Josh Ledgard On MS's Future Open Source Efforts

prostoalex writes "Josh Ledgard from Microsoft, the developer responsible for open-sourcing WiX and WTL, is looking for opinions on what Microsoft should do next in regards to the open source movement that he himself established within the company. "Would you have interest in working on these types of projects with Microsoft? If not, what could entice you? If so, what would be your motivation?", asks Josh." Update: 08/24 19:04 GMT by T : As Ledgard writes on his site, "I am NOT the person responsible for the WIX/WTL projects. I cite them as examples and am working with people who where responsible for those projects to enable more of the same for the groups I work in." Sorry for the misattribution!

Comments

OpenSource IE

[drater]

Open Source IE. That's what they should do.

Office File Formats

[Anonymous Coward]

Open up the file formats for Word documents so that other programs (e.g. Open Office) can correctly decode the formatting.

Make OSS economically sense ?

[vi (editor)]

I wonder what sense it makes for a large company to produce OSS.
Basically there are only three points

1. To push competitors out of the field by offering their products for free. This is a more agressive version of what MS did with Netscape. SAP and IBM are using this to attack Oracle and MS.
2. To establish own standards and push competing approaches out of the way. Stuff like TCP/IP and XML come into mind...
3. To grab control of a competitor's standard base by offering free tools for a modified version. That's the main point of the Eclipse project targeted at SUN.

So, OSS can strenghten/enlarge your own market share greatly but I don't see people making money by OSS directly.
No, RMS getting donated money to get a shower doesn't count.

What would get me interested?

[59Bassman]

Hmm...

1) Microsoft quit funding "independent" bogus TCO research to discredit OS operating systems. Oh yeah, and call off SCO.

2) Microsoft quit attempting to make all of their file formats dependent upon the OS/software that they write. The data is MINE, and I should be able to use other software to read the data. Commit to open file formats and I'd look a lot more favorably on MS.

3) Microsoft quit using draconian EULAs that make me fear that any contribution I made to a MS effort would be locked away for good once MS got a hold of it.

4) Money. Truckloads of it.

Well, maybe not so much 4, but the first 3 would be a good start.

My problem is that I've got such a bad image of Microsoft after working with their stuff for the past 12 years or so that at the end of the day I'd rather contribute effort to a "real" open source effort than anything funded by Microsoft. I just don't trust them to "do the right thing" with anything that came out of an OSS initiative.

Opensource whatever, I'm not interested

[Pingo]

I don't see any point in opensourcing any
Microsoft software except for Windows Media Player series 9 with codecs, perhaps also
Windows Media Encoder could be of some interest. //Pingo

Why ask first?

[houghi]

What I think is distuurbing that they want to askt this question. Just start the project and see who is interested. If it is interesting, people will jump aboard. If not, people will not.

It almost looks as if they know that the OSS comunity will spit them out and then the can play the underdog.

If people are interested, they can first play the nice guy who allowed the OSS comunity what they wanted and then let it blow up to proove that OSS does not work.

This is a win-win situation for them. So my question to them is, if you think that Open Source is so good, when can we help you with other parts that ARE interesting (and who need to be open by European court desision anyway). If you truly believe the project is interesting, it will create followers. If not, it will die a silent dead, as many projects that were started.

I would

[Apreche]

If Microsoft started open source projects, with "real" open source licenses I would be glad to work on them under two conditions. First, the project has to interest me. That's rather obvious that since open source work is volunteer that nobody is going to work on something that doesn't interest them. The second thing is it has to be software I can use. Since I don't run windows there are probably going to be very few MS OSS projects I would work on.

What MS SHOULD do is appeal to all the Windows developers out there. Yes, there are people out there who live in Visual Studio and love windows. They should get these people to fix all the bugs in windows and IE and such. There are people out there, willing and able to do work which the internal MS developers have failed to do multiple times over. Give someone else a try.

Re:I would

[Tarwn]

Apreche raises a good point, mod him up (I'm mod point-les right now).

The best source of Open Source developers your going to find that will work on developiung MS products in their freetime are going to be found in the Windows user groups. These are the people that run into the little problems here and there, consistently, over and over, for weeks and weeks on end. Some of them would probably bribe you to let them get in there and fix the problems. I say some because it doesn't matter wat community your part of, there is always someone who would prefer to whine to doing real work.

So start with a small windows groups of developers. Maybe give them IE as a starting place since it is not an essential product, but rather a bundled one (ie, it doesn't directly have a price). See if you can get them interested in implementing some of the newer standards, or re-implementing CSS or the JScript DOM to match the standards. Don't ask them to help, offer limited access to a SourceShare archive and a tasklist, let them work on it as they please. I'm willing to bet you'll find a lot of work getting done, especially after one of them comes back from an especially painful application install or intranet development.

I have been in the grey area, ie not a fanatic about either Linux or Windows, for years. I develop (for work) primarily with VS.Net, with occasional Web Devel mostly aimed at IE. I prefer Linux these days simply because everything happens for a reason. With Windows it's all black magic even when I kind of know what is going on under the hood based on how the system acts. It's the difference between training Gorillas based on watchng them for several years or training chimps after being given Gods notes on how and why he built them.

-T

No need to open more.

[taj]

Though I would pay attention if they open sourced .net.

What they should try doing is participating with the community rather than trying to harvest/divide it. Ship perl, python, apache... Work with some of these open source projects. Show this isnt headed the direction of mosaic, embraced and extinguished mit licensed works.

Show this isnt just some game. Otherwise, have fun.

MSFT's culture is bankrupt. They have little to bring to the table. Show they are changing their culture, come out and play.

DirectX

[GuyFawkes]

Because it is the one area where MS completely and utterly destroys Linux and the one are where Linux really needs to grow up.

Course, it won't happen, ever.

Code contribution tracking

[afreniere]

I'm being a weenie and posting this to an early reply because I think it's one of the few thoughtful replies. :P

One of the major potential benefits of contributing to open-source projects is that, when searching for a future programming job, one can point to one's open-source contributions and say "Here's some of my code, and people are using it." This works especially if one has contributed to a project with prestige - something that a Microsoft-sanctioned project would certainly have in the closed-source corporate world.

However, it can be difficult to pick out the code that one has contributed from a large project and say, "yeah, download this tgz and look at kluge.cpp lines 377-421, that's my code!" So I would propose, as a carrot to your future open-source contributors, that you design a system that keeps a database of who contributed code, how old it is, and maybe some other statistics about it. You could post a summary page for each contributor with browseable links to the code and statistics.

-Ansel.

Microsoft needs to truly contribute to Open Source

[dmayle]

As I'm sure many of you noticed, Microsoft has been making some 'Open Source' pushes as of late. They started by creating the CPL, getting it certified, and have now been hiring prominent open source developers to work on CPL.

For those of you not in the know, the key distinction between CPL and the GPL, is that I can incorporate a CPL'ed project into my code, whole, without contributing back, as long as I don't modify the CPL'ed code. If I do, I have to publish modifications to the CPL code in the same way I would for the GPL.

I'm of the opinion that the CPL has two serious flaws:

• Transparency
• Free rides

First, the transparency. How often have we all seen companies violating the GPL? I can assure you, it's a whole lot more frequent than even we hear about. I've known many Windows developers, who had no clue about, or respect for, the GPL, and would download bits of code from anywhere to incorporate if it made their lives easier. (Often because they didn't understand the task at hand, and they just wanted to get it to work.) It's not normally a matter of policy, but I've definitely seen it knowingly overlooked at companies I've worked for previously.

Right now, if we find a company using GPL'ed source, we have a smoking gun we can use to ensure license compliance. When someone incorporates CPL code, there's no way to prove that they modified it themselves, and so CPL compliance is based entirely on the goodwill of corporations, and we've seen how generous that goodwill is...

Second, there's the free rides. The CPL is designed so that companies can take advantage of the work of open source developers without having to compensate them in any way. With proprietary code, developers receive fiduciary compensation, and, traditionally, with Open Source, the developers receive the source for their project, and any derivative works.

With the CPL, companies get a free ride, which in the end screws the developers out of any benefit of their hard work.

Now, the reason why I bring this all up is that, as mentioned above, Microsoft has been hiring prominent open source developers, having them release their source under the CPL. It is my belief that Microsoft wishes undermine the Open Source movement. They hope to get the CPL to be popular, so that they can freely incorporate works developed by others without having to contribute back. They've hired 'evangelists' to show the benefits of CPL, without showing the downsides. (Because, of course, Microsoft is currently releasing software, and they want us to believe that's not going to change.)

So, in conclusion, Microsoft, what I'd like to see is Microsoft projects released under the GPL as well, so that we can see a benefit to the Open Source movement, not just to you.

Re:No Interest Whatsoever.

[torpor]

Yeah, and I guess all that mass wealth that gets taken from Cocaine dealers and redistributed is being done by "Communists" too, eh?

Microsoft got this way because they were exceptionally vicious and predatory, destroying all competitors, caring not one single iota about quality, just like your average crack dealer ... should society allow such behaviour to go by, rewarded?

They have ruined the computer industry. Why shouldn't they be punished?

Re:Office..

[nova_ostrich]

How about the 95/98 operating systems? Sure they're more unstable than a schizophrenic, but they're unsupported now anyway. Plus, XP isn't even based on them, so they shouldn't be releasing any secrets... well, in theory. I'm sure the fact that IE is in there might be a problem, but maybe they could release some of the other parts.

Some light on your two items

[hummassa]

1. Id like to see a more rigid Linux Kernel API (so a module doesnt have to be recompiled between kernel maintenace versions),

This is not going to happen. Linus himself already told innumerous times that he thinks this would make unnecessarily hard to change the kernel APIs, and that, notwithstanding the point "2", below, he is not interested in binary compatibility for kernel-space things between versions of the kernel. This is right, and if you did not get it yet, I'll explain it to you: it leads to Big Bad Difficult Bugs, trying to get kernel modules to work in various kernel versions. Many things evolve from one version of the kernel to the next, many assumptions change.

2. and the 'grey' area of binary modules sorted out as well. I dont think it will happen.

This one has already happened, (*) but many people still want to pretend it didn't. Some binary modules are derived works of the kernel, and such, to be distributed at all, they must be distributed under the GPL. Some binary modules are not derived work of the kernel, and as such, they can be distributed under any license that the author seems fit. What determines if a work (in the case, a binary module) is a derived work of another (the kernel) is copyright law.

In the USofA and in Brasil, the copyright law states that a derived work is the result of some non-automated transformation of the original work. USofAn case law established the method of "abstraction, filtration, and comparison" [AFC] to determine derivation of works.

There is a myth, spread by the last paragraph of the "postamble" of the GPL, "How to Apply These Terms to Your New Programs":

This General Public License does not permit incorporating your program into proprietary programs. If your program is a subroutine library, you may consider it more useful to permit linking proprietary applications with the library. If this is what you want to do, use the GNU Library General Public License instead of this License.

I will repeat here the position I have after carefully studying the GPL, copyright law, and case law: the GPL regulates the licensing to derived works of the GPL'd work, but it cannot regulate the licensing of encompassing "anthology" works. Linking does not make a work derived on other work: to see if some work is a derived work, apply the [AFC] method. Some (not all) linking, non-derived, non-GPL'd, works can be even distributed along with a GPL'd work, because they would be covered by the "mere aggregation" clause in the 3rd paragraph of section 2 of the GPL.

And one more funny stuff: the section 6 of the GPL states:

"You may not impose any further restrictions on the recipients' exercise of the rights granted herein."

This basically means that you can't even clarify the license further than copyright law would restrict the rights of the recipient of your work (**), without rendering it undistributable by others (and even by yourself, if your work is derived from another GPL'd work).

But, OTOH, IANAL and TINLA. But I am a paralegal. IMMV and the others TFFLAs :-)

(*) Google for: "linus torvalds" abstraction filtration comparison binary
(**) Google: "hans reiser" derivative plagiarism

what and why

[PlainBlack]

I'll tell you what you should open source, and why it will benefit MS in the long run.

1) Internet Explorer

IE is free anyway, so there's no money to be lost. But what you are losing is marketshare to Mozilla and Opera-based browsers. This is due mainly to the security problems in IE. You could save developer resources AND marketshare by just releasing the source of IE so that it can be fixed/expanded more easily and quickly.

2) Visual Studio

There's nothing that will get developers rallying around your product like allowing them to customize the development tools they have to use every day. In addition, more people could/would write apps for Windows if they didn't have to spend a bunch of money purchasing compilers that they can get for free on Mac/Linux/Unix. You're already dominating on user-level market share. Having free development tools will undoubtedly convert developers, and the universities that are teaching those developers.

3) MSN Messenger

Instant messaging is a hot topic everywhere. AIM obviously dominates the market now, and open source systems like Jabber are starting to take a small foothold in businesses. A quick way for you to crush all of that is to open source your IM client and it's protocols. It's already in use by a lot of people, but it could be in use by a lot more. In addition, you'll likely get a lot of free ports to Mac/Linux/Unix and draw customer-base from there as well. This will give you more ad revenue, and may also lead to people getting MSN internet, hotmail, as well as purchasing your server products.

4) Virtual PC

I know you guys just paid a lot of money for buying it, but the whole purpose of buying it was to get Mac and Linux geeks to start using and relying on Windows apps, wasn't it? By providing it as open source you'll get faster adoption than ever before, which will only increase the number of Windows apps in use by these non-Windows markets.

5) Direct X

This is likely to be the most controversial one internal to MSFT. I know that a lot of people in MSFT believe that the gamer market is the second biggest reason people are holding on to Windows (2nd to MS Office). You're losing that war though. More and more games are being developed to Open GL and other open standards to make them more portable. Giant games like Unreal Tournament, Savage, and America's Army ship with Linux versions these days. And others like Doom 3 and Neverwinter Nights put out Linux versions shortly after their release. In addition, MSFT has become a huge publisher in the gaming market. Releasing Direct X as open source would allow more of your games to be played on more systems, which will increase the profits of that division. In addition, it may help you to hold on to your lead role in the game technology world, instead of losing it to other open standards.

A lot of people fault you for developing your own standards. If you take some of these technologies, and open source them, you can permanently make them become the standard.