• 0 Posts
  • 23 Comments
Joined 1 year ago
cake
Cake day: July 23rd, 2023

help-circle




  • I don’t understand how they can say “overly sexualised voyages into space” and not include Galaxy of Terror. The worm rape scene that didn’t need to be in the movie and Corman forced back in makes it a much better contender for the slot Barbarella is in. On a script, camp, and exploitation level it is worse in every way and it’s even a ripoff of Alien to boot, arguably making it worse than Alien v Predator: Requiem which at least tries to be original.

    I own it on laserdisc. The only reason I’ve seen it more than once is because people are surprised Sid Haig and Robert Englund did a movie together. It’s fucking bad and holy shit is the worm rape scene the fucking peak of Corman being a misogynistic, exploitative walnut. To be totally frank it really disgusts me that he still gets a platform (legacy) after forcing this scene into the movie and directing it himself, even if some people think it contributed to the movie’s commercial success.






  • All of these packaging systems have plenty of tutorials. Speaking from experience, many maintainers were not developers when they started maintaining packages for distros other than the official distros. I have worked with several maintainers who do work in tech and know socially several who had no background. This could be a great place for you to start!

    You bother because FOSS is as much paying it forward as it is getting shit for free.



  • I grew up in a town that had one brewery until the late 00s/early 10s. The food was okay but the beer tasted like rubbing alcohol. It was really fucking bad. The brewers were full of themselves and thought they had first mover advantage. A couple of new breweries opened up and things got way better. Suddenly the community was actually vibrant. A few more opened, a few closed, and that cycle started rolling. The overall quality of brewing improved drastically except at that shitty first brewery who constantly struggles to retain talent and compete (other than location which guarantees a steady supply of drunk businessmen who can’t tell good beer from shit beer).

    Don’t limit your club. Stifling community will only harm the community. You don’t need to trade secrets; you should always trade ideas and be supportive.


  • This doesn’t appear to cover the cost of the electricity it would take to keep your stuff running. There is no way to pay anything out at all. Seems like a pretty straightforward pump-and-dump where the end users are collecting imaginary points while some company abuses their resources. Every blog and Reddit post I looked at to try to understand this was full of referral links. Equally classic sign of pump-and-dump pyramid scheme.




  • If a repo is very popular, it should have a lot of forks. The higher the upstream popularity, the higher the downstream popularity. When a dev makes a claim that there are a ton of malicious forks stealing IP, we can vet that claim by looking at the forks that respect the upstream. Big projects have a big community with big forks with many stars. The popular downstreams drive traffic to the upstream.

    In this case, we have a couple hundred direct forks. That’s not a ton. Out of those, only three have stars. All of them only have one star. At face value, that could imply a few things: the repo is not very popular, the community is centralized around the upstream, or something else along those lines. Comparing this to other open source projects, our initial conclusion is that this is not a hugely popular repo and does not get a lot of development outside of its incredibly niche community.

    Occam’s razor is a tool, not objective truth. Based on the facts as we can see them, this focus on forking from the dev is much more indicative of a burnout spiral, incredibly common in the FOSS community, than nefarious actors. If we see receipts, eg a collection of takedown requests on malicious forks attempting to claim ownership of the code, our analysis falls apart. That’s still a possibility, however remote.


  • There were forks that wanted to hide the fact that they were Floorp forks, forks that did not want to contribute to Floorp at all, forks that used the code for life and just changed the name of Floorp, and many other forks were born.

    There are three visible forks that have any stars. All of them have one star. You’re telling me that a project that is so widely and maliciously repackaged has no normal forks with more than one star? Is this tech that only bad actors want to use and has no following in the open source community?

    Where are these evil forks, how do we actually know they’re forks, and why are they still up if they’re breaking license?




  • My stance has been that, just as long as I’m interviewing with someone, I’m happy to do it, up to an undetermined time threshold. A screening interview, a tech screen, and then a bunch of panels is what I expect from a solid firm. Just as long as I’m interviewing with someone, I have a lot of opportunities to learn myself. I will also occasionally do a take home if and only if there’s a novel problem I want to solve related to that take home (eg I want to learn a library related to the task) but this is very rare.

    As a hiring manager, I try to keep things to a hiring screen, a tech screen, a team interview, and a culture interview. My team is small. I don’t want to spend more than three hours of someone’s time (partially because I can’t really afford to spend more than that myself per candidate or lose more team hours than that). My tech screens are related to the things I actually need people to do, not random problems you’ll never see.

    My assumption is that a good dev has lots of opportunity and I am in competition with everywhere else. I need to present the best possible candidate experience. Big companies with shitty employee experience telegraph that by presenting a shitty candidate experience, which is where the employee experience begins. You can’t have a good customer focus without starting from a good employee focus.