News for hackers. Stuff that matters.
Re:Improving Firefox stability on Linux by jcastro on Wednesday May 19, 2021 @12:59PM (#27211336)
The blog post could be titled "Why the Linux distribution model is totally broken for end user software".

As a linux user I appreciate the work, but it's tough to read this blog post and not think about the wasted effort that could be used to fix the bugs in the upstream software itself.

Re:Improving Firefox stability on Linux by ohthehugemanate on Wednesday May 19, 2021 @01:11PM (#27211485)
Really? Because I just think about how impossible this would be on Windows, where shared or external libs are inscrutable, as are OS build chains, version differences, etc etc.
Re:Improving Firefox stability on Linux by gsvelto on Wednesday May 19, 2021 @01:19PM (#27211584)
We do gather symbol information on Windows too, but it's a lot less detailed than what we get from Linux. It also requires jumping through a lot of hoops. For example we get minimal information from Windows graphics drivers and we have to semi-manually scrape them ourselves.
Re:Improving Firefox stability on Linux by kevingadd on Wednesday May 19, 2021 @02:04PM (#27212205)
To get symbols for basically any MS library I just add the official MS symbol server to my symbol search list. Chrome has one too, there's no real reason other vendors couldn't do it. You can also just ship a .pdb file next to your binaries, and I think the only good reasons for not doing that are keeping download sizes small (use a symbol server...) or paranoid secrecy.

For example, NVIDIA's windows drivers are like 600mb at this point yet they don't even include function name symbols. It's inexcusable.

Re:Improving Firefox stability on Linux by phendrenad2 on Wednesday May 19, 2021 @02:09PM (#27212257)
Has Windows ever broken Firefox? Do you think it's more or less likely to happen in Windows (or Mac) or Linux?