Well, that’s annoying. The freecode.com release-announcement site stopped accepting updates yesterday. The site says it shut down due to low traffic.
Accordingly, I’ve issued a final archival version of freecode-submit and shipped a version of shipper that no longer tries to do freecode notifications, issuing a complaint instead.
Had to revise the How to Become a Hacker, too. I used to advise newbs to watch that site for projects they might want to join. Can’t do that any more.
With this site gone, where can people go for a cross-sectional view of what open source is doing? Bummer….
Freecode/Freshmeat thrived in a time when open source aficionados had to look for software that did what they wanted, download the tarball, and compile it themselves. Those days are gone; we’re living in the era of huge distro repositories and the Ubuntu Software Center. The distros and app stores have taken over the exposure role that Freecode used to fill. “apt-cache search” is your friend.
For keeping abreast of up-to-the-minute developments aggregated across open source projects, I smell a hack opportunity. Maybe a bot that crawls github and reports check-ins or release branches as they happen?
>For keeping abreast of up-to-the-minute developments aggregated across open source projects, I smell a hack opportunity. Maybe a bot that crawls github and reports check-ins or release branches as they happen?
Not good enough. Not all the world is github by any means.
A slimmer replacement for freecode.com might be interesting – a specialized blog, basically, where release notifications can be submitted via web form. The real problem with the design would be spam blocking. Might be addressible with a web-of-trust approach – you have to be sponsored on by somebody who is already on.
> The distros and app stores have taken over the exposure role that Freecode used to fill.
But how do the distribution makers know what’s now?
> But how do the distribution makers know what’s now?
They get petitioned to accept (or create) packages for new software, either by that software’s maintainers or by interested end-users. They then decide whether it should go in the repositories based on whether it seems like a good idea at the time.
They also decide whether to _update_ those packages based on whether it seems like a good idea at the time, but that’s another story.
So what if there were a bot that crawled any active software repository and reported what was new? On a site where you could in turn report any software repo you knew of? How spammable would such a site be, if it required a site to respond to queries as a real repo would? Esp. if that site also occasionally downloaded software from that repo and attempted to compile it or check it for malware?
Paul Brinkley,
Pretty spammable. If it weren’t human-curated, you could poison such a site with links that acted like real repositories but actually contained malware. Other forms of such poisoning have already pervaded OSS community sites: one of the easiest ways to get malware on a Windows machine is to attempt to download Filezilla from its Sourceforge project page. So confusing and ad-encrusted is the layout, that clicking on what you might think is the download link will instead install a useless “toolbar” that mines bitcoins or sends spam in the background.
Turn it the other way.
Have the forges/repos publish RSS feeds of their submits/builds/releases. Then anyone can aggregate.
>Have the forges/repos publish RSS feeds of their submits/builds/releases. Then anyone can aggregate.
That wouldn’t completely solve the problem either. For RSS to work you need to set in advance what feeds you’re monitoring, and the value of a site like freecode is that you get to see what other projects you don’t yet know about are doing. Some people, like me, post releases on personal websites, which makes the discovery problem far worse than if you merely had to scan all the world’s forges.
I’m sure that site had “low traffic” like the Clintons were “broke” after the White House. This is a problem of small operations selling to companies that just don’t care as much as the founders did about the project. The site probably would have sustained a small company by itself.
Outside of searching through the popular VCS repo’s and the obvious Google, I have found Twitter to be helpful and it would be fairly easy to write a watchdog monitor for popular hashtags around OpenSource.
I’m thinking, there’s gotta be something on github or SourceForge to the effect of providing the “cross-sectional view of what open source is doing”.
@Christopher:
>I’m sure that site had “low traffic” like the Clintons were “broke” after the White House.
Except that I find it quite believable that the Clintons were broke after the White House. “Rich” and “broke” are not mutually exclusive if you have a ton of assets and two tons of debt. Given the way that Americans manage our personal finances, and expect our politicians to manage the country’s finances, we should not find it at all startling that our politicians (especially those on that side of the political divide) are unable to manage their own personal finances.
One source of data are the various language specific “package” sites like pypi (python), elpa (emacs) etc. A bot that crawls these sites and maintains a large database of packages along with history would offer a decent cross section of what’s going on. People who want their stuff to be used upload to these sites so that potential users can install and use their libraries using the standard channels.
The downside is that most of the stuff on these sites are “libraries” rather than complete applications and those things might be missed. I don’t see popular things like vlc, audacity ever coming up on such sites.
freecode.com (formerly freshmeat.net) moved to being a static site… and have completely f**ked up CSS styling – the site looks totally broken.
Recommending SourceForge… pffftt…
As to repo crawlers (which won’t catch personal-hosted projects) – there is Ohloh, but it has a different goal…
Honestly, I stopped looking at
FreshmeatFreecode years ago. The site kept getting more and more stale.These days, I use a combination of various things to keep abreast of changes: DistroWatch for the latest in Linux distros, the KDE, GNOME and XFCE sites for information about those desktops. If I’m looking for an application or library to fill a particular need, I will just search do ‘apt-cache search’ or ‘yum search,’ etc, for certain keywords. Also, the Arch Wiki is an endless fountain of information. I use Google Plus and have a big circle with many hackers and open source celebs (but I repeat myself) in it. And there’s always GitHub, Google Code, Ohloh, Berlios, Savannah, etc. Blogs.
Honestly, Freecode stopped being relevant a long time ago.
>And there’s always GitHub, Google Code, Ohloh, Berlios, Savannah, etc.
Berlios is dead.
Right. They died two years ago. Duh. Old habits die hard.
I loved Freshmeat, to be able to look round and see all the different software being introduced and worked on. Being able to see the cutting edge come together in interesting, unexpected ways. Freecode was a sign of it’s downfall, I should have seen it. Why would the people who used Freshmeat care how Politically Correct the name of site they used to find their software was? Oh well, I didn’t see it.
End of an era really. Good bye Freecode! I miss you Freshmeat.
Now I need a new morning routine, no more Slashdot, Freshmeat & good to go with perspective on what the rest of the free software world was up to.
Any suggestions?
How can I export my “following” list? I can’t login. I have saved nearly every interesting projects, even small ones.
This is really not good if this shutdown isn’t announced :-( :-( :-(
This is ridiculous. They could have sent out a notice to ask registered users and offered someone to take it over if it was to much hassle for them – i’m sure there’d be volunteers ready to step up.
Freshmeat/Freecode wasn’t something I used every day but it was damn useful when I needed it.
Travesty just to shut out down without warning.
ESR, some time ago you wrote that you were thinking about “fixing” software forge situation? Did anything came from it, or did the problem vanished in meantime?
About decline of software project directories (well, at least of Freecode / Freshmeat): it might be something similar to how web search went from hand-curated directories, with categories (e.g. DMOZ) to algorithm based automatic search (e.g. Google or DuckDuckGo) – the amount of stuff went way up.
>ESR, some time ago you wrote that you were thinking about “fixing” software forge situation? Did anything came from it, or did the problem vanished in meantime?
I’m still trying to clear my decks so I can do this – it’s a major project, at least 9 months of effort. Making slow progress towards it.
freecode began to die when they introduced their new Web 2.0 interface that replaced troves with tags — and in the process lost all of the discoverability that the old freshmeat site had.
To make matters worse, they blithely ignored the howls of protests from many of their users.
@CorkyAgain: Both free or almost-free tags (think StackOverflow sites, or Gmail) and directory / hierarchy like Trove classification (think Dewey decimal system, or DMOZ) have their place. Tags (folksonomy) catch new things that do not have good classification, and Trove gives completeness and searchability.
I don’t have a problem with tags per se. But what happened with freecode was that they deliberately discarded most of the metadata represented by the old trove system and never reached the same level of detail with tags.
The example I used in comments I made at the time was textmode programs. On the old freshmeat, it was a simple matter to browse through a list of programs written for the console or a terminal emulator, because there was a trove for that environment. There used to be hundreds of programs listed there. But when the site redesign was implemented, there were less than two dozen tagged as “ncurses” or “cli”.
“Those days are gone; we’re living in the era of huge distro repositories and the Ubuntu Software Center.”
Unfortunately the Ubuntu Software Center is no longer an option for Open Source non-mobile software AFAICT. Most of the info on publishing is now geared towards programs for the Ubuntu phones, and publishing traditional Ubuntu apps is only possible for commercial programs (i.e., if you charge a price for it). Trying to publish a free program just gets me the message that they’re working on a new system for free apps, but it’s been that way for over a year now.
Even when Ubuntu Software Center did work for free software, it took weeks to put a new version of an app on there, compared to mere hours for Freecode. Freecode also had the advantage that you didn’t have to upload your app, it was purely for announcements, leaving you to choose where to host your software.
As a developer, I’ve found Ubuntu Software Center to be a complete shambles for free software :/ (if anyone knows a better way to submit to it, please let me know!)
Also I found the exposure to be far less than what I got from Freecode, so I’m not convinced it’s what people are using instead anyway (at least, it’s not a good way that people are using to find new software).
Linux distributions have had respositories for years, but getting stuff on there still seems to be far from straightforward (e.g., on Debian you need an accepted Debian developer to “sponsor” your app).
Oddly, Freecode, despite its later name, was never strictly an Open Source site, it just happened to be mostly that due to its Linux heritage. In fact the only requirement they had was no Windows only software (supposedly because they didn’t have the resources to manage the much larger number of submissions they’d get). They’d gladly accept a closed source app for a closed sourced OS as long as it wasn’t Windows only.
Despite this, I found Freecode was good for publicising Windows software too, which you could add so long as it was cross-platform. Sites like download.com are spammed full of off-topic celebrity ads (as well as malware/spam “download” ads that mislead users as to which button to click), and although popular software gets vast amounts of downloads, I’ve found I got far less exposure than Freecode – either because download.com etc are places people go to to download something specific rather than just browse for something, or because they only advertise the stuff that’s already popular (download.com front page lists most popular downloads of the last week; Freecode front page had most recently updated apps). Speaking as a user, I don’t really have a good way to find new free software for PC platforms, other than just hearing about it, then Googling for the homepage or download link. Which I suppose isn’t necessarily bad – Google Play is so oversaturated that you’re basically at the mercy of Google’s search there – but it feels like there ought to be something more specific for software (or specifically Open Source software) than Googling the entire web.
Something that can automatically report latest releases of Open Source projects sounds a good idea (though ideally not limited to github).
> @CorkyAgain: Both free or almost-free tags (think StackOverflow sites, or Gmail) and directory / hierarchy like Trove classification (think Dewey decimal system, or DMOZ) have their place. Tags (folksonomy) catch new things that do not have good classification, and Trove gives completeness and searchability.
A better example is that scientific articles in physics have both PACS numbers (Physics and Astronomy Classification Scheme) hierarchy, and keywords – free-form with suggested keywords list.