<cole-h>
Maybe because we have emails publicly listed on our profiles :P
<andi->
samueldr: I also have highlights on my github handle ;)
<andi->
and twitter
<samueldr>
now that's on you :)
<samueldr>
I made due diligence not to unjustly ping you :)
<andi->
that is fine
<samueldr>
yes I know
<andi->
There was obvious address on the website of sendgrid.com so I just used abuse@ from the whois record... I somehow have little hopes anyone reads that
<srk>
cole-h: emails are not public if you look someones profile when not logged in
<srk>
git ones are tho :)
<cole-h>
Right. But how hard is it really to login on GH to then scrape emails?
<srk>
but only after you clone the repo
<srk>
proly hard due to rate limiting
<srk>
they sure can detect obvious scraping attempts
<infinisil>
cole-h: Isn't that effectively still public as anybody can create a github account?
<andi->
not if you are signed in then your ratelimiting is probably okay. And maybe you can find some requests that yield many addresses
<infinisil>
Oh, yeah they can detect scraping with that
<srk>
or just clone nixpkgs
<andi->
I never thought it would be private. I've used that a few times myself to send someone an email.
<cole-h>
infinisil: That's my point -- emails are public if signed in, and anybody can get "signed in" by creating an account
* infinisil
nods
<cole-h>
andi-: Just this week I had somebody send me an email using the one on my profile
<cole-h>
I like it
<andi->
Yeah, I wish it would be *more* accessible from the interface.
<andi->
Sometimes there are things that do not fit issues or PRs
<samueldr>
no need to login to github, clone the repo
<samueldr>
oh, srk said that already
<cole-h>
But then you have to clone the repo
<samueldr>
trivial to scrape probably-good e-mails
<infinisil>
If you have nix already, just use lib.maintainers :)
<srk>
cole-h: but you don't need to do captcha / verified email / rate limits :)
<infinisil>
> lib.maintainers.infinisil.email
<{^_^}>
"contact@infinisil.com"
<cole-h>
> lib.maintainers.cole-h.email
<{^_^}>
"cole.e.helbling@outlook.com"
<cole-h>
srk: Fair
<hexa->
> lib.maintainers.mweinelt.email
<{^_^}>
attribute 'mweinelt' missing, at (string):292:1
<hexa->
> lib.maintainers.hex.email
<{^_^}>
attribute 'hex' missing, at (string):292:1
<hexa->
> lib.maintainers.hexa.email
<{^_^}>
attribute 'email' missing, at (string):292:1
<hexa->
sorry, derped
<hexa->
very well
<cole-h>
Derped twice it seems
<cole-h>
:P
<hexa->
master derper
<cole-h>
Keep adding characters until you get there, Ship of Theseus style
<infinisil>
hexa-: Hehe, you're the only person in #82461 I had to add the email for because it was missing :P
<infinisil>
(I used the one from github, I hope that's fine)
<cole-h>
Totally unrelated, but I can't wait until SoloKeys rev2 is released, so I can finally get a security key
<infinisil>
Wait
<infinisil>
There's not one on github, I think I used the one from git or so
<andi->
I tried to open git blame on our maintainers list... GitHub is bailing out big time...
<cole-h>
GitHub is just really bad at handling large files in general
<andi->
s/large files//
<cole-h>
^
<MichaelRaskin>
GitHub is completely failing at handling large .* in UI
<MichaelRaskin>
Large discussions, anyone
<andi->
Discussions on GitHub.
<colemickens>
cole-h: whoa, hadn't seen that update post, that will be an instant buy
<cole-h>
colemickens: Right?!
<andi->
Did anyone actually ever use that teams thing that was made to discuss at?
<cole-h>
colemickens: I'm super hyped to get one and start hacking on it
<hexa->
infinisil: yup, is fine. might change it soon however :)
<cole-h>
Since it's OSS, I can just buy the hacker one, change it to accept fw signed by them OR by me, and then lock it back up. Still be able to hack on it and update to their stuff if I need to
<samueldr>
there's another thing with github, kinda duplicating unfinished features
<andi->
I have only seen them use in one project I am in... and even there they are mostly "dead"
<samueldr>
they're not good imo since it's *only* team members
<samueldr>
probably only good for enterprise users
<andi->
Yep
<samueldr>
I was thinking about those new "like issues, but for asking questions" discussions
<andi->
Yeah, I have seen them. Didn't expect someone would confuse them with taht.
<andi->
I just wish they'd focus on creating a proper git & pr service and not to become just a SPOF for everything.
<srk>
:D this
<cole-h>
You mean you DON'T want notifications that take up the entire page and you have to traverse multiple inches to glean the important information?
<samueldr>
stop maximising your windows
<andi->
I don't open the notifications anymore. I just have a few notmuch queries that do the same.
<cole-h>
samueldr: >:(
<samueldr>
(stop prescribing workflows to users, dummy me)
<cole-h>
I have my Firefox windows maximized and tabbed because I like to actually read stuff (since most websites like to take up the entire page now)
<samueldr>
though it's not an issue, you have white space either left and right, or in the middle, it's choice you have to work with when designing
<samueldr>
doing it like they do in the new UI is better for longer content
<samueldr>
otherwise if you cram everything in a statically-sized column, well, there's no way to expand there
<cole-h>
Just let me choose to narrow it further than it is 😭
<andi->
It probably also depends on the "compatibility" you want. I could imagine the format not being that bad if you unzip, discard most of the tags and just show the text that (hopefully) is left at that point.
<worldofpeace>
cole-h wahh, I was about to tell infinisil that he needs an occlusive too if the skin actually breaks. the emollient won't really do much then
<samueldr>
if you're into cloud services, ms office online and google docs may be other alternatives
<cole-h>
worldofpeace: Sorry :'(
<worldofpeace>
cole-h: 🤣 I'm only ever half serious
<cole-h>
I know :P I only bust out the `:'(` in those semi-serious situations
<infinisil>
It's kind of amazing how our bodies can just fix themselves in many cases. Just wait a couple days/weeks and most wounds disappear on their own.
<infinisil>
Like, these are atoms, molecules, DNA, all that kind of jazz working together to reconstruct skin and such automatically
<samueldr>
infinisil: do you know what scurvy does to those?
<infinisil>
How crazy is that!
<infinisil>
samueldr: Probably no good!
<samueldr>
[citation needed], but from memory scurvy may make old wounds open!
<worldofpeace>
how scary. I wonder how many iu of vitamin c to prevent scurvy?...
<samueldr>
not much iirc
<samueldr>
like it's hard to not consume vitamin C
<worldofpeace>
meanwhile people get vitamin c injectable ascorbic acid
* infinisil
takes a vitamin tablet every day
<samueldr>
I believe jury's still out on the effectiveness, the main issue may be it doesn't have what's needed with the tablet to make it bioavailable (is it the right term?) so that your body can use it
<samueldr>
so it may make it through without doing much :/
<infinisil>
"I believe jury's still out on the effectiveness" <- I have no idea what this is supposed to mean
ZaraChimera has joined #nixos-chat
<worldofpeace>
I think that means it is still to be determined
<samueldr>
yep
<samueldr>
the main thing to remember is to not use that as a replacement for properly eating
<worldofpeace>
But right, anything injected would be 100& bioavailable, while some sort of tablet is less
<samueldr>
not sure about injected, it may end up being "cleaned up" anyway, but I'm not any kind of scientitious person able to tell
<samueldr>
though I figure if the medical establishment uses them, they must be good enough
<infinisil>
Now that's a PR title: #83556
<{^_^}>
https://github.com/NixOS/nixpkgs/pull/83556 (by bhipple, 1 hour ago, closed): <!-- Nixpkgs has a lot of new incoming Pull Requests, but not enough people to review this constant stream. Even if you aren't a committer, we would appreciate reviews of other PRs, especially simple ones like package updates. Just testing the relevant package/service and leaving a comment saying what you tested, how you tested it and whether it worked would be great. L
<cole-h>
infinisil: Please explain why aszlig has -666 karma thanks
<cole-h>
(sorry for ping as*lig)
<samueldr>
some things simply are a mystery
<infinisil>
cole-h: aszlig++ is possessed by a demon
<{^_^}>
aszlig's karma got increased to -666
<infinisil>
Or maybe rather, a demon is possessed by aszlig lol
<aszlig>
infinisil: mhm, gets tedious at some point, maybe i should automate it :-D
<cole-h>
aszlig++
<{^_^}>
aszlig's karma got increased to -666
<cole-h>
😱
<aszlig>
^ that's why
<cole-h>
Maybe that's how they have the power to write those fantabulous commit messages...
* cole-h
G**gles how to get possessed by a demon
<aszlig>
cole-h: well, much simpler than that: if you do ++ on yourself, you get a negative score =)
<samueldr>
(and here I was simply trying to put a good commit message in the spotlight and forgot about that prank/trick :))
<cole-h>
aszlig: I'm too vain to do that. I like seeing my internet points go up whenever I do something especially helpful
<aszlig>
samueldr: mhm, a decade ago i was bugging people to write more exhaustive commit messages but gave up at some point, so is this a thing nowadays?
<aszlig>
("gave up" in terms of "being pesking in code reviews")
<infinisil>
(btw, the bot stores a list of karma entries, both ups and downs, so the number of ups only can be determined still)
<samueldr>
hm? I think I agree with you, commit messages should hold the context that cannot be inferred from the change
<aszlig>
s/pesking/pesky/
<samueldr>
and it would be better if everyone applied, maybe not as much, but at least some, effort in them
<cole-h>
Man, in the past few weeks I've wanted to learn both Haskell and Zig, but then when I have the time to learn, I don't have the motivation...
<cole-h>
:(
<samueldr>
cole-h: stop lurking on IRC
<andi->
i feel like with the big PR templates there aren't many people putting effort in commit messages. It is a shame :/
<andi->
Most of the stuff that is written there should be in the commit(s)
<aszlig>
samueldr: well, at least the reason _why_ the change was done would be quite helpful in most cases =)
<samueldr>
yep
<aszlig>
especially when it comes to reverts
<andi->
One feedback point I got on having proper commit messages was that users were unable to amend their commits and would "usually just squash all of them" (aka using some string defined in the web ui…)
<samueldr>
ugh, squashing as merge makes me shudder
<andi->
I sometimes ask myself what the barrier of contribution should be and where things went wrong that we ended up here :/
<samueldr>
different use cases for the tooling I think
<samueldr>
cole-h: I wasn't telling you off btw, just saying that's one way you might be getting distracted :)
<samueldr>
cole-h: and close that reddit tab
<cole-h>
I know :D It's definitely why I'm distracted
<cole-h>
tbh I'm on IRC more than reddit nowadays
<aszlig>
andi-: "ended up here"?
* cole-h
immediately opens a tab of HN
<aszlig>
you mean the squashing?
<andi->
aszlig: no the average contributor not being able to use git or not even thinking about writing a proper commit message
<samueldr>
I think it's different usage patterns in the tool
<aszlig>
andi-: i think that's not something related to git, saw this in svn, cvs, monotone, whatnot...
<samueldr>
they use it as a kind of "ftp upload to new folder with history"
<samueldr>
and (I at least) we use it as a way to manage changes
<aszlig>
people usually don't think it would make a difference if there would be a message
<andi->
aszlig: It is nothing related to Git. I have just seen it increase with the GitHub adoption. Might not be correlated
<samueldr>
correlation comes only from more adoption I think
<andi->
I spent my days in vim and very often looking at `git blame` just to figure out why a change was amde.
<andi->
That makes me aware of the issue
<samueldr>
in compsci classes here, from a recent student, the basic git workflow they learn is git add everything and git commit
<andi->
"Doh" or "Fixed issue" aren't great messages if you look at stuff that was done :/
<samueldr>
nothing more, no good usage pattern, no fixing commits, not even something as simple as "make small units of changes"
<aszlig>
andi-: mhm, i think github also makes it a lot easier to not provide a message
<aszlig>
that's why i don't use the "merge" button in the webui but rather merge manually
<andi->
aszlig: Yeah, it hides the commits pretty well. You can get around just reviweing the PR message + changes, nobody must see the commits anymore :/
<andi->
I would really like a per commit review process on GH.
<aszlig>
per commit works as well, it's just not the "default" workflow
<aszlig>
but yeah, reviewing the commits one by one actually gets you into the perspective on why and how the change came to be
<aszlig>
and it's especially useful for large changes
<aszlig>
s/$/ets/
<aszlig>
however annoying people to provide better commit messages always has been a fight on the losing end, because after all most people didn't care and maintainers also didn't do it
<aszlig>
so essentially if we want better commit messages, i think everyone needs to do it so it becomes the "norm"
<andi->
I sometimes have told people that I wont review stuff until the commit message is useful once GitHub ceased to exist. That worked... but I only ever do that if a) I know the person well enough and b) the current commit messages aren't helpful at all..
<andi->
aszlig: yeah, I think that is kinda true... I just hope everyone sees the benefit in that.
<aszlig>
the problem is that the benefit manifests with time and so most people would be in "smoker's mode", eg. "ah, well it's not hurting right now" :-D
<aszlig>
difference is just that writing terse commit messages doesn't kill you :-D
<aszlig>
just a few weeks ago i had to dig up a commit message which said something like "update" and when asking the author about the change they didn't remember anymore
<aszlig>
and the change had a few details didn't really make any sense at the time i was looking at it but after refactoring i found why some of those "weird things" (don't remember actually myself anymore) were made
<cole-h>
Like when people say that they can't use `fetchFromGitHub` because "there's a difference between the GitHub release tarball and the tagged tarball" but they don't say what those differences are...
<aszlig>
cole-h: ah, i'd probably would even directly write that as a comment
<cole-h>
There was a comment. But it only said that there was a difference, not what it was
<cole-h>
Luckily, I'm super duper smart and figured it out myself and will be noting that difference in the next release of the package (whenever that may be)
andi- has quit [Remote host closed the connection]
andi- has joined #nixos-chat
waleee-cl has quit [Quit: Connection closed for inactivity]
<cole-h>
In the lambda `λx.x`, do you pronounce it as "lambda x dot x"? Or is there another way to pronounce it?
ZaraChimera has quit [Read error: Connection reset by peer]
<samueldr>
isn't that a funny face with sideburns and funny hair?
ZaraChimera has joined #nixos-chat
ZaraChimera has quit [Read error: Connection reset by peer]
ZaraChimera has joined #nixos-chat
drakonis has quit [Quit: WeeChat 2.7.1]
<cole-h>
I can't unsee that now
<cole-h>
What have you done to em
<cole-h>
s/em/me
lovesegfault has quit [Quit: WeeChat 2.7.1]
tilpner has quit [Quit: tilpner]
cole-h has quit [Quit: Goodbye]
tilpner has joined #nixos-chat
ZaraChimera has quit [Read error: Connection reset by peer]
ZaraChimera has joined #nixos-chat
ZaraChimera has quit [Read error: Connection reset by peer]
ZaraChimera has joined #nixos-chat
<MichaelRaskin>
I When something truly feels like weird magic, I ask for a comment that would make sense if the humanity finally got rid of Git but forgot to convert histories and continued from the latest snapshots.
ZaraChimera has quit [Ping timeout: 258 seconds]
ZaraChimera has joined #nixos-chat
ZaraChimera has quit [Quit: Quit]
<MichaelRaskin>
(reads the complaint at Discourse) Heh, «predictable» device names. I wonder what fraction of systems actually has _more_ predictable interface names, not less, with this naming scheme. Apparently single-if VMs are not in, and single-if-per-type consumer hardware is also out, so the fraction should be quite low
Jackneill has joined #nixos-chat
__monty__ has joined #nixos-chat
<aanderse>
MichaelRaskin: yeah i had to turn them off on my vm network. it was a disaster early on.
<eyJhb>
__monty__: the newer keyboards aren't that great anyways
<__monty__>
Yeah but mine's a T400.
<eyJhb>
Makes sense then. I wouldn't have done it if I didn't have backups
<eyJhb>
But it feels really nice now! Little in-between if I should have left the function keys alone
<gchristensen>
"When you have configured Vault with static credentials, you can use this endpoint to have Vault rotate the access key it used. Once this method is called, Vault will now be the only entity that knows the AWS secret key is used to access AWS." <- this is cool
<__monty__>
gchristensen: Isn't that all an AWS secret key can be used for though?
<gchristensen>
hm?
<__monty__>
"knows the AWS secret key is used to access AWS."
<__monty__>
Isn't that all you can use an AWS secret key for?
<gchristensen>
yeah, but the cool part is the "the only entity"
<__monty__>
But it doesn't say "that knows the AWS secret key that is used". So anyone who knew the key beforehand still knows all you're using it for is AWS, no?
<__monty__>
I guess I simply don't really get it.
<gchristensen>
it isn't about knowing you're using AWS, it is about knowing the extremely powerful key that Vault has
<__monty__>
So it uses the key to change the key?
<__monty__>
Yeah, the phrasing confused me.
<gchristensen>
exactly
<gchristensen>
and the new key cannot be extracted from Vault without compromising it
<joepie91>
that feels like a continuity risk
<gchristensen>
eh?
<joepie91>
nevermind, after I said it I realized I don't have the energy for discussion today :p
<__monty__>
joepie91: Computers never break. : <
<gchristensen>
it doesn't lock the admins out, so in a crisis admins can go create a new key
<gchristensen>
there is no reason for any person to know that specific key :)
neeasade has quit [Ping timeout: 240 seconds]
* gchristensen
does the first deploy which erases /
<__monty__>
o.O
<__monty__>
And this is just a "cause I can?"
<gchristensen>
it turns out erasing / on boot makes it easier to remember how you did things
<__monty__>
Because you *have* to permanently record them?
<__monty__>
Like, explicitly rather than in ambient state?
<gchristensen>
yea
<gchristensen>
although when it doesn't boot up (like now) it is a bit annoying :)
<__monty__>
>.<
<gchristensen>
oh it booted!
<gchristensen>
I guess it probably took a while to delete all the state it accumulated during setup
<adisbladis>
gchristensen: What's the advantage of erasing / instead of using tmpfs for / ?
<tilpner>
adisbladis: You can store larger files without eating up memory
<tilpner>
100GB scratch file on a tmpfs doesn't work, but is fine on larger/slower storage that you delete periodically
<gchristensen>
much cheaper to buy 512G NVMe than 512G ram :)
<adisbladis>
True that :)
<adisbladis>
It's not something I've run into very often, but it has happened.
<gchristensen>
ah, this one it'll definitely happen -- it'll periodically pull and run vagrant VMs
<pie_[bnc]>
MichaelRaskin: so much for getting the rfc OK-d any time soon ;p
<adisbladis>
gchristensen: Makes perfect sense then :)
<MichaelRaskin>
pie_[bnc]: at least now we get a constructive continuation of a discussion
<gchristensen>
which rfc?
<MichaelRaskin>
Named ellipsis
<gchristensen>
ah
<MichaelRaskin>
Some objection/context points I wanted acknowledged zimbatm wants better addressed
<gchristensen>
gotcha
<gchristensen>
I forgot to make it so the SSH host key was preserved between boots ... oops
<__monty__>
Ah, trusty TOEU : >
<gchristensen>
hehe
<etu>
I wonder which files I would miss if I migrated my home directory to tmpfs :S
<etu>
I must test this.
<gchristensen>
andi-: this motherboard seems to log me out of IPMI each time I reboot the box :\
<gchristensen>
etu: I don't erase /home on my laptop, but that is something I wonder too :)
<andi->
gchristensen: meh that sucks
<etu>
I kinda like having tmpfs as fs because it's super fast :p
<etu>
Probably nothing I notice ever, but just the thought is nice :D
<__monty__>
Agreed with andi-, /home needs to *go*, gchristensen ; )
<gchristensen>
hehehe
monsieurp has quit [Remote host closed the connection]
<gchristensen>
andi-: I wonder if my power supply isn't sufficient for it? like maybe it kills power to the bmctoo
<andi->
hexa-: ^ do we have the same issue at the space?
<hexa->
*reading backlog*
<hexa->
bmc logout on host reboot? i'd have to check
<gchristensen>
its that asrock rack thing
<hexa->
yep
<hexa->
X470D4U
<gchristensen>
yea
<hexa->
i'll check
<gchristensen>
thank you <3
vertigo_38 has joined #nixos-chat
vertigo_38 has left #nixos-chat [#nixos-chat]
<hexa->
gchristensen: nope, looks good here
<gchristensen>
hm
<andi->
Does it time out?
<hexa->
ryzen 5 3600 and I think we have a 450W psu
<andi->
Maybe due to inactivity (you are using some storage expansion cards IIRC)
<gchristensen>
what kind of power supply, hexa-? andi- no, it will work fine up until I reboot it