... many new but also old package managers for programming languages directly rely on git repositories?
I know, server space isn't free but that seems downright retarded.
>>53632755
>but also old package managers
Name one
>>53632780
Cocoapod or what it is called.
The rest, like CRAN, CPAN, the Python one etc. are mostly shitty for other reasons, but that can almost be forgiven, since software development is 50% hindsight and it isn't a trivial task.
https://medium.com/@sdboyer/so-you-want-to-write-a-package-manager-4ae9c17d9527#.xe2q3uwp3
(Warning: shitty medium.com page)
>>53632755
>old package mangers
>rely on git
>old
you jackass, git isn't old so how can OLD STUFF RELY ON SOMETHING THAT DIDN'T EXIST
>>53632755
It's more reliable and verifiable than having someone else host binaries.
That said, they're not nearly as good as having packages verified by the distro's repos directly.
>>53632846
Ok, but ~ten years is not young either if you don't consider shitty mainframes etc. and you wouldn't call it medium aged.
>>53632870
>git
>hosting binaries
triggered
Also
>more reliable
It's not like repositories couldn't get renamed, transfered or simply completely disappear from sheer arbitraryness of the hoster.
And let's face it, github is just another repo hoster which at some point will disappear, just like Berlios, Google Code and soon Sourceforge.
>>53632938
>triggered
AUR downloads and compiles the source, when it used to download unsigned binaries.
If you are downloading from github, it's easier to verify than self hosting.
PPAs are garbage tho.
>>53633015
While I agree that PPAs are garbage, let's not get into what some meme bistro does.
Also, let's not get into bistro package managers at all, because the very idea that a bistro should and could provide most packages is flawed.