3 ms·

Great to hear someone other than myself is interested in an all software site. I started working on a site for that in December after getting fed up with the n

Great to hear someone other than myself is interested in an all software site. I started working on a site for that in December after getting fed up with the number of times I've looked for a program to do X, not finding it, starting to build it myself, only to stumble across a random HN comment two months later that links to a program that solves the problem better than I ever could. I'd really like a resource that lets me find this stuff better, so I'm making it. Of course, if the usual pattern repeats again — sometime in March I'll come across such a site that has already done this and been running for 5 years and we never knew

stonogo 3y ago
This used to be freshmeat.net, which rebranded to freecode.org and immediately died. It was an heirarchical directory of software, classified by category, user interface, UI toolkit, license, implementation language, operating system support... I often wonder what happened to that database. It would be a great thing to revive. "I want an ncurses text editor that runs on linux and has a permissive license" or "I want a QT-based email client that runs on Windows and is GPL" or "What Mac OS X software is using GTK these days", all extremely quick and easy to look up. edit: http://freshcode.club/ http://freshcode.club/ is basically the freecode version of the site, revived. Unfortunately, one of the changes late in freshmeat's existence was a removal of the formal categorization heirarchy and a transition to arbitrary 'tagging', which led to almost-identical tags, confusingly conflicting tags, and generally made life slightly harder.
- wild_egg 3y ago
  Thank you for this! I think the information architecture is going to easily be the hardest part of this project. This really gives me some food for thought ... Or I guess I could throw some LLM RAG nonsense on it and see what happens
  - stonogo 3y ago
    The original site (as with sourceforge) was based on "Trove", which I can barely find mention of these days: https://tuxedo.org/esr/trove/ https://tuxedo.org/esr/trove/ the post talking about how it was updated for python 1.5 should be taken as an indication that a new start might be needed!

stonogo 3y ago

This used to be freshmeat.net, which rebranded to freecode.org and immediately died. It was an heirarchical directory of software, classified by category, user interface, UI toolkit, license, implementation language, operating system support... I often wonder what happened to that database. It would be a great thing to revive. "I want an ncurses text editor that runs on linux and has a permissive license" or "I want a QT-based email client that runs on Windows and is GPL" or "What Mac OS X software is using GTK these days", all extremely quick and easy to look up. edit: http://freshcode.club/ http://freshcode.club/ is basically the freecode version of the site, revived. Unfortunately, one of the changes late in freshmeat's existence was a removal of the formal categorization heirarchy and a transition to arbitrary 'tagging', which led to almost-identical tags, confusingly conflicting tags, and generally made life slightly harder.

wild_egg 3y ago
Thank you for this! I think the information architecture is going to easily be the hardest part of this project. This really gives me some food for thought ... Or I guess I could throw some LLM RAG nonsense on it and see what happens
- stonogo 3y ago
  The original site (as with sourceforge) was based on "Trove", which I can barely find mention of these days: https://tuxedo.org/esr/trove/ https://tuxedo.org/esr/trove/ the post talking about how it was updated for python 1.5 should be taken as an indication that a new start might be needed!