« first day (501 days earlier)      last day (228 days later) » 

00:06
rydwolf, listen to me, you gotta go to Kromer st and be a [big shot]
00:41
Okay so apparently I got rejected for that card because the credit report company doesn't even know I exist yet lol
@lyxal LMAO
@lyxal If I gotta go south on 183 any time soon I'll make sure to stop by ig :p
finally, something to [pluck] me from the [Glum Humdrum]
@RydwolfPrograms now see if you went to Kromer St and became a [big shot] this wouldn't have happened
because if you did, the bank would know you were [number1ratedsalesman 1997]
and instantly given you 20 million kromer
for the [Low Low Price Of] too!
00:48
Good news: I'll be passing Kromer Street on November 10th on the way to San Antonio for a concerr
 
2 hours later…
03:13
@DannyuNDos ooh, nice toast
what is that corn mix on the left hash brown though
03:30
Corn-and-cheese
what kind of herb is the green bit
 
4 hours later…
07:34
@Bbrk24 especially with the volatile nature of exchange rates
just saw this on hnq and it is so hard to see "CBOE" and not read it "своё"
08:16
@Ginger it's at least a different kind of terrible taste in music than i usually have
if it qualifies as music
 
2 hours later…
09:59
okay less joking CMSR: youtube.com/watch?v=xOJEL3UPEt4
i swear i've heard most of this album already but i just now ran into this and it's got a really unique vibe even for aran
(but less joking isn't to imply it was just joking to begin with--i genuinely enjoy 1bpm lmao)
10:35
Weird ahh looking car
10:52
Why is the car door melting??!
 
2 hours later…
12:56
hey @RydwolfPrograms
I have a bit of a problem with my VM
specifically:
Filesystem     1K-blocks     Used Available Use% Mounted on
/dev/root      103080888 97683576    138060 100% /
devtmpfs        24675468        0  24675468   0% /dev
tmpfs           24678228     2124  24676104   1% /dev/shm
tmpfs            9871292      492   9870800   1% /run
tmpfs               5120        0      5120   0% /run/lock
I've cleared out some stuff and gotten it down to 94%
but that's still not ideal
13:11
ginger moment
mfw Coursier's cache is 7GB(?!)
oh boy here i go opening a bug on MSE
what's happened this time
see those facebook and twitter buttons
they broke
forver
they relied on a service called addthis.com
which just happened to be sunset in may this year :p
RIP
13:19
@Ginger Sorry, this is gonna have to be a you fix. The one thing I don't have is storage.
uh oh
I'm going to uipgrade but I'm in some analysis paralysis (and I-don't-want-to-spend-money paralysis) right now
okay, what can I nuke
My options are:
hmm
my MC server folder is 2.8 gigs
13:21
Yeah but out of a hundred that's not much right
I'm really not sure how you're using up a hundred gigs
64G of it is in my home folder
and the biggest thing in there by far, taking up almost half the space, is *drumroll* my compiled-from-source copy of LLVM 16
0
Q: "Help us Grow This Site" Twitter/X and Facebook buttons on Public Beta sites broken

lyxalIf you go to any public beta site's homepage, Programming Language Design and Implementation for example, you'll find a box like this: The Facebook and Twitter/X share buttons rely on a service called AddThis which was shut down at the end of May 2023. This means that the share buttons (except f...

1. Add two more 400 GB SSDs (bringing it up to 6×400, 1.2TB at 6× read, with 2 free slots)
2. Add four more 400 GB SSDs (bringing it up to 8×400, 1.6 TB at 8× read, with no free slots)
3. Add two spinning rust hard drives (so 4×400, 800 GB at 4× read, plus 2×4-8 TB, 4-8 TB at extremely slow speeds)
4. Add four spinning rust hard drives (so so 4×400, 800 GB at 4× read, plus 4×4-8 TB, 4-16 TB at extremely slow speeds)
5. Pull the 400 GB drives and replace them with 4-8 800 GB or even 1600 TB SSDs
which I can delete if I can just figure out how to make GDB load the apt-installed LLVM debug symbols
so @RydwolfPrograms do you have any idea how to do ^
@Ginger What's taking up the space, the source or the compiled binary?
I'd assume it's the source
If you've got anywhere near a 32 GB binary you've got bigger issues lol
I'd just grab the binary and delete the source
13:25
yeah uh
> 27G llvm-project-llvmorg-16.0.6/build
it's the binary
tf...debug symbols ig?
correct
and even with that build I still can't get GDB to load it
Still, that's only a quarter of your space. I'm really surprised you've managed to consume so much storage space
since it only outputted .as and not .sos
@RydwolfPrograms well, my home folder is 68G total
I'd check the others too, 32 GB for non-home is huge
Unless you've got a ton of big Docker images or something big in srv or some really big installs
13:28
Hey rydwolf did you see this car?
3 hours ago, by lyxal
user image
Yes, it's beautiful
2.1M    /dev
0       /sys
4.0K    /mnt
4.0K    /boot
8.4M    /etc
70M     /tmp
4.8G    /srv
488K    /run
4.0K    /opt
0       /proc
64G     /home
4.4G    /var
4.0K    /media
11G     /usr
16K     /lost+found
4.0K    /root
84G     /
:thonk:
Wait 84G for /?
@RydwolfPrograms reminds me of you :p
Did you already clear 16 GB?
13:29
yes
Or is the filesystem eating that much
oh
by deleting, among other things, 4gb worth of cached wheels and a partially-installed copy of the built LLVM
Your var's a bit big still
saw that, lemme check
it's mostly the journal
lemme vacuum it rq
Ooh I just found a lot of 6 800 GB drives for $300
13:32
2.7G freed
we're now down to 88% usage
which is probably good enough for now
and once I've figured out this LLVM bug I can delete the rest
so now for the million-dollar question: what are .a files and how do I make GDB load them
I think they're the same as .so right
> '/home/ginger/llvm-project-llvmorg-16.0.6/build/lib/libLLVMCore.a': can't read symbols: file format not recognized.
they are not
I thought one was for dynamically linked stuff and one for statically linked stuff?
@Ginger Answer: files with the "a" extension, and carefully
Gimme my million bucks please
file says they're current ar archives
@user no
@Ginger I swear I've asked this somewhere before
13:37
you're right about the difference tho
.a is static and .so is dynamic
Ah, so I guess .so is shared object?
in The Nineteenth Byte, May 25 at 1:58, by lyxal
What's a .a file?
in The Nineteenth Byte, May 25 at 2:00, by Bbrk24
@lyxal static libraries you link into during compilation
in The Nineteenth Byte, May 25 at 2:00, by Bbrk24
@lyxal static libraries you link into during compilation
Dammit ninja'd
:p
Oop rain time
Everyone make sure you ☂️
okay, I think I got it to load the symbols from the binaries
#0  0x00007fffe0af662d in ?? ()
#1  0x0000000000000206 in ?? ()
#2  0x00007fffe0af6724 in ?? ()
#3  0x00007ffff7d3f490 in ?? () from /usr/lib/jvm/java-17-openjdk-amd64/lib/server/libjvm.so
#4  0x00007ffff6702400 in ?? ()
#5  0x00007ffff795d50a in VM_Version::get_processor_features() ()
   from /usr/lib/jvm/java-17-openjdk-amd64/lib/server/libjvm.so
#6  0x49656e696c65746e in ?? ()
#7  0x01040800000406f1 in ?? ()
#8  0x1f8bfbfffffab223 in ?? ()
#9  0x01c0003f0c000121 in ?? ()
#10 0x000000000000003f in ?? ()
narrator: they did not
13:40
Rain time over
Everyone put their ☂️ as 🌂 please
Thamnk you in advanece
I do not think that this method is working
time to take a long, hard look at my Scala
Scala! :D
> val malloc = LLVMAddFunction(module, "malloc", LLVMFunctionType(ts.pointer, PointerPointer(1).put(0, ts.platformint), 1, 0))
find the null in that line
PointerPointer(1)?
ts.pointer?
13:42
oh right yall don't have context
What's ts?
Is ts.pointer an object representing a pointer type?
Or is it an actual pointer?
given LLVMContextRef = context
there, you should be able to understand everything now
Wow thanks
Such helpful
13:43
is context null? :p
ts is a TypeSupplier, which is a utility class containing lots of LLVMTypeRefs representing types used by the builder code
Ah
Nah context being null shouldn't be impacting anything
Are there any classes with similar names to PointerPointer?
nope
13:44
Like IntPointer?
Oh ok
PointerPointer is a class from the JavaCPP core libs
Is module null?
and I suspect it's the source of the error
I don't get why it's called PointerPointer and not Pointer
If you set it to something other than 0, does it segfault instead?
because the LLVMTypeRefs are pointers themselves
13:44
Oh oh I said that! I said PointerPointer!
Does that mean I got it right?
Yes you did!
Do I get [wild prizes]??!
I was getting errors earlier from PointerPointer about the types being null
@user oh good stuff. Very good stuff.
which were replaced by this segfault after I changed how I was constructing it
13:45
@lyxal As an award, you win 👌
so something is null that shouldn't be
@user oh brilliant
Wonderful
Excellent evem
*even
@Ginger Are you sure the segfault is because it's null and not because it's a pointer to some inaccessible part of memory?
@Ginger print all the values and see which is null
> siginfo: si_signo: 11 (SIGSEGV), si_code: 1 (SEGV_MAPERR), si_addr: 0x0000000000000000
yes
@lyxal I'll try
13:47
Bonus points if you get copilot to generate the println boilerplate for you
Any news?
I have a suspicion that it's not the parameters that are null
the segfault happens when I try to build a call instruction for that function
so maybe the function isn't being added to the module correctly
well, it's being added to the module
maybe I skipped a step...?
...since when has the builder been able to just make mallocs
and frees too???
but not reallocs
Since forever probably
Idk I've never used llvm
it seems to be a utility function
that is so very strange
sadly there's nothing for realloc ;-;
14:25
@lyxal imagine having copilot
14:55
@Dlosc I have no idea why, but this seems to work. Is my guess correct? I'm not quite sure.
alrihgt
(waiting to see how long it takes for someone to get triggered by the typo)
OMG
I fixed it
I was passing the return type of the function to LLVMBuildCall2 instead of the function type itself
and now it works :|
smh can't believe I did that
anyway, the fruit of my labors:
; ModuleID = 'vyxal'
source_filename = "vyxal"

declare i64 @malloc(i64)

declare void @free(i64)

declare i64 @realloc(i64, i64)

define i32 @main() {
entry:
  %alloca = alloca i64, align 8
  store i64 1, ptr %alloca, align 4
  %alloca1 = alloca i64, align 8
  store i64 2, ptr %alloca1, align 4
  %load = load i64, ptr %alloca1, align 4
  %malloc = call i64 @malloc(i64 72)
  %getLengthPtr = getelementptr inbounds { i64, i64 }, i64 %malloc, i32 0, i32 0
  %getNItems = load i64, ptr addrspace(64) %getLengthPtr, align 4
 
1 hour later…
16:16
TIL []=[] is legal JS
(destructuring assignment where you don't bind anything)
ooh interesting
phenomenal
...now i wonder if that also goes in python
it does
I would assume not
both []=[] and ()=() appear to be perfectly legal
Doesn't Python's destructuring assignment use commas without an outer delimiter?
Oh intertesting
16:18
@RydwolfPrograms that's the idiomatic way but it's just tuple syntax
i had no idea it would also work with [] though lmao
Yeah does Python even have destructuring assignment for lists?
it works with arbitrary iterables on the rhs
well, as long as there's no length mismatch
oh that's actually kinda neat
Oh it seems JS does too?
[a, b] = "cd"
I discovered this because I was looking at a new language that compiles to TS, and I got it to generate const [,] = m;
16:22
ahaha they even work in haskell
@noodleman and/or I might start golfing in Civet
16:43
noice
JS does it again
16:57
@emanresuA Ok well first of all, "smaller than a kilobyte" was intended to exclude the tinylisp formatter from consideration. :P But if you're saying you can't use multiple recursion because it'll blow the call stack on too-large inputs, that's true, but I'm not sure how that's different from any other language--you could say the same for my Python code.
@mathscat So close!--but reversed & palindromized doesn't work (should output cbaabc, but gives cbacba). That's because it's parsing as .(R(a.a)), but you want (.(Ra)).a (see the output from verbose mode).
Also, this is on-topic about CGCC, so we should probably move to TNB ;)
17:25
wth...every time I reload I either get Ginger's pronouns or Bbrk's
I guess there's not enough gender to go around
@UnrelatedString Doesn't work in Pip, tho (well, it's still a no-op in any case, but it generates a warning about assigning to a non-lvalue)
 
1 hour later…
18:51
@DLosc oof
I do have a 10+ byte solution, but that’s pretty intuitive
@DLosc I guess that's true, and mostly up to personal preference
19:33
yo @Seggan
do Vyxal arrays store references to their contents or the content objects directly
19:55
It doesn't matter
It stores references to their contents because that's what Python and the JVM do
If for some reason you need to store the contents directly, that would work too, because all Vyxal values are immutable
ighto
Also, there's a dedicated chat room for this :P
what?
@user feel free to move it yourself :p
can strings go in list literals?
20:15
is this about vyxal?
yes
@Ginger Yes, as can list literals
@Ginger ...I'm not an RO
@user then why does [# "amogus" #] cause a MatchError?
@user that's the joke :b
oh wait nvm
I may be stupid lol
No you're Ginger
But I guess you've found out that opening brackets are #[?
20:32
yes
and I said I may be stupid, not that I was stupid
checkmate atheist
 
1 hour later…
21:51
Okay I've resigned myself to the fact that I will never understand girls
The girl I've been talking to hasn't replied to a message I sent like a day ago asking about AP chem, and she set her note on insta at 5 am today to Yellow by Coldplay with a sparkly burrito emoji and I don't even know how to BEGIN to interpret that
22:38
I feel like “I’m not gonna doxx myself what |:<“ is something @Ginger would say
i just installed rust
i now consider myself a rustacean
22:55
Nice
You'll be a true Rustacean the day cat ears magically appear in your closet
7
@RydwolfPrograms Rydwolf incel arc
I'm taking a class on compilers right now, and the thing I was most looking forward to, type checking and other compile-time analysis, isn't a part of it. We're just directly compiling a Lisp to assembly :(
And there's a research group at my university that researches programming languages, but the Zoom link to their weekly meetings doesn't work and there's no other way to find them (I emailed them and was pointed to the Zoom link; the person I emailed isn't even here anymore)
I think I'm gonna have to completely give up on becoming a researcher studying programming languages
Seems like it's too niche to easily get a position
Much easier to become a normal software engineer and make programming languages as a hobby
@user Can confirm
@Bbrk24 it wasn't me I swear
@Ginger But if it was you, you'd still deny it so you don't doxx yourself
I guess we'll never know
Anyway, the main thing we know is it's not Radvylf :P
23:06
I know Rydwolf's:
- full name
- school name
- age
- appearance
- town of residence
and I didn't even have to go look for that stuff, he just posted it all here
Is it just my opinion, or is "Department of Standard & Intellectual Property" too long?
@Ginger references, because JVM and Python are both like that
ended up not mattering
23:36
@user wtf
23:51
Because there's all those jokes about Rust turning people into furries
Not sure how you interpreted it but I wasn't trying to poke fun at furries or something, just joking about Rust's community has a disproportionate number of furries
Oh you meant the other message
That was a joke too. I just found it funny how one person didn't see your message or forgot to respond to it for a day and you jumped straight to all girls being impossible to understand
I'm not actually calling you an incel, I know you're anything but. Should've left a :P
@Bbrk24 38°43'32.0"N 80°48'48.0"W
@Bbrk24 I once heard that home is where the heart is
Which must really suck for heart transplant patients :P
@lyxal the best part is that this is in West Virginia
meaning that the lyric "take me Home, country roads" was about a guy literally asking to go somewhere on country roads

« first day (501 days earlier)      last day (228 days later) »