AlienSoldier has quit [Quit: Vision[]: i've been blurred!]
AlienSoldier has joined #haiku
<AlienSoldier>
just noticed the "all disk" in tracker find only give me my boot partition result now. I need to expressively ask for a particular volume to be able to find file on it. This is a MAJOR problem.
<AlienSoldier>
humm, no it does also show another partition, but not my 4.
<AlienSoldier>
perhaps the other deindexed?
<AlienSoldier>
let see if this problem still exist after a reboot
AlienSoldier has quit [Quit: Vision[]: i've been blurred!]
AlienSoldier has joined #haiku
<AlienSoldier>
nope, still 2 partition that don't give result when "all disks" is selected
<Begasus>
hi AlienSoldier, only use "previously mounted ..." here (and in VM), so can't check
<Begasus>
nightly thing?
janking has joined #haiku
<AlienSoldier>
Begasus can't say, it might be like that since a while
<janking>
hello
<AlienSoldier>
unmounting and remounting does not help
<AlienSoldier>
and those are definitively BFS
HaikuUser has joined #haiku
<Begasus>
hi janking
HaikuUser has quit []
HaikuUser has joined #haiku
<AlienSoldier>
i can't seem to be able to find a setting that would "hide" partition from global search
HaikuUser has quit []
fcn has joined #haiku
<AlienSoldier>
what does template/default does in the file pref panel? it crash for me
<janking>
hi Begasus
hightower3 has joined #haiku
hightower2 has quit [Ping timeout: 480 seconds]
<AlienSoldier>
humm my usb external drive query file correctly.
<AlienSoldier>
i guess i could try to redo those partition but they are full of files
<MonniTheCat>
Would be useful if there was a tool to rebuild the index on existing filesystem...
<AlienSoldier>
there is reindex but it does not to accomplish anything here
<AlienSoldier>
not seem
<AlienSoldier>
and there is that crash that i don't like:
<AlienSoldier>
when i use the default template in the seach file panel bar.
<AlienSoldier>
*search
<MonniTheCat>
I have to do soft reboot at least once before Haiku even sees all my disks... It might have some untested code paths...
<MonniTheCat>
Top most frame in stack trace is usually least interesting... I usually check the next few frames, those tell a lot more why things crash...
<MonniTheCat>
Segment violation can either mean assertion failed or there was null pointer deference...
mmu_man has joined #haiku
<AlienSoldier>
all my 4 partition on my external disk have there files seen by the global find.
<AlienSoldier>
i have space on one of them, i think i will move one of the 2 that have problems there and redo it in bfs to see if it come back to normal.
<MonniTheCat>
Sometimes recreating the file system and copying files back is the only way to fix hidden corruption... Tools for BFS still don't detect all integrity problems...
<nekobot>
[haiku/haiku] 8ada0c0e7c64 - Update translations from Pootle
<AlienSoldier>
i guess the problem is with those partition and not my haiku install. I still find weird that 2 are borked that said.
<AlienSoldier>
perhaps they got too many file at one point and something wrote outside it's variable.
<MonniTheCat>
I've had issues in folders that are mounted using packagefs located on BFS partitions... Ghost files or files that are not found in one Terminal window, but show up in another... Most likely the cached data in memory is out of sync...
<AlienSoldier>
once my copy is done i will try to see if those files are found on their new home.
<AlienSoldier>
i don't think it is a per file problem.
<AlienSoldier>
perhaps a too big fragmentation problem but i don't get why doing a seach just on that partition would find the file fine.
<AlienSoldier>
*partition find the file find.
<MonniTheCat>
Me neither... I think it's more VFS problem...
<AlienSoldier>
**fine
<MonniTheCat>
I think the issue is in the part where it build the list of partitions to query...
<AlienSoldier>
but that would be on my boot partition?
<AlienSoldier>
humm one thing i may try also is to just rename the partition. I guess that will create a new path at least.
<MonniTheCat>
I get funny stuff if I have two Haiku partitions on same PC... Mount point names keep changing....
<AlienSoldier>
rename, unmount, remount, retry
<AlienSoldier>
the "include trash" in the find panel is not kept checked if selected between use of that panel. Is that normal behavior?
<AlienSoldier>
this is in "options"
<AlienSoldier>
yup, my files are seen OK once moved to my external partition.
<AlienSoldier>
just renaming and remounting does not work. I guess this is just a label change and the partition name stay the same.
<AlienSoldier>
it kept the previous name in the tracker mounting selector
<AlienSoldier>
nope, still don't see the file after making a NEW partition...
<AlienSoldier>
Will retry with a different name this time
<AlienSoldier>
MonniTheCat it work now. I needed to not just recreate the partition but to delete it first. I guess the same will work on my other partition. Now that this one is empty i will be able to move the full one on that.
<AlienSoldier>
Some of the thing i am happy to have integrated in my mind is to always do partition of equal size, that way i always know a partition will fit to another.
<AlienSoldier>
that will take time to move :)
<MonniTheCat>
AlienSoldier: That almost suggests that there was something wrong with the partition table entry...
<AlienSoldier>
humm, perhaps it did something that fixed the other partition at the same time, let me check
<AlienSoldier>
no other partition is still borked
mmu_man has quit [Ping timeout: 480 seconds]
<AlienSoldier>
that was redefining saying not able to find a good movie to watch. I was not even seeing the movies with my query :)
<AlienSoldier>
for big move like that, the tracker could introduce wait period to not make the drive get hot :)
<AlienSoldier>
it take a lot more cpu to do that copy than i was expecting, that is no amiga DMA transfer!
<AlienSoldier>
the thread that hit the cpu is called "move task"
<AlienSoldier>
about 75% of one of the "core"
hightower2 has joined #haiku
<AlienSoldier>
now cpu is back to normal
<AlienSoldier>
and it still transfer.
<Begasus>
cp/mv is still an issue I guess with Tracker (on larger volume)
janking has quit [Quit: Vision[]: i've been blurred!]
hightower3 has quit [Ping timeout: 480 seconds]
<AlienSoldier>
now it move small files, will try to notice if it get hit once it reach big file again.
mmu_man has joined #haiku
janking has joined #haiku
HaikuUser has joined #haiku
HaikuUser has quit []
<AlienSoldier>
Begasus it indeed take more cpu for big files but only thing 1G or up and it get worst the longer that file is moved. It is like if it was not able to keep up with the moving and it activate some kind of cache mechanism, the ram usage does not increase but i think the file cache ram is not shown in the replicant.
<Begasus>
seen it also when moving larger source archives that are unpacked AlienSoldier, although I don't do that as much as I used to :)
<AlienSoldier>
the transfer speed that the tracker report does not seem to vay much between the high and low cpu requirement, so at least it does not seem to slowdown the transfer operation.
<AlienSoldier>
*vary
_-Caleb-_ has left #haiku [#haiku]
<AlienSoldier>
just noticed that there is more info in the pop up window (when we let the pointer on the tracker status) that the always alrernating annoying data under the progress bar.
_-Caleb-_ has joined #haiku
tqh has joined #haiku
<AlienSoldier>
1h:40m left
kinkinkijkin has joined #haiku
<Begasus[m]>
ready for the lazy afternoon :)
<AlienSoldier>
You take the tea like the English?
<Begasus>
nah, coffee! (and occasionaly a coockie with it) :)
<Begasus>
never got the taste for tea
<AlienSoldier>
i started to do pear juice this fall, so this is my vice latetly :)
<Begasus>
home made?
<AlienSoldier>
yep
<Begasus>
nice!
<AlienSoldier>
Taste the same thing as the looza juice i was drinking kid.
<fcn>
hi. I'm new to Haiku. I installed it on a Thinkpad T470s. Touchscreen and USB mouse work but the trackpoint and the touchpad do not. Is this a known issue or is there something I need to do to fix this?
<Begasus>
k, this probably will bite me ... launched build for qt6-base
draKon has joined #haiku
<Begasus>
reboot, fresh start ...
Begasus has quit [Quit: Vision[]: i've been blurred!]
Begasus_32 has joined #haiku
OscarL has joined #haiku
<Begasus_32>
hi OscarL!
<OscarL>
'lo Begasus_32! :-)
<Begasus_32>
still alive! :)
<OscarL>
Apparently :-/
<OscarL>
:-P
<Begasus_32>
tsss
<Begasus_32>
how's it going there?
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
<OscarL>
somewhat in better mood, but with a few nasty migranes lately, so no much will/concentration to do computer stuff. Thanks for asking. All good on your side, I hope?
<OscarL>
looking at updating serf today (to see if it resolves the issue with missing symbols)
<Begasus[m]>
so far so good here, was in the process of building the rc for upcomming framework
<Begasus[m]>
then I saw they started to integrate pyside :)
<Begasus[m]>
yeah, saw something passing by on that
<Begasus[m]>
checking pyside revealed the error in qt6-base
<OscarL>
better pyside than PyQt (IMO), but still might require quite some work :-(
<Begasus[m]>
it's only Terminal checks so far, and not a hard requirement for the frameworks
<Begasus[m]>
seen some comments from upstream being not to happy with that :)
<OscarL>
Ah, kind of frameworks just providing PySide bindings type of thing?
<Begasus[m]>
yep
<Begasus[m]>
at KDE they had a student working on that for GSoC
<OscarL>
hopefully they'll make/keep the build/install of such things optional (to make packager's life a bit easier at least at first :-D)
<Begasus[m]>
same here, doesn't stop me from looking into it though :P
mmu_man has quit [Ping timeout: 480 seconds]
<OscarL>
right, that was what I was thinking... always good to know how dificult they would be to build/integrate, even if just to decide to drop them for now :-D
<OscarL>
speaking of dropping... /me starts to sharpen the axe...
<Begasus[m]>
heh
<OscarL>
do we really need gcc2 stuff for apr/serf?
<OscarL>
svn already is x86 (only user I can see for those)
<Begasus[m]>
don't think so, but I could be wrong
<Begasus[m]>
if it would end up in $commandBindir or the likes I guess it should be fine?
<OscarL>
"gir libserf_1" only shows serf and subversion :-)
<Begasus[m]>
1 down :D
<OscarL>
gir libapr_1... shows: apr, serf, apr_util (no need for those commands to be gcc2 that I can think of), subversion and apache (last two already !x86_gcc2)
<OscarL>
I'mma just chop off gcc2 for serf then, and wait for korli to shout at the review if I missed something :-D
<AlienSoldier>
This time recreating my partition did nothing, still can see files i query :(
<OscarL>
mmm, hit a misleading error message on Haikuporter :-/
<AlienSoldier>
retry failed
<Begasus[m]>
bogus disk?
<AlienSoldier>
i don't think so
<AlienSoldier>
I will recreate it as fat with a different name just to be sure all is earased and then recreate it
<OscarL>
"Error: Error: Patch file serf-1.3.10.patchset is not referenced in PATCHES, so it will not be used" (I don't have a file named "serf-1.3.10.patchset" anymore there)
_-Caleb-_ has left #haiku [#haiku]
<Begasus[m]>
touch the recipe
<OscarL>
AlienSoldier: or just "dd if=/dev/zero" the few first MBs?
mmu_man has joined #haiku
_-Caleb-_ has joined #haiku
<OscarL>
Begasus[m]: that worked. Darn cached recipes :-/
draKon has quit [Quit: Vision[]: i've been blurred!]
<Begasus[m]>
it's a feature! :P
x10z has joined #haiku
<OscarL>
Begasus[m]: now I restored the .patchset file... and get no complains :-P
<Begasus[m]>
keeps you sharp :D
<OscarL>
you mean mad(der), right? :-P
<OscarL>
let's see if "-f" is enough (instead of having to use touch)
<OscarL>
noup. lol.
<Begasus[m]>
heh
<OscarL>
from serf's changelog for 1.3.10: "Build changes to support Python 3 (r1875933)"... good think, because I was already commenting out all the rest of .patch files.
janking has quit [Quit: Vision[]: i've been blurred!]
<OscarL>
Uh.. also: "Support for OpenSSL 3 (r1901937, ...)" :-D
<OscarL>
chances of success went up.
<AlienSoldier>
damn it is frustrating :) was able to fix the partition 2 but not the 1.
HaikuUser has joined #haiku
HaikuUser has quit []
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
<OscarL>
AlienSoldier: part 1 and 2 are of different, or equal capacity and blocksize?
<AlienSoldier>
same
<OscarL>
I had some issue with partitions being mistaken as "being the same" when size/blocksize matched, albeit my issue was with auto-mounting, not with queries.
<OscarL>
Maybe try making one a bit smaller, and see if that change things? or if changing blocksize does? Might help narrow down the issue at least?
<AlienSoldier>
indeed
<OscarL>
I did had (or have?) a partition that gives me some unexpected results for queries. I wasn't sure if it was due to broken indexes, too much fragmentation, or what. Data was ok, though, as moving all files to a different partition worked fine, and results are the expected ones.
<OscarL>
*query results (on the new partition)
<AlienSoldier>
OscarL if i do the queries for the volume all is fine, it is only invisible if i do it for all partition that the file does nto appear.
<AlienSoldier>
*not
<OscarL>
Ah.. then it *might* be a similar case as I had with auto-mounting then.
<OscarL>
somehow it sees the two partitions being "equal" and it thinks... "I've added this before, skip" ?
<AlienSoldier>
OscarL ok now it work, but because i am an idot without patience i did 2 thing at once :) I made the partition 1 byte smaller and named it to a different name at creation (not simply a tracker rename of the partition).
<OscarL>
Begasus[m]: serf-1.3.10 build without patches just fine (only have some minor warrnings about initialized vars, and such).
<Begasus[m]>
niceeee Oscar-L!
<Begasus[m]>
if it works ok now it should be fine :D
<AlienSoldier>
i think i will keep it that way, i can live with a byte less :)
<OscarL>
AlienSoldier: my time as a QA-monkey trained me to think "only change one thing at a time!" (too bad I'm also not very patient sometimes :-D)
<OscarL>
AlienSoldier: in any case, please file a ticket!
<Begasus[m]>
lol, sounds familiar
<AlienSoldier>
well now i have to rename it to the old name anyway, i need consistent partition name, i will see if that re-introduce the problem.
<Begasus[m]>
about 600 objects ...
<OscarL>
installed the new serf with "pkgman -H", but that doesn't seems to work as expected. pkgman still wants to uninstall svn if I try to uninstall older serf from /system :-(
<AlienSoldier>
OscarL i scared myself, named it as haiku2 instead of haiku one and it said partition haiku2 is formated! Seem you can have 2 partition with the same name. Something should prevent this.
<OscarL>
severely cripples the usages of "-H" as install location, at least for me.
<AlienSoldier>
OscarL in any case, queries still are fine it seem after returning to normal name.
<Begasus[m]>
I don't think it's intended to be used like that Oscar-L?
<jmairboeck>
OscarL: system packages can't depend on user packages
<OscarL>
AlienSoldier: You can have multiple volumes with the same label, yes. Mountpoits would differ though.
<AlienSoldier>
Yes that is how i knew what one i toyed with
<OscarL>
jmairboeck, Begasus[m]. Understood. but still a bit annoying for the way *I* want to use it :-D
<Begasus[m]>
some of you out there just want to do it the hard way :P
<Begasus[m]>
maybe you can have subversion installed in -H too? :P
<jmairboeck>
that should work
<AlienSoldier>
OscarL i notice i use 8 similar naming character for my partition name before putting number, if the mounting skip happen like you say, perhaps it is because it only use the first character of the name as differentiator.
<OscarL>
Begasus[m]: if pkgman were able to "cleanup" the crap it installs as depedencies when I'm just trying things... I would use the easy way gladly :-D
thelounge819 has joined #haiku
<Begasus[m]>
ps, for some like those I keep a backup from the original one and install the newer one, that just didn't work for zlib :P
thelounge81 has quit [Read error: Connection reset by peer]
<Begasus[m]>
OscarL we already live in dependency hell :D
<Begasus[m]>
those are not the hardest :)
<AlienSoldier>
fentanil injected library.
* Begasus[m]
had to look up fentanil
<OscarL>
mmm, uninstalled subversion, just to install it with "pkgman install -H"...
<OscarL>
instead of reusing the .hpkg from system administrative states... it just re-downloaded it :-/
<Begasus[m]>
lol
<OscarL>
so much for caching :-P
<OscarL>
Begasus[m]: svn doens't errors out anymore.
<Begasus[m]>
+1
<AlienSoldier>
OscarL i made a file i copied of all partition as a query on my destop. I will launch it from time to time to see if the problem come back. The result should show me my 4 partition (one named boot).
<AlienSoldier>
*on all partition
tqh has quit [Quit: Leaving]
<OscarL>
AlienSoldier: alright, but don't forget to open that ticket if you can anyway :-)
<Begasus[m]>
qmplay2 still running fine (has some country music in the background from the 32bit install) :)
x10z has joined #haiku
<x512[m]>
OscarL: It will need to track which packages are installed by user request and which are installed as dependencies that is not currently implemented.
dovsienko has joined #haiku
mr_lou has joined #haiku
<OscarL>
x512[m]: indeed. a pain in the rear, and thus why I even atempted some half-assed "pkgman search --not-required" implementations.
<x512[m]>
OscarL: Applications are usually not required to any other package, but needed to user.
mattlacey has quit []
<AlienSoldier>
why do i always find computer problem just before going to bed and find a satisfying working solution each time only when the sun come up...
<AlienSoldier>
at least it is the weekend this time!
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
<OscarL>
x512[m]: right. mine was just my poor attempt at having at least *something" rather than nothing. Proper solution would be, as you suggest, keeping track of "auto" vs user requested installs, of course. (waaay out of my skill level :-D)
MisthaLu has quit [Ping timeout: 480 seconds]
<x512[m]>
At least it need to record names of manually requested packages to some file.
<x512[m]>
Should be not so difficult.
<OscarL>
AlienSoldier: I keep finding small but annoying/distracting things while trying to do other small things :-( (not great for my short attention spans :-D)
<Begasus[m]>
AlienSoldier, close down the machine earlier? ;)
<Begasus[m]>
about 300 to go .... getting there
<Begasus[m]>
OscarL I would be happy of those here were only small and annoying :)
janking has quit [Quit: Vision[]: i've been blurred!]
<OscarL>
janking: https://streamable.com/bob445 nice :-) (reminds me of hearing some friends' band playing live... but better tunned and with more hair in this case :-D)
<OscarL>
PR open for serf (fixes the missing symbol on 32 bits too)
MsInput has quit [Quit: WeeChat 4.4.3]
<Begasus>
+1
MsInput has joined #haiku
<Begasus>
guess I should do this again for 6.7.2 :P grabbing qt6_base-6.8.1-3-x86_64.hpkg and moving it to /Share/haikuports/packages/qt6_base-6.8.1-3-x86_64.hpkg
mmu_man has quit [Ping timeout: 480 seconds]
<OscarL>
alright, cleaned up old apr_util and gcc2 apr. Testing build for updated apr 1.7.5.
<OscarL>
mmm, might need to update "libVersion" :-/
<Begasus[m]>
not equal to portVersion?
pch has quit [Ping timeout: 480 seconds]
<OscarL>
recipe currently has: libVersion="0.7.4"
<Begasus>
got this for the rc's :) libVersion="${portVersion/\~rc1/}"
<OscarL>
(port version was 1.7.4)
<Begasus>
still could be valid
<OscarL>
final .so is named: "libapr-1.so.0.7.5"
<Begasus>
see :P
<OscarL>
so I guess I should change that to libVersion="0.7.5" then.
<OscarL>
forgot to reset revision anyway :-D (also... need to run a new build to test if a hack is still needed)
<OscarL>
"hp -c" on RAMFS is fast at least :-D
HaikuUser has joined #haiku
HaikuUser has quit []
Begasus has quit [Quit: Vision[]: i've been blurred!]
<Begasus[m]>
rebooting vm
<Begasus[m]>
switched back to not using RAMFS, too many downloads lost after reboot :P
<Begasus[m]>
jikes ... Reconnecting 1
<Begasus[m]>
64bit buildmaster :/
<OscarL>
Begasus[m]: maybe you set OUTPUT_DIR but forgot to also set REPOSITORY_PATH ? (the latter makes sure the downloads are placed just as before, next to the recipes.
<Begasus[m]>
didn't fidle with repository path (don't have a dual setup), so no real need for that
<OscarL>
I mean... if you set only OUTPUT_DIR (to ramfs)... both downloads and work dirs get placed on RAMFS. that's not ideal as you have already found out.
Begasus_32 has quit [Ping timeout: 480 seconds]
<Begasus[m]>
also, while checking the frameworks etc I sometimes need those 'work' directories after a reboot :)
<Begasus[m]>
the idea is great, and I still have those lines commented out if I want to switch back, but not for now, don't want to start from scratch on this qt6-base source :)
<OscarL>
yeah, for longer work I need that "move to SSD/HDD" script I've mentioned last time :-)
<Begasus[m]>
for now I'm good on the scripting parts :)
<OscarL>
Also... it still doesn't works reliably for all builds (due to some bugs on RAMFS/Haiku).
<OscarL>
seems that apr hack is not needed any more. Build went fine without it.
<OscarL>
(was creating a libtool symlink under $sourceDir for some reason :-D
<Begasus[m]>
nice, we're shaping up! :D
<OscarL>
let's see what "hp --test apr_x86" says.
<Begasus[m]>
give a yell when it finishes after 3 hours :P
<OscarL>
:-/
<Begasus[m]>
jk ;)
<Begasus[m]>
don't know how long that test runs :)
<OscarL>
I remember the svn ones taking a loong time.
<Begasus[m]>
some large projects finish very quick, some small projects take like forever
<OscarL>
this one measures time in microseconds... yet one of the tests took 3 minutes :-(
<OscarL>
(and no indication of current vs total number of test :-/)
<Begasus[m]>
can't use the new "cmd:time"? :D
<OscarL>
CPU usage is low on the VM, yet, cpus on host are maxed out :-/
<Begasus[m]>
yeah, same here :(
<Begasus[m]>
err .. not on host*
<OscarL>
might be one of those "VirtualBox things" then.
<Begasus[m]>
but CPU usage on the VM isn't what it should be, sometimes hangs around %20
<Begasus[m]>
VMWarre here
<Begasus[m]>
and sometimes a reboot helps out
<OscarL>
I would like to use VMware now that I have a separate SSD for Haiku... but darn VMware takes almost 1.5 minutes to boot, against < 25 seconds on VBox.
<Begasus[m]>
that's still pretty OK here
<OscarL>
(tried all the settings I could on VMWare... ide/sata/scsi, all do the same :-/)
<OscarL>
even nvme
<Begasus[m]>
heh
<Begasus[m]>
been fidling with the CPU numbers in there this week
* Begasus[m]
pokes waddlesplash ... 64bit buildmaster down again :)
<Begasus[m]>
32bit almost done for qt6-base
x10z has joined #haiku
bbjimmy has quit [Quit: Vision[]: i've been blurred!]
draKon has joined #haiku
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
bbjimmy has joined #haiku
linuxmaster has quit [Ping timeout: 480 seconds]
<OscarL>
for tests that take this long, not offering at least an ETA, or "X of N total" counter is why people usually don't even bother running tests :-)
<Begasus[m]>
yeah, one tend to nuke them :)
<Begasus[m]>
check gentoo's ebuild, maybe they disabled it for that :)
bbjimmy has quit [Read error: No route to host]
<Begasus[m]>
build completed on 32bit
<OscarL>
seems like this is just repeating the tests, each time increasing the number of concurrent threads.
<OscarL>
up to 6 now.
<Begasus[m]>
OscarL something like [750/1240]? :)
<Begasus[m]>
6 tests?
<OscarL>
6 threads.
<Begasus[m]>
ow
<Begasus[m]>
it's just waiting on you to run out on :P
bbjimmy has joined #haiku
<Begasus[m]>
biab
<OscarL>
apparently it is just 2 tests: "apr_thread_mutex_t" and "apr_thread_rwlock_t". The first is run "unnested", "nested" and "timed". And it runs it repeatedly with 1, 2, 3, 4, etc threads. :-/
x10z has joined #haiku
<OscarL>
just when I was about to nuke it... it finally progressed into running new tests :-)
illwieckz has quit [Remote host closed the connection]
illwieckz has joined #haiku
jagtalon has quit [Remote host closed the connection]
thelounge819 has quit [Ping timeout: 480 seconds]
<waddlesplash>
OscarL: open the VMX file and delete all the stuff related to sata
<waddlesplash>
even if you remove the sata controller from the GUI it doesn't remove it from the VMX often
<OscarL>
Will do waddlesplash. Thanks!
<OscarL>
right... sata stuff still there, even in files where IDE or NVMe are the one in use :-/
<OscarL>
mmm I think apr tests finally froze my VBox :-D
<OscarL>
can't enter KDL, cursor frozen. Oh well.
thelounge819 has joined #haiku
<waddlesplash>
if that's reproducible let me know
<waddlesplash>
I'll want to investigate of course
<OscarL>
very slow tests, I'll see how it behaves on VMware, next, I guess :-)
<waddlesplash>
yeah if it's thread contention VMware will be better
thelounge819 has quit []
HaikuUser has joined #haiku
HaikuUser has quit []
mmu_man has joined #haiku
<OscarL>
VMware didnt liked my SCSI .vmdk :-/ "Panic: did not find any partitions!" :-D
<waddlesplash>
we don't support VMware 's scsi controllers
<OscarL>
will try IDE next
<waddlesplash>
the nvme will perform the best probably
<OscarL>
alright... let me see.
thelounge819 has joined #haiku
<OscarL>
30 seconds to boot (32 bits). Much better. Thanks waddlesplash :-)
<OscarL>
clipboard integration on 32 bits at last :-D
<waddlesplash>
:)
<OscarL>
running apr tests on VMware. Guest is using about 20% of a couple of cores, but host has all cores at about 60/70% (was more like 80/90% with VBox).
melnary has quit [Remote host closed the connection]
melnary has joined #haiku
<OscarL>
funny. I capped the cpu freq to 2.1 GHz (2.8 max), and host cpu usage went down 10/15% :-)
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
<Begasus[m]>
re
<Begasus[m]>
OK, need to test that too OscarL and waddlesplash :)
<Begasus[m]>
no sata in my vmx ...
x10z has joined #haiku
<OscarL>
waddlesplash: Haiku froze again running apr's tests (same place as before). Mouse cursor moves, but GUI is non-resposive. Alt+SysRq+d does nothing.
<OscarL>
this time on VMware, took about 30/35 minutes before freezing.
Begasus has joined #haiku
<OscarL>
this is with apr-1.7.5 recipe I was just trying to update.
<OscarL>
waddlesplash: mmm, serial output shows: Keyboard Requested Halt. / Welcome to Kernel Debugging Land... / Thread 3 "idle thread 3" running on CPU 2
<Begasus>
k, errors on pyside: /boot/system/develop/tools/bin/../lib/gcc/x86_64-unknown-haiku/13.3.0/include/c++/cstddef:50:10: fatal: 'stddef.h' file not found ?
HaikuUser has joined #haiku
HaikuUser has quit []
<OscarL>
was able to blindly type "reboot", and it worked.
<jmairboeck>
Begasus: sounds like haiku[_x86]_devel is missing?
<Begasus>
KDL gives light enough for that OscarL, no need to do it blindly :)
<Begasus>
jmairboeck, with all the building here that would sound strange?
<Begasus>
ps 64bit
<OscarL>
Begasus: I didn't got any signs of KDL being active (other than the output of that serial.txt file :-D)
<jmairboeck>
stddef.h should be in haiku_devel
<Begasus>
yeah, should be kinda default around I though
<Begasus>
I got /boot/system/lib/clang/18/include where it is included also
<Begasus>
oh well, putting that aside :)
<Begasus>
maybe those headers for clang should be in $developLibDir ...
Anarchos has joined #haiku
<Begasus[m]>
ok, that should fix (atleast did here) qt6_base
<Begasus[m]>
it's not something that's widely used since it never popped up before
<Anarchos>
hello
<Begasus>
hi Anarchos
carlomon1e has quit [Ping timeout: 480 seconds]
<Anarchos>
Begasus grok.nikisoft.one/opengrok is down ?
<Begasus>
I wouldn't know Anarchos, OscarL might know
<OscarL>
Seems so. Anarchos you should poke nipos about it :-)
<OscarL>
"502 Bad Gateway". Last time that happened, nipos was quick to fix it.
<Anarchos>
OscarL how can i develop without it ??? :)
<Begasus>
local backup? :)
<OscarL>
Anarchos: right... what are we to do? grep like filthy animals? :-P
<Anarchos>
OscarL it seems so
<nipos>
Thanks for the notice, will be fixed as soon as my laptop has booted
<OscarL>
thanks again to you for providing that service nipos. Much appreciated.
<nipos>
No problem, I'm happy that it's useful to you all
<OscarL>
"
<OscarL>
all builders lost".... with qt6_base and haikuwebkit on the pipeline. Ouch.
<OscarL>
32 bits still up it seems.
<Begasus[m]>
yep, 32bit still good
<nipos>
It's working again. "sudo service tomcat101 restart" was all it needed.
<Begasus[m]>
64bit since this morning
<OscarL>
github commit comments are ununsable on Firefox :-(. Only show if I use incognito mode.
<waddlesplash>
OscarL: VMware video driver doesn't update screen in KDL. I just disable it and use VESA
<OscarL>
Inconvenient, but good to know.
<Forza>
Hi :)
<OscarL>
I use vmware driver on VBox too, so that explains why I didn't got a visible KDL there then :-)
<OscarL>
hello Forza.
<Begasus>
Hi Forza
<Forza>
:) hope you are well
<Forza>
I noticed the rsync server isnt responding since a while. Is there an issue with it?
zardshard has left #haiku [Error from remote client]
zardshard has joined #haiku
<OscarL>
No idea about infra stuff, Forza, sorry. Try paging Dr. kallisti5[m] ? :-)
<Forza>
The rsync.haiku-os.org gives me "rsync: server sent "HTTP/1.1 400 Bad Request" rather than greeting"
<Forza>
Thanks
* OscarL
rethinks the issue of vmware driver and KDL... because I surely had lots of visible KDLs on VBox even with vmware driver :-D
kinkinkijkin has quit [Quit: Leaving]
<OscarL>
Let's see if "hp --test apr" also freezes on 64 bits.
<OscarL>
Mmm, I probably should have tried setting up putty or something into vmware's serial output :-/
fcn has quit [Quit: Vision[]: i've been blurred!]
<OscarL>
VMware offers named pipe or "physical serial port" (besides file output). Not sure if Putty can use named pipes. Might need to setup com0com, I guess.
Anarchos has quit [Quit: Vision[]: i've been blurred!]
<OscarL>
Begasus[m]: seems not, as instead of a system freeze, I got the test app to crash instead.
<OscarL>
Debugger also crashed while trying to save the .report :-/
<Begasus>
progress! :D
<OscarL>
segment violation on "testall", but the rest of .report is just empty :-D
<OscarL>
crash on debugger: "state: Exception (Invalid opcode exception)"
<Begasus>
ok, know you know what, you can fix it :P
<OscarL>
not even in dreams :-)
mr_lou has quit [Quit: Leaving]
Anarchos has joined #haiku
<Begasus>
well, now you're done, check llvm18? ;)
<OscarL>
and you tackle moving things away from llvm12, deal? :-P
<Begasus>
maybe we could poke zardshard for pyside, already did the bindings for Haiku, shouldn't be too hard for Qt6? :P
<Begasus>
already did ICU and boost :P
dqk_ has joined #haiku
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
dqk has quit [Ping timeout: 480 seconds]
<Begasus>
ok what "is" using llvm12_libs ?
<Begasus>
you don't want to see the list for "pkgman uninstall llvm12_libs" :)
<OscarL>
mouse integration is acting up on VMWare, sigh.
<Begasus>
I guess mesa is a major part in there: requires: lib:libllvm_12>=12.0.1
<OscarL>
right.
<OscarL>
more than acting up... not working at all. it just captures the input, and need to git ctrl+win+alt to free it :-/
<OscarL>
s/git/hit/
<Begasus>
shit?
<Begasus>
I guess moving llvm version for mesa isn't that easy, probably would have been done otherwise
<waddlesplash>
OscarL: iirc there's a USB tablet pointing device you can enable in the vmx
_-Caleb-_ has left #haiku [#haiku]
<waddlesplash>
which works better than the vmwaretray mouse integration
<waddlesplash>
I think that's what I have enabled
_-Caleb-_ has joined #haiku
<OscarL>
Thanks waddlesplash, will look into it. I use USB tablet on VBox because it works better (mostly) too.
<OscarL>
except for the broken drag an drop on Qt, but that just makes me avoid Qt, so that's fine :-)
* zardshard
looks at Begasus's message, decides not today, and goes back to sleep
<Begasus>
heh
<zardshard>
cya :)
<Begasus>
worth a try
<Begasus>
cya! :)
<OscarL>
sleep well zardshard :-)
<zardshard>
OK, only figuratively sleeping ;-)
Anarchos has quit [Quit: Vision[]: i've been blurred!]
<OscarL>
liar. now I better hear some snoring from your direction, or else!
<zardshard>
Np. Zzzz :D
<OscarL>
*trout-slapping averted!*
* Begasus
calls in DaaT
janking has joined #haiku
<Begasus>
Plasma Wayland Protocols 1.15.0 is now available for packaging. It is needed for the forthcoming KDE Frameworks.
<Begasus>
<Begasus>
should update that one here, didn't cross it yet in the frameworks yet though
<Begasus>
but could also be for the plasma frameworks :)
<janking>
hi Begasus
<Begasus>
wb janking :)
<Begasus>
grabbing plasmawayland_protocols-1.15.0-1-x86_64.hpkg and moving it to /Share/haikuports/packages/plasmawayland_protocols-1.15.0-1-x86_64.hpkg
<Begasus>
so far still good
<jmairboeck>
that is probably mostly no-op on Haiku because Qt doesn't use Wayland.
<Begasus[m]>
good thing it's only xml files (if I looked correctly), so far not used anywhere
<jmairboeck>
ah, ok
<Begasus[m]>
Haiku isn't the only one not using wayland or x11 :)
<Begasus[m]>
only one referenced a few times is plasma-activities(stats), was kactivities in KF5
<Begasus[m]>
should even review that if it "is" required at all
<Begasus[m]>
same with globalaccelaration, not in use
<Begasus[m]>
(on KF6)
Begasus has quit [Quit: Vision[]: i've been blurred!]
AlienSoldier has joined #haiku
x10z has joined #haiku
<Begasus[m]>
closing down here
<Begasus[m]>
cu peeps
KapiX has joined #haiku
<janking>
cu
janking has quit [Quit: Vision[]: i've been blurred!]
<OscarL>
Mmm, usb tablet was already enabled. So, I just need to NOT use vmwaretray then :-)
_-Caleb-_ has left #haiku [#haiku]
_-Caleb-_ has joined #haiku
OscarL has quit [Ping timeout: 480 seconds]
x10z has quit [Quit: My Unrecognized Mac has gone to sleep. ZZZzzz…]
AD_Haiku_onPC has quit [Ping timeout: 480 seconds]
AD_Haiku_onPC has joined #haiku
draKon has quit [Quit: Vision[]: i've been blurred!]
janking has joined #haiku
janking has quit []
x10z has joined #haiku
jmairboeck has quit [Quit: Konversation terminated!]
KapiX has quit [Quit: KapiX]
HaikuUser has joined #haiku
HaikuUser has quit []
AlaskanEmily has joined #haiku
_-Caleb-_ has left #haiku [#haiku]
_-Caleb-_ has joined #haiku
_Dario_ has joined #haiku
higen6 has joined #haiku
higen has quit [Ping timeout: 480 seconds]
higen6 is now known as higen
vdamewood has quit [Quit: My Mac has gone to sleep. ZZZzzz…]