2006 Aug 07 08:33:53  <csant-> upgrading to the last released ion3 should be safe, shouldn't it?
2006 Aug 07 08:35:29  <tuomov> yeah, but if you're going to do any customisation, you may have to do it again soon
2006 Aug 07 08:35:38  <tuomov> (I'll try to keep the layout save files compatible, however)
2006 Aug 07 08:35:53  <tuomov> (for the next release, not in darcs)
2006 Aug 07 08:37:20  <tuomov> if all your customisations are going to be a few new bindings in cfg_ion.lua or cfg_user.lua, then it should be ok to upgrade
2006 Aug 07 08:37:34  <tuomov> but if you're going to do bigger modifications, I suggest waiting
2006 Aug 07 08:38:04  -*- csant- is waiting for a usable darcs
2006 Aug 07 08:38:13  <-- maxauthority (n=maxautho@193.170.48.242) has quit (Read error: 110 (Connection timed out))
2006 Aug 07 08:38:13  <csant-> then I'll re-do all my customization
2006 Aug 07 08:38:16  <opqdonut> any larger feature improvements, tuomov?
2006 Aug 07 08:39:00  <tuomov> not sure about improvements, but the architechture is quite different
2006 Aug 07 08:39:07  <tuomov> there are no WIonWS, WFloatWS and so on anymore
2006 Aug 07 08:39:13  <opqdonut> wow, sounds cool
2006 Aug 07 08:39:30  <tuomov> there's just WGroup(WS|CW) (workspace, clientwin)
2006 Aug 07 08:39:48  <tuomov> which are sort of improved floating workspaces
2006 Aug 07 08:40:08  <opqdonut> and tiling will be done on top of that?
2006 Aug 07 08:40:14  <tuomov> normally all client windos are contained in WGroupCW
2006 Aug 07 08:40:21  <tuomov> and all workspaces are WGroupWS
2006 Aug 07 08:40:45  <tuomov> WTiling may be set as a 'bottom' of a WGroupWS like the main window can be the 'bottom' of a WGroupCW
2006 Aug 07 08:41:25  <opqdonut> that latter case would be a full-screen app?
2006 Aug 07 08:41:27  <tuomov> (the bottom is an object on the workspace with some special handling)
2006 Aug 07 08:41:39  <opqdonut> ho
2006 Aug 07 08:41:41  <opqdonut> *oh
2006 Aug 07 08:42:05  <tuomov> client windows are in a WGroupCW in tiled frames too
2006 Aug 07 08:42:34  <tuomov> not in floating frames atm, however, and atm full screen mode doesn't work for them (this has to be resolevd..)
2006 Aug 07 08:44:11  <tuomov> (it would be easy to add dummy fs support that doesn't do anything for the transients, but I'd like full-screening a window in a floating frame to include the transients on the fs WGroupCW)
2006 Aug 07 08:44:45  <tuomov> so, before this all can be released, I have to fix that issue
2006 Aug 07 08:45:19  <tuomov> there's also the issue of navigation, and I've been working on reworking the internal structure of WMPlex too do make that simpler
2006 Aug 07 08:45:41  <tuomov> then there's some other minor issues
2006 Aug 07 08:46:09  <tuomov> but after that, the detach.lua functionality should be implementable much more neatly
2006 Aug 07 08:46:35  <tuomov> and it should also be possible to convert between "floating" and "tiled" workspaces
2006 Aug 07 08:47:50  --> maxauthority (n=maxautho@193.170.48.242) has joined #ion
2006 Aug 07 08:48:25  <csant-> nice :)
2006 Aug 07 08:48:47  <csant-> looking forward to it.
2006 Aug 07 08:49:09  <tuomov> the primary reason I set out to do this was the issue of transients and such in tiny frames..
2006 Aug 07 08:49:40  <tuomov> it seems that the best way to deal with them is to simply "detach" them
2006 Aug 07 08:50:21  <csant-> but will transients that fit into the parent window's frame still be treated as now?
2006 Aug 07 08:50:38  <tuomov> yep, and perhaps detaching won't be automatic
2006 Aug 07 08:50:47  <tuomov> at least it probably shouldn't be automatic for inactive windows
2006 Aug 07 08:51:19  <csant-> one of the things I really love about ion is its way of treating transients
2006 Aug 07 08:52:03  <csant-> it probably is the most sensible way of treating them that I have come across in WMs
2006 Aug 07 08:52:09  <tuomov> you shouldn't notice much difference between old ion and new ion unless you choose to use the new features, aside from possible automatic detach of transients in small frames
2006 Aug 07 08:53:32  <tuomov> mostly what I'm doing is de-hierarchifying the architechture
2006 Aug 07 08:54:03  <tuomov> to some extent
2006 Aug 07 08:54:59  <csant-> but configuration will be radically different?
2006 Aug 07 08:55:27  <tuomov> to the extent that there's no WIonWS and WFloatWS class, and many of WMPlex functions will also be different
2006 Aug 07 09:43:23  <tuomov> hmm... 29C... and freezing in the office
2006 Aug 07 09:43:57  <tuomov> I don't get why, if there's air conditioning, they always keep it at so low temperatures
2006 Aug 07 09:45:48  <csant-> open the window - it works fine in this office :)
2006 Aug 07 09:46:10  <tuomov> no window..
2006 Aug 07 09:47:09  <tuomov> well, there are windows and a door to an interior walkway with sauna temperatures
2006 Aug 07 09:47:35  --> markues (n=worker@dslb-082-083-253-039.pools.arcor-ip.net) has joined #ion
2006 Aug 07 09:48:10  <tuomov> [offices][cubicles][window][walkway][open space to three floors down][sky windows\
2006 Aug 07 09:48:27  <mainzelM> did you complain too much about conventional window managers and hence they decided to put you in a room without windows?
2006 Aug 07 09:48:40  <tuomov> and of course the windows always have curtains on them, so there's no sunlight
2006 Aug 07 09:48:59  <tuomov> stupid glass architechture...
2006 Aug 07 09:49:30  <tuomov> wherever there's a window in the interior of a building, there's a curtain
2006 Aug 07 09:52:12  <tuomov> it's so lame to have to use a flourescent lamp on this weather for light
2006 Aug 07 09:52:27  <tuomov> but opening the curtains to let some light in would be too distracting for the people next to the windows to the walkway
2006 Aug 07 09:52:51  <tuomov> architects... they're just like web and UI architects
2006 Aug 07 09:53:02  <tuomov> no regard for functionality, just cool design
2006 Aug 07 09:54:40  <tuomov> and glass buildings in this climate... what a waste of energy!
2006 Aug 07 09:54:59  <tuomov> -30 is just as common as +30
2006 Aug 07 09:57:46  <tuomov> the early average temperature must be somewhere close to zero
2006 Aug 07 10:00:40  <tuomov> about 3C it seems
2006 Aug 07 10:02:26  <tuomov> s/early/yearly/
2006 Aug 07 10:12:46  <-- nemo (n=nemo@220-253-116-146.ACT.netspace.net.au) has quit ("rebootarific")
2006 Aug 07 10:27:06  --> Etn40ff (n=VulK@host191-151.pool8251.interbusiness.it) has joined #ion
2006 Aug 07 10:27:09  <Etn40ff> re
2006 Aug 07 10:27:35  <-- Etn40ff (n=VulK@host191-151.pool8251.interbusiness.it) has quit (Client Quit)
2006 Aug 07 10:27:56  --> Etn40ff (n=VulK@host191-151.pool8251.interbusiness.it) has joined #ion
2006 Aug 07 10:34:02  <mainzelM> ./configure of ion3 says
2006 Aug 07 10:34:02  <mainzelM> checking for dlopen in -ldl... no
2006 Aug 07 10:34:02  <mainzelM> configure: error: *** missing library dl
2006 Aug 07 10:34:07  <mainzelM> on FreeBSD 5.4
2006 Aug 07 10:37:17  <mainzelM> (dlopen is in libc on FreeBSD)
2006 Aug 07 10:46:17  <-- Talisker (n=shai@IGLD-80-230-179-212.inter.net.il) has quit (Connection timed out)
2006 Aug 07 10:52:50  <-- tyler (n=tyler@ARennes-356-1-151-43.w86-220.abo.wanadoo.fr) has quit (Read error: 110 (Connection timed out))
2006 Aug 07 10:58:37  --> tyler (n=tyler@ARennes-356-1-141-23.w86-220.abo.wanadoo.fr) has joined #ion
2006 Aug 07 11:10:16  --> nemo (n=nemo@220-253-116-146.ACT.netspace.net.au) has joined #ion
2006 Aug 07 11:12:26  --> Talisker (n=shai@IGLD-80-230-104-141.inter.net.il) has joined #ion
2006 Aug 07 11:21:39  --> shai_ (n=shai@IGLD-80-230-170-72.inter.net.il) has joined #ion
2006 Aug 07 11:23:40  <-- Etn40ff (n=VulK@host191-151.pool8251.interbusiness.it) has quit (Read error: 104 (Connection reset by peer))
2006 Aug 07 11:26:44  --> Etn40ff (n=VulK@host191-151.pool8251.interbusiness.it) has joined #ion
2006 Aug 07 11:30:42  <-- nemo (n=nemo@220-253-116-146.ACT.netspace.net.au) has quit ("reboot again?!")
2006 Aug 07 11:33:24  <-- eno-away (n=eno@adsl-70-137-173-220.dsl.snfc21.sbcglobal.net) has quit ("[BX] With a BitchX here and a BitchX there, here a BitchX there a BitchX everywhere a BitchX")
2006 Aug 07 11:35:56  <-- Talisker (n=shai@IGLD-80-230-104-141.inter.net.il) has quit (Read error: 110 (Connection timed out))
2006 Aug 07 11:58:05  --> nemo (n=nemo@220-253-116-146.ACT.netspace.net.au) has joined #ion
2006 Aug 07 12:05:49  <-- tyler (n=tyler@ARennes-356-1-141-23.w86-220.abo.wanadoo.fr) has quit (Read error: 110 (Connection timed out))
2006 Aug 07 12:07:30  -=- shai_ is now known as Talisker
2006 Aug 07 12:08:30  --> tyler (n=tyler@ARennes-356-1-119-152.w86-220.abo.wanadoo.fr) has joined #ion
2006 Aug 07 12:09:14  <tuomov> hmm.. google banging day
2006 Aug 07 12:09:57  <tuomov> some 20 hits today from google to the non-blog main page
2006 Aug 07 12:18:30  <-- mulligan (n=user@e178021222.adsl.alicedsl.de) has quit (Read error: 110 (Connection timed out))
2006 Aug 07 12:21:57  --> mulligan (n=user@e178007094.adsl.alicedsl.de) has joined #ion
2006 Aug 07 13:04:22  <amv> how huge a pain is it to maintain a cvs/svn fork repository to which you can easily merge things from upstream?
2006 Aug 07 13:05:07  <deryni> Just use darcs and you can do what you want.
2006 Aug 07 13:05:35  <opqdonut> amv: why would anyone need that
2006 Aug 07 13:05:37  <amv> i'm not sure if i can get that through to the other people here if it can be done in cvs/svn
2006 Aug 07 13:06:36  <deryni> Right, but see darcs is distributed as soon as you grab a copy it is a 'fork repository'.
2006 Aug 07 13:06:39  <deryni> No magic involved.
2006 Aug 07 13:06:46  <deryni> And certainly no cross-propagating.
2006 Aug 07 13:06:47  <amv> opqdonut: we will probably need to make some changes to the project which are somewhat specific to our use and thus will not be incorporated upstream
2006 Aug 07 13:07:24  <amv> deryni: i understand that darcs would be nice. i'm jsut asking how huge a pain it is to do with cvs/svn ;)
2006 Aug 07 13:07:41  <tuomov> impossible, with plain cvs and svn
2006 Aug 07 13:07:45  <amv> is there something you just can't do with them or is it just much easier with darcs?
2006 Aug 07 13:08:06  <tuomov> possible with svk, but far more painful than with darcs/mercurial/bzr2/monotone/...
2006 Aug 07 13:08:07  <deryni> That's not true, it is just a lot of work. Assuming all you want is a local repository that is up-to-date and has your own patches.
2006 Aug 07 13:08:09  <-- Talisker (n=shai@IGLD-80-230-170-72.inter.net.il) has quit (Read error: 110 (Connection timed out))
2006 Aug 07 13:09:09  <tuomov> but if you insist on svn/cvs, svk is what you want
2006 Aug 07 13:09:21  <tuomov> it can track upstream svn repository locally
2006 Aug 07 13:09:49  <tuomov> but since svn isn't designed for distributed operation, it isn't as great as vcs's designed for distributed use
2006 Aug 07 13:10:20  <amv> the project i need to fork doesn't even seem to have have a public repository
2006 Aug 07 13:10:24  <amv> (TinyMCE)
2006 Aug 07 13:10:50  <deryni> Then you probably don't want your own repository for the software, just a repository for your patches to it I'd imagine.
2006 Aug 07 13:12:08  <amv> well setting up my own repository would make it possible to track the upstream changes at some level
2006 Aug 07 13:12:26  <tuomov> what you want to do is import a release into a local repository, tag/branch it, record patches to it, and when a new release is available, patch the original import tag/branch, and pull the changes back to your patch repo
2006 Aug 07 13:14:33  --> riwa (n=riwa@c-31eb70d5.016-163-6c756e10.cust.bredbandsbolaget.se) has joined #ion
2006 Aug 07 13:20:14  <-- mulligan (n=user@e178007094.adsl.alicedsl.de) has quit (Remote closed the connection)
2006 Aug 07 13:20:41  <amv> yeah i figured out something like that but it's still a bit fuzzy hwo the actual merging will work and when i should branch and when will simply taggin be enough.. 
2006 Aug 07 13:20:57  --> mulligan (n=user@e178007094.adsl.alicedsl.de) has joined #ion
2006 Aug 07 13:21:15  <opqdonut> http://www.albinoblacksheep.com/flash/llama.php
2006 Aug 07 13:23:56  <philips> amv: just use quilt and port the patches on each new release?
2006 Aug 07 13:27:10  <-- elmex (n=ryzrk@dslb-088-064-032-033.pools.arcor-ip.net) has quit ("=>weg")
2006 Aug 07 13:30:00  <riwa> Where do I set "statusd_*" to get added in the statusbar?
2006 Aug 07 13:30:49  <opqdonut> in cfg_statusbar.lua, the template line
2006 Aug 07 13:31:35  <amv> philips: what will quilt offer me over using svn/cvs?
2006 Aug 07 13:31:38  <riwa> opqdonut: k. thx :P
2006 Aug 07 13:32:16  <amv> riwa: if you put some template slot there, a script which corresponding name will be loaded.
2006 Aug 07 13:34:51  <riwa> amv: I'm note sure I get you.
2006 Aug 07 13:36:31  <riwa> Besides, I have mod_statusbar.create and *.launch_statusd. Should I just add "statusd_batt.lua" somewhere in the latter one?
2006 Aug 07 13:37:12  <deryni> riwa: In the template='...' line you need to add the meter name for the script you want to be using.
2006 Aug 07 13:37:13  <philips> amv: quilt is designed for maintaining patches on top of a changing code base, svn/cvs are not
2006 Aug 07 13:41:26  <amv> riwa: if you add a %batt template, statusd tries to find statusd_batt.lua and load it
2006 Aug 07 13:42:53  <amv> philips: yeah i got that, but what benefits do i get? =)
2006 Aug 07 13:44:11  <philips> amv: I am confused, I thought that you were looking for a system to manage patches on top of an upstream source.  Quilt does that.  The advantage is that the patches are kept seperate for rebasing on top of a new version...
2006 Aug 07 13:45:26  <tuomov> quilt is just a poor man's version control system, afaik
2006 Aug 07 13:45:52  <deryni> In some senses it is.
2006 Aug 07 13:46:17  <deryni> But I do find it a tad more convenient if all I am doing is keeping a small set of patches current against an upstream repository.
2006 Aug 07 13:46:25  <amv> i need to have something to show if i intend to justify adding yet another system for our developers to learn to use
2006 Aug 07 13:48:19  <amv> so how does quilt work? does it keep the original source somewhere and store the patches somewhere and then give me a "up to date" branch to work with for the changes?
2006 Aug 07 13:50:30  <deryni> No, it's a bit stupider than that. You make a patches directory in the source directory (you might be able to put it somewhere else) and then you push/pop patches onto the working directory.
2006 Aug 07 13:51:51  <philips> tuomov: I don't think it is a version control system at all.  It is simply a way to maintain patches on top of another code base.
2006 Aug 07 13:52:42  <tuomov> that's what the darcs way of version control is...
2006 Aug 07 13:52:49  <tuomov> patch maintenance and propagation
2006 Aug 07 13:53:02  <tuomov> perhaps "version" is the wrong word
2006 Aug 07 13:53:48  <tuomov> cms (configuration management system) might be more appropriate, but that term is already taken and I think it means something more than "version" control
2006 Aug 07 13:55:00  <opqdonut> pcs?
2006 Aug 07 13:55:16  <amv> so it does not help me get a better idea of what changes happen in the upstream? does it help me keep track of patches that are working for a previous version of the upstream software if something breaks horribly with the newest?
2006 Aug 07 13:55:19  <opqdonut> dvcs is good enough tho imo
2006 Aug 07 13:56:24  <amv> for example if i want to work simultaneously on one "patch set" to fix my patches to apply to the newest one and still push some new changes to the older, still functionin version, and then merge these at some point?
2006 Aug 07 13:56:32  <tuomov> most dvcs's do still have versions, sort of
2006 Aug 07 13:56:47  <tuomov> the patches must be applied linearly, although there may be multiple heads
2006 Aug 07 13:57:19  <tuomov> they don't support cherry-picking properly
2006 Aug 07 13:57:31  <philips> amv: www.suse.de/~agruen/quilt.pdf 
2006 Aug 07 13:57:44  <tuomov> you can't just choose to take one patch from another repository, you must take everything that it depends on
2006 Aug 07 13:58:00  <tuomov> you have to merge the whole version that that patch identifies
2006 Aug 07 13:58:07  <riwa> how do I make ion create the cfg_statusbar.lua? I incidently removed the file and would like it documented again..
2006 Aug 07 13:58:48  <tuomov> in darcs, a patch can not be identified with a state of the repository
2006 Aug 07 13:59:04  <opqdonut> riwa: you can copy it from /etc/X11/ion3 if you deleted only your own copy in ~/.ion3
2006 Aug 07 13:59:04  <tuomov> patches may be in any order, modulo conflicting patches
2006 Aug 07 13:59:19  <riwa> opqdonut: thanks
2006 Aug 07 13:59:30  <opqdonut> np
2006 Aug 07 13:59:48  <riwa> opqdonut: But it's not there..
2006 Aug 07 14:00:14  <opqdonut> oh? somewhere in /usr/local then if you installed by hand
2006 Aug 07 14:00:16  --> SeeChoas (n=lars@p5496C021.dip0.t-ipconnect.de) has joined #ion
2006 Aug 07 14:00:28  <riwa> opqdonut: I installed with my package manager.
2006 Aug 07 14:01:02  <opqdonut> riwa: then ask it where it put the global config files
2006 Aug 07 14:01:11  <riwa> opqdonut: found it.. Was directly in /etc/ion3/.
2006 Aug 07 14:01:16  <opqdonut> okay
2006 Aug 07 14:02:32  <-- hnaz (n=hnaz@5.161.119.82.versanetonline.de) has quit ("hardware")
2006 Aug 07 14:03:42  -=- kylem_ is now known as kylem
2006 Aug 07 14:08:08  <riwa> Wouldn't the template look like this? template="[ %date ][ %cpufreq ][ load:% %>load ] %filler%systray",
2006 Aug 07 14:08:33  <opqdonut> yup
2006 Aug 07 14:09:03  <riwa> But it doesn't work. I get a "?" in the %cpufreq" space and it doesn't work with neither one (iv'e tried several).
2006 Aug 07 14:09:18  <opqdonut> have you got statusd_cpufreq.lua?
2006 Aug 07 14:09:35  <riwa> opqdonut: yes
2006 Aug 07 14:09:37  <opqdonut> and check your .xsession-errors, it might've thrown some errors
2006 Aug 07 14:09:44  <deryni> Is statusd_cpufreq.lua looking in the right place for your system?
2006 Aug 07 14:09:49  <opqdonut> (or the vc if you used startx)
2006 Aug 07 14:10:03  <riwa> opqdonut: where do I do that? (I use startx)
2006 Aug 07 14:10:10  <deryni> startx will still log to .xsession-errors if you used a .xsession file instead of a .xinitrc fil.e
2006 Aug 07 14:10:14  <deryni> (Most people don't do that).
2006 Aug 07 14:10:25  --> honkykong (n=nik@cpe-24-195-247-12.nycap.res.rr.com) has joined #ion
2006 Aug 07 14:10:36  <opqdonut> riwa: ctrl-alt-f1 most probably
2006 Aug 07 14:10:43  <riwa> deryni: It should be. Wouldn't that be where all the look_* and cfg_* are located?
2006 Aug 07 14:10:52  <riwa> opqdonut: Oh.. Like that...
2006 Aug 07 14:11:05  <deryni> riwa: What? I mean is the script itself looking at the right data source for your machine.
2006 Aug 07 14:11:07  <opqdonut> riwa: look inside the script, at what place it reads the info
2006 Aug 07 14:12:28  <riwa> opqdonut: It certainly spits out errors already. I keeps complaining about "dopath" and "skipping unnamed C functions".
2006 Aug 07 14:12:36  <opqdonut> :o
2006 Aug 07 14:12:41  --> daniel_ (i=daniel@unaffiliated/daniel) has joined #ion
2006 Aug 07 14:12:59  <opqdonut> sounds like some conf file isn't found
2006 Aug 07 14:13:09  <opqdonut> does it give the filename in the dopath error?
2006 Aug 07 14:13:22  <riwa> opqdonut: several maybie.. I need to scroll up to read all the output.
2006 Aug 07 14:13:50  <riwa> opqdonut: I haven
2006 Aug 07 14:14:12  <riwa> 't configured scrolling in the console yet but as far as I can see. It'
2006 Aug 07 14:14:32  <riwa> s only cfg_ion.lua.
2006 Aug 07 14:16:59  <XTL> puuh