3 ms·
So this limit is technically per repo, right? I've created a small organization for myself to store mirrored government sites. None of them (so far) take up mor
by danso 10y ago
So this limit is technically per repo, right? I've created a small organization for myself to store mirrored government sites. None of them (so far) take up more than a gig, but I'll probably refrain from pushing giant mirrors on to Github. It's not that it isn't trivial to push these mirrors on S3, it's just that Github pages provides so many user-friendly endpoints.
A researcher recently asked me for help in mirroring one of the CMS.gov subdomains. Of course, by default, Github Pages serves up the mirror as if it were the real site:
https://wgetsnaps.github.io/marketplace.cms.gov/ https://wgetsnaps.github.io/marketplace.cms.gov/
But the researcher was obviously more interested in the documents published on the site than the site itself. So I told her to just checkout the subdirectories which are easy to nav via Github.com's standard repo listing: https://github.com/wgetsnaps/marketplace.cms.gov https://github.com/wgetsnaps/marketplace.cms.gov
But even better, I told her she could just download it as an archived zip, another endpoint that Github provides automatically, and peruse the file tree on her own operating system: https://github.com/wgetsnaps/marketplace.cms.gov/archive/master.zip https://github.com/wgetsnaps/marketplace.cms.gov/archive/mas...
And of course, there's the option to git clone it. These are all features that can be reproduced, but I use Github not to save money on AWS, but just because of how discoverable the repos are.