Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)DA
Posts
0
Comments
130
Joined
2 yr. ago

  • PostgreSQL is definitely a boost to performance, especially if you offload the DB to a dedicated server (depending on load, can even be a cluster)

    Nevertheless, it probably has much to do with how it's deployed and how many proxies are in front of it, and/or VPN. If you have large numbers of containers and small CPU/low memory hardware, and either running everything on one machine or have some other limitations, it'll be slow.

    Admittedly, I'm not very familiar with the codebase, but I feel Apache isn't improving the speed either. Not exactly sure how PHP is nowadays with concurrency and async, but generally a microservice type architecture is nice because you can add more workers/instances wherever a bottleneck emerges.

  • Beware of the W̷̞̬̍̌͘͜ĭ̴̬̹̟͕̒̆̈́n̸̢̧̙̈́̅̂̆̕͜ͅd̵̟̟̪͎̀̀ő̴̼̺̺́̐̂͘w̵̨͊̀s̵̡͎̭̊ ̸͔̬͔̜̊́̈́̌̈́ͅŬ̴͉͚̳̌̉͘͝p̸̼̅̆͐̃̑d̸̜͂ǎ̵̛̯̏͝ť̷̰é̸͇͝ as it can screw up/overwrite your other bootloader completely.

    Kinda sucks, when you've got a meeting/work and you find out that forced update made your system unbootable/partially unbootable and you now get to live boot in and go fixing the EFI partition manually, in the CLI.

    That happened to me once and that's when I decided feature parity was less important than a reliable system that "just works" for getting things done on a schedule. (I removed windows completely, in case that wasn't clear)

    Anyhow, make sure you install windows to a separate drive that can't see any others during the windows install, then will keep the bootloader separate.

  • I tend to agree.

    Nevertheless, some unknown implementation can have bugs and things can go wrong and there's nothing you can do about it, short of "rebooting" or d̷o̶w̸s̸i̷n̴g̸ ̴t̶h̸e̷ ̸h̵a̵r̵d̷ ̵d̷r̶i̴v̶e̷ ̵w̶i̴t̸h̷ ̸̞̺͠h̵̺͙̎̍o̸͔͠ͅḻ̷̀̇y̵͚͍̎ ̷͉̅̅w̸͎̔a̷̧̫̒́t̶̼̉̓ę̵̾͗r̶̫͑͑ ̴̣̿͒(̷͙̎a̸̬̺͝͝n̸̞̓̓d̴̬͌̍ ̸͇͕͌͝s̷̡̯̓͝u̸̡̳̇͝b̴̳͜͠s̷͍̘̽ë̵̜q̷̝͐̄ȕ̵̞̐e̷̲̠̐́ń̴̨̙͝t̸̛̬͝l̶̮̔͠y̴͕̪̑͝ ̵̖̆ḃ̴̪̟u̶̢͓͑̌y̵̜̤͌̏i̵̦̋ň̴̨͚̀g̸͓͑ ̴͍̬̽à̶͜ ̴͇͔̓n̴̬͂͜ì̷̢̛̯c̴̤̖̈́e̶̼̫̐̊ ̵̹̏͝f̸̙̀̑r̷̪̩͆͆e̸̤̫͛͋s̷̢̙̏h̷͇͔́ ̸̭̆͝N̷̰͗͛͜V̶͇͒̚M̸̟̍͜ě̷̛̟ ̸̢̞́͝a̷͙͔͒͒n̷̻͇͝d̸̘̥͌̾ ̴̜͓͑p̷̬͑͊ŭ̸̮̏t̸̲̀t̴̡͚̽í̶͎͓̑n̴͕̘̒̈́g̴͓̰̓͝ ̵͓̎a̴̻̼͗ ̷̦̍̈́s̷̥̅̈l̴̝̂e̴̞̅͊ḛ̴̊̅k̷͚̕ ̵̛̼̬͗D̴̻̾̽e̵̙͂̊b̷̝͘ī̵̢͇ą̵̂n̴͖̑ ̶̼̚h̴̼͂͑e̷̲͆̆a̵̡̋d̸̢͔̈l̶͕̍̍e̸̛͕̙̒s̶̞͔̀͠s̸̯͖̕ ̵͍̦̈́̉ ̸̨̨̓i̸̙͖͗̌ņ̶̯̍s̸̡̖͗̇ṯ̷́̒ä̵̦́̎l̶̼̄l̵̨͊̊ ̴̳͑͗ó̵͎̅ǹ̴͈̚ ̷͖͊͝i̷̠͇̊t̷̼̞͒͘)̵͎̤̔͌

  • I had a .bin where the change hadn't been implemented and one after. Was using file carving tools as I was just trying to figure out what was going on. Probably spent too much time, but once I found out what it was doing, I was pretty mad. I tried to just corrupt the firmware to force a fallback to a "safe" state, but eventually I had to look into reverse engineering the binary that seemed to be controlling different things, such as the genuine ink check and things like that. Many hours of trial and error staring at the xxd and gdb output, semi randomly breaking things, until I seemed to break the right thing. Was bit tricky to get around the firmware signing, but eventually got that worked out too by tricking it into not checking (very old firmwares for that printer weren't signed) and accepting the 'new' firmware, with a much higher version number, as that's also one thing it checks to prevent downgrading.

    Tools used as far as I remember were

    • binwalk,foremost, autopsy
    • radare2
    • This page I believe was helpful, at least i had it bookmarked
  • Especially when printers ask you to waive your class action rights just like this.

    Makes sense, when they illegally push straight up malware that sets the ink flow rate to 0 should non-geuine ink cartridges be detected. This will destroy/clog the print head if attempting to print for too long.

    Yes, I reverse engineered your fucked up Linux 2.4 (!) based firmware, Epson. Your printer is printing nicely offline with refillable cheap ink. Fuck you, I won.

    Sorry about the rant, it had become personal at one point.

  • Virtual high five! Keep it up, I'm always happy to see someone find joy and a sense of accomplishment in programming things.

    Check out polymorphism and concurrency when you feel you're solid enough on the other things. Concurrency/threading can be a bit weird to wrap your head around, but essential to build powerful things. (See Reader-/Writer problem to start)

  • But like, shouldn't she be taking it up with the FAA then? I'm sure you can apply for delay or exemption in extraordinary situations like this one undoubtedly is?

    That is the source of the data after all. If you don't actually stop the data source from publishing your data, others will continue to use it.

    It's a dumb approach that makes it look like a hurt ego thing rather than a legitimate concern.

  • Yeah, I've come across one or two papers of questionable origin presenting performance metrics nowhere near what I could replicate.

    Needless to say, I could only replicate the results when I introduced a very mundane procedure error nobody reputable would make. So yeah, lots of garbage out there

  • But even then, there were issues plaguing Quora that would continue to fester. First, an anonymous former Quoran told me, the site started “shortening the length of questions.” The professed reason was to increase Quora’s visibility on Google, but that brevity came with a cost: It also made it difficult for users to ask the types of complex questions that could be addressed by specialists

    Ah, I see they started the enshittification very early. It might've been a good LLM database, but the good quality content would be outdated by now and the more recent is infested with troll and bot garbage and AI writing. Sad.