They seem to be doing everything that was asked of them at the height of the sleazy days:<p>- They removed all bundled adware<p>- They scan the hosted packages for adware bundled by the developer and warn accordingly<p>- They removed deceptive download buttons<p>This is really very good news, and I'm willing to give them another shot. To quote from later in the Reddit thread:<p><i>It brought in quite a lot of revenue, but obviously that strategy is not sustainable and SourceForge was/would have been a sinking ship. The previous owners were a publicly traded large corporation and SourceForge was not a core part of their business. We are a lean web company with talented developers that has the ability to do things more efficiently. The site is monetized via advertising, but we believe it can be profitable and sustainable without throwing users and developers under the bus. At over a million unique visitors per day, we don't think we need to trick people into clicking on ads in order to turn a sustainable profit.</i><p>That had me nodding my head in understanding at "publicly traded corporation." I get it. I work for one. They will throw their workers, their customers, and their business into a meat grinder without batting an eyelash if it means being able to add a couple more cents to their shareholders' dividends.
This is a very good development, but it's difficult to see what SF could do better than Github/Gitlab/Bitbucket, they are bogged down by 15 years of technical debt and a ruined reputation.<p>Maybe they could retool as a Cloudflare/Akamai competitor? Their network of mirrors is probably the one thing still distinguishing them. Or they could go full-FOSS and somehow integrate with distributions, like a cross-distro Launchpad, but that's a very very <i>very</i> niche market. Or they could find a mobile-oriented spin (I honestly don't know anything about mobile dev).<p>Or, and I say this very seriously, they could find a way of getting bought by Microsoft. A lot of SF projects are legacy win32 apps that people still find essential (FileZilla etc); MS could buy them and build an appstore that actually has the stuff people want, with real developer tools and workflow powering it all. MS backing would remove the malware stigma, at the very least.
SourceForge spent many years making decisions that sank its reputation. I don't think that's going to dramatically turn around overnight.<p>HOWEVER... it bums me out when a bad actor makes legitimate efforts to turn their act around, and get mostly shit and snark for it. Paraphrasing:<p>* "<i>I would never download something from them no matter who owns the company now, because the previous owners sucked and I'm emotionally invested in disliking the brand.</i>"<p>* "<i>This speed test service that they just made available for no ulterior motive at all doesn't work outside the U.S. yet. Pffffft.</i>"<p>* "<i>Why would anyone use them instead of GitHub?</i>"<p>* Etc... far more nasty stuff over in Reddit discussion.<p>Look. Years ago I switched over to GitHub (and later GitLab)... because SF was slow to adopt Git, and the interface was pretty weak once they did. There are more steps and complexity involved in setting up a full SF project than creating a simple GitHub/GitLab repo. So even now, I wouldn't consider SF for hosting a personal project Git repo if I had no intention of distributing binaries.<p>However, SF has <i>always</i> been geared more toward hosting full project sites directed at end-users, rather than simply hosting a source code repo for developers. That's why they "lost" to GitHub, because it turns out that most developers just need the best source code host and don't care about distributing binaries to end-users.<p>But if you have a project that you <i>want</i> to share with the world in binary form, rather than just a resume item to be seen by other developers, then SF has never had a serious challenger. GitHub, GitLab, Bitbucket... none of those guys really care about competing in THAT space. So for the couple of projects that I've wanted to make end-user facing, I've continued to use SF even though my source code is over on GitLab. Bandwidth is expensive, and hosting a reliable end-user facing website is a pain in the ass, so I'm grateful that SF is there as a free option.<p>So if you just need Git hosting, maybe you don't care. But it's fantastic that someone has stepped in and tried to right the ship, so there will continue to be a viable option for end-user facing binary hosting. I'm grateful for these guys, and hope they succeed.<p>Seriously... we've largely started embracing <i>Microsoft</i> of all companies after their recent about-face. There's no reason not to be positive about SF trying to become a good actor, even if their strengths don't happen to fit your own use case.
Their new advertised speed test has servers only in US. Completely unusable for the rest of the world (unless you need to measure your interned speed against US servers).