the actual topic of my post: Microsoft’s attack on the open source movement. You are entitled to your opinion, of course, but I’d prefer to keep discussion here on topic, so if you really want to sound off more about how much Adobe is/is not contributing to WebKit on a daily basis, well, I’d encourage you to get your own blog and do exactly that.
One minor factual correction: I’ve done contract work for Microsoft in the past, and I’ve worked almost exclusively with Microsoft technologies since roughly the DOS 2.0 era, but I’ve never been a Microsoft employee.
Mike: my mistake, sorry about that.