Slashdot Mirror
SourceForge Drifting
Zocalo sent us a story running at FSF Europe talking about SourceForge's Drifting. Talks about the fact that they are releasing a closed-source version of the code commercially and various copyright related things. Obviously VA owns both SF and Slashdot so I'm skewed, but my personal opinion is that VA is doing what they need to do to make a buck while still providing the SourceForge.net website to the Open Source community. And I think their decision to sell a closed-source proprietary version of the code would be hypocritical, except that they aren't a 100% open-source company any more. And *that* is the part that makes me the most sad.
Step 1: Start an open-source based company
Step 3: Profit!
Apparently Step 2 is "completely change the business model of this corporation so that it may actually make money."
Bitter pill to swallow, but giving away IP just doesn't work.
Considering that no one is exactly sure if VA can make it as a business selling proprietary extensions to Source Forge has anyone thought about what will happen to Freshmeat and Source Forge if (or is it when) VA goes under?
I know that a couple of projects have started mirroring their Source Forge content in case anything happens but are there any credible replacements being worked in case both these extremely useful sites lose their their parent company? Specifically are there any sites that are viable replacements to either Freshmeat or SourceForge? Currently we have multiple Linux distros so the death of one, two or more companies in that area would be sad but not devastating on the other hand the dissappearance of VA considering how much of a central repository for Open Source apps SourceForge and Freshmeat have become would be devastating.
If people out there take serious issue with Source Forge's turn to the proprietary, then take the last release of open source code and start your own Source Forge. I mean isn't that supposed to be one of the magical things about open source, that folks who want to go proprietary cannot because the community will hijack it.
Of course if you want to set up your own Source fFrge you have to have the money to run all of the servers, bandwidth, etc. Don't have the cash? Well I guess that's what Source Forge was running into as well.
Personally I think that Source Forge being open source itself was cool but rather secondary to the fact that source forge provides a great place for people to collaborate on projects. If they have to close the source to make it financially feasible to continue to provide the service, so be it. Which would be worse for the community: Source Forge running on proprietary software or Source Forge shutting down?
Unless the FSF is going to fund an open alternative to Source Forge they should get off their high horse.
This sig has been temporarily disconnected or is no longer in service
Note that the FSF, which does like all things free, is more concerned about the possible GNU GPL violations that might be occuring, and "appropriation" of contributors' work. While I'm not an GPL-junkie, this does seem to be a valid point from the FSF, with SF walking a thin, grey line.
After visitng linuxworld and drilling their sales reps we came to the conclusion that Sourceforge can't compete with free alternatives. (by 'we' I mean the software Co. I'm working for)
Bugzilla/bonsai/tinderbox provides a more complete solution. We were even able to modify the trio to deal with java, our many different build scripts (make is rather lacking for java), and our test automation.
What we found was that Sourceforge provided discussion groups which we got using exchange or INND, bug tracking which wasn't nearly as feature rich as bugzilla, and cvs integration which bonsai provided just as well. It was still lacking the automated builds, and by the time they got back to us after linuxworld we had allready deployed the bugzilla solution (partly thanks to some nice debian packages put together by Remi Perrot).
One large drawback is that bonsai relies on glimpse as its fulltext indexer. Glimpse used to be free but since then has gone commercial. We were, however, able to find some old glimpse source (which may have been GPL or artistic license - perhaps we should redistribute the old code as GNUlimpse).
We have made our own tweaks to bugzilla/tinderbox/bonsai and contributed a few of them back to the mozilla developers (in the future probably all will be recycled into the public implementation).
I'm not the first to mention it, but this bears repeating: this isn't a sign of VA * abandoning their ideals; they are doing the best under the circumstances. It's really a sign of them struggling for their lives in a hostile environment.
The recently posted their quarterly income statement, and my analysis is that it looks very bad. They posted a net profit of negative $290 million. Most of that is imaginary money, so let's look at the rest of the figures to get an idea what is actually going on.
To project the long term viability of the company, we will look at the burn rate, and try to extend that against their short term assets, accounting for any factors that will change their revenues or expenses.
The balance sheet shows that their current assets continue to drop. Particularly disturbing is the continued drop in cash and equivalents and short term investments. These have gone down by about $17 million, indicating that as their burn rate. Inventory has also decreased, presumably as they sell off what remains from their hardware business. This provides a revenue stream that has basically finished this quarter. Since the $8 million drop there is about half of the total revenue, we can expect revenue next quarter to be about half of what it was this quarter.
Long term assets are also dropping. Reductions in long term capital are likely due to exiting the hardware business and getting rid of associated facilities. They are also writing off huge amounts of goodwill and intangibles. Neither of these is important, since the money was already spent and does not affect their long term viability. The only thing to note is that the poor economy now means that the money spent acquiring these assets is not giving much of a return, and they would have been better just sticking it in the bank.
Although their liabilities are increasing, they do not explain why, categorizing the increase as "other liabilities". We can't factor this into any calculations directly.
It appears that the current burn rate is $17 million per quarter, against reserves of about $97 million. With revenues expected to fall to half of the $16 million they are now once the remaining hardware inventory is sold, we expect the burn rate to increase to $25 million. At this rate we can expect the company to survive four quarters, just one year.
In that time frame, there really isn't anything that we can expect to make them viable. Revenues from SourceForge On Site will likely ramp up, but that will be a slow process that can not offset much of the projected loss. Further, aggressive cost cutting measures will reduce the burn rate, but it is unlikely they can cut it enough to survive long, particularly with the conflicting goal of building the SourceForge brand and ramping development and sales.
I really don't see a future for VA. Look for them to sell off unprofitable assets (likely including Slashdot, unless the changes Rob discussed can make it profitable). Developers with projects on SourceForge should make offsite backups just in case they remove it suddenly and don't give developers sufficient time to withdraw their code. Think also what the rush on the site will be when they announce its closing and everybody tries to checkout their projects at the same time.
Even Slashdot wants to hide some things
So it's a service, not a software. That's interesting, because it implies that VA won't make its source code available, as a service doesn't have any source code. It is based on a software, but you didn't say that you'll continue to make its source code available as free software.
So I assume that the SourceForge software will become proprietary. Correct me if I'm wrong.
Read the FSF website some day -- free software does not refer to GPL only, it's basically a different term for open source (with some philosophical differences). E.g. the BSD license is an open source license as well as a free software license, but it is not "copyleft." I guess about 40% of the projects hosted by SF are free software, or open source, whichever term you prefer. (The other 60% are status 1, planning.)
The more interesting question is, will they continue to do so?
Sig (appended to the end of comments I post, 54 chars)