3 ms·
The thing about just pulling the number of packages on Hackage is a large number of them are actually applications. I am lazy so the first section that proves
by stonemetal 16y ago
The thing about just pulling the number of packages on Hackage is a large number of them are actually applications. I am lazy so the first section that proves my point is AI 9 entries. 3 libraries, 3 programs, 3 that list themselves as both program and lib. Other sections are similarly app full. Depending on how well separated the app and lib are in the app+lib category you could be looking at more than half of that package count being applications not libs.
- dons 16y agoYour point is well taken. Here's a quick script. Libs that also ship programs are counted twice, but we get: 2139 libraries 455 programs So, no, its more like 25% programs, 75% libs.
- stonemetal 16y agoI still don't think we are in quite the situation you imagine. As a for instance check out distributed computing in hackage 18 entries, great right? Number of actual projects 4, 13 entries are for Happs. Opengl has a separate entry for its test cases. There is a large number of unimportant random junk packages in hackage. The main problem with hackage is that I need to figure out which of the 30 GTK packages I need before I can use GTK.