<micro_O>
i thought there would be some slick way to split and join, i.e. node -e 'process.env.KISS_PATH.split(':').map(p => `${process.env.rep}/${p}`)'
<acheam>
n..n..n..node?
<acheam>
what does that do? I dont have node
micro_O has quit [Ping timeout: 268 seconds]
<Guest47>
Is that java script
<acheam>
yes
Guest47 has quit [Quit: Client closed]
kawics11 has quit [Ping timeout: 252 seconds]
micro_O has joined #kisslinux
kawics11 has joined #kisslinux
<msk[m]>
kiss packages in node?
zenomat has quit [Ping timeout: 255 seconds]
zenomat has joined #kisslinux
zenomat has quit [Changing host]
zenomat has joined #kisslinux
<micro_O>
acheam msk[m] that line does a more fleshed out version of the `sed` command i put up earlier
Guest47 has joined #kisslinux
<acheam>
i dont know what the sed command does either :)
<acheam>
because it goes into a loop
<Guest47>
wait wut
<Guest47>
are you talking about the colon?
soliwilos has quit [Quit: nyaa~]
soliwilos has joined #kisslinux
noocsharp has quit [Quit: noocsharp]
noocsharp has joined #kisslinux
Guest47 has quit [Quit: Client closed]
<testuser[m]>
Hi
<micro_O>
jeez this is annoying
<micro_O>
echo "warning: something dumb" | sed -E 's|(warning:error) |::\1|'
<testuser[m]>
sed 's/^warning/error/' ?
<testuser[m]>
What the ::1 do
<micro_O>
im so dumb, neverming
<micro_O>
:: doesnt do anything, \1 does first capture group
GalaxyNova has joined #kisslinux
<GalaxyNova>
website's down
<micro_O>
okay, that was like way too much work. still needs a little update to the logic (warnings will 'fail' the pr check), but its almost there
<GalaxyNova>
also kiss seems to not want to install programs under root
<micro_O>
GalaxyNova was changed to .org again
<GalaxyNova>
oh nice
<micro_O>
done for the night, see yall tomorrow
micro_O has quit [Quit: micro_O]
gtms has joined #kisslinux
kawics11 has quit [Ping timeout: 252 seconds]
michalsieron has joined #kisslinux
GalaxyNova has quit [Read error: Connection reset by peer]
schillingklaus has joined #kisslinux
schillingklaus has quit [Quit: ERC (IRC client for Emacs 26.3)]
necromansy has joined #kisslinux
Uks2 has quit [Ping timeout: 240 seconds]
Uks2 has joined #kisslinux
soliwilos has quit [Quit: nyaa~]
claudia has joined #kisslinux
soliwilos has joined #kisslinux
<claudia>
o/
<claudia>
toybox grep seems to fail building python.
<claudia>
clear speaking, it fails to detect the endianess.
<soliwilos>
o/
<soliwilos>
I didn't really look into it, but I had some trouble with otools grep while building gtk+3.
<soliwilos>
Seemingly a grep process was infinite, never finishing, at lest took so long that my system started using swap. I eventually just killed the build and changed grep.
<soliwilos>
s/lest/least/
<claudia>
soliwilos: I build gtk+3 with otools grep without problem.
<soliwilos>
Hm, thanks. I'll try again.
<soliwilos>
It's apparently pausing at the same spot again.. "CCLD gtk-encode-symbolic-svg".
<soliwilos>
Eating more and more memory.
<testuser[m]>
check arguments for grep in top
<soliwilos>
Found this "/usr/bin/grep -o \bg[td]k_[a-zA-Z0-9_]*_get_type\b"
<testuser[m]>
Try it manually on some data, could be a bug in otools itself
<claudia>
dilyn: Yeah I found your fix. thanks. When I grep the configure file for the 'ax_cv_c_float_words_bigendian', there is just a simple grep without any special flags.
<dilyn>
i believe I saw an explanation for what python is doing in a stack overflow post a while back; i'll see if i can dig it up
<dilyn>
but basically some greps can grep binaries and others can't, python is just assuming gnugrep (ggrep, I should probably clarify that) is grep for this test
<dilyn>
the problem you notice when you do dumb things is that people make big assumptions about the roots of their problems. for instance, these fine folx just assume it's a solaris problem, not a grep problem
<dilyn>
i also don't know why they would use grep here lol
<claudia>
Appreciate the detailed explanation :p
<dilyn>
;)
<claudia>
I guess the solaris fraction also has the interpretation of the operating system as a whole, so theri grep is tightly bound to it.
<sad_plan>
finally, firefox built. I had to disable webrtc, and there I fixed it. no patch needed :D
<testuser[m]>
That's just hiding the problem :p
<sad_plan>
perhaps, but I have webrtc disabled in about:config anyway, so its really no need for me to build firefox with webrtc to begin with :p
necromansy has quit [Quit: necromansy]
claudia has quit [Quit: zzz]
<cem>
openbsd grep seems to be much stricter in terms of POSIX BRE
<cem>
like "what even sbase grep supports this" type of strict
<testuser[m]>
Bruh
<cem>
for instance I've always thought grep 'something\|otherthing' to be POSIX
<cem>
turns out it's not in BRE
<cem>
yet it's the first grep implementation that I saw to not support this expression
<testuser[m]>
We wouldn't have these issues if nonstandard versions of utils didn't exist in the first place :v
<cem>
yeah, and if only people tried to make their build systems portable
<cem>
ahem landley
Guest5669 has joined #kisslinux
<Guest5669>
So, would anybody here happen to know why sed isn't working with variable numbers above ten? The file is this https://pastebin.com/05rHQQCF
<Guest5669>
The command is for i in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16; do sed -i "s/#define file${i}.*/#define file${i} something_else/" "test.txt"; done
<Guest5669>
Output should replace something with something_else
<testuser[m]>
Do $(seq 16)
<testuser[m]>
Instead of i
<Guest5669>
It works for the most part except the file10 - file16 all are called "file1"
<Guest5669>
See how file10 - file16 are called "file1"
<Guest5669>
Sorry the loop is supposed to be 0 - 15 but that doesn't really matter
<Guest5669>
Issue still remains
<dilyn>
interestingly, if you change the sed to /#define file$i.*/#define file$i something_else/ it puts the leftover ones digit at the end of something_else
<dilyn>
i imagine . might mean something special to sed...
<dilyn>
considering it's all regex you probably want to do \.* instead?
<testuser[m]>
ang:
<Guest5669>
Yeah idk this is so baffling
<dbrooke>
when i is 1 it will also match the file10 to file15 lines
<dilyn>
lmfao
<Guest5669>
oh ok.
<Guest5669>
Is there any way to get an exact match with sed?
<Guest5669>
I thought sed was exact match, no?
<dilyn>
without g, the s option will replace the first match in every line
<dilyn>
you can alter the command to only replace once in the file iirc
<dbrooke>
for that file maybe add a space before .*
<ang>
why not simply do 's/\(#define file[0-9]\{1,\}\).*/\1 something else/' ?
<ang>
no need for the for-loop shenanigans
<Guest5669>
Literal wizard, I didn't know sed could do that
<dilyn>
the power of regex
<dilyn>
I think i may have potentially identified a resolution to that whole 'commit is not signed' error that plagues kiss u when forcing verification on commits...
<dilyn>
should've taken better notes on when this error appears with the repo tbh -- it's probably unrelated
<Guest5669>
Also, since POSIX doesn't define the array like in bash, how would one work with large amount of variables with a similar name? E.g. file_1, file_2, ... file_n
<Guest5669>
Ok something like this may work
<Guest5669>
set -- this is "a list" of "several strings"
<Guest5669>
Then looping over them for string in "$@"; do ...
<Guest5669>
Ha, it works >:D
Guest5669 has quit [Quit: Client closed]
michalsieron has quit [Ping timeout: 272 seconds]
akira01 has quit [Ping timeout: 252 seconds]
ax has quit [Quit: ax]
GalaxyNova has joined #kisslinux
kawics11 has joined #kisslinux
ax has joined #kisslinux
kawics11 has quit [Ping timeout: 265 seconds]
jslick has quit [Quit: Ping timeout (120 seconds)]
jslick has joined #kisslinux
kawics11 has joined #kisslinux
Guest2 has joined #kisslinux
<Guest2>
Hi. Can you tell me please where to start a deep learning of Linux to understand the meaning of compiler flags, etc?