Docs.snap.berkeley.edu is a still-far-from-release online version of the Reference Manual, which we hope will be easier to update.
Because of a security failure in our hosting of this site on Github Pages, for a few days the link was taken over by so-far-unknown bad guys, and for part of that time it was showing pornography.
We apologize sincerely for this having happened, and for how long it took for us to fix it.
None of your personal information is stored on that server. So you don't have to worry about identity theft as you would if this happened at your bank.
People shouldn’t be able to take over other people’s sites! Maybe it’s something we did wrong in setting it up, but yes, we should tell Github about it.
In ~10 years when I retire from getting paid to code, one of the things I want to do is pitch in with Snap. It sounds like a relaxing, fulfilling project to me to convert the existing documentation to LaTeX (with a framework to autogenerate html from the same source on every update)
I wrote four books in TeX (not LaTeX; I needed better control over formatting than it offers), so I should have known better, but back then it was really painful to insert pictures in the text and control the size and position in TeX, so for the manual I wimped out and used a WYSIWYG editor.
Long ago now, someone actually did translate the manual to TeX, but I dropped the ball on using it.
When we started the recent conversion I proposed to use TeXInfo, so we could get both beautiful PDF to print and online HTML, but the young people who are doing the actual work are using young people tools, namely Jupyter Notebook.
If GitHub is hosting the subsubdomain, does this mean they discovered a vulnerability that could be potentially used on other pages, and thus should be reported?
All the information on custom github pages domains are here
I'm not entirely sure how it works, but it does talk about verifying you domain to avoid takeover attacks, which id what happened with docs.snap.berkeley.edu.
It’s been reported, but this seems to be a long-standing GitHub pages issue… I am not sure how it was possible for someone to re-register the subdomain, but I have reported the repo which is taking over hundreds of domains. I’ve also enabled their DNS verification which should prevent this from happening. (And, argh, GitHub is frustrating here because the have multiple types of DNS verification which apparently don’t all offer the same protections…)
Sadly, this seems to be a huge semi-known problem they haven’t fixed. Though, it’s not 100% clear to me why someone was able to ‘reclaim’ a domain that was supposedly in use…
how is it not in LaTeX?
