<Not-5726>
[haikuports/haikuports] threedeyes pushed 1 commit to master [+1/-2/±0] https://git.io/JPKI9
<Not-5726>
[haikuports/haikuports] threedeyes aecb863 - KWrite: bump version
ClaudioM has quit [Quit: leaving]
<Not-5726>
[haikuports/haikuports] threedeyes pushed 1 commit to master [+1/-2/±0] https://git.io/JPKZF
<Not-5726>
[haikuports/haikuports] threedeyes f5d14f7 - Kate: bump version
deksar has quit [Quit: Going offline, see ya!]
Dusan_Krehel has joined #haiku
Dusan_Krehel has left #haiku [#haiku]
AlaskanEmily has quit [Remote host closed the connection]
bitigchi has joined #haiku
bitigchi has quit []
bitigchi has joined #haiku
<Not-5726>
[haikuports/haikuports] threedeyes pushed 1 commit to master [+1/-1/±0] https://git.io/JPKXO
<Not-5726>
[haikuports/haikuports] threedeyes 8a651c0 - Marble: bump version
WishBoy has joined #haiku
<Not-5726>
[haikuports/haikuports] threedeyes pushed 1 commit to master [+1/-1/±0] https://git.io/JPK1M
<Not-5726>
[haikuports/haikuports] threedeyes 55e68f6 - Blinken: bump version
x10z has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
bitigchi has quit [Ping timeout: 480 seconds]
WishBoy has quit []
<Not-5726>
[haikuports/haikuports] threedeyes pushed 1 commit to master [+1/-1/±0] https://git.io/JPKQE
<Not-5726>
[haikuports/haikuports] threedeyes f6b4838 - libkdegames: bump version
<Not-5726>
[haikuports/haikuports] threedeyes pushed 1 commit to master [+2/-2/±0] https://git.io/JPK5I
<Not-5726>
[haikuports/haikuports] threedeyes a209cff - Konsole: bump version
bitigchi has joined #haiku
Begasus has joined #haiku
<Begasus>
g'morning peeps
Begasus_32 has joined #haiku
<Begasus>
PulkoMandy, there was a mail about extending the 80char limit in Haiku's source (iirc) does that apply to haikuports recipes too now (for DESCRIPTION)?
<PulkoMandy>
I don't know, what do people at haikuports think of it?
<Begasus>
Or maybe I should open an issue at haikuports about it ...
<PulkoMandy>
personally I have set all my editors to 100 columns now and I would prefer to not use different settings for haikuports
<Begasus>
So far nothing has been said there
<jessicah>
morning Begasus
<Begasus>
Some source_uri lines already exceed that limit, so I would be in favor of going the same way Haiku does, just my 2 cents ...
<Begasus>
evening jessicah :)
<jessicah>
may as well make it uniform across repos
<Begasus>
issue created at haikuports
<Not-5726>
[haiku/haikuwebkit] pulkomandy pushed 1 commit to haiku [+0/-0/±1] https://git.io/JP6U0
<Not-5726>
[haiku/haikuwebkit] waddlesplash 9c22566 - Gigacage: We need to commit after reserveUncommitted.
<Begasus>
on R1B3 I'm still good it seems mrentropyvcs
BrunoSpr has joined #haiku
<mrentropyvcs>
Interesting. I'm using the nightlies.
jmairboeck has quit [Ping timeout: 480 seconds]
fjmorazan has quit [Quit: fjmorazan]
fjmorazan has joined #haiku
<Begasus_32>
[system] Changes applied. Old activation state backed up in "state_2021-11-01_14:15:20"
Begasus_32 has quit [Quit: Vision[]: Gone to the dogs!]
<Not-5726>
[haikuports/haikuports] threedeyes pushed 1 commit to master [+2/-2/±0] https://git.io/JPiCU
<Not-5726>
[haikuports/haikuports] threedeyes 406545c - KWave: bump version
<Begasus>
reboot
Begasus_32 has joined #haiku
mrentropyvcs has quit [Quit: Vision[]: i've been blurred!]
<Begasus_32>
The CXX compiler identification is GNU 11.1.0
<Begasus_32>
nice, let's see it in action :)
HaikuUser has joined #haiku
HaikuUser is now known as mrentropyvcs
<mrentropyvcs>
My time stopped updating on boot up, too.
<mrentropyvcs>
It's been a day, and it's only just started.
bbjimmy_64 has quit [Quit: Vision[]: i've been blurred!]
beaglejoe has joined #haiku
BrunoSpr has quit [Quit: Vision[]: Ich wurde verwaschen!]
<beaglejoe>
mrentropyvcs: Software Updater worked for me. hrev55603 -> hrev55606 (x64_86)
<mrentropyvcs>
I'm on 55602.
<beaglejoe>
* hmm gcc didn't update still 8.3.0_2019_05_24-13
<Begasus>
only on 32bit for now beaglejoe
<beaglejoe>
Ah. Thanks Begasus
maquak has joined #haiku
<mrentropyvcs>
Ah. I had to disable the BeSly repo.
<beaglejoe>
*Updating 32bit now..
mrentropyvcs has quit [Quit: Vision[]: i've been blurred!]
<Begasus>
wanted to see if using ffmpegthumbnailer created thumbnails with gerbera, doesn't seem to ...
bbjimmy_64 has joined #haiku
HaikuUser has joined #haiku
HaikuUser is now known as mrentropyvcs
<Begasus>
I guess it worked now mrentropyvcs ? :)
<mrentropyvcs>
Begasus: Yup. Not sure why besly was holding it back, but there it is. But now I'm back in action and can start doing my NaNoWriMo thing.
<mrentropyvcs>
As long as I can resist rebooting to play Black Widow: Recharged....
Maturion has quit [Ping timeout: 480 seconds]
<Begasus>
;)
maquak has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
BrunoSpr has joined #haiku
x10z has joined #haiku
PepsiMan has joined #haiku
KapiX has joined #haiku
rennj has quit [Ping timeout: 480 seconds]
jmairboeck has joined #haiku
BrunoSpr has quit [Quit: Vision[]: Ich wurde verwaschen!]
<jmairboeck>
Begasus: sorry, I was away. I just tested 64 bit for now. Should I test 32 bit as well?
rennj has joined #haiku
<Begasus>
if you can jmairboeck ... still poking around gerbera :)
<Not-5726>
[haikuports/haikuports] Begasus pushed 1 commit to master [+1/-0/±0] https://git.io/JPipC
<Not-5726>
[haikuports/haikuports] Begasus 90393f3 - ffmpegthumbnailer, new recipe (#6355)
<Begasus>
side track :)
<Begasus>
cool, buildmaster for 32bit also uses gcc11 :)
<Begasus>
whoot!
<Begasus>
no crash when building gerbera with gcc11?
<Begasus>
time to update (64bit probably will still crash ... (?))
<Not-5726>
[haikuports/haikuports] Begasus pushed 1 commit to master [+2/-2/±0] https://git.io/JPPUf
<Not-5726>
[haikuports/haikuports] Begasus ecffcec - gerbera, bump version (#6356)
tqh has joined #haiku
<jmairboeck>
PulkoMandy: If I am not mistaken, guile1 (the legacy version) needs a patch to build with gcc >= 10. How should such a case be handled? Should it be patched when gcc11 becomes "official"? I suppose the current version (built with gcc 8) still works fine probably. See https://lists.gnu.org/archive/html/lilypond-devel/2020-06/msg00047.html
<PulkoMandy>
gcc11 is official for haikuports now, so add the patch (or tell people to stop using legacy libraries)
<jmairboeck>
ok, I'll check that when I'm finished with ghostscript
x10z has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
bitigchi has joined #haiku
WishBoy has joined #haiku
<Begasus>
win8linux[m], just commented on the PR for spout
<jmairboeck>
Begasus: it seems to be fine on 32 bit :)
<Begasus>
ok, ready to merge then?
<jmairboeck>
I think so. It is at least no regression to the old version
<Begasus>
I'm ok with leaving tesseract and leptonica in there, it just adds time to build it :)
<Begasus>
OK :)
<jmairboeck>
except ghostscript8, but that is something completely different
<Not-5726>
[haikuports/haikuports] Begasus pushed 1 commit to master [+2/-4/±0] https://git.io/JPP3z
<Not-5726>
[haikuports/haikuports] jmairboeck 4839399 - ghostscript: update to version 9.55.0, remove old versions (#6337)
<jmairboeck>
thanks Begasus :)
<Begasus>
Thank you :)
<jmairboeck>
what about fig2dev?
<Not-5726>
[haikuports/haikuports] Begasus pushed 1 commit to master [+1/-0/±0] https://git.io/JPPsI
<Not-5726>
[haikuports/haikuports] jmairboeck 4de46a1 - new recipe: fig2dev (#6283)
<Begasus>
done :)
<jmairboeck>
thanks :)
<beaglejoe>
bbl
beaglejoe has quit [Quit: Leaving]
WishBoy has quit []
topfuel has joined #haiku
maquak has joined #haiku
<Begasus>
only blender left in the report.txt now, not touching that one ...
BrunoSpr has joined #haiku
<Begasus>
biab
maquak has quit [Remote host closed the connection]
countryboy has left #haiku [#haiku]
<win8linux[m]>
Begasus: Can confirm that your suggestion for Spout works, but there are bugs which render the game unplayable.
<win8linux[m]>
Some later investigation will be needed to figure out what's needed to fix those.
<win8linux[m]>
In the meantime, could you please look at the Race Into Space PR if your schedule permits?
<win8linux[m]>
It's just been sitting there for weeks now.
<nekobot>
[haiku/haiku] 3457006a9adc - WebPositive: fix linking with static libraries
<Not-5726>
[haikuports/haikuports] pulkomandy pushed 1 commit to master [+1/-2/±0] https://git.io/JPPuP
<Not-5726>
[haikuports/haikuports] pulkomandy b59cc4c - HaikuWebKit: new version.
<jmairboeck>
PulkoMandy: I tried rebuilding guile1 and it seems to work okay. I don't have Lilypond in my 32 bit VM right now, so I can't really test with that. But I suppose there is something different between Haiku and Linux. So I will keep the recipe as is. If some problems come up later, we can still patch it.
<Begasus>
win8linux[m], the game builds, but yes, playing it is a pain, doesn't really seem to respond on the arrow keys (or atleast not the one moving upwards)
<Begasus>
checking out openblas atm on 32bit
<julicenri>
Yeah, that's not intended behaviour at all.
countryboy has joined #haiku
<Begasus>
I'll mark the recipe as wip julicenri win8linux[m] :)
PepsiMan has quit [Quit: Vision[]: i've been blurred!]
<julicenri>
Alright then, just uploaded the Spout recipe fixes to the remote.
bitigchi_ has joined #haiku
tqh has quit [Remote host closed the connection]
tqh has joined #haiku
bitigchi has quit [Ping timeout: 480 seconds]
<Begasus>
win8linux[m], lib:libpng$secondaryArchSuffix that should be libpng16
<Begasus>
(for raceintospace)
<Begasus>
no specific boost library specified also there?
<Begasus>
ps, didn't check the build (yet)
<Not-5726>
[haikuports/haikuports] Begasus pushed 1 commit to master [+1/-1/±0] https://git.io/JPPwn
tqh has quit [Read error: Connection reset by peer]
tqh has joined #haiku
<Not-5726>
[haikuports/haikuports] Begasus pushed 1 commit to master [+1/-0/±0] https://git.io/JPPoS
<Not-5726>
[haikuports/haikuports] hoanga 64338b8 - abcl: new recipe (#6343)
<Begasus>
hi Vidrep_64
<Vidrep_64>
Hi there Begasus
<Vidrep_64>
It looks like we've got a new HaikuWebKit and WebPositive to try out tomorrow
<Begasus>
yeah, buildmaster still running on it atm, probably will take some time :)
euandreh has quit [Remote host closed the connection]
<Vidrep_64>
It's an improvement in several ways from what we have currently. There's still a few outstanding bugs yet to be resolved.
euandreh has joined #haiku
<Vidrep_64>
Thanks to PulkoMandy, nephele, waddlesplash and whoever else contributed
<PulkoMandy>
thanks for your help with testing and bugreports
<Vidrep_64>
I contribute complaints ;)
<Vidrep_64>
PulkoMandy, are we planning on resolving the outstanding bugs in our current webkit, or will the next step be to migrate to webkit2?
<mmu_man>
ok, let's see we close the jitsi channel
<mmu_man>
I thought it'd get more use this week-end but eh
<mmu_man>
at least we had some interested chats
<mmu_man>
thanks jessicah for staying until now :)
<PulkoMandy>
Vidrep_64, I do not really work on anything at the moment, I just update to latest webkit versions and make sure things still compile
<PulkoMandy>
I did not manage to get webkit2 compiling in the last few months, may try that again later
<PulkoMandy>
for other things I hope people will join and work on fixing bugs. Nephele and Waddlesplash are doing some of it, which is nice
euandreh has quit [Ping timeout: 480 seconds]
<Vidrep_64>
PulkoMandy, did you have a chance to review your comments about a memory leak in StreamRadio?
<PulkoMandy>
not yet
<Vidrep_64>
OK. Just asking
<PulkoMandy>
my computer was busy building webkit this weekend and ran out of disk space multiple times. So now I'm making backups and I will be trying to use the BFS resizer
<nephele>
My next points of interests for webkit are: fix the icon database crash; fix button drawing with controllook; fix svg drawing (not sure what the problem is there yet)
<PulkoMandy>
then I can get back to work on other things
<nephele>
So atleast for me webkit2 is not /that/ high on the priority list, I do certainly want that api (mainly to use in other haiku apps) but i'd rather work on removing crashes, removing deadlocks, and fixing some rendering stuff first
<nephele>
(that should also all be work that carries over to webkit2 regardless)
<Vidrep_64>
nephele, I had a look this morning at my open tickets for WebPositive. Only 3 remaining - icon database crash, Google Maps (ports), and download thread continues running after Web+ killed
_____ore has quit [Ping timeout: 480 seconds]
BrunoSpr has quit [Quit: Vision[]: Ich wurde verwaschen!]
<nephele>
Probably all haikuwebkit... not sure how a thread would continue if you killed it? you mean asking it to quit?
<Vidrep_64>
At one time I had a dozen open tickets, so progress is being made
<nephele>
I'm not going to investigate google maps issues though, or any $google issues for that matter, I don't want to bother going through their privacy eula stuff
<Not-5726>
[haikuports/haikuports] Begasus pushed 1 commit to master [+1/-0/±0] https://git.io/JPPDE
<nephele>
Yes Begasus, but the forum header is an svg, and it disapears if you scroll down
<nephele>
if you open a svg and resize the window smaller than the svg and scroll down it will also disapear, either in part or fully
<Begasus>
ah, will have a look into that when I get the time
<nephele>
andrewrk: I'm curious a bit, as i understand it zig does the C abi, but most of the interesting stuff in Haiku is all C++ abi, is interopability possible there?
<andrewrk>
oh interesting, the OS ABI itself is based on the C++ ABI?
<andrewrk>
I thought the C++ ABI wasn't consistent between different compilers though
<nephele>
We do the posix stuff for C support, but our "kits" (as in system components) are all C++, like the interface kit for the gui, the application kit for connecting to the app_server and such
<andrewrk>
I see
<nephele>
Basically all our gui applications are written in C++ against these libraries
<andrewrk>
I'm afraid the best we can do on the zig side of things is the typical .h/.cpp C wrapper for C++ APIs
<andrewrk>
with the upshot at least being that zig is capable of compiling the .cpp files
<nephele>
(we have a stable abi for the 32bit version to match BeOS, with gcc 2.95 or so, gcc3 changed the abi, we use a modern gcc for the 64bit version though)
<nephele>
andrewrk: ah, no direct integration to link against dynamic libraries or so?
<nephele>
I'm not too good with the whole ABI internals anyhow, but our symbols should be mostly stable to link against, i think
<andrewrk>
I mean, zig has no problem with dynamic libraries, but C++ ABI is not something any language supports except for C++ and D (kind of)
countryboy has joined #haiku
<andrewrk>
and not even different C++ compilers support compatible C++ ABIs
<waddlesplash>
well ... outside of gcc2 wer'e just on the standard itanium abi
<waddlesplash>
which everyone except MSVC uses
<waddlesplash>
andrewrk: yeah, you can CC me I suppose. not sure who else would review such things, plenty of us are more than knowledgeable enough but we're generally busy :-p
<andrewrk>
appreciate it, as a generally busy person myself :)
<waddlesplash>
but yes, the C++ ABI instability is really not fun. if we had all the money in the world we'd probably design our own programming language or something
<waddlesplash>
Zig seems cool but you don't seem to be about the RAII that we love so much :-p
<andrewrk>
if you did that, wouldn't you run into the problem that haiku ABI is C++ based? wouldn't you want to make the ABI not tied to C++ so that you had that option possible in the future?
<waddlesplash>
I would, yeah
<waddlesplash>
swift has some interesting things going on in terms of ABI stability apparently
<waddlesplash>
the worst part of it is mostly just virtual inheritance
<waddlesplash>
and object sizing, but that's less of an issue honestly. if vtables were just built at RTLD time instead of compile time, that would save a massive amount of headache
<andrewrk>
interesting
<andrewrk>
I wouldn't have guessed that you wanted that
<waddlesplash>
what?
<waddlesplash>
virtual inheritance?
<andrewrk>
vtables built at runtime instead of compile-time
<waddlesplash>
well, it would solve the problem that changing vtable size breaks aBI
<waddlesplash>
there isn't really any other way to do it
<andrewrk>
I guess with virtual inheritance it makes sense- I code with a mental model that excludes the possibility of virtual inheritance
<waddlesplash>
and we code with mental models that assume virtual inheritance everywhere
<andrewrk>
yeah. interesting!
<waddlesplash>
a large portion of the kernel is virtually inherited actually
<waddlesplash>
massively improves code reuse and makes code *much* easier to read and understand
<waddlesplash>
virtual memory code in most kernels is a near-incomprehensible maze of branches, and to make things worse it's often poorly commented, though the BSDs do somewhat better here
<andrewrk>
isn't a virtual function call equivalent to a branch?
<waddlesplash>
yes
countryboy has left #haiku [#haiku]
<andrewrk>
anyway I won't come in here trying to convince anyone to switch languages or coding styles :)
<waddlesplash>
I mean the code has those branches explicitly
<waddlesplash>
in Haiku ... C++ virtual inheritance and RAII objects mean that the code can much more clearly communicate what it is trying to do at the same time it is actually doing it
<andrewrk>
I do think that there is a misunderstanding that we don't care about RAII principles in zig though. I mean yeah you have to manually `defer cleanup();` but the coding model is the same
bitigchi has quit [Ping timeout: 480 seconds]
<waddlesplash>
but what happens if you decide you don't want to clean something up?
<waddlesplash>
for instance we have a very common paradigm in Haiku:
<waddlesplash>
if (error) return error; /* whatever is deleted */
<waddlesplash>
whateverDeleter.Detach(); return ENOERR; /* success, we want to keep whatever */
<waddlesplash>
you can't undo a "defer cleanup", but you can tell a RAII object that you actually want to keep the thing
<waddlesplash>
and yes, moving unique_ptrs would handle this case ... but only in code we manage. we can use this paradigm even when e.g. implementing POSIX functions that must return raw pointers
<waddlesplash>
similarly, in core kernel code, we may need to lock and unlock some lock a lot of times, so we don't want to just "defer unlock", we want a MutexLocker object that will keep track of whether the thing is locked or unlocked, and act appropriately on "return error"
<nephele>
comptime seems like a major advantage zig has, though :)
<andrewrk>
if you show me a simple example code I can show equivalent zig if you're curious
<waddlesplash>
andrewrk: well, the above template for instance allows me to specify an arbitrary function as a template argument, as well as what type it accepts and what return type it has
<waddlesplash>
and now I've got a new "deleter" type that I can liberally sprinkle through the code
<andrewrk>
yeah that is directly translateable to zig, although you might make some people squint their eyes and accuse you of unnecessary abstraction
<waddlesplash>
well, we have quite a lot of these typedefs, because we can and do create them when interacting with random C libraries :)
<andrewrk>
I'm not sure what the types or usage of CObjectDeleter is, but if I guessed correctly it would look something like this: https://clbin.com/EeKTz
<waddlesplash>
yep, that's about how it works
<waddlesplash>
some more nuance, but not a lot
<waddlesplash>
well, if zig had virtual inheritance or an equivalent thereof, maybe it'd be something we'd consider :-p
<waddlesplash>
or maybe if you add support for language extensions or something, someone will invent "Object Zig" :D
<andrewrk>
hahaha well I'm only here to ask for help reviewing a haiku PR, not to try to convince anyone to use zig :)
<nephele>
I would be interested in seeing how well zig could cache haikus build though
<waddlesplash>
nephele: we need less build caching, not more, honestly
<nephele>
though, that implies clang can build haiku fully before... heh
<andrewrk>
waddlesplash, you will probably be horrified to know that in zig we build our vtables manually, without any language feature
<nephele>
waddlesplash: what makes you say that?
* waddlesplash
really needs to get back to "ham"
<Begasus>
OK, guideline for the recipes has been updated to use "100" characters now
<waddlesplash>
andrewrk: if by "manually" you mean you have structs with pointers in them ... that's what most C programs do, yes. pretty primitive :-p
<waddlesplash>
FreeBSD even has a "kobj" system with some vague concepts of inheritance
<andrewrk>
well if you ever massage the OS ABI to be more friendly to your own potential future language, that would help us over in zig land too (and pretty much every other language that isn't C++)
<waddlesplash>
there are projects out there that bind Haiku's APIs already I think
<waddlesplash>
there's Python bindings at least, or used to be
<nephele>
Don't those also use the C abi?
<waddlesplash>
yes, not sure how they work
<nephele>
(so basically the same header wrapper stuff andrewrk mentioned above to work, i think)
<waddlesplash>
Haiku is much more friendly to bindings than Qt is, I will say that
<waddlesplash>
because you can just intercept all messages and call it a day
<waddlesplash>
well, for the most part anyway.
<waddlesplash>
the signal/slot system is also a huge pain to bind, at least, and we don't have that at all
<nephele>
andrewrk: Well, even if we standardize the abi for our kits nicer, i doubt C++ would follow, probably we just drift even further from them than we already are
<waddlesplash>
yeah
<waddlesplash>
already our conception of what "C++" is has wildly diverged from basically anyone else out there
<andrewrk>
oh interesting, I did not know there was already a difference in C++ ABI and Haiku ABI
<waddlesplash>
no, there isn't, we use the standard ABIs
<waddlesplash>
this is referring to our use of language features and how we do (or don't) employ them
<andrewrk>
ahh I see. yeah that sounds typical, every C++ codebase uses its own subset
<nephele>
More like we implement our own subset for a lot of stuff
<waddlesplash>
andrewrk: now the real question is, can you implement BStackOrHeapArray in Zig :-p
<nephele>
we also use gcc2 still, but that is mostly because of it's abi, and while in theory LLVM could probably be made to emit that ABI nobody has investigated that seriously or made a patch for it :g
<waddlesplash>
avoid all alloca() stack overflows with this one cool trick! linux developers hate him
<Not-5726>
[haikuports/haikuports] korli pushed 1 commit to master [+0/-0/±1] https://git.io/JPPFI
<Not-5726>
[haikuports/haikuports] volo-droid 2e7801a - coreutils: show creation time in stat and ls output (#6326)
<nephele>
(That's only for ABI compatibility for applications and addons written for BeOS though)
<nephele>
I should patch the build though that it also uses -fpic by default
<nephele>
for clang, that is
<waddlesplash>
andrewrk: obviously not in plain old C++
<waddlesplash>
I think we may have some sort of trickery by which we do something similar, though it probably involves macros, and is only used in like one place if at all
<waddlesplash>
andrewrk: I would be interested in performance numbers on that, though. oftentimes struct padding is extremely critical to preformance, partially so on x86 but probably moreso on arm
<waddlesplash>
and by extracting things into arrays you may have to add padding anyway to get dword alignment
<waddlesplash>
if the struct is entirely power-of-two-sized types then maybe not
countryboy has joined #haiku
<andrewrk>
in the use cases for this, we have a 1-byte enum and a couple of 4-byte integers, so it saves 3 bytes per element
<andrewrk>
I have some perf numbers, one sec
<andrewrk>
I'm actually doing a talk on this at handmade seattle in a couple weeks :)
<Begasus>
(7e09091, 3219b08) Fixed a few other (less severe though) HTML injection discovered since #789 - thanks to nephele for spotting those.
<Begasus>
nice one nephele :)
<nephele>
Aparently you can put html into matrix user id's, who would have thought :D
<Begasus>
should probably update libquotient also for it?
<nephele>
no, why?
<Begasus>
current 0.6.6 in haikuports, new one is 0.6.11
<Begasus>
Prebuilt binaries at GitHub Releases use libQuotient 0.6.11 with its own set of fixes.
HaikuUser has joined #haiku
HaikuUser has quit []
<nephele>
anything interesting there in those fixes?
<Begasus>
Just after 0.6.10 was out another nasty bug has been discovered, preventing from proper usage of URLs with an unescaped double-hash (##), which may come when a matrix.to URL is generated for a bridged IRC channel (and it even got a very nice number #512). Aside from this fix, quotest will not return with success code 0 any more if it failed to even start testing for some reason (#496).
<Begasus>
from the release page
<nephele>
ah, yes that makes sense...
<nephele>
should update, then
<Begasus>
first the lib then the app :)
<nephele>
Trying to figure out why my patch doesn't work anymore, only a minor version increase .-.
<nephele>
Huh, i don't think it applied it, what's the syntax to add a patchset? :g
<Begasus>
use the force (in this case repology) ;)
<nephele>
bah
<nephele>
I'm no web interface user :P
<Begasus>
building openblas atm extrowerk_ , should have thought it would take a while to build (last time was a while ago) ;)
<jmairboeck>
nephele, Begasus: you can also just use the shell: cd */libquotient
KapiX has quit [Quit: KapiX]
<jmairboeck>
assuming there is only 1 package with that name :)
<nephele>
jmairboeck: I could have, if i had known the directory was named libquotient
<Begasus>
ah nice one jmairboeck , didn't know that one :)
<extrowerk_>
nephele: ```inrecipe quotient``` could help
<nephele>
I used find, it didn't find it
<nephele>
slow tool
<nephele>
not that nice
<extrowerk_>
not many recipe contains this word i assume.
<extrowerk_>
nephele: if you adjust it to use ag instead of grep it becomes pretty quick.
<nephele>
Although I doubt your shell globbing would have resolved that much faster than find... although it would only be the equivalent of a depth of 1 for find
<Begasus>
we still follow gentoo's layout, and it's put/called there
<nephele>
extrowerk_ ?
<extrowerk_>
nephele: it gives me the recipe name in 1 sec.
<nephele>
ripgrep is somewhat faster on linux, I don't think anyone did any quantitive tests though
<Begasus>
that one is an easy one to keep my fork uptodate :)
<nephele>
but the main reason i use it is simple: I found this before i ever used the silversearcher, and it was nicer than egrep at the time :)
<nephele>
ever know of, to be more precise
<nephele>
so dunno how good or bad the silversearcher is, I haven't used it
<Begasus>
used it a few times, fast ... :)
x10z has joined #haiku
<nephele>
cmd:nvim -> not found cmd:nvi -> not found cmd:vi
<nephele>
eh
<extrowerk_>
here is my profile with every small bits i mostly use. Nephele, you could use query to find recipes, see "searchrecipe" alias: http://0x0.st/-n38.txt
<nephele>
cmd:vi installs vim .-.
<nephele>
I'm not a fan of such stuff, i just write scripts if i want to have a specific function, much easier to copy to my laptop
<nephele>
Would be cool to port cmd:ee "Easy edit" from FreeBSD... nano has such a bad UI
<Begasus>
extrowerk_, coreutils not working for you?
<nephele>
extrowerk_: you exposed a bug in Renga, btw .-.
<nephele>
nickname matching for highlights is case sensitive, should be not so, i think
<nephele>
Okay Begasus
<nephele>
I have the commits on git, want me to help you pull them properly?
<Begasus>
I guess I can use an extra tab in Terminal for this?
<LegendaryAmerican>
hmmm still too early for bt mice.
<LegendaryAmerican>
omoton bm003 bt3/bt5
<LegendaryAmerican>
shows up however as bt 3.0 mouse.
<LegendaryAmerican>
just doesnt work.
<nephele>
Bluetooth <*> doesn't work in haiku
<LegendaryAmerican>
well.. I chose a mouse that was a little flexable. and thats a tip for anyone testing it.
<nephele>
If you have something like a unifying receiver or something, you can use that for logitech mouse, anything where the receiver pretends to be a normal usb mouse, but connecting anything to bluetooth by pairing it won't do anything
<nephele>
like, we support bluetooth pairing... nothing else :)
<LegendaryAmerican>
well a mouse would be the first one theyd want to try.. and I found a dual version, might be their best shot.
<nephele>
whose?
<LegendaryAmerican>
thats cool. I have a logitech for now.
<LegendaryAmerican>
omoton bm300. it is a bt v 3 and bt v 5 mouse. It does show up, and pairs, but nothing else.
<extrowerk_>
<nephele> don't switch branches while you are building your recipe though <- it should not make any trouble as hp creates a copy about the recipe before it starts to do the build. at least i never had problem with it.
<LegendaryAmerican>
amazon sells the omoton cheap.
<nephele>
git branch switching could always remove random files you need, like .rdef, patches etc.
<extrowerk_>
LegendaryAmerican: only pairing is implemented. _Nothing_ else.
<nephele>
LegendayAmerican: again, to be expected, I don't think /anything/ works with bluetooth
<LegendaryAmerican>
tips for getting bluetooth mice to work(when the stack is mature enough!) go for a dual version bt mouse.
<LegendaryAmerican>
it might increase the likelyhood of making a hit.
<extrowerk_>
Ow, nice, calibre needs libstreamer now, which is an unversioned lib, provided in an unversioned zip. It doesn't even have a copyright or any info about the author, license. Nice. And it builds only a static lib. Heh.
<LegendaryAmerican>
itll matter later bigtime. the macbook pros going way back supports bt 5 mice
<extrowerk_>
LegendaryAmerican: i stay with my Logitech MX Anywhere 3, which works very well with the supplied receiver. But it is $$$
<extrowerk_>
LegendaryAmerican: do you mean a bt mouse doesn't supports some mackbook variants?
<extrowerk_>
maybe you holding it wrong...
<LegendaryAmerican>
nonon..bt mouse works fine either as bt 3 or bt 5, back to 2015 macbooks.
<nephele>
with 5G MMwave antenas you really will be holding it wrong, unless you get double size phone with double the battery just to get three of those stupid antenas in there
<nephele>
Logitech Unifying receiver doesn't use bluetooth to talk with the mouse afaik
<LegendaryAmerican>
as apple gets more into the m1 procs over the years.. youll start to see support for duos fade slowly
<nephele>
(also Bluetooth low energy is something different to Bluetooth, go figure)
<LegendaryAmerican>
apple supports their computers a stupid long time.
<LegendaryAmerican>
you look at it also.. macs at least, used to be higher specced than their pc counterparts. my 2015 macbook pro.. 3ghz duo.. 8 g ram.
<LegendaryAmerican>
stock.
<extrowerk_>
"We have not bothered to insert the licensing arrangement into the text of the Snowball software."
<extrowerk_>
Thank You!
<extrowerk_>
LegendaryAmerican: macs at least, used to be higher priced than their pc counterparts. stock.
<LegendaryAmerican>
yes. but. their support level for new was second to none.
<nephele>
Atleast they don't have windows on it
<LegendaryAmerican>
and I hate a bloody mac you know.
<LegendaryAmerican>
not gonna trash them. haiku doesnt make machines afaik.
<LegendaryAmerican>
and we all heard mac support is so good for new because the rich dummies who buy them
<nephele>
They also have good resale value, i definetely understand why people buy those pcs
<nephele>
especially when the alternative is the random whims of microsoft randomly making a new UI again
<LegendaryAmerican>
yeah. pcs are cheaper to get into. but. unlike macs.. pcs take a lot more work from haiku side to get into.
<Begasus>
nephele, pulling the changes from you works, but I guess that when I create a PR I still need to add you as co-authored?
<nephele>
Begasus: why? did you remove the commits?
<LegendaryAmerican>
the macbook pros.. oh gee. we support duo 3ghz. 8 g ram. not so much bt 5 yet.
<nephele>
I am not that familiar with github, but why would you need to alter the commits, can't you just push the commits you pulled from me to the pr?
<Begasus_32>
spam ....
<LegendaryAmerican>
all day long however. pretty sure tho haiku handles everything else on a macbook pro..
<Begasus>
that was with "git pull nephele review-quaternion-95"
<nephele>
That means you just merged that branch with whatever you were on
<nephele>
which aparently /wasnt/ the upstream head
<nephele>
git fetch origin should catch you up, but only the two top commits are from me
<jmairboeck>
it fast-forwarded, there was no merge
<jmairboeck>
you can simply push it to your github remote now
<Begasus>
stupid .... (me) :)
<jmairboeck>
I think github does this correctly when creating the PR then
<nephele>
Why do you have to push it to another remote... is there no straightforward way to make a github pr? it sounds way too convoluted
<jmairboeck>
you create the PR in the web interface after having it pushed to github. Begasus could theoretically also skip the whole PR stuff and push it directly to upstream haikuports, but probably wouldn't want to do that
<nephele>
You have to use a webrowser just to create the thing .-. jeez
<jmairboeck>
or the github CLI
<Begasus>
I got the basics covered, but this one not yet ;)
x10z has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<jmairboeck>
if you have pulled it from nephele, you have it in your local clone, you can just use your normal workflow like always
<Begasus>
I pushed it to my fork, but can't seem to find it there? ;)
<jmairboeck>
or maybe it is just a setting of the haikuports repo
<nephele>
I guess it makes sense with those wierd commit chains many prs on github accumulate
<nephele>
I much prefer the gerrit way where you see exactly what you will be getting finally, but you also see the history of how the change got to be how it is now
<jmairboeck>
yes, but after the "merge" only 1 combined commit ends up in master
<nephele>
yes, but that isn't what it put in the repo
<jmairboeck>
I think you can select what it should do when merging the PR, but I am not sure. I have never done this yet myself, at least not in github
<Begasus>
anyway, thanks for stepping in jmairboeck :)
<jmairboeck>
no problem Begasus
<Begasus>
at least one step closer there nephele ;)
<extrowerk_>
wasnt this possible before? "ls > /dev/tt/p1"
<extrowerk_>
to send outputs between terminals.
<extrowerk_>
it returns "bash: /dev/tt/p1: Operation not allowed" now
bitigchi has quit [Ping timeout: 480 seconds]
<nephele>
Begasus: almost done, heh
<nephele>
seems done now
<Begasus>
yep :)
<Begasus>
for 32bit we'll have to wait for the webkit to finish
bitigchi has joined #haiku
nhtello has quit [Quit: Vision[]: i've been blurred!]
<Begasus>
k, fork back on par
phaedros_ has joined #haiku
jmairboeck has quit [Quit: Konversation terminated!]
<Begasus>
heading down here, will leave the build for openblast running, will see tomorrow when it's finished ...
<Begasus>
g'night peeps
Begasus has quit [Quit: Ik ga weg]
<nagerst>
Begasus: are you the legend that ported scummvm?
<nagerst>
i know you ported lots of games for zeta
<nagerst>
thank you
<nagerst>
Has there been a 3dmix app on haiku?
gouchi has quit [Remote host closed the connection]
barctor has quit [Quit: Vision[]: i've been blurred!]
bitigchi has quit [Quit: Lost terminal]
<nagerst>
I wish i had a BeOS system with suport for more cpu than 2147mhz and 768mb ram
<nagerst>
I am building a ultimate BeOS system
<mmu_man>
the 2GHz bug I had to fix in ZETA…
<mmu_man>
it's just a bad cast somewhere in the calibration, might actually be possible to binary patch the R5 kernel
<nagerst>
i know, but i can not buy zeta any more, and it is limited to the device you bought it for
<nagerst>
I have a legal copy of both 1,1.1,1.2,1.21 and 1.5*
<nagerst>
i do not hav ethe machine signed to it anyt more
x10z has joined #haiku
<nagerst>
I do not even know if it is legal at all any more
<nagerst>
mmu_man: please dont. Let it be
<nagerst>
mr francois, may i ask you about some stilly questions?
<nagerst>
I bought a dreamblaster from belgium. Does haiku support it? BeOS does.
<nagerst>
granted i planted the dreamblaster on a aureal vortex 2
<nagerst>
dang it sounds well
<PulkoMandy>
it should work, I think, but not sure anyone has done much testing
<nagerst>
PulkoMandy: It is firmware based so i should play midi
<PulkoMandy>
I have one here but not sure I have a PCI soundcard with a waveblaster port so I could plug it in to one of my haiku machines
<nagerst>
or to be frank override
tqh has quit [Quit: Leaving]
<PulkoMandy>
yes, I know :) but we need support for the MPU401 serial port to communicate with it
<nagerst>
oh ok
<nagerst>
PulkoMandy: Are you in the EU? I could send you a pci card
<PulkoMandy>
yes, but I have enough piles of hardware as it is now
<nagerst>
ok. sorry
<PulkoMandy>
and probably would not have time to look into it for a few years
<nagerst>
Not that it is on my haiku wishlist either (just though it might be simpler). My wishlist is "hdmi sound" and 4k reolutionö.
<nagerst>
resolution*
<jessicah>
hdmi sound is not going to happen anytime soon, from what I've observed of the linux driver
<nagerst>
neither is probable
<PulkoMandy>
I would be happy enough with 1080p over DVI or VGA on my laptop :/
<PulkoMandy>
what is your videocard?
<jessicah>
and I'm pretty sure it also requires the hda driver communicating with the graphics driver
<jessicah>
not to mention the driver specifics for HDMI in the linux driver
<jessicah>
which is bloody huge
<PulkoMandy>
yes, it needs drivercollaboration
<nagerst>
PulkoMandy: I am using Haiku on a 142ft projector.
<jessicah>
it's been something I've looked into in the past, because I'd love it working too, I use digital audio everywhere in my setup :-/
<nagerst>
ice hockey forever!
<jessicah>
I'd be happy with toslink as well, but the realtek codec in my nuc needs a pretty hefty quirks implementation from what I've seen of the linux driver
extrowerk_ has quit [Quit: Vision[]: i've been blurred!]
<jessicah>
I'm also thinking a thunderbolt dock with toslink might be another way to go for dev if can find one with an hda codec that doesn't require much in terms of quirks
<nagerst>
thunderbolt carry lectical current.
<jessicah>
uh, both hdmi + toslink uses the same spdif signal, as I understand it
<nagerst>
electrical*
aeryndunham has joined #haiku
<jessicah>
it should look like a PCIe bridge to haiku, it's still digital all the way
<nagerst>
if i need to convert from 3.5 to anything else....
<nagerst>
it should like like a "file"
<nagerst>
Every operation and action shoud be a file