Re: so, what now?
On Mon, 2010-04-05 at 03:52 +0400, Mikle Krutov wrote: > On Mon, Apr 05, 2010 at 12:54:31AM +0200, M Rawash wrote: > > now that tuomov has finally severed his last remaining connection with > > ion, what do we (committed ion users) do now? > > > > I spent the last few days committedly trying out other WMs, and i can > > say this with absolute certainty: there's nothing out there that's even > > comparable! (the closest thing i could find, in terms of > > what-you-can-do, was 'awesome', but it subscribes to a very different > > philosophy than ion's, and is kind of a mess) > > > > so, I need to hear some ideas about the future of this project (not > > simply suggestions for other WM's; i will NOT switch to anything that's > > currently on the market, and that's my final decision!), I do have a few > > ideas of my own, but i want to know what everybody else thinks; after > > all, I'm just new-comer compared to the majority of the remaining users > > here (been using Ion for only a few months now), I don't know anything > > about the community (or what's left of it), how it works, or how the > > author's actions affected it. so i think it's best of somebody else took > > the initiative and got this conversation going... > Firstly, why not to fork ion3 to add needed functionality? (i do not > know is there any 'no forks' etc in it's license because the ion3's > page is down) The last version of the licence can be seen at: http://git.decadent.org.uk/gitweb?p=ion3.git;a=blob;f=LICENSE;h=fdfa652c68cd5d06ec2383e7ea4429e6dc5cb4d0;hb=e3aec18706513a87eaa7839dfdaf7e0fcd0d8d2a You can fork it but then you must rename it. Tuomo said in the past that this included functional use of the name 'Ion' such as in the installation directories. This restriction was introduced in the copyright licence of ion-3rc-20070506; if you fork an earlier version then you should only have to consider trademark rights. > Secondly, there was some project like 'ion3 with scripts in python but > without tuomo', unfortunately i can not give you it's name or link to > it. It was dead when i lurked about it, but it has no licensing > problems - so you can just fork it and write the code you need. [...] Qtile? wmii? There's a long list of tiling window managers at <http://en.wikipedia.org/wiki/Tiling_window_manager#List_of_tiling_window_managers_for_X>. Ben. -- Ben Hutchings Once a job is fouled up, anything done to improve it makes it worse. signature.asc Description: This is a digitally signed message part
Re: Youtube fullscreen not working
On Sun, 2009-11-22 at 15:10 -0500, Timandahaf wrote: > When I hit the fullscreen button inside a youtube video in firefox, the > video just disappears, and I'm left with the browser without the video. > There's no new full-screen workspace created. Pressing either alt-tab or > Escape has the same effect: the video comes back into its original > position embedded inside the browser. I don't think I've ever had this > work correctly in the past either. > > Any idea why this is happening/how to fix this? Flash seems to do stupid things with window focus. I had a workaround for an earlier version of Flash[1] but it no longer works. Ben. [1] http://git.decadent.org.uk/gitweb?p=ion3.git;a=blob;f=debian/cfg_kludge_flash.lua -- Ben Hutchings Unix is many things to many people, but it's never been everything to anybody. signature.asc Description: This is a digitally signed message part
Re: Youtube fullscreen not working
On Sun, 2009-11-22 at 22:16 +, Ben Hutchings wrote: > On Sun, 2009-11-22 at 15:10 -0500, Timandahaf wrote: > > When I hit the fullscreen button inside a youtube video in firefox, the > > video just disappears, and I'm left with the browser without the video. > > There's no new full-screen workspace created. Pressing either alt-tab or > > Escape has the same effect: the video comes back into its original > > position embedded inside the browser. I don't think I've ever had this > > work correctly in the past either. > > > > Any idea why this is happening/how to fix this? > > Flash seems to do stupid things with window focus. I had a workaround > for an earlier version of Flash[1] but it no longer works. > > Ben. > > [1] > http://git.decadent.org.uk/gitweb?p=ion3.git;a=blob;f=debian/cfg_kludge_flash.lua It seems that Flash no longer sets the initial window size to 200x200 sizes but to the current screen size. So changing: match = function(prop, cwin, id) local geom = cwin:geom() return is_fullscreen(cwin) and geom.w == 200 and geom.h == 200 end, to: match = function(prop, cwin, id) return is_fullscreen(cwin); end, makes it behave, though this might have false positives. Ben. -- Ben Hutchings Unix is many things to many people, but it's never been everything to anybody. signature.asc Description: This is a digitally signed message part
Re: Bug#492906: ion3 does not preserve the aspect for tiled windows
On Tue, 2008-07-29 at 20:41 +0200, Berenyi Mihaly wrote: > Package: ion3 > Version: 20080707-3 > Severity: normal > > since the last update (or thereabouts, i cannot be sure, but certainly > since the last few days), tiled mplayer windows no longer have the > correct aspect ratio. this used to work (the usual black bands around > the "extra" space). > it still works with floated windows or fullscreen windows, but tiled > mplayers just fill the entire split they occupy. I think this was an intended effect of this upstream change: Fri May 2 20:14:56 BST 2008 Tuomo Valkonen <[EMAIL PROTECTED]> * WGroup needed "whatever" attach geometry option for proper WGroupCW/WClientWin initial size setup. Tuomo, does this seem correct to you? Ben. -- Ben Hutchings If God had intended Man to program, we'd have been born with serial I/O ports. signature.asc Description: This is a digitally signed message part
Re: How to include Debian menu in to Ion3 installed from source?
On Sat, 2008-07-26 at 23:55 +0200, Tomáš Ebenlendr wrote: > There is ion3 debian package for debian unstable: > http://packages.debian.org/sid/ion3 > download it and get some files from it e.g. by midnight commander. Or: 1. Get the source package from unstable 2. In debian/control, change "x11-utils" to "xbase-clients" 3. Run "dpkg-buildpackage -uc -us -rfakeroot" to rebuild it Ben. (Debian ion3 maintainer) -- Ben Hutchings Everything should be made as simple as possible, but not simpler. - Albert Einstein signature.asc Description: This is a digitally signed message part
Re: Warnings for missing keys for non-critical key bindings
On Wed, 2008-07-23 at 17:01 +, Tuomo Valkonen wrote: > On 2008-07-20, Ben Hutchings <[EMAIL PROTECTED]> wrote: > > Some of the standard bindings refer to keys that aren't in all keymaps > > (e.g. Home in this case). Thankfully they are redundant with other > > bindings. So it seems like it would be OK to suppress the warnings for > > those bindings. > > > > I came up with a patch that makes that change, which I'd like to apply > > to the Debian package. What do you think of this approach? > > Too hacky, and complicates the config files. > > Better to just have a switch to make all the warnings non-logged > (ioncore_warn_nolog), or something. Or maybe it should be the default; > it's not that critical that keysym to keycode conversion fails, while > string to keysym conversion is a clear config error that should be > more prominently notified of. OK, then I'll remove the warning. Ben. -- Ben Hutchings Usenet is essentially a HUGE group of people passing notes in class. - Rachel Kadel, `A Quick Guide to Newsgroup Etiquette' signature.asc Description: This is a digitally signed message part
Re: Bug#490601: fails with many xio errors when started from xdm
On Wed, 2008-07-23 at 16:52 +, Tuomo Valkonen wrote: > On 2008-07-21, Ben Hutchings <[EMAIL PROTECTED]> wrote: > > I don't understand how the server coordinates window changes with the > > WM. > > They don't. The window might disappear any moment. That's what I suspected. So is my patch to ignore BadMatch a reasonable workaround? Ben. -- Ben Hutchings Usenet is essentially a HUGE group of people passing notes in class. - Rachel Kadel, `A Quick Guide to Newsgroup Etiquette' signature.asc Description: This is a digitally signed message part
Re: Bug#490601: fails with many xio errors when started from xdm
I got a report of a weird and unreproducible error in Ion3 startup:
http://bugs.debian.org/490601
BadMatch is not a documented error code for XGetWindowProperty which
sends X_GetProperty requests. However, the implementation in Xorg
appears to return BadMatch if and only if it's passed a reference to a
drawable that isn't a window:
http://gitweb.freedesktop.org/?p=xorg/xserver.git;a=blob;hb=8822110d7d6b684f373fc883aeb7cab9734e9ddb;f=dix/dixutils.c#l197
I don't understand how the server coordinates window changes with the
WM. Is it possible that a window can be completely deleted before the
WM has handled it? In that case there is presumably a race condition
where a window could disppear and another drawable be created with the
same id.
This seems like a bug in Xorg, but we presumably need to work around it.
The obvious patch is below.
Ben.
--- ion3.orig/ioncore/rootwin.c
+++ ion3/ioncore/rootwin.c
@@ -58,10 +58,11 @@
static char msg[128], request[64], num[32];
/* Just ignore bad window and similar errors; makes the rest of
- * the code simpler.
+ * the code simpler. Due to a Xorg bug, window lookups may fail
+ * with BadMatch instead of BadWindow.
*/
if((ev->error_code==BadWindow ||
-(ev->error_code==BadMatch && ev->request_code==X_SetInputFocus) ||
+ev->error_code==BadMatch ||
(ev->error_code==BadDrawable && ev->request_code==X_GetGeometry)) &&
ignore_badwindow)
return 0;
--- END ---
--
Ben Hutchings
Once a job is fouled up, anything done to improve it makes it worse.
signature.asc
Description: This is a digitally signed message part
Warnings for missing keys for non-critical key bindings
if(!ioncore_parse_keybut(ksb_str, &mod, &ksb,
(action!=BINDING_KEYPRESS && action!=-1),
- init_any)){
+ init_any, extra)){
goto fail;
}
}
--- ion3.orig/ioncore/ioncore_bindings.lua
+++ ion3/ioncore/ioncore_bindings.lua
@@ -125,9 +125,12 @@
-- Creates a binding description table for the action of pressing a key given
-- by \var{keyspec} (with possible modifiers) to the function \var{cmd}.
-- The \var{guard} controls when the binding can be called.
+-- The \var{extra} flag signals whether the binding is a non-critical extra
+-- binding. If this is true and a named key does not exist, the binding will
+-- be quietly ignored.
-- For more informationp see Section \ref{sec:bindings}.
-function ioncore.kpress(keyspec, cmd, guard)
-return putcmd(cmd, guard, {action = "kpress", kcb = keyspec})
+function ioncore.kpress(keyspec, cmd, guard, extra)
+return putcmd(cmd, guard, {action = "kpress", kcb = keyspec, extra =
extra})
end
--DOC
--- ion3.orig/ioncore/conf-bindings.h
+++ ion3/ioncore/conf-bindings.h
@@ -19,6 +19,6 @@
extern bool ioncore_parse_keybut(const char *str,
uint *mod_ret, uint *ksb_ret,
- bool button, bool init_any);
+ bool button, bool init_any, bool
quiet_absent);
#endif /* ION_IONCORE_CONF_BINDINGS_H */
--- ion3.orig/etc/cfg_ioncore.lua
+++ ion3/etc/cfg_ioncore.lua
@@ -284,7 +284,7 @@
defbindings("WMoveresMode", {
bdoc("Cancel the resize mode."),
-kpress("AnyModifier+Escape","WMoveresMode.cancel(_)"),
+kpress("AnyModifier+Escape","WMoveresMode.cancel(_)", nil, true),
bdoc("End the resize mode."),
kpress("AnyModifier+Return","WMoveresMode.finish(_)"),
--- END ---
--
Ben Hutchings
You can't have everything. Where would you put it?
signature.asc
Description: This is a digitally signed message part
Re: Minor issue when using alternate xkb groups
On Thu, 2008-05-29 at 01:32 +0300, Tuomo Valkonen wrote: > On 2008-05-28 14:02 +, Tuomo Valkonen wrote: > > The original complainer > > should therefore have no problems on the version of X I'm running > > (from Etch), because the server simply masks the groups, or at > > least the one tied to Mode_Switch, and for some reason Ion actually > > masks the state with KNOWN_MODIFIERS_MASK in the binding search code > > although it does not ask as to grab things with other modifiers set. > > Actually, making Ion print things instead of relying on xev > alone, it appears that the X server does not set the Mode_Switch > bit (0x2000) on key events for grabbed keys, although it does set > it in events for non-grabbed keys (that xev can catch). Thus the > stray KNOWN_MODIFIERS_MASK code actually does nothing, and the > X server completely hides Mode_Switch (and any modifiers bound to > it) in events for grabbed keys -- but only for them. On my system, > that is. Smells like a rotting pile of kludges... Yeah, grabbed keys are kluged so you don't see group modifiers. This is all spelled out in <http://www.xfree86.org/current/XKBproto.pdf> section 12.3. I still haven't worked out how to ignore group locking when looking at non-grab events. Ben. -- Ben Hutchings Everything should be made as simple as possible, but not simpler. - Albert Einstein signature.asc Description: This is a digitally signed message part
Re: Minor issue when using alternate xkb groups
On Sat, 2008-05-24 at 10:58 +0200, Andreas Hübner wrote: > I did have a very minor issue with ion3 when using alternate xkb groups. > You may want to read the bug report I filed in the DBTS, if you are > interested in this. > [http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=414974] > > Summary (from someone who is not really familiar with X programming): > X sets some additional bits in the XKeyEvent.state when you use an > alternate xkb group. Good work. I looked into the Xkb header and found this: /* * BuildCoreState: Given a keyboard group and a modifier state, * construct the value to be reported an event. * GroupForCoreState: Given the state reported in an event, * determine the keyboard group. * IsLegalGroup: Returns TRUE if 'g' is a valid group index. */ #define XkbBuildCoreState(m,g) g)&0x3)<<13)|((m)&0xff)) #define XkbGroupForCoreState(s) (((s)>>13)&0x3) #define XkbIsLegalGroup(g) (((g)>=0)&&((g) signature.asc Description: This is a digitally signed message part
Re: New Ion User - Focus Question [+ Gentoo braindamage]
On Wed, 2008-05-21 at 08:36 +, Tuomo Valkonen wrote: > On 2008-05-18, Kevin Monceaux <[EMAIL PROTECTED]> wrote: > > Thanks for the pointer. I'm running Ion3 via the ion3 Debian package. > > Debian shows it's version as 20080411-1. By latest version do you mean > > the Ion3+ development branch available via darcs? If so, I'll give it a > > try. > > There was a problem with the mousefocus setting being ignored, but > it should be fixed in 20080411. Maybe Debian have broken it again, > or you're doing something wrong -- such as expecting what would > normally work to work with Debian's fucked up config that probably > overrides everything you set in Debian-specific files. Tuomo, you should know very well that I don't make such changes. You reviewed all the patches, remember? Ben. -- Ben Hutchings If the facts do not conform to your theory, they must be disposed of. signature.asc Description: This is a digitally signed message part
Re: New Ion User - Focus Question
On Sun, 2008-05-18 at 09:06 -0500, Kevin Monceaux wrote: > On Sat, 17 May 2008, Tuomo Valkonen wrote: > > > On 2008-05-16, Kevin Monceaux <[EMAIL PROTECTED]> wrote: > >> I've tried adding > >> mousefocus='disabled' to the ioncore.set section of the cfg_ion.lua file > >> in my ~/.ion3 directory, but it seems to be ignored. How does one disable > >> the focus following the mouse. > > > > Probably by upgrading to the latest version. > > Thanks for the pointer. I'm running Ion3 via the ion3 Debian package. > Debian shows it's version as 20080411-1. By latest version do you mean > the Ion3+ development branch available via darcs? If so, I'll give it a > try. If you've modified some of the configuration files in /etc/X11/ion3 and then upgraded, it's possible that you've missed some significant changes there. "debsums --config ion3" will show you which configuration files are different, if any (listed as "FAILED"). The default configuration files will have been installed with the extension ".dpkg-dist". Ben. -- Ben Hutchings The two most common things in the universe are hydrogen and stupidity. signature.asc Description: This is a digitally signed message part
Re: Bug#462690: flashplugin-nonfree's new fullscreen mode doesn't work with ion3
On Sun, 2008-01-27 at 15:44 +, Tuomo Valkonen wrote:
> On 2008-01-27, Ben Hutchings <[EMAIL PROTECTED]> wrote:
> > -- Note that we can't create a timer in the match function, and we can't
> > -- change switchto in the hook function, so we need both.
>
> You can ioncore.defer(). Also, in the clientwin_do_manage_alt handler,
> it would be possible to check for ioncore.getwinprop(cwin).flashy_shite
> or so, instead of duplicating code.
OK, here's a cleaner version. It also increases the focus delay a
little, and adds a check for full-screen state, to reduce the risk of a
false positive. (Is there a simpler way to check this?)
-- Flash player opens full-screen windows from Firefox with size
-- request 200x200. It closes them if they lose focus, but also if
-- they get focus too quickly. Therefore set focus 200 ms after
-- such a window is mapped (this may need to be adjusted on slower
-- computers).
_NET_WM_STATE = ioncore.x_intern_atom('_NET_WM_STATE', false)
_NET_WM_STATE_FULLSCREEN = ioncore.x_intern_atom('_NET_WM_STATE_FULLSCREEN',
false)
function is_fullscreen(cwin)
local state = ioncore.x_get_window_property(cwin:xid(), _NET_WM_STATE, 4,
1, true)
if state then
for k, v in pairs(state) do
if v == _NET_WM_STATE_FULLSCREEN then
return true
end
end
end
return false
end
defwinprop {
class = "Firefox-bin",
match = function(prop, cwin, id)
local geom = cwin:geom()
return is_fullscreen(cwin) and geom.w == 200 and geom.h == 200
end,
switchto = false,
flash_fullscreen = true,
}
ioncore.get_hook('clientwin_do_manage_alt'):add(
function(cwin, table)
local winprop = ioncore.getwinprop(cwin)
if winprop and winprop.flash_fullscreen then
local timer = ioncore.create_timer()
timer:set(200, function() cwin:goto() end)
return true
else
return false
end
end
)
Ben.
--
Ben Hutchings
One of the nice things about standards is that there are so many of them.
signature.asc
Description: This is a digitally signed message part
Re: Bug#462690: flashplugin-nonfree's new fullscreen mode doesn't work with ion3
On Sun, 2008-01-27 at 12:24 +0100, Johannes Weiss wrote:
> Hi,
>
> > > Package: ion3
> > > Version: 20080103-1
> > > Severity: normal
> > >
> > > the fullscreen window opens shortly but closes immediately...
> > > In GNOME it works.
> >
> > I noticed a similar problem myself on one web site, but assumed it was a
> > site-specific scripting bug. If it affects Flash full-screen mode in
> > general, this is probably a bug in Flash - it's very unlikely that Ion3
> > is telling the window to close. However it may be possible to configure
> > a kluge-rule for Ion3 that will work around the bug. Can you point to a
> > specific web site that demonstrates this problem?
>
> It affects the NEW (since flash 9.0.115.0) fullscreen-mode in general.
> In older flashplayers it was not possible (under Linux) to switch to
> fullscreen-mode on an existing window. That was very bad, because you had
> to watch e.g. youtube-videos from the beginning when you wanted to switch to
> fullscreen.
> That is now fixed with a real flash fullscreen mode, but it does work with ion
> at all. Some sample sites:
>
> - http://youtube.com -- choose any video and press fullscreen
> - http://myvideo.de -- the same thing
> - http://flowplayer.org/demo/h264
>
> be sure to have installed
> ii flashplugin-nonfree 9.0.115.0.1 Adobe Flash Player plugin
> installer
This seems to be a bug in Flash. I was able to workaround this by
adding a nasty hack to cfg_ion.lua:
-- Flash player opens full-screen windows from Firefox with size request
-- 200x200. It closes them if they lose focus, but also if they get focus
-- too quickly. Therefore delay focus for 100 ms.
-- Note that we can't create a timer in the match function, and we can't
-- change switchto in the hook function, so we need both.
defwinprop {
class = "Firefox-bin",
match = function(prop, cwin, id)
local geom = cwin:geom()
return geom.w == 200 and geom.h == 200
end,
switchto = false,
}
ioncore.get_hook("clientwin_do_manage_alt"):add(
function(cwin, table)
local ident = cwin:get_ident()
local geom = cwin:geom()
if ident.class == "Firefox-bin" and geom.w == 200 and geom.h == 200 then
local timer = ioncore.create_timer()
timer:set(100, function() cwin:goto() end)
return true
else
return false
end
end
)
Ben.
--
Ben Hutchings
One of the nice things about standards is that there are so many of them.
signature.asc
Description: This is a digitally signed message part
Re: Nagware version
On Thu, 2008-01-03 at 18:42 +, Tuomo Valkonen wrote: > On 2007-12-29, Tuomo Valkonen <[EMAIL PROTECTED]> wrote: > > I've been considering releasing (as a patch) a separate nagware version > > of Ion, which would have a license without the 28-day clause. > > I take it there's no interest for such a "free" nagware version? Any version that goes into a Debian stable release would be outdated and therefore unsupported by you. So how about a free version that simply disclaims upstream support without bothering to check? Ben. -- Ben Hutchings Beware of programmers who carry screwdrivers. - Leonard Brandwein signature.asc Description: This is a digitally signed message part
Re: Life, universe, and ION3
On Tue, 2007-10-09 at 15:07 +0200, ,,, wrote: > Hi all! > > > Some week ago i read about ION window-manager at the Hungarian Unix > Portal. Due, I'm an engineering monkey, and i really hates wasting my time > playing with the windows locations, i > thought the tiling wm would be a good thing for me. So, i pulled down the > ION2 from the Debian respository and give > it a shot. > My first thoughts: Simple, efficient and fast. > After that, i dowloaded the ION3 source, and compiled it. There is also a Debian package of Ion3 (which I maintain) but due to Tuomo's trademark policy it is in the non-free section so you will have to add that to your APT sources to see it. Ben. -- Ben Hutchings I haven't lost my mind; it's backed up on tape somewhere. signature.asc Description: This is a digitally signed message part
Re: kpdf
On Tue, 2007-09-11 at 20:31 +0200, Albert Shih wrote:
> Le 11/09/2007 à 10:31:25+, Tuomo Valkonen a écrit
> > On 2007-09-11, Albert Shih <[EMAIL PROTECTED]> wrote:
> > > Sometime I need to use the "presentation mode" of kpdf, well it's not work
> > > very well with ion3.
> >
> > That piece of crapware (version below) seems to create a new window
> > for the FS mode instead of asking the window manager to FS the present
> > one, and the most brain-dead of all: marks the window as transient of
> > the main window. So Ion handles it as a transient.
> >
> > $ kpdf --version
> > Qt: 3.3.4
> > KDE: 3.5.3-0.2.fc4 Red Hat
> > KPDF: 0.5.3
> >
>
> Hummm..thanks for your answerbutplease don't shoot me, ..how can I
> define this ?
>
> How can I tell ion the kpdf windows is a transient windows ?
It *is* a transient window, but it shouldn't be. You can override that
by adding this to cfg_ion.lua:
defwinprop {
class = "Kpdf",
role = "presentationWidget",
transient_mode = "off"
}
Ben.
--
Ben Hutchings
Nothing is ever a complete failure; it can always serve as a bad example.
signature.asc
Description: This is a digitally signed message part
Re: ion-3rc-20070902
On Sun, 2007-09-02 at 18:28 +0400, Alexander Shishkin wrote: > On 9/2/07, Nick Murdoch <[EMAIL PROTECTED]> wrote: > > Alexander Shishkin wrote: > > > On 9/2/07, Tuomo Valkonen <[EMAIL PROTECTED]> wrote: > > >> This release features a few minor fixes. If no major problems are > > >> discovered, this is likely to be the last "rc" release before the > > >> first and hopefully final "stable" release. > > > Sorry, haven't been following recently (vacation and stuff), but does > > > anybody (Ben?) still keep packaging ion3 for debian? The last version > > > that I see is 20070720-2. > > > > It's a little behind at the moment, but yes, he does seem to be. It'll > > be nice when ion3 is stable for Debian. mod-xinerama-for-ion3 seems to > > be missing from apt all of a sudden, however. (I know this has nothing > > to do with Tuomo, though) > Yes, mod-xinerama is part of my concern indeed. It's unmaintained in Debian. I've been trying to convince Holger Levsen to pick it up. Ben. -- Ben Hutchings For every action, there is an equal and opposite criticism. - Harrison signature.asc Description: This is a digitally signed message part
Re: configuration files not respected on ubuntu?
On Sun, 2007-08-05 at 14:54 -0700, Red Daly wrote: > I removed the ubuntu package and manually installed from the ion3 > tarball. My configuration changes now seem to work correctly from the > ~/.ion3 directory. It is a shame that the Ubuntu package is so out of > date, though :( > > The only tricky part was adding a new file to start ion3 from GDM: That's not really necessary (at least it isn't on Debian). I just put ion3 in ~/.xsession and use GDM's "Default System Session". Ben. -- Ben Hutchings Reality is just a crutch for people who can't handle science fiction. signature.asc Description: This is a digitally signed message part
Re: configuration files not respected on ubuntu?
I am not responsible for the Ubuntu package, but it is based on the Debian package that I now maintain. On Sun, 2007-08-05 at 00:14 -0700, Red Daly wrote: > I have recently installed ion3 and am trying to use the windows key > rather than the alt key for META. xmodmap shows the windows key is > Mod4, so I made the following edit to cfg_ion.lua in my ~/.ion3/ directory: > > -- Set default modifiers. Alt should usually be mapped to Mod1 on > -- XFree86-based systems. The flying window keys are probably Mod3 > -- or Mod4; see the output of 'xmodmap'. > META="Mod4+" > --ALTMETA="" > > from > > -- Set default modifiers. Alt should usually be mapped to Mod1 on > -- XFree86-based systems. The flying window keys are probably Mod3 > -- or Mod4; see the output of 'xmodmap'. > -- META="Mod1+" > --ALTMETA="" > > I restarted the session and nothing doing. I restarted my computer and > still windows does nothing. META and ALTMETA are set in /etc/default/ion3 so that they are easy to change on a system-wide basis without modifying the default configuration files. There is a comment about this in cfg_ion.lua though it is possibly not as obviously as it should be. > p.s. I am running the ion3 that I apt-got with my ubuntu feisty machine. > > $ ion3 -V > > 3ds-20061029 Please be aware that this is quite outdated and will not be supported by the author (and probably not by Ubuntu either). You should be able to build the current version from the Debian source package <http://ftp.debian.org/debian/pool/non-free/i/ion3/ion3_20070720-1.dsc> (<http://www.debian.org/doc/FAQ/ch-pkg_basics.en.html#s-sourcepkgs> explains how to do this) or from the original tarball <http://modeemi.fi/~tuomov/ion/download.html>. Ben. -- Ben Hutchings [W]e found...that it wasn't as easy to get programs right as we had thought. ... I realized that a large part of my life from then on was going to be spent in finding mistakes in my own programs. - Maurice Wilkes, 1949 signature.asc Description: This is a digitally signed message part
Re: error questions
On Fri, 2007-07-13 at 10:03 -0500, David Polk wrote: > On startup I get this error message - "Unable to find > 'debian-menu-i18n' on search path" - Ion seems to work fine, so how do > I eliminate this error message? I'm the maintainer of the Debian package and should be the first person you ask about it. The debian-menu-i18n.lua file is generated automatically on installation of Ion3 or another package that provides a menu entry, and is loaded by cfg_ioncore. It lives in /var/lib/ion3 but sym-linked from /usr/share/ion3 so that Ion3 can find it. Please report the output of: $ dpkg -l ion3 $ ls -l /usr/share/ion3/debian-menu-i18n.lua $ ls -l /var/lib/ion3/debian-menu-i18n.lua I can't answer your other question. Ben. -- Ben Hutchings Nothing is ever a complete failure; it can always serve as a bad example. signature.asc Description: This is a digitally signed message part
Re: Some kind of "proxy-tool" for Apps running inside Xnest?
On Tue, 2007-07-10 at 13:25 +0200, Andreas Bernauer wrote: > Hi Christian, > > Christian Walther schrieb am Di, 10. Jul 2007 um 10:49 (+0200): > > With ion, the Citrix client is nearly unusable. Even if you create a > > new workspace with an "empty" layout it's a mess. I won't blame ion > > for this. There's no need for an application to handle window > > manager tasks... If it needs to provide Win32 semantics to remote applications, perhaps there is. (I don't know whether the Citrix client uses one local window for a whole Windows desktop session or for each remote window.) > This does not actually answer your question, but did you consider > using rdesktop (http://www.rdesktop.org/)? We were using the Citrix > ICA Client, too, but we switched to rdesktop several months ago and I > had no problems with it so far. rdesktop also works well for me. It doesn't want to be resized, but you can configure its "screen" size to exactly fit the frame. However, it doesn't provide any activity indication for remote windows. Ben. -- Ben Hutchings Absolutum obsoletum. (If it works, it's out of date.) - Stafford Beer signature.asc Description: This is a digitally signed message part
Re: Dock padding_pixels suggestion
On Wed, 2007-07-11 at 09:03 +, Tuomo Valkonen wrote: > On 2007-07-11, Roy Lanek <[EMAIL PROTECTED]> wrote: > > Just try with a notebook--then you might become interested > > in monitoring a few things; and I am not sure ion's own > > statusbar is i) "quite enough," ii) the best place to put > > them on ... > > The laptop I have is a 486/25MHz/8MB with battery live of zero > seconds. I use it as a serial terminal to fix hung X occasionally. > (Recently it's been crashing more -- a lot -- though.) And the > newer laptops also aren't good enough to bother with either. > Too Fucking Expensive, There exists a second-hand market, where laptops are much cheaper. > Fragile And Battery-Hungry or Too Fucking > Bulky To Be Considered Even A Laptop, Let Alone A Mobile Computer > (which few laptops are). The OLPC is probably the best laptop out > there, supposing the screen really works in low-power mode too... The screen is a thing of beauty, but I doubt you'd get on with the keyboard. Ben. -- Ben Hutchings Absolutum obsoletum. (If it works, it's out of date.) - Stafford Beer signature.asc Description: This is a digitally signed message part
Re: ion3 RC backports for Debian etch
On Sat, 2007-06-16 at 16:40 +0200, intrigeri wrote: > Hello, > > I’m maintaining backports for Debian etch of Debian unstable ion3-* > packages : https://intrigeri.boum.org/debian/ > > I’ll update them reasonably quickly when Debian unstable packages are > themselves updated. Which should happen this week. I am currently busy with DebConf video coverage. > They are built for i386, a bit later for powerpc, > and the source packages are of course available. > > I’ll go on with this work as long as I materially can, until Ben or > another Debian developer takes responsibility of maintaining backports > on backports.org, volatile.debian.org, or wherever. Only packages in testing should have backports on backports.org. Ion3 is blocked from propagating into testing (by #415287) until there is a stable release of Ion3 that can be supported in a stable release of Debian. However, since I normally run stable myself, I normally ensure that my packages will build on both stable and unstable, making it trivial to backport them. Ben. -- Ben Hutchings Power corrupts. Absolute power is kind of neat. - John Lehman, Secretary of the US Navy 1981-1987 signature.asc Description: This is a digitally signed message part
Re: ion-3rc-09f911029d74e35bd84156c5635688c0.tar.gz
On Sat, 2007-05-12 at 19:36 +0400, Alexander Shishkin wrote: > Except, probably the case when I discovered that ion3 no longer does > xinerama due to you being somehow upset with xinerama API. Just like > you rolled back to an older kernel, I had to roll back to a > known-to-work snapshot of ion3. Have you tried mod_xinerama? There is a Debian package (currently ion3-mod-xinerama; may need to be renamed) and it has just entered the archive. Ben. -- Ben Hutchings Klipstein's 4th Law of Prototyping and Production: A fail-safe circuit will destroy others. signature.asc Description: This is a digitally signed message part
Re: ion3 segfault
On Mon, 2007-05-07 at 14:18 +, Tuomo Valkonen wrote: > On 2007-05-07, Ben Hutchings <[EMAIL PROTECTED]> wrote: > >> This is ion3 20070318-2 debian package. > > > > You should use reportbug. > > Debian will DI! MUAHAHAHAHAHHA. > > I noticed from your posting that Xen has also added trademark terms > to their license. I wonder how long Debian can keep up with the trend, > until they will have to change their policies to respect the authors > a bit. Debian takes the position that authors read and understand the licences they place software under and that Debian should therefore act within the terms of those licences. It is not our practice to guess that the author might have meant something else. Ben. -- Ben Hutchings If God had intended Man to program, we'd have been born with serial I/O ports. signature.asc Description: This is a digitally signed message part
Re: ion-3rc-09f911029d74e35bd84156c5635688c0.tar.gz
On Mon, 2007-05-07 at 06:08 +, Tuomo Valkonen wrote: > On 2007-05-06, csant <[EMAIL PROTECTED]> wrote: > >> Also known as ion-3rc-20070506. > > There seem to be no darcs changes to pull, and also the online changelog > > doesn't show the last changes. > > I'm not going to make it easy for all the forkers... So next time there's a bug in a release like the one with size requests, we'll just have to wait for your next release rather than retrieving the fix from your repository? Or make a local fix? This deliberate inaccessibility seems like a good way to deter (l)users from following and getting involved in Ion development. This doesn't seem to fit in with your apparent wish for "a progression of users of different levels between them and the developers". Ben. -- Ben Hutchings If God had intended Man to program, we'd have been born with serial I/O ports. signature.asc Description: This is a digitally signed message part
Re: ion3 segfault
[I thought I already sent this, but apparently it was left in my drafts folder.] On Wed, 2007-05-02 at 16:49 +0200, Konstantin Kletschke wrote: > Well, I experienced an ion3 segfault is this the right place to post > kind of those to, may be, improve code quality? Given that: > This is ion3 20070318-2 debian package. You should use reportbug. Ben. -- Ben Hutchings Never put off till tomorrow what you can avoid all together. signature.asc Description: This is a digitally signed message part
Re: dock and transients
On Thu, 2007-05-03 at 15:58 +0200, Michael Gwin wrote: > Ben Hutchings writes: > > > You need to apply this patch from darcs: > > > > Mon Mar 19 18:37:26 GMT 2007 Tuomo Valkonen <[EMAIL PROTECTED]> > > * Oops, return in wrong place in fs rq check was breaking other cfgrqs > > Right. I thought I'd have that though, having done the darcs pull yesterday. Sorry, I thought by snapshot you meant the last tagged and tarred version. Looking at the download web page, I see the word is used also for the latest repository version, which is confusing. > However, looking at the log at > http://modeemi.fi/~tuomov/repos/_log/ion-3.html.gz, it seems like there > haven't been any changes since 2007-03-18. There certainly were some, but I don't see them any more. > Am I missing something here? > Also, not being very knowledgeable w/ respect to darcs, I'm simply doing > "darcs get --partial http://modeemi.fi/~tuomov/repos/ion-3"; - is that enough > to get the latest and greatest? I believe it should be. Ben. -- Ben Hutchings Never put off till tomorrow what you can avoid all together. signature.asc Description: This is a digitally signed message part
Re: To beat the ugly
On Wed, 2007-05-02 at 12:12 +, Tuomo Valkonen wrote: > On 2007-05-02, Evgeny Kurbatov <[EMAIL PROTECTED]> wrote: > > The Iceape 1.0.8 (from Debian 4.0 Etch) has a strange dialog "Save As" > > which tryes to change it's window size. For some reasons Ion3 don't > > want to allow the window's size to change so that dialog remains > > shrinked and unusable. The ugly method is to make it fullscreen but its > > not beauty... > > There's a typo/copy-paste error in the latest release, which causes > size requests to be ignored.. Do you mean the error in handling size request changes, or another bug? I applied your fix for that to the Debian package. Ben. -- Ben Hutchings Never put off till tomorrow what you can avoid all together. signature.asc Description: This is a digitally signed message part
Re: dock and transients
On Wed, 2007-05-02 at 22:05 +0200, Michael Gwin wrote: > I just upgraded to the latest darcs snapshot. Transient windows aren't > resizing properly (e.g. the firefox 'save as' dialog). A recent post > mentioned a "typo/copy-paste error in the latest release" which affects > this - is that relevant to the darcs code too? > > For quite a while now, I've been using "docker" in a floating dock in > order to swallow a few icons/systray stuff. Since the upgrade, they > aren't showing up anymore. I haven't upgraded docker or any of the apps > that display an icon. Has something changed in ion? You need to apply this patch from darcs: Mon Mar 19 18:37:26 GMT 2007 Tuomo Valkonen <[EMAIL PROTECTED]> * Oops, return in wrong place in fs rq check was breaking other cfgrqs Ben. -- Ben Hutchings Never put off till tomorrow what you can avoid all together. signature.asc Description: This is a digitally signed message part
Re: License change
On Fri, 2007-04-27 at 19:33 +0300, Tuomo Valkonen wrote: > (Hmm... I wonder where the first try wen't.) > > As a further note, I consider the restrictions on the name use > retroactive: this is also an issue of trademark law, and as the > author, I am the (unregistered) trademark holder for Ion, Ion3, > etc. IANAL, but I think you are too late to make a trademark claim against modified or outdated versions of Ion3 not labelled as such, having allowed them in the past. > The new license and copyright law merely act as additional > enforcement devices. They certainly allow you to give rather more specific permissions and restrictions to distributors than trademark law supports. Ben. -- Ben Hutchings If God had intended Man to program, we'd have been born with serial I/O ports. signature.asc Description: This is a digitally signed message part
Re: Dodgy function pointer casts
0On Thu, 2007-04-26 at 07:12 +, Tuomo Valkonen wrote: > On 2007-04-25, Ben Hutchings <[EMAIL PROTECTED]> wrote: > > Is this change likely to be acceptable, Tuomo? > > Umm... any kind of OO shit in C absolutely positively depends on > function pointer casts. I haven't noticed this being necessary in the GObject system. It does rely on data pointer casts, but those are fine. > Or are you going to wrap all the _hundreds_ > of pseudo-dynamic functions? Over my dead body. > > More uglier is the case of function pointers to data pointers used > somewhere in the code. I haven't checked if there's a way around it, > but then again, any sane architechture will have matching function > and data pointers. (Functions should be nothing special! They should > be data like everything else! Stuff of the first class!) While these casts aren't allowed in standard C, casting between void * and function pointer types must work on any system which has dlopen(). Ben. -- Ben Hutchings If God had intended Man to program, we'd have been born with serial I/O ports. signature.asc Description: This is a digitally signed message part
Dodgy function pointer casts
When checking over the Debian package of mod_xinerama, I noticed that the code was casting a pointer to a callback function because the function's argument types didn't quite match those expected for the callback. When I objected to this, the package's maintainer pointed out that there were similar casts throughout Ion3, which is indeed true. While this obviously works on most platforms, calling through a function pointer that doesn't match the type of the function it points to has undefined behaviour. I would like to clear this up by adding wrapper functions and changing function parameter types so that only data pointer casts are needed. Is this change likely to be acceptable, Tuomo? Ben. -- Ben Hutchings All extremists should be taken out and shot. signature.asc Description: This is a digitally signed message part
Font rendering oddity in status bar
A Debian user reported a cosmetic bug in Ion3 where the top pixel of a vertical bar character (U+2502) in the status bar could be left behind when the text changed. This suggests some sort of error in finding the text extents (but at what level, I don't know). The bug report is at http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=417651 and includes a screenshot of this behaviour. I identified the font for most characters as being fixed 6x13, but the vertical bar appears to be 14 pixels tall; perhaps it comes from another font in a fontset? I have spent some time investigating this (explanations attached to the report) but have not been able to work out where the mismatch is. In particular I have not been able to construct a fontset that includes fixed 6x13. Can anyone help me work out what's going on? Ben. -- Ben Hutchings All extremists should be taken out and shot. signature.asc Description: This is a digitally signed message part
Re: If it is still unclear to someone...
I wrote: > I agree, actually; I think Debian has very variable levels of support > for released packages and this is not made clear to users. Since you > say ion3 is not in a state to be supported for the lifetime of a Debian > release, I must accept that. Therefore I will try to keep it up-to-date > in unstable and backports. Sorry, not backports. I forgot that only packages in testing are eligible for backports, so ion3 is now gone from there. Ben. -- Ben Hutchings I'm not a reverse psychological virus. Please don't copy me into your sig. signature.asc Description: This is a digitally signed message part
Re: Documentation bughunt
On Tue, 2007-03-20 at 11:32 +, Tuomo Valkonen wrote: > It'd be nice if people would report any significant omissions > and bugs in the reference documentation [1]. There's a longstanding bug report about the meaning of "stretchable space" which is mentioned briefly in the default configuration but not in the reference: <http://bugs.debian.org/319870>. Ben. -- Ben Hutchings The generation of random numbers is too important to be left to chance. - Robert Coveyou signature.asc Description: This is a digitally signed message part
Re: ion-3ds-20070318
On Sun, 2007-03-18 at 21:02 +, Tuomo Valkonen wrote: > This may be the last "ds" release if Ion3; if all goes well, the next > release is likely to be an "rc". Some minor improvements will still > be done, and minor requests may be accepted, so be quick. I'm just about to upload a Debian package of this. Some things I noted while preparing it: - Context menus for frames now affect the current object. Previously (certainly in 20061223; I'm not sure about 20070203) they would affect the tab that the context menu was opened on. This change doesn't seem to have been documented, but I've noted it in NEWS.Debian. Was it an intentional change? - Size request changes by docked applications (I use "docker") seem to be ignored. I narrowed this down to the patch 'Full screen mode fixes and improvements'. I don't understand quite what this is doing, so for now I've simply reverted it in the Debian package. - Minor nit: rootwin_init() does: LINK_ITEM(*(WRegion**)&ioncore_g.rootwins, (WRegion*)rootwin, p_next, p_prev); and similarly rootwin_deinit() does: UNLINK_ITEM(*(WRegion**)&ioncore_g.rootwins, (WRegion*)rw, p_next, p_prev); I think these casts violate the aliasing rules, but I'm not certain. Ben. -- Ben Hutchings compatible: Gracefully accepts erroneous data from any source signature.asc Description: This is a digitally signed message part
Re: If it is still unclear to someone...
Hello everybody. I've agreed to adopt the Debian packages of ion3, but have not yet made any changes to them. On Wed, 2007-03-07 at 19:00 +, Tuomo Valkonen wrote: > 2. But if you want my help, or otherwise come complaining within my >sight, you'd better be using or have tried things out on a supported >release, i.e. the latest release on a supported "branch". (Ion3 or >Ion2; Ion1 available merely as a historical curiosity. Ion2 support >will also cease when "stable" Ion3 is released. Not that there's >been much of a need to support it.) I'm quite happy to include a clear notice that bugs and queries should be reported to me as the maintainer rather than you. However, this is what users should do for any package, and there's not much I can do about (l)users that don't know about reportbug or how to find the maintainer's email address. They probably won't read the Debian documentation either. > 3. Distros should provide only supported releases (with a reasonable >delay, which two years isn't), or at the very least let users know >(in a manner that they can not ignore) when they're installing an >unsupported release. ion3 has been removed from the testing suite and will not be included in the etch release. >The reasons are simple: a) if your distro is supposedly "stable", it >is easy for you to think that the software it provides is "stable" >too, and hence supported. I agree, actually; I think Debian has very variable levels of support for released packages and this is not made clear to users. Since you say ion3 is not in a state to be supported for the lifetime of a Debian release, I must accept that. Therefore I will try to keep it up-to-date in unstable and backports. > b) And if your distro supposedly provides the latest software, No-one accuses Debian of that, surely? ;-) Ben. -- Ben Hutchings Make three consecutive correct guesses and you will be considered an expert. signature.asc Description: This is a digitally signed message part
Re: Debian sucks // Ion may become (DFSG-) "non-free"
On Tue, 2007-03-06 at 09:49 +, Tuomo Valkonen wrote: > Sucks that the first question to everyone must be "which > version and distro you're using" "Which version" has always been a critical field for bug reports. Even if you expect everyone to compile from the latest source, they're still not going to do it every day. > -- and it's usually the people who've > installed Ion from a distribution, that are using old unsupported > development snapshots. (Actually, maybe I should just stop supporting > anyone who has not installed Ion from the official tarball. Let the > distros support their own lusers.) My understanding is that a Debian maintainer must at the absolute *minimum* provide first-line support to Debian users. > I'm also considering extending the license (LGPL) with a "Distributor > timely response clause", something like the following (D). It could > make Ion "non-free", but I don't care about these idealists' definitions > of freeness. Please don't do this. We'd have to come up with some funny new name, possibly involving the word "ice". Ben. -- Ben Hutchings Make three consecutive correct guesses and you will be considered an expert. signature.asc Description: This is a digitally signed message part
