<cracken>
Figured I would give it a go on an old hp I had lying around
<coolcoder613_32>
I remember my first time using Haiku... I wanted to be quadruple booting..
<coolcoder613_32>
Also on an old HP laptop
<cracken>
Oh lol
<coolcoder613_32>
So.. now i have 3 computers that mainly run Haiku
<phschafft>
coolcoder613_32++
<coolcoder613_32>
That HP laptop
<cracken>
Im thinking of booting this with icaros/aros also to have fun with that, if you're familiar
* coolcoder613_32
is not
<coolcoder613_32>
An old 32-bit desktop (What i am using now)
<phschafft>
aros is the one of kitty?
<coolcoder613_32>
and a thinkpad
<cracken>
Yeah its the one with the cat mascot
* phschafft
nods.
<cracken>
I was wondering why apps kept jumping behind vision but I realized I turned on auto raise by accident lol
<cracken>
Was about to ask how to change that
<cracken>
Has anyone here used emacs on haiku?
<cracken>
If I imported my config from linux would everything work like I expect?
<coolcoder613_32>
try it
<cracken>
That I will
cracken has quit [Quit: Vision[]: i've been blurred!]
HaikuUser has joined #haiku
HaikuUser is now known as cracken
tuaris has quit [Read error: Connection reset by peer]
<cracken>
Huh, weird. Emacs just crashes after loading. But weirdly if I copy my init.el into the emacs folder it'll launch and execute everything until after my calls to different repos it seems, then it crashes. Replacing my config with a blank init.el just crashes at start, same as if I had no init.el in the first place. I'll have to look into this a bit more later
<cracken>
Maybe because I'm on a 32 bit system
<cracken>
Dont know if that effects some packages here
<coolcoder613_32>
why 32-bit?
x10z has joined #haiku
tuaris has joined #haiku
mmu_man has quit [Ping timeout: 480 seconds]
tuaris has quit [Read error: Connection reset by peer]
coolcoder613 has joined #haiku
<cracken>
The computer I'm using has an athlon in it
Maturi0n_ has joined #haiku
OscarL has quit [Read error: Connection reset by peer]
Maturi0n has quit [Ping timeout: 480 seconds]
x10z has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
scanty has quit [Remote host closed the connection]
x10z has joined #haiku
scanty has joined #haiku
cracken has quit [Quit: Vision[]: i've been blurred!]
tuaris has joined #haiku
tuaris has quit [Read error: Connection reset by peer]
x10z has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
cracken has joined #haiku
tuaris has joined #haiku
kinkinkijkin has joined #haiku
cracken has quit [Ping timeout: 480 seconds]
x10z has joined #haiku
AlaskanEmily has quit [Remote host closed the connection]
cracken has joined #haiku
cracken has quit []
Yonle has quit [Remote host closed the connection]
chhavi has joined #haiku
Yonle has joined #haiku
cracken has joined #haiku
cracken has quit []
cracken has joined #haiku
chhavi has quit [Quit: Page closed]
<cracken>
Hi guys! I've been messing around with haiku for a bit now and I've been experiencing a major issue, that being I can't boot in without using VESA. I'm using a hp compaq 5750 which has an amd xpress 1150 graphics card, both I've read should be officially supported in the 64 stable release though correct me if that's wrong. I considered making a bug report but I want to know if theres anything I can do before I do that.
ChhaviTekriwal[m] has joined #haiku
<augiedoggie>
can't boot? it hangs or it gives you a kernel panic?
<cracken>
I just get a blank screen after the haiku screen when loading
<dovsienko>
it probably didn't make it into the nightly build window: suggests installing hrev57581-1
garndalf has joined #haiku
garndalf has quit []
OscarL has joined #haiku
tuaris has joined #haiku
<Begasus>
Hola OscarL
<OscarL>
Hola Begasus :-)
nhtello has joined #haiku
bbjimmy has quit [Quit: Vision[]: i've been blurred!]
bbjimmy has joined #haiku
HaikuUser has joined #haiku
HaikuUser has quit []
louis77 has joined #haiku
<Begasus>
k, let's see if we can get swift-lang bumped to python3.10
<OscarL>
good luck with that! :-D
<OscarL>
"Fetching package for haiku >= r1~beta4_hrev57578-1" <<< LOL Wut?!?!
<nephele>
OscarL?
<Begasus>
something messed up versioning?
<OscarL>
I'm running beta4... no idea why it wants me to download a nightly ?
<Begasus>
jikes!
<Begasus>
that's plain wrong ...
<Begasus>
OscarL, is that with updating the systtem or something pulling it in?
<Begasus>
maybe a package build on nightly*
Oscar-L has joined #haiku
<Oscar-L>
Oops! Power went down :-D
OscarL has quit [Ping timeout: 480 seconds]
Oscar-L is now known as OscarL
<OscarL>
a storm is comming this way, it seems.
<Begasus>
stay safe!
<OscarL>
Begasus: was trying to run a build for uncrustify 0.78.1
<Begasus>
so some dependency pulled it in?
<OscarL>
Will have to double check... once the VM decides to start :-/
<Begasus>
"boot" it :)
* OscarL
kicks the thing with both legs at the same time. Did not ended well for him.
* OscarL
runs his "system_installed_packages_built_by_me()" bash function.
<Begasus>
:D
<Begasus>
which: no ninja in (.:/boot/home/config/non-packaged/bin:/boot/home/config/bin:/boot/system/non-packaged/bin:/bin:/boot/system/apps:/boot/system/preferences)
<Begasus>
who wrote that recipe?
<jmairboeck>
missing BUILD_PREREQUIRES cmd:ninja?
<Begasus>
yeah, but still no go (seems it's using "cmake --build")
zdykstra has joined #haiku
<jmairboeck>
then you'll need to unset cmake because of haikuporter's alias
<jmairboeck>
or switch to invoking ninja directly
<Begasus>
did you have a look there? ;) no wonder it's not enabled :D
<Begasus>
k, manual calling build-script in the directory seems to launch it
<jmairboeck>
what are you trying to build? (missed that I think)
HaikuUser has joined #haiku
<Begasus>
swift_lang 4*
nosycat has joined #haiku
<louis77>
perhaps naïve question, in which C++ standard is Haiku written in?
<nosycat>
Dunno, but that's a legit question. Considering the project's age, the oldest parts could still be C++98.
HaikuUser2 has joined #haiku
HaikuUser2 has quit []
<Begasus>
currently on gcc13, not sure if gcc2 is still used in building 32bit
<jmairboeck>
everything that is written with gcc2 as target is using only C++98 because gcc 2 doesn't understand anything newer
HaikuUser has quit []
HaikuUser has joined #haiku
<nosycat>
gcc 2.x is still preinstalled in 32-bit beta 4
HaikuUser has quit []
nhtello has quit [Quit: Vision[]: i've been blurred!]
<jmairboeck>
and C89, which is probably even "less convenient" (no declarations in the middle of blocks, no line comments)
<nephele>
louis77: depends. We use gcc2 for some system components that need ABI compat with BeOS
<jmairboeck>
the weird thing is, it does support these in C++ mode, but not for C
<nephele>
for everything else it's pretty modern compiler toolchains
<nephele>
64bit is compiled with the normal toolchain only for example
<nosycat>
But there's no explicit requirement on the C++ standard to use?
<jmairboeck>
I think the new netservices use newer C++ features as well
<nephele>
for the parts that don't need BeOS compat, no.
<nephele>
If the ABI can be kept compatible with a newer toolchain then it would also not matter for that part
<nephele>
it's just currently only gcc2 that can. But there were some ideas to for example add support in llvm to support the ABI
louis77 has quit [Ping timeout: 480 seconds]
<nosycat>
That would be an interesting development.
<nephele>
some patterns like exceptions are not used at all, even if the toolchain supports them
<nephele>
the new libnetservices is kind of a testbed for supporting some newer features
<nephele>
figure out what works for us and doesn't :)
<nosycat>
But that's more of a tacit convention?
<nephele>
hmm. Well, new code in the kits follows the style that was there before. And there exist no exceptions in the api because they werent around then
<nephele>
that could change if the experiemtns in new netservices are positive anyhow ;)
<nephele>
mostly stuff that would be changed after R1 though, probably
<nephele>
first getting a solid release out, then change everything xD
<nosycat>
Agreed. There's a similar situation with the FLTK library.
<nosycat>
It was first released literally in the same year as the '98 standard.
<nosycat>
That means pre-standard C++. No STL, no nothing.
<nephele>
haiku uses the c++ standard lib only sparingly. We have equivalent things for many things
<louis77>
I'm sorry I think I got cut off from IRC without noticing... so which was it? :)
<nosycat>
And there's always the question of what new language features to adopt without changing too much.
<nephele>
hi louis77, you can check the log. There is a link in the room topic
<nephele>
not sure how much you missed
<louis77>
oh thx I'll check it
<nephele>
nosycat: indeed! some stuff maybe just doesn't work for us
<OscarL>
no idea why I'm still getting that "haiku >= r1~beta4_hrev57578". /me runs a pkgman full-sync, and cries seeing noto_*_cjk in that list.
OrangeBomb has quit [Read error: Connection reset by peer]
<nosycat>
louis77, short version: some system components stick to C++98 or even older on 32-bit editions, otherwise there's no hard rule.
<nephele>
OscarL: that "crying" should soon be less
* OscarL
crosses fingers :-)
<nephele>
if i understand correctly that patch on gerrit will load variants from otf files too
OrangeBomb has joined #haiku
<nephele>
i looked at the freetype api and it exposes them quite wierd... but the TL;DR is that for us as an api consumer we can just pretend it's a normal "old" non-variant font
<OscarL>
can't we "pin" some packages :-D (thinking I should create fake noto_*_cjk packages, so Haiku doesn't tryies to unninstall itself completely, just because I want that package gone :-P)
<nephele>
I don't understand why you would want it gone
<nosycat>
There are quite a few nice-to-have features one could wish were in pkgman.
<OscarL>
I have no need for those.
<nephele>
proper font coverage is definetely one thing the OS should ensure, they don't take that much space, and i have no idea how the app_server would behave if you delete this hardcoded font from the drive
<Begasus>
OscarL, I guess those using noto-cjk have no need for the regular noto package also :)
<nephele>
haikus font selection really is quite narrow, but very good coverage
<nephele>
Begasus: thats definetely wrong
<Begasus>
no idea nephele :) just a though
<nephele>
unless they don't use arabic numbers, and don't ever see english text
<Begasus>
afk for a bit
<OscarL>
nephele: I'm not advocating for their removal from Haiku... just from MY system.
<nephele>
OscarL: ¯\_(ツ)_/
<nephele>
^ Noto CJK character used
<nephele>
I understand. But my question is "why"? The power of defaults is strong, and especially in Haiku i'm trying to rather upstream any behaviour changes so i don't have to redo everything all the time ;)
<nephele>
But i don't understand why you want that package gone so badly. If it's the case of "full-sync" that won't work. Because it will delete your fake package every time since it is unobtainable
<OscarL>
if the package wasn't so DAMN big, my network so slow...
<OscarL>
OR if I had a local "update proxy" so keeping many Haiku installs properly updated wasn't such a PITA, I wouldn't mind so much.
<OscarL>
regarding the removal of fake packages... thus my question about pinning packages.
<nephele>
don't use full-sync
<nephele>
easy
<nephele>
full-sync is ment to synchronize your package state with the server, even by deleting packages
<nephele>
so if that isn't what you want use pkgman update, and potentially only with the packages you want to update as arguments
<nephele>
pkgman full-sync is, in debian equivalent to a "apt-get dist-upgrade"
<OscarL>
whatever... I'm just trying to solve this weird "letme install a random nigthly over you beta4" that haikuporter is pulling on me.
<nephele>
you have some package that wasn't build on beta4
<OscarL>
no shit.
vdamewood has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
B2IA has quit [Quit: Vision[]: i've been blurred!]
<OscarL>
ran `package list -i $f | grep "requires: haiku"` for every single package I have installed... not a single match for "hrev57578"
B2IA has joined #haiku
<nephele>
well, yes. if you had any installed you could not run beta4?
<OscarL>
same by searching for that on "haikuports/repository".
<OscarL>
will nuke that repository, and let haikuporter run dependencies again. Sigh :-(
<Begasus>
this works from Terminal but not with haikuporter ...
<augiedoggie>
do you have the llvm16_lld package in the requirements?
<Begasus>
nvm, found it :)
<Begasus>
right missing that augiedoggie
hightower3 has quit [Ping timeout: 480 seconds]
<nephele>
Begasus: ld.lld is the linker from llvm
<nephele>
oh, already solved. nvm then
<nephele>
:)
<Begasus>
k, build launched :)
<OscarL>
too bad ActivityMonitor doesn't has graphs for I/O activity. Would be fun to document my HDD's pain while doing `haikuporter --repository-update` :-P
<OscarL>
"real 7m15,246s", welp... not THAT bad.
<Begasus>
the lld packages don't conflict with eachother, packaging error?
<jmairboeck>
don't they all provide commands with the same name?
<OscarL>
Still getting that "Fetching package for haiku >= r1~beta4_hrev57578-1"... build finished OK. No idea what's going on.
v_harkonnen has quit [Ping timeout: 480 seconds]
<Begasus>
used the one from LLVM17 that's installed just now, so that's ok, but still
<OscarL>
"chroot has these package" only lists "haiku-r1~beta4_hrev56578_95-1", as expected, lol.
<Begasus>
ah, seen that passing by once or twice too OscarL
<Begasus>
maybe switched between them at one point?
v_harkonnen has joined #haiku
<OscarL>
Not that I recall. I even have a nightly on bare metal on this same machine. use the same haikuports tree from both, but even after nuking "haikuports/repository" and "haikuports/packages", that should cause no issues, AFAIK.
<Begasus>
OscarL, any idea on this? ModuleNotFoundError: No module named 'AttributeNodes'
<OscarL>
yup. Seems that Fedora had at some point a patch for that, but even that got removed (from seeing the link above).
<OscarL>
maybe it was needed for older versions?
hightower4 has quit [Ping timeout: 480 seconds]
<Begasus>
could be, also keep in mind this is still the 4* version, upstream is already at 5.9.2
<OscarL>
only mention of "hrev57578" I can find... in a comment on the python3.10 recipe. And now, after doing a clean build of uncrustify... no more "Fetching" of that haiku version.
<OscarL>
wonder if I need an exorcist for this install ;-)
<Begasus>
heh
<OscarL>
"Running the tests is as simple as: > ctest" <<< LIAR!! >>> "No test configuration file found!" is what that gives :-P
<nephele>
> wonder if I need an exorcist for this install ;-)
<nephele>
If you'd like, you can give me ssh access. I will make that problem appear tame by comparison
janjo has joined #haiku
<nephele>
;)
<OscarL>
I'm behind 7 proxies, thou! :-P
* phschafft
waves to nephele.
<nephele>
hi phschafft
<phschafft>
all good (all of you)?
<nephele>
I am fine
<nephele>
playtesting a game with someone
<nephele>
(and soon package it for haiku)
<phschafft>
:)
<nosycat>
I'm good. Been learning Vala.
<Begasus>
now we need a vala IDE? :P
<nosycat>
Nah.
<OscarL>
Mmm, will need to add more RAM to this VM to run uncrustify tests :-D
extrowerk has joined #haiku
<OscarL>
"Out of memory" at last for one of the tests... after 572 seconds of floating around 2040 MB out of 2048 total.
<OscarL>
This is actually holding up way better than I expected, given that memory usage.
HaikuUser has joined #haiku
HaikuUser has quit []
janjo has quit [Read error: Connection reset by peer]
OscarL has quit [Quit: "Rebooting into bare-metal"]
BiPolar has joined #haiku
<BiPolar>
Damn! peaked at 6200 MiB, and one of the test still died with OOM. I'm *not* enabling swap for you uncrustify!
janjo has joined #haiku
x10z has joined #haiku
* nephele
laughs in 32GB physical memory
<BiPolar>
Oh yeah? /me does a sum() of ALL his DIMMs.... eh... can't even complete 24 GB :-D
<nephele>
well, i have a good justification
<nephele>
this machine is supposed to compile haikuwebkit
<nephele>
does a clean build in 48minutes weeee
nosycat has quit [Quit: Leaving]
<nephele>
my main computer i use haiku on has like 4gb and an i3
<BiPolar>
I'm happy when a Python build takes me less than 30 minutes :-D
* phschafft
goes back to his 2kB.
<BiPolar>
damn, now even git is laughing at my slow PC: "It took 23,70 seconds to enumerate untracked files / See 'git help status' for information on how to improve this."
janjo has quit [Quit: Leaving]
<Begasus>
k, let's save this small patch before I forget :)
hightower4 has joined #haiku
<nephele>
I get that often too
<nephele>
it's related to how fast git can access files on disk iirc
<nephele>
for big repos i just turn this off entirely
<BiPolar>
I have "commit.status=false" and usually use status with -uno, but yeah.. slow HDD+big repo... first gits ops are ALL slow here :-D
<nephele>
sadly it's slow even with M.2 SSD :(
<nephele>
kind of hoping waddlesplash has some time to improve that one of these days, maybe :D
<nephele>
#makegitfast
<nephele>
it's like one of the major pains of developing on haiku that git is so slow
hightower3 has quit [Ping timeout: 480 seconds]
<BiPolar>
not only git, but deleting a "work" dir can take A LOT of time when the file count is, say > 10k
<nephele>
i think it's mostly a problem with many small files
<BiPolar>
and sometimes it freezes, you can't open other apps, or even close them, until something "unfreezes" :-D
<nephele>
(though, on the other hand that this isn't a single api call for the FS where it deletes the folder and cleans it up in it's own time is beyond me)
<nephele>
maybe you need a tree based fs for that
* BiPolar
remember using deltree on DOS :-(
HaikuUser has joined #haiku
HaikuUser has quit []
Dario has joined #haiku
Dario is now known as Guest168
<Begasus>
closing down here
<Begasus>
cu peeps!
Begasus has quit [Quit: Vision[]: i've been blurred!]
Guest168 has quit []
<bitigchi[m]>
What kind of CPU load bar widget Slayer uses?
<bitigchi[m]>
I don’t understand anything at all from the source code
<waddlesplash>
nephele: git isn't too bad for me... faster than on Windows I think
<nephele>
waddlesplash: really? Well I've never used it on windows but it's waaaaay slower than on Linux, FreeBSD, Macos for me :(
<nephele>
running git status in the webkit checkout takes 90 seconds or more for untracked changes alone
<botifico>
[haikuports/haikuports] OscarL d224508 - uncrustify: update to version 0.78.1 (#10121)
B2IA has joined #haiku
kescher has quit [Remote host closed the connection]
kescher_ has joined #haiku
kescher_ is now known as kescher
kescher has quit []
<AlienSoldier>
nephele native game?
<nephele>
Somewhat. It uses irrlichtMT, and thus sdl2
<AlienSoldier>
Got rayman Origin today, quite late to the party, very good platformer.
<AlienSoldier>
nephele seem a nice engine, give at least dreamcast level 3d engine.
<nephele>
well, it's a 2d game :)
<AlienSoldier>
at one point 3d game almost become playable, i surely should be able to lay playstation 1 3d only qiht my cpu.
v_harkonnen has quit [Remote host closed the connection]
nephele has quit [Quit: Vision[]: i've been blurred!]
HaikuUser has joined #haiku
kescher has joined #haiku
mr_lou has quit [Quit: Leaving]
HaikuUser has quit [Quit: Vision[]: i've been blurred!]
hightower3 has joined #haiku
BiPolar has quit [Quit: Vision[]: i've been blurred!]
<dovsienko>
waddlesplash: I did a few greps in Haiku repository to get the hang of interface flags, and it seems IFF_RUNNING isn't used, but IFF_UP is
<dovsienko>
also there is IFF_LINK, which ifconfig prints as "link" and which directly corresponds to Ethernet link status on my virtual Intel 1000 in the VM
hightower4 has quit [Ping timeout: 480 seconds]
<dovsienko>
so I thought this could be mapped into libpcap to replace the current "Connection status unknown" with the link status where applicable
<dovsienko>
however, tap interface has the same IFT_ value and it seems to have IFF_LINK nailed up. is there a better way to tell a tap interface from an Ethernet interface in the POSIX space?
<dovsienko>
or at least in the Haiku C space
<dovsienko>
that is, better that comparing the first two characters of the interface name with "tap"
v_harkonnen has joined #haiku
Ellenor has quit [Ping timeout: 480 seconds]
Ellenor has joined #haiku
B2IA has quit [Quit: Vision[]: i've been blurred!]
B2IA has joined #haiku
jmairboeck has quit [Quit: Konversation terminated!]
x10z has joined #haiku
v_harkonnen has quit [Ping timeout: 480 seconds]
dby has joined #haiku
janjo has joined #haiku
v_harkonnen has joined #haiku
* coolcoder613_32
waves
jnn is now known as jn
v_harkonnen has quit [Ping timeout: 480 seconds]
janjo has quit [Quit: Leaving]
janjo has joined #haiku
Srdran has joined #haiku
janjo has quit []
x10z has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
Srdran has left #haiku [#haiku]
janjo has joined #haiku
xet7 has joined #haiku
dcatt has quit [Quit: Vision[]: i've been blurred!]
janjo has quit [Ping timeout: 480 seconds]
<dovsienko>
waddlesplash: I tried adding some code to manage IFF_PROMISC on the network interface. one thing that does not make sense is that SIOCGIFFLAGS works on both AF_INET and AF_LINK, but SIOCSIFFLAGS works only on AF_INET (no problem, I use both)
<dovsienko>
another thing is that after closing the socket the interface still retains IFF_PROMISC (I gather that's what allows ifconfig to manage interface flags and then to exit without restoring the old values)
<dovsienko>
so I tried a call to clear IFF_PROMISC in pcap_cleanup_haiku(), and it works, but it seems it would not work well when the user has run "ifconfig xxx promisc" beforehand and wanted it to stay this way, but now tcpdump would unset IFF_PROMISC on exit
<dovsienko>
also multiple tcpdump/Wireshark copies would fight to set/unset the flags and cause unwanted effects to each other
<dovsienko>
do you think it would be better not to manage IFF_PROMISC from libpcap at all, or to accept the potetial disadvantages of multiple processes trying to manage the same flag?
<dovsienko>
it would not be difficult to remember what IFF_PROMISC was at the activation time and to restore it during cleanup, if it helps