3 ms·
One day too early for Aprils Fools, ain't it? I'm not really one to kiss up Microsoft most of the time, but the fact GitHub is becoming the new SourceForge is
by kemonocode 10y ago
One day too early for Aprils Fools, ain't it?
I'm not really one to kiss up Microsoft most of the time, but the fact GitHub is becoming the new SourceForge is not too keen.
- xanderstrike 10y agoI can't see Github going the way of SourceForge. SF was exclusively a download portal for FOSS, it needed ads (and, eventually, malware) to survive because it had no other revenue stream. Github, on the other hand, really only hosts FOSS to further cement its position as the standard in git hosting. Their bread and butter is large organizations who are paying per seat to keep their code hosted there.
- dredmorbius 10y agoThere was an effort at VA Linux to convert SourceForge into an enterprise project. That failed. I believe that's when SF was spun off (along with Slashdot) as an independent. Or was it that SF became what was pretty much all that was left of VA after the hardware business left? I'm a bit fuzzy on the history, and wasn't there at the time, though I do know people who were. Point being that SourceForge was not intended to be exclusively FOSS.
- ajross 10y agoSourceForge really only became a problem very late in the game when they had mostly been abandoned for new development and started trying to monetize the (often semi-abandoned) projects still hosted there. Their behavior was bad, but honestly I think a little too much was made of it. People got all wound up over the ethics issues (leeching off of other people's open source work to sell creepy adware), but the actual work itself remained accessible and supported. Honestly even in its fading days, if you wanted a free host for a CVS archive and mailing list, SourceForge remained a good choice. It was providing more value than it derived even then. If that's what happens to github, I think we can deal. Someone would just start a giant project to pull and mirror all the junk locked inside github (git makes that process distributed, you don't have to actually crawl their site and can use external archives) and we'd just start the cycle all over again.