R.O.S.I.E 5448
@rosie@0x4d4f5448.systems
Location: 0,0
it's incredible people still recommend delta chat while it's not capable of adequate security (forward secrecy, completely forbid unencrypted messages, no metadata leakage, right now who sent and to whom is leaked, that's already way too fucking much)
until that's solved, i can't consider delta chat to be secure
Do you remember how to do long multiplication AND long division using pen and paper
| Yes: | 62 |
| No: | 28 |
If you never leaned one or both of these pick no
I personally had to teach myself long division because they never taught it to me in school
@CenTdemeern1@eepy.moe they tried teaching me but i never understood either of them and was never able to do division at all lol
ruby looks so good in theory if only it was actually readable
i feel like kotlin really is the best of all worlds. reasonably concise syntax. non-godawful runtime. i should use it more often
Growing increasingly frustrated as I attempt in vain to explain how one can modify the browser chrome of a browser that isn’t Chrome to look like chrome. That is, to change the chrome of a non-Chrome browser to look like chrome, without making it look like Chrome.
the mischievous and devious voice in my head is telling me to work on an electron project of some kind
whyyyyy do so many people still wanna do GPG shit? Someone just said "please don't retire gnupg v1.4 from fedora, I still need it" WHYYYYY
@gwenthefops yea i really need to switch to ssh signing for git
@gwenthefops ok done that was surprisingly painless
@SRAZKVT@tech.lgbt yay! git shouldn't be painful, only dates should
@gwenthefops the fact author name and email addresses are a part of commits is the part of git i dislike the most
and i love git
@gwenthefops like to me that's just a lack of forward thinking and a fundamental design flaw, if people change mail provider or name, then the old email address might not be valid anymore, and at this point, the value of it being present at all within the commit object is basically 0
@SRAZKVT@tech.lgbt yeah no, I totally agree with you there
@gwenthefops yea
i love git but that part is a mistake
how to resolve it ? idk, but just because i don't know how to fix a problem doesn't mean it doesn't exist
@SRAZKVT@tech.lgbt you could have identity files in the .git, and the author/email could be referencing that file. I just don't know how you'd avoid someone defacing that
@gwenthefops ssh signatures ? maybe have file be commit signature fingerprint, and signed by that same fingerprint
problem is a lot of devs use different ssh keys on different devices. in some cases, fair enough, email may change (for example, work and personal computers), but for several personal computers, yea, it's a bit pointless and updating them on every repo you might have contributed to is going to be a chore
speaking of, how would you update it ? how would it even be given to clients ? objects are immutable, so can't use the usual object pack system, so something custom for identity instead ?
@SRAZKVT@tech.lgbt yeah, those are definitely hard problems...
@gwenthefops i like how git is structured, but i don't think an immutable object system is the correct way to go because of this
i do think it is reasonable enough for the commits themselves, file data, trees, signed tags, etc, as those shouldn't be changed ever, but for things that do change it just doesn't work
i guess mailmap works ? it would kinda require having a mailmap file in the repo though, and it'd be clunky
do ipv6 mfs just go like "oh just join my server, it's [646b:4883:a577:aff5:8185:c42d:9aff:2890]:25565"
History