<kallisti5[m]>
taking a pause to eat. repos healthy. haikuporter buildmaster stopped for all architectures while I wait on some information. (I can power forward without the info... but the risks of doing something dumb increase)
<Habbie>
enjoy your food :)
Nasina has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
Aedil has quit [Quit: leaving]
Nasina has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
Nasina has joined #haiku
HaikuUser has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
HaikuUser has quit []
HaikuUser has joined #haiku
HaikuUser has quit []
HaikuUser has joined #haiku
HaikuUser has quit []
HaikuUser has joined #haiku
Gues__________________________ has joined #haiku
<Gues__________________________>
hi there, I am just wondering if icons HVIF format support some sort of Gaussian blur effect which is available in SVG format?
<waddlesplash>
I'll deal with it after kallisti5[m] switches everything
<OscarL>
thank you! (to both of you)
<kallisti5[m]>
waddlesplash: you can do it now. The new buildmasters are running, they're just halted
<waddlesplash>
ah
<kallisti5[m]>
just don't run any builds for now... i have the fear of erasing all of our repos
Nasina has joined #haiku
<kallisti5[m]>
(which is why I sent that mmlr email)
<waddlesplash>
I'm just away from my PC atm
<kallisti5[m]>
oh, ok
<waddlesplash>
I'll sign in and investigate later
<waddlesplash>
kallisti5[m]: isn't there a pause file or something!
<waddlesplash>
*?
<kallisti5[m]>
yup. That's the one I'm using
<kallisti5[m]>
that way it's running, just not doing anything
Nasina has quit [Read error: Connection reset by peer]
Nasina has joined #haiku
Nasina has quit [Ping timeout: 480 seconds]
Gues__________________________ has joined #haiku
KENZ has quit [Quit: Vision[]: i've been blurred!]
Gues__________________________ has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
Nasina has joined #haiku
mmu_man has quit [Ping timeout: 480 seconds]
* OscarL
curses Linux (at least this EndeavourOS install ) for not being able to survive a motherboard transplant.
KENZ has joined #haiku
Aedil has joined #haiku
KENZ has quit [Quit: Vision[]: i've been blurred!]
tetrislife has joined #haiku
duncsauce_ has joined #haiku
duncsauce has quit [Read error: Connection reset by peer]
<kallisti5[m]>
can't stop, won't stop
* kallisti5[m]
is getting tired. going to fix the download page and call it a night
<OscarL>
updating from haikuports is broken at the moment, but I rather have you sleep at least a few hours :-D
<kallisti5[m]>
yup. buildmaster is tomorrow morning. There's a chance it could eat the repository, and that's not something I want to manage lol
<kallisti5[m]>
(manage this evening at midnight)
<kallisti5[m]>
https://download.haiku-os.org/ that's all the nightly images y'all get. Nobody needs 200GiB of nightly haiku isos
<kallisti5[m]>
ok. Heading to bed. bbl
<OscarL>
'Refreshing repository "HaikuPorts failedValidating checksum for HaikuPorts...: I/O error' <<< we even have a missing space after "failed" there... seems I found my next "patch" :-P
<OscarL>
Sleep well kallisti5[m]! Thanks for your work!
<kallisti5[m]>
OscarL: wait... is that still happening?
<OscarL>
just did, yes.
<kallisti5[m]>
I guess that cached file mmu_man mentioned causes it?
<kallisti5[m]>
it shouldn't happen anymore moving forward
<kallisti5[m]>
the cdn is dead to me and out of the repo
<kallisti5[m]>
(nginx is doing a little caching though on the s3 server, if it's still an issue I can turn that off)
<kallisti5[m]>
aah... ok. bed. make a ticket
<OscarL>
sweet dreams!
Nasina has quit [Read error: Connection reset by peer]
tetrislife has left #haiku [Error from remote client]
the_sea_peoples has quit []
Nasina has joined #haiku
Nasina has quit [Remote host closed the connection]
<OscarL>
Tried doing pkgman drop-repo HaikuPorts`, and re-adding it, just in case... it does "Fetching repository from ..." then "Activating repository config from ..." and errors with: "*** faied! : Bad data".
<OscarL>
Oh well... /me goes play with something else for a while :-)
Nasina has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
OscarL has quit [Quit: Page closed]
<Begasus[m]>
g'morning peeps
Nasina has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
<nephele>
I wanna take a bit of a break from webkit, so now i want to take a look at the imapfs thing. Do we perhaps have an outline how this could work=
<phschafft>
imapfs?
<nephele>
phschafft: basically a filesystem exposed by the mail_daemon over userlanfs (similar to fuse) for a remote imap location, instead of putting "plain" files on the disk
<nephele>
so that you know that the state of the exposed filesystem actually matches what the imap server thinks
<nephele>
and you don't have to download the entire archive if you don't want to :)
<nephele>
phschafft: what do you think about that?
Retropikzel has left #haiku [#haiku]
<PulkoMandy>
nephele: There is an imap fuse filesystem out there called imuse. May be a good starting point to try building that. It is not from beos/haiku so it will not use the native mail format but itys a starting point
<PulkoMandy>
You don't need to involve mail_daemon at all
<nephele>
what makes you say that? isn't mail_daemon the thing actually speaking imap? so it is the one that knows about the remote state, no?
<phschafft>
nephele: my thoughts are basically what PulkoMandy said.
<nephele>
If so, is there a newer version? that one has basically... no code, at all. like maybe 50 lines in total.. i really doubt it implements a fusefs in any useable capacity
<PulkoMandy>
My review of it didn't go that far :)
<nephele>
okay, so assuming this simply isn't written yet, what design do you envision, and why would that not involve mail_daemon? or am i misunderstanding you?
<PulkoMandy>
The filesystem pretty much exactly replace what mail_daemon tries to do
<PulkoMandy>
What would mail_daemon do? The filesystem is connected to the imap mailbox, shows it as files, you can use queries and all the usual things to access your mail
<PulkoMandy>
Everything else (mail client, queries, etc) can be left unchanged
<PulkoMandy>
maybe you still need mail daemon for email sending, but for receiving, it is not needed anymore
<nephele>
Now i'm really confused. Yes, this needs rework. But we still need the actuall IMAP implementation, and that is in mail_daemon, so why not extend it to provide the filesystem instead of dumping them on the disk?
<nephele>
Is your idea to split out the IMAP support out of mail_daemon completely and transplant it into the new filesystem driver?
<phschafft>
side thought: what about offline operation?
<nephele>
in both cases we need a long running process, as that matches imaps design, and we should have atleast something that has the socket open for imap IDLE (so we can receive mails immidiently)
<phschafft>
many MUAs use a download on demand, keep locally until deleted style (more or less).
<nephele>
we don't only have delete as things to worry about, but also mv operations, and tagging of things :)
<PulkoMandy>
nephele: A filesystem is not a daemon and a deamon is not a filesystem
<PulkoMandy>
So, yes, all the imap code must be in the filesystem, how else would it work?
<nephele>
can't a daemon provide a filesystem over userlandfs?
<nephele>
well, the "other" way around would be simply sending messages from the filesystem to the daemon on what work you want to have done
<PulkoMandy>
userlandfs works with add-ons
<PulkoMandy>
So I don't see how a daemon would be involved. The code will run in the userlandfs process as an add-on, not in a separate daemon
<nephele>
can the add-on for userlandfs then keep something like a socket open for long running operations? (in essence, do stuff even if no FS operations are requested)
<PulkoMandy>
Yes, if you want to make things complicated for no reason
<nephele>
IMAP is complicated....
<nephele>
How else would you receive new emails?
<PulkoMandy>
Imap with an extra layer on uop is more complicated than just imap?
<nephele>
IMAP is build around a persistent connection
<nephele>
and if we want IMAP IDLE we need to have one
<PulkoMandy>
yes, how is that a problem? Sshfs does this just fine with ssh connections for example?
<nephele>
??
<nephele>
I don't know what the problem is, you called it complicated for no reason...
<PulkoMandy>
you want to have the filesystem talk to mail daemon instead of just directly talking to the imap server
<PulkoMandy>
I don't understand what problem that tries to solve, and instead of "just do imap" now you have to design a whole new thing to make the filesystem and mail daemon talk to each other
<nephele>
I see, your statement of it beeing complicated was a response to a previous message, and not on the question if userlandfs can keep a socket open
<PulkoMandy>
ah, yes
<nephele>
that makes more sense :)
<phschafft>
nephele: my point is that you should keep in mind that you need a local persistent storage backend.
<PulkoMandy>
So, the filesystem code in userlandfs can do whatever it wants, including opening a socket, starting a worker thread if needed, etc
<nephele>
mail_daemon does need some imap still, we should save send emails in imap. Maybe it can "just" place the file there if everything works as intended
<phschafft>
therefore just dumping stuff on the filesystem on-demand and then letting the filesystem handle the rest seems like a (one of many) sound ways to do it.
<nephele>
phschafft: stupid question. but, can I use the folder behind the mount point?
<PulkoMandy>
phschafft: Not really, if you want to keep things locally you can copy them manually. I would rather have my mail client not try to keep a local copy of the mails in syc with the server by itself (that always ends badly with lost data for me)
<phschafft>
I have no idea about the deails on Haiku, but most other operating systems support doing that in a number of different ways (depending on what exactly you want to do)
<nephele>
PulkoMandy: I think this is in relation to caching and offline use, those should probably still be possible
<phschafft>
PulkoMandy: then make it an option!
<PulkoMandy>
nephele: You can, but then strange things can happen if the filesystem is not mounted (the internal dat becomes visible in its place)
<phschafft>
PulkoMandy: also, again, this is about mails that got requested by the user. not all of them.
<phschafft>
so it doesn't try to keep things in sync. it just makes sure you only download things once unless the user asks for downloading it again.
<PulkoMandy>
phschafft: Your option is to continue using the existing mail daemon if ycu want automatic local backups. The filesystem approach is for when you really don't want that
<phschafft>
(such as by a clear-cache button)
angrystar17043 has quit [Remote host closed the connection]
angrystar17043 has joined #haiku
<phschafft>
PulkoMandy: again, this is not about backup nor keeping in sync.
<PulkoMandy>
So, that's just a cache. The normal in-ram filesystem cache can be used?
<nephele>
I think you are talking around each other. I don't think pulkomandy ment that you can't cache data...
<phschafft>
but if you have two different implementations that do exactly the same (from a user perspective) just that one is using a lot of disk and the other is using a lot of network and there is nothing in the middle and you want to maintain both, fine with me ;)
<phschafft>
PulkoMandy: I prefer mails not to be cached in-ram only. as this requires a lot of ram given modern e-mails with hardly any performance gain over disk based caching.
<PulkoMandy>
network usage for my current ways to get mail (mutt through ssh or webmail) seem fine, and this should be a ot better even without caching
<nephele>
one is using a lot of network and disk, and the other is not using much of either ;)
<phschafft>
also a disk based cache can be maintained over reboots/restarts.
<phschafft>
nephele: !?
<nephele>
phschafft: the one downloading *all* your email uses a lot more network than the one not doing that
* phschafft
has a very strong feeling that PulkoMandy nor nephele seems to actually understand what he tries to say. maybe lost in other models...
<PulkoMandy>
Yes, it can, but I don't want that. I will manually copy the mails I want to access offline to a separate partition, I don't want some algorithm to decide what gets stored there and more importantly, what gets deleted
<phschafft>
nephele: yes, and what does such an implementation to do with anything I said?
<nephele>
phschafft: you claimed one was using more network and the other more disk, while I think the mail_daemon will use more of both compared to the imapfs
<nephele>
anyway, yes, we will have a cache. How that looks exactly will be an implementation detail we can hammer out in code review...
<phschafft>
nephele: I fail to see how something that downloads everything once uses more traffic than something that downloads everything every time it is requested again.
<phschafft>
unless you just don't access anything. ;)
<phschafft>
that said, time for T&K.
<phschafft>
happy hacking!
<PulkoMandy>
Depends how full your mailboxes are, I guess
<nephele>
No, i don't acess everything. And also imap_daemon will re-download the stuff in all folders (except the actuall file content) on each refresh to compare it with what we have now
<PulkoMandy>
Idon't really need toever downldad all these mails from 2006 :)
<phschafft>
so imap_daemon has bugs and that makes it bad. whuha!
<phschafft>
whatever, see you after T&K.
<nephele>
It's not a bug
<nephele>
PulkoMandy: personally I would like to mark some imap folders as important, and to download those permanently. But still respect whatever the imap server said (i.e in regards to deletion)
<nephele>
But that to me is GUI stuff that can be implemented later without too much effort, i think
<PulkoMandy>
That could be done by any generic directory sync tool
<nephele>
No?
<nephele>
This isn't a backup or an archive stuff
<nephele>
I just want some emails ready for quick access even if the network goes down temporarily
<nephele>
Or say, for the case of rebooted the computer and the email programm is open, and will be reopened, that that email would be cached over a reboot
cassisian has quit [Remote host closed the connection]
<nephele>
PulkoMandy: where can i find sshfs?
<PulkoMandy>
I think it's in fuse examples?
<nephele>
where is that? :)
<PulkoMandy>
You can look at the sshfs_fuse recipe in haikuports
<nephele>
ah okay, i thought this was in-tree :)
cassisian has joined #haiku
cassisian has quit [Ping timeout: 480 seconds]
KitsunePrefecture has joined #haiku
KitsunePrefecture has quit []
stux|away has joined #haiku
stux- has quit [Ping timeout: 480 seconds]
cassisian has joined #haiku
cassisian has quit [Ping timeout: 480 seconds]
<Begasus[m]>
PulkoMandy: no idea where tokio would be used in anythin gui I could find, but I guess the issue can be closed
<kallisti5[m]>
ok. If you were having trouble getting to Haikuports from a stable release of Haiku, it should be resolved
janking has quit [Quit: Vision[]: i've been blurred!]
<kallisti5[m]>
tldr; i forgot we symlink r1beta5 -> current on the old repo which means requests to r1beta5 were failing. The solution involved feature additions to hpkgbouncer so it took me a bit to address 🙃 sorry for the issues this morning
tetrislife has left #haiku [Error from remote client]
* Begasus[m]
checks ...
<Begasus[m]>
*** Failed to download package qt_creator: Resource not found
<Begasus[m]>
is there something in some cache blocking this kallisti5 ?
<kallisti5[m]>
package_repo shows a valid repo file
<kallisti5[m]>
We can resume transfers. If you interrupt the package download, and give wget -c to resume, it resumes just fine
<nephele>
> tldr; i forgot we symlink r1beta5 -> current on the old repo which means requests to r1beta5 were failing
<nephele>
For maintaining the "one" mirror problems like that are expected, unfortuntately :)
kinkinkijkin has joined #haiku
<kallisti5[m]>
nephele: yeah, the goal is to have a bunch of mirrors... getting this done was one of the blockers to ti
<kallisti5[m]>
*it
<kallisti5[m]>
we were way too spread out. Now with everything consolidated the steps to make it happen are a lot easier
<Begasus[m]>
provided it works* :)
<nephele>
less work for the sysadmin team is always good ;)
<kallisti5[m]>
team == ~2 people
<Begasus[m]>
Refreshing repository "HaikuPorts" failedValidating checksum for HaikuPorts...: I/O error
<nephele>
in german we say TEAM = "Toll, Ein anderer machts" a.ka "Nice, someone else has to do it"
<Begasus[m]>
lol nephele
<nephele>
hmm, codeberg issue is open for a week, so far no "official" response
<nephele>
we will see i guess :/
mmu_man has joined #haiku
<kallisti5[m]>
PulkoMandy: are the repo files cached anywhere on haiku systems by the package kit?
<PulkoMandy>
kallisti5[m]: Yes, but pkgman does a HEAD request on the online version everytime it starts to check if there is a newer version available
<kallisti5[m]>
PulkoMandy: wait... it's a HTTP 303
<kallisti5[m]>
eh. That's how the Haiku repo has worked
<kallisti5[m]>
Begasus: yeah, that's an old checksum
<kallisti5[m]>
8ca5e36cf9f5e8e02d8f5feaa6f210ef116e2302adb7554f86f1655a5542571d is latest
<Begasus[m]>
I didn't change it here :)
<kallisti5[m]>
so, why is Haiku failing to realize the checksum differs
<PulkoMandy>
I don't remember what we check in the reply, supposedly there should be a last-modified header or something, but if there isn't the repo file will be re-downloaded each time
<kallisti5[m]>
date: Sun, 29 Jun 2025 15:10:44 GMT
<PulkoMandy>
Seems quite fresh
<kallisti5[m]>
wait
<kallisti5[m]>
last-modified: Sat, 28 Jun 2025 23:02:25 GMT
<kallisti5[m]>
OK
<kallisti5[m]>
I need to touch it
<kallisti5[m]>
it served the wrong file before... and this one is dated 28th
<kallisti5[m]>
not sure why... caches should expire in 10mins
<kallisti5[m]>
Begasus: thank you for testing... that was two unexpected issues layered 😆
<Begasus[m]>
np kallisti5 , nothing important to pull in, but having it working is always a good thing, so credits and a big thumbs up for the quick fix! :)
HaikuUser has quit [Remote host closed the connection]
<Begasus[m]>
oh wait!
<kallisti5[m]>
runs
<Begasus[m]>
I still have master in repositories :(
<kallisti5[m]>
oh, that's ok. swap it out and it should still work
<kallisti5[m]>
you mean current?
<Begasus[m]>
still fails enabling r1beta5
<kallisti5[m]>
wait...
<kallisti5[m]>
OH NO
<kallisti5[m]>
i'm a dumb ass 😢
<Begasus[m]>
should I take the credits back? :P
<kallisti5[m]>
I aliased the version not the branch
<Begasus[m]>
774 in .obsolote too here, need to clean those too now and then
<Begasus[m]>
grabbing gst_plugins_bad-1.26.3-1-x86_64.hpkg and moving it to /Opslag/haikuports/packages/gst_plugins_bad-1.26.3-1-x86_64.hpkg
<Begasus[m]>
k, that was the last one in the list :)
<Begasus[m]>
kallisti5 (@kallisti5:matrix.org) what's up with zstd, still needs work?
Skipp_OSX has joined #haiku
<Begasus[m]>
errr .... need some more time to get back on par there :)
HaikuUser has joined #haiku
HaikuUser has quit []
<kallisti5[m]>
oh, no. I definitely erased repos this time
<kallisti5[m]>
back to the drawing board 😆
<Begasus[m]>
cmake4 still OK without libarchive dep
<Begasus[m]>
so far for a lazy Sunday :P
bbjimmy has joined #haiku
<kallisti5[m]>
OK so, the inventory of packages in the repository needs to be in /var/buildmaster/haikuports/packages, NOT /var/buildmaster/repository/packages
<Begasus[m]>
getting there :)
<Begasus[m]>
I'm closing down here, thanks for your work on this again kallisti5 (@kallisti5:matrix.org) !
<Begasus[m]>
cu peeps
r6fej has quit [Ping timeout: 480 seconds]
r6fej has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
HaikuUser has joined #haiku
HaikuUser has quit []
arjen has joined #haiku
bunkermatty has quit [Remote host closed the connection]
Nasina has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
Nasina has joined #haiku
Nasina has quit [Remote host closed the connection]
mmu_man has quit [Ping timeout: 480 seconds]
mmu_man has joined #haiku
Nasina has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
arjen_ has joined #haiku
arjen has quit [Read error: Connection reset by peer]
cassisian has quit [Remote host closed the connection]
cassisian has joined #haiku
HaikuUser has joined #haiku
HaikuUser2 has joined #haiku
HaikuUser2 has quit []
moparisthebest has joined #haiku
HaikuUser has quit []
HaikuUser has joined #haiku
HaikuUser has quit [Remote host closed the connection]
HaikuUser has joined #haiku
Nasina has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
HaikuUser has quit [Quit: Vision[]: i've been blurred!]
HaikuUser has joined #haiku
HaikuUser has quit []
jmairboeck has quit [Quit: Konversation terminated!]
dodo75 has quit [Quit: Vision[]: i've been blurred!]
neoncortex has joined #haiku
xet7 has quit [Remote host closed the connection]
nephele has joined #haiku
Nasina has joined #haiku
AndrevS has quit [Quit: umount /dev/irc]
Nasina has quit [Read error: Connection reset by peer]
Nasina has joined #haiku
Nasina has quit [Read error: Connection reset by peer]
e1z07567351 has quit [Read error: Connection reset by peer]
e1z07567351 has joined #haiku
Aedil has quit [Quit: leaving]
Skipp_OSX has joined #haiku
<Skipp_OSX>
Beta6 soon or do we have more time?
<nephele>
we have some time i think
<nephele>
what are you planning? :D
<Skipp_OSX>
nothing, just fixing bugs
<nephele>
we have the control colors fix in now by the way :)
<nephele>
still some lingering things, but mainly fixed not
<nephele>
now*
<Skipp_OSX>
yeah I saw
<nephele>
by the way, now that you are here, what do you think about the controllook api as it is now?
<Skipp_OSX>
I don't know why you want to remove all my updates to rounded buttons but I guess if it must be done it's fine.
<nephele>
Well, you want some rounded buttons, right?
<nephele>
the current api makes this the responsibility of the caller, and it means every controllook has to implement it
<nephele>
i guess it is a question of if you consider this part of the visual style of a controllook or not, i would argue that it is, and that therefore the api should have no effect on this.
<Skipp_OSX>
Sure, isn't that how it should work?
<nephele>
Well, it makes controllooks a bit harder to implement having these options, BeControlLooks code looks a lot worse because of all the indirection
<Skipp_OSX>
Idk it's an API you implement the parts you want.
<Skipp_OSX>
Yeah but idk where else that would go
<nephele>
My suggestion would be that this rendering detail would be an option of the controllook itself, like a config option
<nephele>
then you could actually turn on rounded buttons aswell, i mean, like it is now you'd have to recompile the interface kit for this, no?
<Skipp_OSX>
Well I thought we were going to go all rounded at the time but people didn't like that idea.
<nephele>
I like rounded buttons. It's a nice accent, but at the same time i think HaikuControlLooks "default", as it is now is a nice middle ground
<Skipp_OSX>
I guess
<nephele>
What would be your idea?
<Skipp_OSX>
The way it is now was my idea
<Skipp_OSX>
and then put 4 in BButton to get rounded corners
<Skipp_OSX>
but like I said, people didn't like that idea
<nephele>
Well, that needs recompiling the interface kit then? hmmm. Or do you think this should be an option in the interface kit?
<nephele>
but then, if it is an option it's still wierd how HaikuControlLook implements this, and the other two ignore it
<Skipp_OSX>
That's how options work.
<nephele>
... I have an alternative idea, I could move the Rounded constructors to BControlLook, that way atleast BControlLook and FlatControlLook do not have to have this indirection in their own code
<nephele>
so we atleast hide some of that nasty stuff untill you reimplement it
<Skipp_OSX>
I guess, it's not really my focus right now.
<nephele>
(though HaikuControlLooks code is not that clear because of the indirection either)
<nephele>
Well, I wanted to run it by you in any case. Don't want to step on your toes :)
<Skipp_OSX>
I did a deep dive into BMenu, I don't like how nav works.
<nephele>
what do you mean by nav?
<Skipp_OSX>
Using arrow keys to move around menus
<Skipp_OSX>
it doesn't work if you're hovered over a supermenu
<Skipp_OSX>
It never did, but I still don't like it
<nephele>
Huh, you are right. I did not notice, but that is bad
<Skipp_OSX>
BeOS is super broken too
<Skipp_OSX>
well ok not super broken, but broken
<nephele>
My main problem with keyboard nav has been offscreen menus not scrolling when using keys, only with the mouse
<Skipp_OSX>
offscreen menus?
<nephele>
I mean a menu that extends out of the screen
<nephele>
like a long list of options
<Skipp_OSX>
like too tall to fit you mean
<nephele>
yeah
<Skipp_OSX>
bc when mouse is idling over menu all keystrokes are ignored
cassisia_ has joined #haiku
<Skipp_OSX>
if you wiggle your mouse you can use kb but I guess it doesn't scroll :/
<nephele>
I just mean phyiscally scrolling, if you aproach the end of the list with the mouse it will scroll to reveal more options, but it will not do that if you aproach the end of the list using keyboard keys
<Skipp_OSX>
yeah true
<Skipp_OSX>
if you do page up/page down in scrolls
cassisian has quit [Ping timeout: 480 seconds]
HaikuUser has joined #haiku
<Skipp_OSX>
it's an easy fix, but we should really rework the whole thing based on menu layout
<Skipp_OSX>
meaning rows, columns or matrix layout
HaikuUser has quit []
<nephele>
preferences->screen has a matrix layout if you have a huge list or resolutions
<Skipp_OSX>
yeah and arrow keys don't work
<Skipp_OSX>
arrow keys don't work at all over submenus on BeOS... and it locked up
<Skipp_OSX>
so could be worse
<nephele>
well, but it might be a start to work off, so just wanted you to be aware of it
arjen_ has quit [Read error: Connection reset by peer]
<nephele>
I still want to write some docs for the menu layout for windows at some point... I used that in StyledEdit and i think one other app, but there was no docs at all on how to use this
arjen has joined #haiku
<nephele>
(and ideally it would be cool that if a menubar is added in a layout, that this actually could go into a global menu bar... ;)
Nasina has quit [Read error: Connection reset by peer]
HaikuUser has joined #haiku
HaikuUser has quit []
Nasina has joined #haiku
Nasina has quit [Read error: Connection reset by peer]