ChanServ changed the topic of #haiku to: Open-source operating system that specifically targets personal computing. | https://haiku-os.org | Nightlies: https://download.haiku-os.org | Bugtracker: https://dev.haiku-os.org | SCM: https://git.haiku-os.org/ | Logs: https://oftc.irclog.whitequark.org/haiku | Matrix: #haiku:matrix.org | XMPP: #haiku%irc.oftc.net@irc.jabberfr.org
diver has quit [Quit: Leaving.]
<adrian_> ok, toolchain building, haiku later
<adrian_> fun
diver has joined #haiku
<adrian_> waddlesplash: how do i get the device list? so i can see what touchpad driver its using?
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
<scantysnax> good evening.
diver has joined #haiku
diver has quit []
diver has joined #haiku
Al2O3 has quit [Quit: Textual IRC Client: www.textualapp.com]
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
<scantysnax> hm, solid connection
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
diver has joined #haiku
diver has quit []
<AlienSoldier> red-green-red-greed, it is still xmax
<scantysnax> lols AlienSoldier
<Yoke> holy fuck did someone's internet have a stroke or something?
<scantysnax> net might not be good where ever in russia he his...
<Yoke> I mean that's..... fair
<Yoke> good lord though
<Yoke> it's christmas all over again with the amount of red and green
<augiedoggie> it's not uncommon for an IRC client to freak out for a variety of reasons
<augiedoggie> just need to ping one of the operators to kick him temporarily, like waddlesplash
<scantysnax> speaking of, are there any devs in the house?
<scantysnax> experiencing a maybe bug.
<Halian> New keyboard and replacement trackball get \:D/
<Yoke> Nice
<Halian> augiedoggie: can confirm; when I used HexChat under Windows, it would randomly crash soemtimes
<Halian> *sometimes
<Yoke> I've been meaning to try a trackball, are they any good or is a bit of a gimmiky thing?
<scantysnax> wouldn't want to get used to a trackball when all other systems i will use will not have one.
<scantysnax> if that's your thing, fine.
<Yoke> looking at them I don't really understand, is it just so you don't move your arm about? (or fingers with how mice seem to be going)
<scantysnax> you don't really need too much arm movement with a regular mouse.
<Yoke> yeah that's what i mean
<Halian> Yoke: I like them
<Halian> scantysnax: you can plug the trackball into your other systems :þ
<Halian> The main draws of a trackball for me are ① minimal wrist movement, lessening RSI risk ② since you don't move the whole thing, it takes up less desk space, and my desk is cluttered
<scantysnax> you could, but what a pain in the ass. it's not like i'm going to leave my house and take it with me if i know i will be using a computer
<augiedoggie> my old laptop had a trackball built into it that was nice
<Yoke> I totaly don't do that with my heavy ass mx master..... yeah
<scantysnax> sure, i'll work on your system... but first let me get my trackball... i don't get it.
<augiedoggie> why would a trackball be a requirement?
<Saijin_Naib[m]> Ploopy has amazing trackballs
<Yoke> then you gotta ask if it's alright to plug something into their system in the first place, I'd guess some places would be pretty iffy about that
<scantysnax> that is basically my point.
<Habbie> my wife absolutely brings her trackball places
<Halian> Nice
<AlienSoldier> i only like trackballs for games
<Halian> scantysnax: if it's not your thing, that's perfectly fine
<Halian> but I'm trackball-pilled >w<
<zdykstra> Where are you going that you're constantly using random computers?
<scantysnax> i do a lot of a lot of things.
<zdykstra> When I actually worked at an office, I just expensed a trackball for my desk ...
<Yoke> lucky
<zdykstra> How's that lucky? Most places with desk jobs will do that, I imagine
MsInput has joined #haiku
<Yoke> only if you have some sort of accesability need over here, other than that it's just suck it up and deal with it
Halian| has joined #haiku
Halian has quit [Ping timeout: 480 seconds]
dalme has quit []
nipos has left #haiku [Disconnected: Replaced by new connection]
nipos has joined #haiku
erysdren has joined #haiku
erysdren has quit [Quit: Konversation terminated!]
<waddlesplash> adrian_: if it's "PS/2 Mouse" then it's just the PS2 driver
<waddlesplash> all the support for PS/2 is in one driver and published as separate devices
AlienSoldier has quit [Read error: Connection reset by peer]
angrystar170 has quit [Quit: The Lounge - https://thelounge.chat]
angrystar170 has joined #haiku
<scantysnax> i think i might have stumbled upon a bug in BView/BBitmap drawing
<scantysnax> screenshot: https://0x0.st/8soH.png
<scantysnax> as you can see, i am trying to draw the whole screen one colour but the first few pixels of the first few scanlines are black
<scantysnax> relevant code:
<waddlesplash> hmm
<scantysnax> sometimes is skips 8 scanlines or so.
<scantysnax> it's a roll of the dice.
<waddlesplash> oh
<scantysnax> sometimes it works as expected
<waddlesplash> I think ViewBitmaps may be drawn asynchronously
<waddlesplash> they aren't drawn in the BView::Draw method
<waddlesplash> so it's random whether the view bitmap is drawn before you modify it
chilledfrogs has quit [Quit: connection reset by purr]
<waddlesplash> if you want to use view bitmaps in this way, you will need to use double buffering probably
<waddlesplash> otherwise, just modify the bitmap whenever and call Invalidate()
<waddlesplash> or, don't call SetViewBitmap and instead just call DrawBitmap inside Draw
<scantysnax> i just need something simple. the data for the bitmap is calculated once only in advance in most cases
<waddlesplash> then you don't need to do anything in Draw at all
<waddlesplash> you can load the data into the bitmap and call SetViewBitmap once
<waddlesplash> you can delete the Draw method altogether
Al2O3 has joined #haiku
Al2O3 has quit []
<scantysnax> removed SetViewBitmap and added DrawBitmap to Draw() seems to be working properly
chilledfrogs has joined #haiku
<scantysnax> thanks waddlesplash :)
<waddlesplash> np :)
vdamewood has quit [Quit: Textual IRC Client: www.textualapp.com]
<adrian_> waddlesplash: how do i get the list of devices?
<scantysnax> listdev
artemisroadrunt has joined #haiku
<artemisroadrunt> Hi this is my first time trying IRC. The discord link didn't work. Hopefully I'm doing this right. I need help installing Haiku on VM.
<phschafft> then, welcome. and what is your question?
<waddlesplash> adrian_: listdev lists PCI
<waddlesplash> there's some argument to get more than just PCI
<waddlesplash> also Devices app
<waddlesplash> however not all drivers are properly tied to the device tree yet unfortunately
<waddlesplash> and the "driver" field may not be properly filled in even if they are
<waddlesplash> there's a bunch of TODOs here
<artemisroadrunt> Thank you phschafft for the welcome. I am curious if I have to do anything specific with VM or if I can just choose the ISO I downloaded and thats all.
<phschafft> last time I did that I just downloaded the nightly. but naturally those are not stable in any way.
mmu_man has quit [Ping timeout: 480 seconds]
<phschafft> people seem to have different opinions. but I always had luck with qemu based virtualisation.
<artemisroadrunt> ok. Next question, is 32 bit a better option or should I try the 64 bit version?
<artemisroadrunt> my pc is 64 bit, but I dont think it matters in vm
<zdykstra> If you have legacy software to run, 32bit. Otherwise 64bit
<phschafft> again, personal opinion here: we're in the 21th, 32 bit should be *way* in the past ;)
mmu_man has joined #haiku
<artemisroadrunt> Ill vm the 64 bit then. I read on haikus FAQ about it being very different from BE OS inspired coding in 32 bit
<phschafft> there is no 32 binary compatibility as you find in other systems.
<phschafft> but that does only matter if you want to run old beos binaries.
<phschafft> for all software compiled for Haiku that shouldn't really matter (someone correct me if I miss anything)
<artemisroadrunt> Ok. still going to use 64. I use linux btw lol
<phschafft> I hear virt-manager even detects Haiku correctly as a guest OS.
<Yoke> It stands to reason if you even know about the existance of Haiku you probably daily drive linux or a BSD derivative
<artemisroadrunt> What is BSD like? I only know Linux, Windows, and MAC. the first one being what I currently use the most. the last being my least used OS
<Yoke> BSD is derived from Unix same as Linux is, so at it's base it's for the most part the same, I haven't messed with it much myself but when you use something premade it feels pretty much the same as any linux distribution, this is just my opinion though and i feel like it would piss off people on both sides of the fence.
<phschafft> it likely does. and that is why I skipped answering.
<phschafft> I mean I don't agree with you like, more than 10%, but at least I enjoy that you have a different opinion. ;)
<Yoke> I'm tired to the point of being somewhere between don't care avenue and insomnia way, I'm gonna let opinions fly lol
<artemisroadrunt> Ok sorry I mentioned it.
<artemisroadrunt> I need to install Haiku on vm now. I'll be back if I have questions. If not, Ill respond my reaction to it later. :)
<phschafft> I would say that BSD and Linux ae both connected via POSIX (and so is Haiku!). So while they are completly different systems a good number of software is available for all of the as-is.
<phschafft> so naturally a user may not see that much of a difference. a power user, or higher may/will.
<Yoke> this one is sure to annoy someone but MacOS at it's core is a super modified version of NextStep, which in itself is Unix based (if i got that wrong just say)
<Yoke> For most people having something that can get on the internet and not crash every time they close a tab (if they ever do)
<Yoke> is enough
<phschafft> all fruitware is basically a copy of something else with a ton of marketing added. I don't have the biggest contact surface. but I must say that for the modern (non-classic) ones I have not seen a single thing that I have not seen somewhere else often years earlier.
<Yoke> yep, and whenever they have done something first it's buggered everything up for everyone else
<Saijin_Naib[m]> <phschafft> "I hear virt-manager even detects..." <- Very easy guest OS
<artemisroadrunt> I actually gtg. Do I just close this window?
<andreasdr[m]> Hi there.
<artemisroadrunt> hello
<artemisroadrunt> Im gonna try to close this window. I gtg now
<artemisroadrunt> thank you guys
artemisroadrunt has quit [Quit: Leaving]
<Yoke> brb gotta go dark for a min
flag has quit [Ping timeout: 480 seconds]
<adrian_> waddlesplash: ok, listdev and devices
<adrian_> waddlesplash: got it
<waddlesplash> there's also listusb for USB devices specifically
<adrian_> waddlesplash: i had haiku build crap out early tonight, i gotta get irssi installed on it so i can chat
<adrian_> waddlesplash: whilst tinkering
<adrian_> also
<adrian_> do y'all have laptop acpi stuff?
<waddlesplash> if you didn't disable building bios loader then it may quit early yeah
<adrian_> like s3 suspend?
<waddlesplash> we don't have sleep if that's what you're asking
<waddlesplash> at all
<waddlesplash> unfortunately
<adrian_> ah ok
<waddlesplash> been on the TODO forever
<adrian_> yeah
<waddlesplash> at this point we need a big refactor of the device manager before we can do that
<adrian_> *nod*
<waddlesplash> but that's also been kicked down the road ...
<adrian_> not today, i just want the touchpad working better
<waddlesplash> yeah that's more doable lol
<waddlesplash> seeing as there are touchpads that do work
<adrian_> yeah
<adrian_> and i unfortunately, sigh, know the synaptics stuff
<waddlesplash> so getting your touchpad to work in touchpad mode with gestures and all that should be easier
<adrian_> god i wish i didn't
<waddlesplash> hmm I thought synaptics generally worked already
<waddlesplash> and it was elantech that didn't
<waddlesplash> but idk
<adrian_> i'll go look
<waddlesplash> we do have elantech code too tho
<waddlesplash> it's just disabled and doesn't work quite right
HaikuUser has joined #haiku
HaikuUser has quit []
mmu_man has quit [Ping timeout: 480 seconds]
TMM has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
TMM has joined #haiku
linuxuser01 has joined #haiku
HaikuUser has joined #haiku
HaikuUser has quit []
lsd_82 has joined #haiku
lsd_82 has quit [Quit: Chat-O-Matic[0.1]: i've been liquified!]
HaikuUser has joined #haiku
lsd_82 has joined #haiku
lsd_82 has quit []
HaikuUser has quit []
lsd_82 has joined #haiku
linuxuser01 has quit [Quit: WeeChat 4.4.2]
Begasus has joined #haiku
<Begasus> morning peeps
<lsd_82> Morning Begasus
<Begasus> morning lsd_82
probono98 has quit [Quit: The Lounge - https://thelounge.github.io]
probono98 has joined #haiku
HaikuUser has joined #haiku
HaikuUser has quit []
flag has joined #haiku
MisthaLu has joined #haiku
HaikuUser has joined #haiku
HaikuUser has quit []
draKon has joined #haiku
<nekobot> [haiku/haiku] autocommitter pushed 1 commit to master [hrev58476] - https://git.haiku-os.org/haiku/log/?qt=range&q=bd0f4983c8ac+%5Ec5874ee4286f
<nekobot> [haiku/haiku] bd0f4983c8ac - Update translations from Pootle
Al2O3 has joined #haiku
Al2O3 has quit []
OrngBomb has joined #haiku
OrangeBomb has quit [Ping timeout: 480 seconds]
_-Caleb-_ has left #haiku [#haiku]
_-Caleb-_ has joined #haiku
<Begasus> KERN: runtime_loader: /boot/system/lib/mlt-7: Troubles reading ELF header
<Begasus> guesss that's not a good thing?
HaikuUser has joined #haiku
HaikuUser has quit []
yann641 has joined #haiku
yann641 has quit []
yann64 has joined #haiku
tqh has joined #haiku
lsd_82 has quit [Quit: Vision[]: i've been blurred!]
yann64 has quit [Quit: Vision[]: i've been blurred!]
FreeFull has quit [Quit: rebooting]
HaikuUser has joined #haiku
HaikuUser is now known as kantraa
<kantraa> Hello everyone
<Begasus> Hi kantraa
<kantraa> How are you doing on this fine day
<Begasus> fine as usual :)
<Begasus> How is it going there?
FreeFull has joined #haiku
<kantraa> Pretty good
<kantraa> Exams are starting in a few days I'm a bit stressed
<Begasus> Just be sure to take your time to prepare for it :)
jmairboeck has joined #haiku
tqh has quit [Remote host closed the connection]
<Begasus> hi jmairboeck
<Begasus> thanks for spotting that one :)
<jmairboeck> hi Begasus
<jmairboeck> another thing I found weird when looking at this patchset: on Windows it links ws2_32 (winsock), but not libnetwork on Haiku. Or does it somewhere?
<Begasus> I had that in, but reverted it, doesn't seem to error on it
<jmairboeck> but it could well be that the Windows code is quite different to the Unix code, so that doesn't have to mean anything
<jmairboeck> ok
<Begasus> from an older patch
<Begasus> had a MR at KDE declined for simular ws2_32/network not that long ago :)
<Begasus> k, no more segfault atm
<Begasus> let's see if opencv is the one ..
<jmairboeck> fun fact: in OpenCASCADE I made the build system pretend that libnetwork is the same as winsock: https://github.com/haikuports/haikuports/blob/0f041819cd7efc268fc59ab3c8633648afe97564/sci-libs/opencascade/patches/opencascade-7.8.1.patchset#L17
<Begasus> in ohter patches upstreamed it was accepted, but those gave undefined referrences without it, so that made sense
<Begasus> so in case of OpenCASCADE that probably made sense too :)
<Begasus> k, opencv it is, if I enable that it gives the segmentation violation
<Begasus[m]> k, first need to push frei0r opdate :)
Anarchos has joined #haiku
<Begasus> kdenlive: ../haiku-git/src/system/libroot/posix/pthread/pthread_mutex.cpp:95:status_t __pthread_mutex_lock(pthread_mutex_t*, uint32, bigtime_t): mutex->owner == -1
<Begasus> one of the last line in Terminal when kdenlive crashes opening one of the examples :(
<Begasus> hi Anarchos
<Anarchos> hello Begasus !
<Anarchos> Begasus it's my birthday today.
<Begasus> Happy Birthday then Anarchos!!!
<Anarchos> thanks
* Begasus blows the wistle and blows up some balloons
<Anarchos> :)
<Begasus> How young if I may ask?
<Anarchos> 46
<Begasus> Prime time still :D
<Begasus> but at that age I was starting to push some tasks to the younger ones ;)
dalme has joined #haiku
<Anarchos> Begasus :)
<draKon> anyone good at evaluating a site?
wicknix_ has joined #haiku
DKnoto has joined #haiku
<draKon> i dont know where to go to get help on the web with something like this
<Begasus> no good here
<draKon> ok
<draKon> it runs great with some haiku browsers
<draKon> :)
<draKon> except for that cool font i wanted, seems i gotta install it
<Begasus> put it in ~/config/non-packaged/data/fonts/ttfonts?
wicknix has quit [Ping timeout: 480 seconds]
<Begasus> reboot ...
Begasus has quit [Quit: Vision[]: i've been blurred!]
Begasus has joined #haiku
mmu_man has joined #haiku
<Begasus> k, rebuild for kdenlive
_-Caleb-_ has left #haiku [#haiku]
_-Caleb-_ has joined #haiku
<Begasus> My BeOS polo is getting old :)
kieren has joined #haiku
kantraa has quit [Quit: Vision[]: i've been blurred!]
draKon has quit [Read error: Connection reset by peer]
x10z has joined #haiku
<Begasus> lot's of these still: plugin not available: "frei0r.alpha0ps" :/
<Begasus> frei0r not listed in: melt -query filters
MisthaLu has quit [Quit: Leaving]
kieren has quit [Quit: Vision[]: i've been blurred!]
andreaa71 has joined #haiku
andreaa71 has left #haiku [#haiku]
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
MsInput has quit [Quit: WeeChat 4.5.0]
<Begasus[m]> latest shotcut still crashing :(
deneel has joined #haiku
MisthaLu has joined #haiku
<Begasus> not getting any wiser with "strace shotcut" :P
<Begasus> last line: [ 22410] --- SIGSEGV (Segmentation violation) {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_errno=0x80001301, si_pid=22410, si_uid=0, si_addr=0x1, si_value=(nil)} ---
tqh has joined #haiku
HaikuUser has joined #haiku
<dovsienko> Begasus: if that reproduces reliably, the usual next step is to have a debugger produce the stack trace to see if there's a runaway pointer somewhere
<dovsienko> (that's how I debug things on Linux)
HaikuUser has quit []
<Begasus> dovsienko, it is, has been crashing on launch since day one, been fiddling with mlt a bit more, but no changes with those
<Begasus> now if I knew how to debug this (noob with that) :)
<dovsienko> when Clang was crashing with SIGSEGV in my tests, it always popped up a GUI dialog with a "debugger" button
<Begasus> with strace I don't get that, but a normal launch in Terminal pops up in a debug window
<Begasus> I could save the report, but wouldn't even know where to look (or what to look) for :)
<Begasus> this pops up in syslog https://bpa.st/PTJA
<dovsienko> there should be the stack trace, but that likely requires the binary to have the debugging symbols (the "-g" flag of GCC and Clang)
<dovsienko> the segfault is certainly in QT, the question is whether QT code in this case is trying to use a pointer that comes from QT itself (which would be a QT bug) or from the application (which would be the application bug)
<Begasus> this is where it stops in Debugger: https://0x0.st/8sMM.png
<dovsienko> hmmm, although QArrayDataPointerIDs is in /boot/system/bin/shotcut, so that's some kind of a callback from QT back into the application, which means it would be easier to debug it
<Begasus> could run a build with debug enabled, it's using "Release (cmake)" atm
<dovsienko> QArrayDatapointer<char32_t>::~QArrayDat... looks like a destructor of a class template to me
<dovsienko> so I would take the source code of shotcut and add a couple debug prints to warnIfNotWritable()
<Begasus> ok, now you got me :)
<dovsienko> maybe it is trying to validate a file name that is a NULL, or some such
<Begasus> ok, first do a build with RelWithDebInfo
<Begasus> whoot, base package almost the same, debug package +32MiB :)
<dovsienko> typically with debug information you should be able to see the exact line numbers, but I am not skilled with Haiku debugger
<Begasus> Not sure if this is getting any closer (debugger window) https://0x0.st/8sMJ.png
<dovsienko> perhaps if the "Variables" tab showed the variables, something could be comprehended even without debug printf()s
<dovsienko> hmm, "this" is not supposed to point to 0x1, unless it is some idiom I am not familiar with
xet7 has joined #haiku
<dovsienko> Begasus: if you select the previous function on the stack, does it have "this" set to a more normal-looking pointer?
<dovsienko> note that this time warnIfNotWritable() is not on the stack
<Begasus> err .. previous function in the debugger?
<dovsienko> FilterController::qt_static_metacall()
<dovsienko> or MainWindow::showEvent()
<Begasus> 0x1236556154c0
nephele_xmpp has joined #haiku
<dovsienko> oh, that does not look unrealistic
<Begasus> remind you I'm a noob
<dovsienko> "this" in C++ is a pointer to the current object instantiated from the class, which in normal circumstances cannot be one byte away from the NULL pointer
x10z has joined #haiku
<dovsienko> it should be in the data segment of the process, or maybe on the stack, but not at 0x1. if it is, that will surely trigger a SIGSEGV
<dovsienko> so I would guess a pointer to an object gets wrong somewhere in MainWindow::showEvent() or shortly afterwards
<Begasus> should have done better at school when I was young(er) :)
<dovsienko> where's the source code of this thing?
<Begasus> first comment on that :) // This is needed to prevent a crash on windows on startup when timeline
talos has quit [Quit: Ping timeout (120 seconds)]
talos has joined #haiku
<dovsienko> a competent user of a debugger would set a breakpoint at the beginnng of the method and see which line crashes it, but I do not know how to do that
<dovsienko> it would be really nice to see the line numbers in the stack trace
<Begasus> np, thanks for thinking along
<dovsienko> the most practicable thing is to put a printf() before every statement of the method (this would be just 13) and see which specific statement triggers it
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
<nekobot> [haiku/haiku] waddlesplash pushed 3 commits to master [hrev58477] - https://git.haiku-os.org/haiku/log/?qt=range&q=fccefbf34729+%5Ebd0f4983c8ac
<nekobot> [haiku/haiku] b1a913c3aa6b - kernel/vm: Free pages via vm_page_free only, not vm_page_set_state.
<nekobot> [haiku/haiku] 0c2ad1411faa - kernel/vm: Unreserve memory as well as unreserving pages.
<nekobot> [haiku/haiku] fccefbf34729 - kernel/vm: Use allocate_early instead of passing a get_free_page method to early_map.
nephele_xmpp has left #haiku [Error from remote client]
imrahil has joined #haiku
nephele_xmpp has joined #haiku
imrahil has left #haiku [Error from remote client]
imrahil has joined #haiku
imrahil has left #haiku [Disconnected: Replaced by new connection]
imrahil has joined #haiku
imrahil has left #haiku [Disconnected: Replaced by new connection]
imrahil has joined #haiku
imrahil has left #haiku [Disconnected: Replaced by new connection]
imrahil has joined #haiku
tqh has quit [Quit: Leaving]
imrahil has left #haiku [Error from remote client]
imrahil has joined #haiku
HaikuUser has joined #haiku
HaikuUser is now known as kantraa
<kantraa> Is there an unofficial Discord client for Haiku that works on 32 bit? The web client is so heavy that it maxes out my cpu
<Begasus> afaik there is "no" discord client for Haiku :)
* Begasus fails misarably trying to add printf() somewhere :)
imrahil has left #haiku [Error from remote client]
imrahil has joined #haiku
<Begasus> out of my skills (well was anyway) :)
<dovsienko> oh well, then report the bug to the upstream and if it has trivial steps to reproduce, maybe they will fix it
<waddlesplash> could be a Haiku bug too
<waddlesplash> or a bug in our Qt port at least
<Begasus> or mlt ...
imrahil has left #haiku [Error from remote client]
imrahil has joined #haiku
<kantraa> A bug? In MY beta operating system? How is this possible
<Begasus> ah, think I got atleast one line :)
imrahil has left #haiku [Disconnected: Replaced by new connection]
imrahil has joined #haiku
<Begasus> the crash has been there for years, hence it was never enabled (shotcut)
<dovsienko> printf ("DEBUG: %s:%u %s()\n", __FILE__, __LINE__, __func__);
<dovsienko> and "#include <stdio.h>" at the top of the file
imrahil has left #haiku [Error from remote client]
imrahil has joined #haiku
imrahil has left #haiku [Disconnected: Replaced by new connection]
imrahil has joined #haiku
imrahil has left #haiku [Disconnected: Replaced by new connection]
<Begasus> the printf() in the statement?
<Begasus> still builds, so didn't brake it yet (further) :)
kantraa has quit [Quit: Vision[]: i've been blurred!]
nephele_xmpp has left #haiku [Error from remote client]
<Begasus> should take some crash cource on debugging once :) https://0x0.st/8suu.png
HaikuUser has joined #haiku
<HaikuUser> s
<HaikuUser> x
<HaikuUser> x
<HaikuUser> x
<HaikuUser> x
<HaikuUser> x
HaikuUser has quit []
nephele_xmpp has joined #haiku
<Begasus> DEBUG: /sources/shotcut-24.11.17/src/widgets/newprojectfolder.cpp:59 showEvent()
<Begasus> DEBUG: /sources/shotcut-24.11.17/src/mainwindow.cpp:2650 showEvent()
<Begasus> getting this now when launching from terminal
<Begasus> those 2 contain the printf() statement
nephele_xmpp has left #haiku [Error from remote client]
HaikuUser has joined #haiku
HaikuUser has quit []
<dovsienko> Begasus: what is on line after 2650 now?
<Begasus> Q_UNUSED(event)
<Begasus> well, those 2 commented out lines and then Q_UNUSED(event)
nipos has left #haiku [Error from remote client]
nipos has joined #haiku
MsInput has joined #haiku
<dovsienko> Begasus: and after Q_UNUSED there's a printf(), correct?
<dovsienko> I am looking at the usage of Q_UNUSED() in other places of the source tree and so far I am seeing it applied to a reference argument rather than a pointer
<dovsienko> in src/models/multitrackmodel.cpp:2013 it is applied to an int
<Begasus> it's deffenitly in that block :) https://0x0.st/8sSm.png (nuked it out) :)
<Begasus> that was before I nuked it out
<dovsienko> aha...
* Begasus ducks
<dovsienko> so if it is indeed "#define Q_UNUSED(x) (void)x;", then I cannot understand why doing or not doing it triggers a segv because there's no pointer dereference here
<phschafft> there is a printf() after it?
<Begasus> nope phschafft, atm like it's in the link presented earlier
<phschafft> 16:35 < Begasus> https://bpa.st/SZVQ
<phschafft> that one?
<Begasus> did have the next lines wrapped in printf() before though
<Begasus> yes phschafft
<phschafft> the difference is that the optimizer may have a different opinion depending on such hints.
<phschafft> I'm not into C++ to really tell what is going on however.
<Begasus> I'm not into anything :) but appreciat the help +1
<phschafft> but you could try adding something like: printf("DEBUG: event=%p\", event);
<Begasus> appreciate(?)
<phschafft> that will *force* a use of the pointer without a dereference.
<Begasus> inside that statement?
<dovsienko> phschafft: the matter is, the argument is indeed not used in the method
<phschafft> (because the compiler can just not optimise that the printf() away no matter what)
<phschafft> yes, so my idea here is to force the usage and then see if it changes things.
<phschafft> also keep in mind that with C++ if you change anything in a class you need to recombile all code that includes that class (also indirectly).
<Begasus> like this? https://bpa.st/YKGA
<dovsienko> make it "\n" at the end, like the other debug prints
<Begasus> so far I'm happy to find the root cause, and got it launching (even with a hack) :)
<dovsienko> that's not the root cause yet, just a workaround
<Begasus> printf("DEBUG: event=%p\\n", event);
<Begasus> hence the "hack" thingy :)
<Begasus> root cause thinking about the place where it fails
<dovsienko> I still do not understand how that relates with ending up in a destructor of an object with this == 0x1
<dovsienko> "\n", not "\\n" please
<Begasus> just in time :)
<Begasus> should save this later in the patchset
<phschafft> what dovsienko said about \n
<Begasus> changed just in time to: printf("DEBUG: event=%p\n", event);
<phschafft> generally errors like that are somewhere else. but forcing it to do something some way might give a better picture.
<phschafft> I also recommend using valgrind. it will detect a lot of memory access errors. even those that are within correctly mapped memory.
<Begasus> 0_°
<phschafft> so that can sometimes find the problem or at least part of the problem.
<Begasus> do we even have valgrind?
san2ban has joined #haiku
<dovsienko> address sanitiser and undefined behaviour sanitising would also be helpful, as well as static code analysis, but all of that belongs to the responsibilities of the original developers rather than a package maintainer
<dovsienko> short version: make a bug report to the upstream with the steps to reproduce and the workaround
<dovsienko> anyway, what did the latest debug print output?
<Begasus> checking, build just finished
<Begasus> beside the two lines earlier this one: DEBUG: event=0x7fa3aa294370
<dovsienko> does not look immediately invalid to me
<dovsienko> this supposedly comes from QT
<phschafft> I must say I strongy agree with dovsienko: that IS a problem of the software so it should be a problem of the authors. still a mainatiner should be helful. e.g. sometimes it's only visible on some platforms. then good communication and a helping hand can speed up a solution a lot.
<phschafft> sounds valid, yes.
<Begasus> maybe 3dEyes could have a look later if he has some time then
<Begasus> I can report upstream also
<Begasus> but first food :)
<Begasus> thanks guys!
<andreasdr[m]> Hi there.
<dovsienko> I do not understand how voiding a pointer in the source code would influence the actual code path at run time
san2ban has quit [Quit: Vision[]: i've been blurred!]
<phschafft> dovsienko: my understanding is that it marks it as used and therefore might not be optimised out.
<phschafft> but surely that is somewhere wthin the (black) magic of compilers/optimisers.
<phschafft> and if it changes what the program does then there is a problem somewhere.
<phschafft> even if it generates different code, it should not generate code that does different things (like die)
<dovsienko> phschafft: no code in that method uses the "event" argument, correct?
<phschafft> correct.
<phschafft> depending on the call convetions it might be passed as a register or via stack.
<phschafft> I mean if you really want to know what the difference is, look at the compiler output for both cases.
<dovsienko> so the Q_UNUSED line is just the tip of the iceberg, there's something else at play
<phschafft> my guess is that the stack is already corrupted and the compiler generates a little bit different boiler blade code, one triggering the segfault and the other one doesn't.
<phschafft> and exactly that is my point. it might be the tip of the stack that is slightly different. ;)
AlienSoldier has joined #haiku
<Begasus> output from build with changes done, nothing specific in there: https://bpa.st/PXZQ
<dovsienko> the warning about src/util.cpp is genuine
<Begasus> thought so, afaict not really related
<Begasus> and nothing on mainwindow.cpp
<dovsienko> Util::isMemoryLow() is empty because it does not have a preprocessor branch that would match Haiku. having an "#else #error" at the end would help to fail earlier
<dovsienko> please make it another bug report to the upstream
<Begasus[m]> k, nothing upstream about Haiku, so this hasn't been reported before
<dovsienko> your chance to introduce this feedback loop
<Begasus> hope I can write up something useful :)
<dovsienko> on Haiku Util::isMemoryLow() would need to use the same means as "About Haiku" does
<Begasus> I'll wrap up the build, add the patchset for debug info, push the changes to haikuports (build disabled) first
<dovsienko> and it would be helpful to fail at the compile time if no OS matched
* phschafft nods.
<phschafft> however I must say a function called Util::isMemoryLow() sounds a little bit... strange to begin with.
HaikuUser has joined #haiku
HaikuUser has quit []
<dovsienko> what the developers could do is run "git grep Q_OS_" and check which parts of the code require Haiku-specific blocks
<dovsienko> because there may be other bugs that do not prevent the software from starting, but could be triggered later
<phschafft> dovsienko++
<Begasus> you guys make it sound so easy :)
<Begasus[m]> do you guys mind if I link the log in the issue upstream, I guess they could grab more info than I could explain there :)
MsInput has quit [Quit: WeeChat 4.5.0]
<dovsienko> it is public anyway
<Begasus[m]> just asking, don't want to step on anyones feet :)
<phschafft> I don't know if anything I said would help, but I hink it would at least wise to include that printf() of the event pointer returned a value that seems valid.
<Begasus[m]> +1
<dovsienko> the steps to reproduce the problem are important
* phschafft nods.
<Begasus[m]> aside from building and launching there isn't much else involved?
<phschafft> maybe there is also a build log.
diver has joined #haiku
<Begasus[m]> err ... could grab that, means a clean build though :)
<phschafft> maybe something that includes the compiler flags.
<dovsienko> opinions are subject to interpretation and can be taken as starting points if that makes sense
<Begasus> reboot ..
Begasus has quit [Quit: Vision[]: i've been blurred!]
<dovsienko> Begasus[m]: maybe tell the developers how to get a VM into the required initial state
<dovsienko> and quote the compiler and QT versions
Begasus has joined #haiku
<Begasus> k, build launched
<Begasus[m]> me is bad at reporting issues when not so comfortable :)
<Begasus[m]> k, that didn't work as expected ...
<Begasus[m]> Nheko doesn't want to save it :(
<Hanicef[m]> anyone else having issues booting haiku on latest nightly? i get into the bootloader but can't boot the os due to it saying that the boot volume is not valid (i'm using qemu)
<Begasus> not me, and I'm not going to check latest nightly now :)
<Begasus> k, enough for one day
AlienSoldier has quit [Read error: Connection reset by peer]
<Begasus> closing down here, cu and thanks peeps!
Begasus has quit [Quit: Vision[]: i've been blurred!]
x10z has joined #haiku
<dovsienko> Hanicef[m]: hrev58475 has booted fine in my VirtualBox
TMM has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
TMM has joined #haiku
B2IA has quit [Quit: Vision[]: i've been blurred!]
B2IA has joined #haiku
<Hanicef[m]> <dovsienko> "Hanicef: hrev58475 has booted..." <- Hm, wonder why :/
<Hanicef[m]> Thanks for the info, though
dpirate has quit [Quit: Connection reset by beer]
dpirate has joined #haiku
MsInput has joined #haiku
zardshard has left #haiku [Error from remote client]
zardshard has joined #haiku
dqk has joined #haiku
dqk_ has quit [Ping timeout: 480 seconds]
OrngBomb is now known as OrangeBomb
nagerst has joined #haiku
<nagerst> dingleding
<scantysnax> good afternoon.
<nagerst> What version of debian do i have to use for amiwm to work? I am prepared to go VERY far.
<nagerst> no errors, but it just crashes
<scantysnax> i keep debian 10 around just in case i need to do something i can't do in haiku. never tried amwim before, however.
<nagerst> the third most used embedded UI not working is rather poor showing for debian.
<nagerst> or perhaps i am just doing it wrong
<scantysnax> some day i will switch back to freebsd.
<scantysnax> things are running well here, so i don't want to poke the bear.
<scantysnax> OS Uptime [Haiku]: 16 days 38 mins 53.45 secs
<nagerst> most banks in sweden, norway, finland, estonia and latvia uses Claes implementation of amiwm
<scantysnax> neat
<nagerst> only for the terminals where you get money though
<nagerst> in olden times they used either OS2 or NT4
<scantysnax> i always that OS2 was interesting.
<scantysnax> it was like windows, but not
<nagerst> In the way that you swallow the bad to get the good?
<nagerst> 1.2 was better than dos, but shit
<nagerst> 2.0 was NT but with a shitty filesystem
<nagerst> windows today is a good kernel (sadly not free) on a VERY old flie system.
<nagerst> old does not mean bad, but even david Cutler who wrote it said naah!
<scantysnax> i keep a copy of windows 7 on this PC *just in case*
<nagerst> 30 years later he hates his old design
<nagerst> the best filesystem designers are Giampalo and the people behind XFS.
<scantysnax> sometimes i have to do audio work on this PC, and that means windows.
<scantysnax> so my hands are kind of tied.
<nagerst> why=
<nagerst> What part does not work with jack nowadays?
<scantysnax> i use reason, a daw. it's not available for haiku obviously.
<scantysnax> that is why i need windows sometimes.
<x512[m]> nagerst: Windows kernel is bad too. It is overcomplicated and inefficient.
<nagerst> yes your particular hardware is unlikely to work with pulse or pipewire.
<scantysnax> when i feel like composing something.
<x512[m]> Linux works better on servers.
<nagerst> x512[m]: correct
<nagerst> x512[m]: but it is rather fast
<x512[m]> NT kernel have overhead caused by its overengeneering.
<nagerst> x512[m]: Any stray kernel works better on any CPU. Examples are Amiga Exec** Haiku-newos, mach7(ios) steamos(linux)
<x512[m]> And NT kernel overingenerring is meaningless because it is not used by its own outside of Windows.
<nagerst> NT kernel has some very nice advantages
<nagerst> Linux have workarounds for all of them
<nagerst> Portability?
<nagerst> no
<x512[m]> NT kernel is actually portable. There were a lot of versions for Alpha, SPARC etc. Now ARM[64] versions are available. The only problem with portability if closed source code.
deneel has quit [Quit: deneel]
<nagerst> good point it WAS portable. tru compiling any XP code for ppc or alpha
<nagerst> ill wait
<nagerst> =D
<x512[m]> I have no access to NT kernel source code.
<nagerst> Do you want it?
<x512[m]> It can't be done legally for me.
<nagerst> i have all cutler source code.
<x512[m]> I am not working in Microsoft.
<nagerst> leak 2004
jmairboeck has quit [Quit: Konversation terminated!]
<x512[m]> Leaks shouldn't be used.
<nagerst> thank you
<nagerst> the beos leaks is different for me
<nagerst> the company died
<x512[m]> BeOS leaks are even more dangerous because it may cause Haiku contributions to be illegal.
<x512[m]> BeOS assets are still owned by someone.
<nagerst> geist said it can not be illegal
<nagerst> geist is the one that wrote the beos kernel v2
HaikuUser has joined #haiku
HaikuUser has quit []
<x512[m]> Can that Geist give a legal promise that nobody will be sued for this?
<nagerst> and that is why no assetts at all is used by haiku- Mclintock is redesigned to haiky foundation
<x512[m]> BeOS assets are currently owned by Access Co. if I am correct.
<nagerst> yes
<nagerst> there is no mention of BeOS other than inspiration in the documentation
<nagerst> access corp doues do not own your feelings
diver has quit [Quit: Leaving.]
<nagerst> no i wont share
<nagerst> you are right
<nagerst> makes me sad
MsInput has quit [Quit: WeeChat 4.5.0]
<nagerst> Begasus[m]: please interject and tell me i am wrong
<nagerst> I am wrong. I admit and subit.
xet7 has quit [Remote host closed the connection]
mmu_man has quit [Ping timeout: 480 seconds]
mmu_man has joined #haiku
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
nagerst has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
<Skipp_OSX> I don't think there's much there to steal anyway except serial printers and floppy drivers
imrahil has joined #haiku
MisthaLu has quit [Ping timeout: 480 seconds]
imrahil has left #haiku [#haiku]
B2IA has quit [Quit: Vision[]: i've been blurred!]
<geist> hmm?
<Skipp_OSX> (from BeOS)
x10z has joined #haiku
<Skipp_OSX> talking about BeOS source code leak and legality of sources, probably not much to see here
B2IA has joined #haiku
diver has joined #haiku
Anarchos has quit [Quit: Vision[]: i've been blurred!]
_justin_kelly16 has quit [Ping timeout: 480 seconds]
AlienSoldier has joined #haiku
<adrian_> heh
<adrian_> cutler code, I love it
<adrian_> good clean fun times right there
Skipp_OSX has quit [Ping timeout: 480 seconds]
Skipp_OSX has joined #haiku
<Skipp_OSX> can somebody help me create a hpkg package?
vdamewood has joined #haiku
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
mmu_man has quit [Ping timeout: 480 seconds]
bbjimmy_laptop has joined #haiku
x10z has joined #haiku
mmu_man has joined #haiku