<{^_^}>
nix-community/impermanence#9 (by lovesegfault, 3 days ago, open): nixos: implement full bind/link modules
<lovesegfault>
Can you see if this evaluates for you?
<lovesegfault>
Like, if a config using environment.persistence."/foo".{bind,link} works
<cole-h>
lovesegfault: So, something like `environment.persistence."/foo".link.files = [ "asdf" ];`, where `/foo/asdf` is a file and exists?
<lovesegfault>
It doesn't have to exist, I just want to see whether you can build the cfg at all
<lovesegfault>
I get while setting up the build environment: executing '/nix/store/jyqd2s6w5wmcvskphpx18k2pjp379awp-bash-4.4-p23/bin/bash': Argument list too long
<lovesegfault>
(error ...)
<cole-h>
error: attribute '/foo' missing, at /home/vin/.config/nixpkgs/hosts/scadrial/test.nix:267:42
<cole-h>
where test.nix is the nixos.nix from your most recent commit
<lovesegfault>
Ugh, you might need to point it to a path that exists, my bad
<cole-h>
Referenced line is part of `markedNeededForBoot`
<cole-h>
so probably because I just `mkdir /foo`'d instead of having it be a partition lol
<lovesegfault>
:D
<lovesegfault>
I don't think the part needs to exist
<lovesegfault>
it just needs to be in the fileSystems clause
<lovesegfault>
Grr, I need to get groceries, brb
<cole-h>
lovesegfault: I get "Bind source '/media/asdf' does not exist!", when /media/asdf is the same as /foo/asdf
<cole-h>
lovesegfault: So I think that files are maybe fscked, but bind-mounting a directory worked.
<cole-h>
linking a directory didn't work -- it gave me "Bind source '/media.' does not exist; it will be created for you." and "ln: failed to create symbolic link '': No such file or directory"
<cole-h>
And also when I bind-mount a directory, I got "Bind source '/mediaasdf' does not exist; it will be created for you." yet it bind-mounted the `/media/asdf` directory
<cole-h>
But when I change "asdf" to "/asdf", it does the same thing, sans the warning.
<cole-h>
When I change it to "/jkl", `environment.persistence."/media".link.files` works, but `environment.persistence."/media".bind.files` does not
<cole-h>
So, summary is: with "/file_or_dir", bind works for directories and not files, whereas linking works for files and not directories.
<colemickens>
anyone else following the zstd zfs pr?
<cole-h>
Yeah lol
<cole-h>
It's fun
<cransom>
train wreck fun?
<cole-h>
It's honestly strange. Sometimes, the major forces behind it get along well, have good input, etc. And then other times, they start fighting lol.
<colemickens>
The saga is weird to follow. Not at all what I would've expected for something like ZFS.
<colemickens>
"This is not the first PR that has been opened for this..." is just scratching the surface it seems.
<colemickens>
Not trying to be dramatic, I'm not even trying to pass judgement necessarily.
<gchristensen>
what do you mean?
<gchristensen>
like they'd be faster to merge?
<colemickens>
I thought there would've been a more coordinated effort for such a large feature. There's a lot of what I would consider design-time decisions being made about how to vendor libs, how to handle zstd metadata, etc, for something that has seemingly been through reviews and multiple submissions already.
<gchristensen>
ah
<gchristensen>
link to the PR?
<gchristensen>
it is possible somebody just ... sent it
<colemickens>
Yes! Totally! Which is why I'm hestitant to be throw too much shade, and also don't want to misrepresent something I've been skimming along wiht.
<colemickens>
It's really not even shade, just surprising to me. I think it's a neat story if it just "somebody" deciding to take a crack and then pushing through the (seems like layers) of process.
<samueldr>
unless there is something specific you wanted to point to?
<colemickens>
Oh hey, this isn't in the logs yet: https://www.youtube.com/watch?v=oPymb2-IXbg. I haven't watched it all, but seems like it contains good information and has been well received (great thumbs up ratio). 20K views to an intro to NixOS video is cool to see, too! :)
<ldlework>
samueldr: just interesting that the people who own the subreddit have had issues before
<ldlework>
samueldr: googling the social media accounts that the person listed, none of them exist anymore.
<samueldr>
I still haven't received a reply to the modmail I sent
<samueldr>
more and more it makes me think it's hostile or that they mostly left
<ldlework>
I only found one other page on the internet mentioning this person and I'm not comfortable linking it.
<colemickens>
for kxra? I found a slashdot profile?
<ldlework>
I think I was looking for kxxra and kxxxra which they've also used
<cole-h>
colemickens: That Dorian guy has a voice like one of those TTS apps lol
<colemickens>
cole-h: Heh, I kind of agree. I was kinda wondering if it was a bad mic or over compression? But I also kind of liked it in an almost asmr-y way?
<cole-h>
Haha
cjpbirkbeck has quit [Quit: cjpbirkbeck]
rajivr has joined #nixos-chat
cole-h has quit [Quit: Goodbye]
<eyJhb>
Why in the loving god, does FF not have a --verbose switch?
<eyJhb>
Currently it won't start, and there is no reason why
<etu>
strace
<srk>
eyJhb: looking at nixos-container maybe you could reuse what it does for chroot
rardiol has quit [Ping timeout: 264 seconds]
<pie_>
eyJhb: i feel your pain, generally speaking
<eyJhb>
Well, firefox starts now, there is just no audio :(
<eyJhb>
But everything else works
<qyliss>
eyJhb: there's a MOZ_DEBUG environment variable or something
EatThem has joined #nixos-chat
EatThem has quit [Client Quit]
lopsided98 has quit [Quit: No Ping reply in 180 seconds.]
lopsided98 has joined #nixos-chat
<eyJhb>
I got AUDIO!
<eyJhb>
What audio deamons do you guys use?
<eyJhb>
Pulseaudio, alsa, etc.?
__monty__ has joined #nixos-chat
parsley936 has joined #nixos-chat
nckx has quit [Ping timeout: 256 seconds]
<monsieurp>
I use systemd-audio-http-qr-code
<monsieurp>
daemon
<monsieurp>
each time I listen to audio, I open up a browser to http://localhost:8080 where a QR code is displayed for me to scan so that the music gets played on my iPhone
<monsieurp>
eyJhb: you should try
<gchristensen>
monsieurp: I'm not sure this is useful or on-topic
<clever>
gchristensen: and isnt this the off-topic channel?
<monsieurp>
nah I'm kidding :) I use FreeBSD which has a built-in audio mixer
<gchristensen>
trolling is off-topic imo
* clever
hides the google search for systemd docs
<joepie91>
it would actually be legitimately neat to have QR code pairing for audio on Linux
<qyliss>
Aww yeah I assumed that was a real thing
<qyliss>
Sounded entirely plausible and actually pretty cool
<qyliss>
an iPhone would have better sound quality than lots of laptops, etc
<gchristensen>
you can probably do it with one of PA's fancy sinks :)
<clever>
qyliss: in the past, i have messed with bluetooth, and i was able to make the laptop claim to be a bluetooth speaker
leah2 has quit [Ping timeout: 258 seconds]
<clever>
qyliss: the cellphone audio then came out of the laptop (it auto-configured a loopback in pulse)
<qyliss>
that's cool
<qyliss>
I'm quite excited about PipeWire
<joepie91>
I think all the pieces are already there for this sort of usage
<joepie91>
just not the UI flow :P
<clever>
my problem with bluetooth is the limited pairing slots on a lot of crap, and the very short range
<joepie91>
(as usual, on Linux..)
<qyliss>
not like this is a feature on any other platform P:
<qyliss>
*:P
<joepie91>
qyliss: it mostly just frustrates me that there's so much cool shit on Linux, that *could* be readily used to do some neat stuff that blows everything else out of the water... but it's undocumented and/or the UI for it is trash, so noone actually uses it
<joepie91>
applies more broadly to the general Linux ecosystem, I don't just mean the kerne
<joepie91>
kernel*
leah2 has joined #nixos-chat
rajivr has quit [Quit: Connection closed for inactivity]
<infinisil>
"To unsubscribe your email from our mailing list, please contact Customer Service."
<infinisil>
"Please note, you will still receive e-mails regarding licensing, new releases, and prereleases if you are associated to a license."
<infinisil>
-.-
<infinisil>
That's from mathworks.com (the people who make Matlab)
emilazy_ has joined #nixos-chat
manveru_ has joined #nixos-chat
jw358 has quit [Ping timeout: 256 seconds]
jw358 has joined #nixos-chat
manveru has quit [Read error: Connection reset by peer]
emilazy has quit [Write error: Connection reset by peer]
tazjin has quit [Read error: Connection reset by peer]
tazjin has joined #nixos-chat
tazjin has joined #nixos-chat
tazjin has quit [Changing host]
emilazy_ is now known as emilazy
manveru_ is now known as manveru
pasukon has quit [Ping timeout: 256 seconds]
pasukon has joined #nixos-chat
pr06lefs has joined #nixos-chat
rardiol has joined #nixos-chat
drakonis_ has joined #nixos-chat
drakonis has quit [Ping timeout: 265 seconds]
drakonis has joined #nixos-chat
drakonis_ has quit [Ping timeout: 260 seconds]
{^_^} has quit [Remote host closed the connection]
cole-h has joined #nixos-chat
{^_^} has joined #nixos-chat
<cole-h>
lovesegfault: Ping
mog has joined #nixos-chat
{^_^} has quit [Ping timeout: 256 seconds]
FRidh has joined #nixos-chat
drakonis_ has joined #nixos-chat
drakonis has quit [Ping timeout: 240 seconds]
<pie_>
still pondering whether theres any way to sell software in practice while providing the source code without having to make everything depend on support contracts
<pie_>
cant do high tech stuff without funding
<__monty__>
I suppose you're not satisfied with open code, closed assets/everything else?
drakonis has joined #nixos-chat
<joepie91>
pie_: depends on how broadly you interpret "support contract"
<joepie91>
eg. for personal-use-targeted software, a license will come with support from the vendor
<joepie91>
it's not typically considered a "support contract" but it definitely works like one, and is a reason for people to buy it
<joepie91>
other options include custom feature development (with or without funds pooling mechanism), working off a Patreon-esque model, etc\
<joepie91>
ultimately "selling software" just doesn't make a lot of sense, economically, selling something with zero replication cost, and essentially what you are asking money for is the *development process*, not the end product
drakonis_ has quit [Ping timeout: 272 seconds]
<joepie91>
that can be difficult to get money for, so often people try to sell different 'related services' instead, like support, or some sort of additional convenience
<joepie91>
but if your base model is "I'm looking for money for development, not for a product", then it makes it easier to reason about what can work and what can't :P
<Valodim>
the crux is that money for development doesn't scale, while money for a product does
<joepie91>
well, yes and no... they scale differently
<Valodim>
or rather: money for development will always scale linearly with your time, you can only work on the constant factor
<joepie91>
right
FRidh has quit [Ping timeout: 256 seconds]
FRidh has joined #nixos-chat
{^_^} has joined #nixos-chat
<cole-h>
Man, debugging Go is probably one of the worst things I've ever tried to do
<pie_>
__monty__: i dont want to be obnoxious and my pipe dream is working on tooling lol
<pie_>
pretty niche tooling at that maybe
<pie_>
stuff like llvm has apple behind it or whatever
<pie_>
yeah. and high tech stuff needs a lot of dev <joepie91> ultimately "selling software" just doesn't make a lot of sense, economically, selling something with zero replication cost, and essentially what you are asking money for is the *development process*, not the end product
<pie_>
oh right, exactly, as you continue <joepie91> that can be difficult to get money for, so often people try to sell different 'related services' instead, like support, or some sort of additional convenience
<pie_>
Valodim: right
<pie_>
i feel like the economics just doesnt really work :/ but im not too clever
<joepie91>
pie_: the unspoken underlying general problem here is that capitalism does not reward for contribution to society (ie. work), it rewards for salesmanship and monopolistic business practices primarily
<pie_>
it just feels wrong to sell closed source reverse engineering tools
<pie_>
lol
<joepie91>
pie_: the economics aren't *easy*, but they can work
<pie_>
joepie91: uh, right
<pie_>
(shhhh dont tell any of the acquisition departments that its open source! ;P)
<pie_>
well it not like companies can pirate stuff right :P so you just give it a license that prohibits redistribution right? :P
<__monty__>
The sorry state of license adherence begs to differ.
<joepie91>
companies pirate stuff constantlyu
<joepie91>
constantly*
<joepie91>
the copyright system, bluntly summarized, is "your work is protected if you have a bag of money to sue people"
<joepie91>
most people do not
<pie_>
__monty__: thatsthejoke
<joepie91>
hence all the GPL infringements that litter the embedded world for example
<pie_>
as a developer it drives me nuts if i cant see the source and i dont want to impose that on other people but damn
<pie_>
on a small scale, ransomcode coul work, but i doubt so on a larger scale
<pie_>
you cant build a tech business on people being charitable
<joepie91>
you can
<pie_>
._.
<pie_>
oh right mozilla n stuff
<pie_>
i thought of that like 5 minutes ago
<pie_>
but i was like, nah im too niche?
<pie_>
otoh maybe not but like,,,,idk
<cransom>
i'm not sure mozilla is the best case there. google was the main source of income for a while. it came down to who would bid more to be the default search engine.
<pie_>
mozilla has a very large user base
<pie_>
oh right
<pie_>
so idk who you were thinking about joepie91
<pie_>
something something mozilla is google antitrust insurance :P
<joepie91>
pie_: I'm not thinking about a specific anything or anyone. I'm looking at behaviour I've observed across many initiatives, companies, etc. and how that extends to a business model
<joepie91>
business model in the "I can live off this" sense, not the "tech giant" sense
<pie_>
gonna need a "several people can live off this" at least
<pie_>
how2 less shitty autodesk also.
<joepie91>
start with 1
<joepie91>
figure out how to scale as you go
<joepie91>
this is a largely unexplored field
<pie_>
probably because it doesnt
<pie_>
dslkfjgljk i still kind of want to give it a shot for one of my smaller projects
<joepie91>
pie_: I've spent some 7 years of my life trying to find a way to make a living within my ethical constraints and open-source requirements. throughout those 7 years, I have *continuously* been told that it cannot be done, because if it could, other people would be doing it.
<joepie91>
I now make a very respectable living within my ethical constraint
<joepie91>
constraints(
<pie_>
*nod*
<joepie91>
so I don't subscribe to the "people must not be doing it because it's not possible" perspective :)
<__monty__>
Tell us your secrets!
<joepie91>
the stark reality is that most people just have never put in the work to figure it out
<gchristensen>
I have written almost zero non-OSS code in the last couple years and I recommend it heartily
<pie_>
i feel like it might work, but i cannot fathom it working at a larger scale unless you get buttloads of government money or idk
<joepie91>
__monty__: for me, the key insight that took me entirely too long to gain, was that companies do not actually *care* about the license of their internal tooling
<joepie91>
__monty__: that is, they need a thing that works, that fits their requirements, and that they are allowed to use however they see fit. whether *someone else* can use it too, who gives a shit
<pie_>
i dont care about the license either
<joepie91>
so now most of my paid work is internal software dev, the results of which are open-source
<pie_>
it just breaks the model if the can just download it and be done with it
<pie_>
*they
<joepie91>
(fully)
<joepie91>
__monty__: the exact insight is going to be different depending on what you're trying to accomplish, exactly. but this is an insight that is documented basically nowhere, and it surprises many people when I tell them about it - which tells me that people just *didn't consider it*
<joepie91>
I have no difficulty believing that this is the same elsewhere; the non-standard path just hasn't been explored very much
<pie_>
joepie91: im somewhat surprised they ust let you oss internal ode
<pie_>
code
<pie_>
ive done some of that myself for the one gig ive done so far, but i assumed the default is no with a probable yes
<pie_>
modulo management
<gchristensen>
were you an employee or contractor?
<pie_>
and non business critical code
<joepie91>
pie_: it initially surprised me, too. but I think that's largely because I, like many people, at one point assumed that people would be irrationally possessive about their tools
<joepie91>
this turned out to be false (the same is true for many other assumptions about people being irrationally ______)
<joepie91>
instead, it was just a case of "I didn't understand the business considerations behind it"
<gchristensen>
contractors are also often able to bypass internal politics
<joepie91>
after learning those, it became apparent that the license just isn't very relevant; they care that they can solve a company problem in time, on budget, and as-needed
<joepie91>
the license is just a contract term like any other
<joepie91>
if the template contract you offer them says "work will be released as open-source", and they do not have a specific reason to object against that, they're just going to say "okay, sure, if that's a requirement from your side"
<joepie91>
gchristensen: not sure if the question was aimed at pie_ or me?
<gchristensen>
the question was for pie, w.r.t. my next message
<pie_>
something i see is companies asking to get paid for some enterprise features
<joepie91>
ok :)
<srk>
I'm on the sabe boat, I refuse to write stuff which is not oss for years, RedHat was kind-of compatible but there was some pressure later to write internal only-code which I refused to do :)
<srk>
*same
<joepie91>
srk: crucially, I don't do work for the typical "open-sourcey" companies :)
<joepie91>
usually more SMB, people who might have never heard of open-source to any serious degree, and who I first have to explain what it means and why
<pie_>
gchristensen: oh. i was an employee, but it was "just" some tooling rnd that wasnt too business relevant anyway because the project i got hired for (thankfully) got postponed
<joepie91>
which makes the pool considerably bigger
<srk>
joepie91: cool, I like to work with people who understand the benefits already :)
<gchristensen>
pie_: yeah, that is very hard to jump around the politics on "just"s like that
<pie_>
gchristensen: it was also a very CCC-tied companz
<pie_>
company
<gchristensen>
yeah but still it is a whole thing
drakonis1 has joined #nixos-chat
<srk>
pie_: what langs do you prefer btw?
<pie_>
srk: considering that im not good at any of them, probably haskell and rust and erlang
<pie_>
but actually i can only do python lol
<__monty__>
joepie91: I could see "If no one else uses it, it doesn't matter. If another company wants to use it they might pay me to refine it further and you could benefit from these improvements." as a decent argument.
<srk>
pie_: I went thru mainly Pascal, PHP, C, Python and settled on Haskell for now, also bits of erlang / c++, mostly during uni
<gchristensen>
no argument needed if your contract says it is a requirement
<joepie91>
__monty__: I explicitly avoid selling open-source as a business benefit
<joepie91>
doing so would be dishonest
<joepie91>
because that isn't the point
<__monty__>
pie_: I think restricting the languages you want to work with can really constrain your opportunities. I assume joepie91 does most of this work in JS which is enormously pervasive compared to haskell and rust.
<joepie91>
(it's also generally unnecessary, people are generally totally fine with "it benefits society such and such, and doesn't cost you anything")
<__monty__>
joepie91: Hmm, then how do you convince companies who are reluctant at first?
bqv has joined #nixos-chat
<joepie91>
__monty__: as per above; it doesn't cost anything, and it benefits society. if that is insufficient for them, then I don't do business with them :) but that is rare
<joepie91>
the vast majority of people I deal with are totally reasonable
<joepie91>
the ones who continue making a problem out of it are 90% of the time people I've already identified as a potential problem customer for other reasons
bkv has quit [Ping timeout: 246 seconds]
<joepie91>
(being rude or unreasonably demanding, being in a morally.... ambiguous line of work, etc.)
<gchristensen>
it is also useful if the company already knows they want to hire *you*
<joepie91>
very occasionally someone has business goals that just do not match up with open-sourcing of tools, in which case I wish them the best and I am not the right person for them
<joepie91>
right
<joepie91>
I sell my competence, not my open-sourceness; that is just a contract term :P
<pie_>
<gchristensen> it is also useful if the company already knows they want to hire *you*
<pie_>
<gchristensen> it is also useful if the company already knows they want to hire *you*
<pie_>
ugh
<pie_>
* <joepie91> I sell my competence, not my open-sourceness; that is just a contract term :P
<pie_>
both relevant points :P
<__monty__>
gchristensen: Do you align 100% with joepie91 or do you have another strategy for being able to work almost exclusively on foss?
<srk>
pie_: what does interest you? or .. what would you like to do?
<pie_>
me: the tooling *is* the business
<srk>
+1
<pie_>
srk: i guess i like engineering tools, like, tools for engineerng i men
<pie_>
mean
<gchristensen>
__monty__: people typically hire me because they want me to help them with Nix, and they typically also want to improve the Nix ecosystem at the same time
<pie_>
srk: also frustrated by the state of a lot of tools in general and i think we should do better
<srk>
pie_: asking because I kind-of do that already, for robotics, hw, 3d printing..
<joepie91>
pie_: honestly, a better option for that case may be to sell commercial support that is expensive enough to also pay for the other hours put into the software
<pie_>
srk: good to know, i dont actually know much about cad
<pie_>
joepie91: thats the idea but like, what does that even mean
<pie_>
*an idea
<joepie91>
pie_: basically, rate goes x2, means you can now spend half your time on whatever you want :D
<pie_>
most of the people using the reveng tools would probably be highly competent to begin with
<joepie91>
plugins for other tools, priority bugfixes, etc.
<pie_>
i guesssss
<pie_>
joepie91: they can probably build it themselves?
<joepie91>
pie_: if you can become known as "that RE vendor who, unlike the others, has top-notch support" - which, let's face it, is currently a hole in the market where RE is concerned - then that is a very viable business model, even if you ask for a lot of money
<joepie91>
pie_: of course they can. but RE is a highly-paid industry
<pie_>
apparently IDA has good support if you pay for it lol
<joepie91>
it's probably still cheaper to pay for a few expert hours than to have them spend twice the hours themselves by someone whose expertise is not "working on that software"
<joepie91>
pie_: I've heard mixed stories :P
<srk>
no need to pay for IDA when there's Ghidra :D
<pie_>
srk: right that too
<pie_>
i wish i didnt have so much tunnel vision lol
<joepie91>
"people who can sort shit out themselves" should not be your target market here :P
<pie_>
srk: I want to: imagine openscad on steroids. now imagine it on even more steroids.
<joepie91>
rather, you should probably target "organizations who want to keep software dev out of the org tree"
<pie_>
srk: now imagine it on even MORE steroids.
<srk>
pie_: lol, like Implict? :)
<pie_>
srk: also add anime girls to all our marketing
<srk>
* ImplicitCAD
<pie_>
joepie91: hm. right.
<evanjs>
*tries to use JDK downloaded from IDEA*
<evanjs>
why won't it.... ohhhhhhhh lol
<joepie91>
pie_: one crucial part of commercial support is that you have an inherent advantage over everyone else; you know every corner of the software you are providing support for
<pie_>
inb4 its cheaper for them to makee a bug bounty than hire or support lol
<pie_>
joepie91: right
<evanjs>
I do wish there was some sort of hook for when ldd fails tho lol
<joepie91>
pie_: this is what lets you compete on economic efficiency with in-house software dev
<pie_>
joepie91: i kind of pondered that
<pie_>
joepie91: it seems kind of rickety
<pie_>
?
<evanjs>
or at least something better than "no such file or direcetory" when it can't resolve symbols or blah
<joepie91>
pie_: how do you mean?
<pie_>
joepie91: i dont have anything concrete in mind
<pie_>
i definitely dont want to be writing it in haskell then lol
<pie_>
or actually i definitely do
<pie_>
xDD
<joepie91>
pie_: well, like I said, unexplored territory. best you can do is understand the economic processes (talk to people at your potential target customer organizations!), and play it be ear, and adjust the process as-needed to make it work
<pie_>
i cant find any devs to hire but i can also ask people for support contracts beause noone knows haskell
<pie_>
x'D
<joepie91>
wouldn't worry about that too much. you can always hire "competent developer who does not know Haskell but is interested in learning it"
<joepie91>
a good developer can pick up new tools :)
<pie_>
idk does haskell seem kind of hard
<srk>
at first it does :)
<pie_>
also funny thing, galois probably already built the hard part
<__monty__>
I sometimes wish I picked up a "good developer" certificate rather than a degree.
<pie_>
they have some 10 million dollar government contract and are now building a ton of this stuff
<srk>
pie_: no worries, they'll abandon it when the grant ends
<joepie91>
__monty__: as in, real-world experience on a resume? :P
<pie_>
srk: haha yepp
<pie_>
i was like ok this is probably all relly cool and i wont understand any of it, bu muh ego
<pie_>
*but muh ego
<pie_>
srk: if the contract ends
<joepie91>
don't undersell yourself :P
<pie_>
also arent all those guys basically programming languages researchers
<pie_>
phds or whatever
<pie_>
thats a pretty high bar
<__monty__>
joepie91: Yes, it looks a lot like you have to start out doing one or more jobs you don't want to do for anyone to actually take a serious look at your resume.
<pie_>
srk: now take the openscad on steroids and put a gui on it
<pie_>
cool sadly this doesnt actually say much to me , im just a version control fetishist
<pie_>
my cad experiene is basically 0
<srk>
yeah, that's exactly why you should get your hands dirty :D
<pie_>
hmhm:P
<pie_>
noted
<pie_>
now if only i could unblock my haskell qt project
<pie_>
but first i need to survive friday
* pie_
goes back to studying
<pie_>
gchristensen: joepie91thanks for the input
<srk>
I've learned way more stuff by just doing things than uni could ever teach me. it's something completely different to passively learn things and learning because you need the knowledge to proceed :)
<pie_>
yeah
<pie_>
im extremely conflicted on finishing my bachelors
<srk>
good luck with the exams :) what's the topic btw?
<__monty__>
Yes, all job advice appreciated : )
<srk>
or ^
<pie_>
university has not been going well for me exacerbated by family stuff between the interval 1-2 years ago
<pie_>
and now im like ffffuuuu i guss i jut dont care enough about this material to study it properly
<cransom>
fwiw, i'm glad i finished my bachelors. i could have had a good job without it, but i was forced to learn things that were applicable later that i might otherwise not have known.
<pie_>
i dont dislike learning
<pie_>
and also yeah would be good to hae a bachelors but...
<srk>
I've only finished bachelors CS degree, the higher started boring me too much, wanted to do more practical stuff
<pie_>
i keep trying to convince myself and telling people maybe i should switch to CS , but the university cs department here is super meh
<srk>
I would go for EE now :)
<pie_>
srk: right now i "do" physics and fridays exam is my N+1th attempt at the materials science course exam
<srk>
cool!
<srk>
don't switch! :D
<pie_>
its not even a bad topic
<pie_>
i just rather bang on the computer than studying, and i dont have the self control (anymore?) to jsut study what they give me
<pie_>
other pipe dream: do a youtube series where we litarally build up a computer but first we construct the universe
<pie_>
ok maybe not THAT ddeep
<FRidh>
pie_: ugh materials science was my least favorite course during physics studies. Good luck with it! In retrospect though, the things I did pick up from it have been worth it.
<joepie91>
pie_: is the pipedream to make this, or for this to exist?
<pie_>
joepie91: more the former but the latter would also probably be neatr
<joepie91>
pie_: right, so I *think* Ben Eater's breadboard computer series kinda goes like this? not quite invent-the-universe, but pretty deep on the foundations
<pie_>
not sure how to do it without compressing an entire universiy education into it, but i didnt thik too hard :D
<pie_>
(im also lacking said education, but hey, books! i guess)
<drakonis1>
that monorepo has some real cool stuff
<drakonis1>
this bit about nix native build systems is the best
<tazjin>
drakonis1: yeah, we've now got people using those for actual production stuff, too
<drakonis1>
excellent
<tazjin>
I have WIP versions of a Rust one (this is complicated by the fact that there are no ABI stability guarantees, and artefacts aren't reusable even across things like builds on different core counts)
nckx has joined #nixos-chat
<tazjin>
and a C++ one (this is complicated by everything being a mess)
<drakonis1>
nice!
<drakonis1>
the rust builder is available on your repository, yeah?
<drakonis1>
i'm only asking because i haven't looked up
<drakonis1>
i'm very curious about how it looks like and how it handles crates
parsley936 has quit [Read error: Connection reset by peer]