[e-users] a small bug on winlist
hi I found a small bug on winlist, after I have set winlist can display iconified windows, that is: $enlightenment-remote -winlist-list-show-iconified-set 1 $enlightenment-remote -winlist-list-uncover-while-selecting-set 1 if there is only one window iconified, when I use ALT+TAB select, it will switch itself, the behavior is strange, and when it switch to another itself, the iconified window will be eaten, and I can never find it in winlist, even with mouse middle key. but if I iconified 2 windows, this is ok, it can switch between 2 different windows, and no windows will be lost. I got source from 7.23cvs brs yakie --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] Dual display w/e17 question
Carsten Haitzler (The Rasterman) wrote: On Sun, 31 Jul 2005 05:16:16 +0200 xnowfall [EMAIL PROTECTED] babbled: Carsten Haitzler (The Rasterman) wrote: 2. If I have an application open on the second display and right click the title bar the close/iconify menu appears on the first display. Probably just a bug though. just testing here - it works fine with mulithead and xinerama. menu comes up where it should. i've noticed this when opening a window (on screen 2) from an xterm (on screen 1). the menu will then slide onto the screen of the xterm. if you just move the new window a bit though e seems to catch on, and the menu will appear on the proper screen can you describe in more detail? i can't reproduce here. this is e17's window border menu not the application menu? as in slide it actually animates (it starts on screen 2 and goes and slides onto screen 1) ? i have 2 screens. on the left screen is an xterm where i run say xclock -geometry 48x48-0+0 and the clock window opens at the top right of the rightmost screen. i alt-right click the clock and the pointer jumps to the right border of the left screen causing the menu, which pops up at the point where i clicked, to start sliding towards the left screen. if i move the cursor away from the edge the sliding stops as you'd expect. if i move the window a bit, everything works as it should. this also happens in reverse, with the xterm on the right screen opening the xclock on the left. also the middle-click window menu on the desktop shows the window on the xterm's screen and the pager doesn't seem to show the window at all; move the window and both are updated to reflect its true position. the little label that pops up to show the window position while you drag it also appear on the xterm's screen the first time. so it seems that e isn't registering that the window is opened on a separate screen from the window used to open it. my build of e is a couple of weeks old, so i could try upgrading if you still don't see it, but it has always behaved this way for me as long as i've used it (several months now). --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] Dual display w/e17 question
On Sun, 31 Jul 2005 09:33:25 +0200 xnowfall [EMAIL PROTECTED] babbled: Carsten Haitzler (The Rasterman) wrote: On Sun, 31 Jul 2005 05:16:16 +0200 xnowfall [EMAIL PROTECTED] babbled: Carsten Haitzler (The Rasterman) wrote: 2. If I have an application open on the second display and right click the title bar the close/iconify menu appears on the first display. Probably just a bug though. just testing here - it works fine with mulithead and xinerama. menu comes up where it should. i've noticed this when opening a window (on screen 2) from an xterm (on screen 1). the menu will then slide onto the screen of the xterm. if you just move the new window a bit though e seems to catch on, and the menu will appear on the proper screen can you describe in more detail? i can't reproduce here. this is e17's window border menu not the application menu? as in slide it actually animates (it starts on screen 2 and goes and slides onto screen 1) ? i have 2 screens. on the left screen is an xterm where i run say xclock -geometry 48x48-0+0 and the clock window opens at the top right of the rightmost screen. i alt-right click the clock and the pointer jumps to the right border of the left screen causing the menu, which pops up at the point where i clicked, to start sliding towards the left screen. if i move the cursor away from the edge the sliding stops as you'd expect. if i move the window a bit, everything works as it should. this also happens in reverse, with the xterm on the right screen opening the xclock on the left. also the middle-click window menu on the desktop shows the window on the xterm's screen and the pager doesn't seem to show the window at all; move the window and both are updated to reflect its true position. the little label that pops up to show the window position while you drag it also appear on the xterm's screen the first time. so it seems that e isn't registering that the window is opened on a separate screen from the window used to open it. my build of e is a couple of weeks old, so i could try upgrading if you still don't see it, but it has always behaved this way for me as long as i've used it (several months now). aaah now i get it. ok - much better description now. i could reproduce it. it was because the app explicitly asked for geometry that was OUTSIDE the current zone (where the mouse was) that confused e - basically all new windows get put on the current desktop in the current zone. ie where the mouse is OR the focused window. anyway - e tracks what zone (xinerama screen) the window is in when u move it etc. BUT it doesn't get it right at the start as at the start it goes well go to the current screen thus when u pop up a menu e THINKs its on the first screen even tho its on the 2nd and thus the menu handling goes weird (as it grabs the mouse and confines it to an invisible window the size and location of just the screen the window is on). fixed in cvs now - but u can fix this by just moving xclock from the 2nd to the first screen and back again then e will get its internal state right. -- - Codito, ergo sum - I code, therefore I am -- The Rasterman (Carsten Haitzler)[EMAIL PROTECTED] 裸好多 [EMAIL PROTECTED] Tokyo, Japan (東京 日本) --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] Background for E16
Is there a way to get the png/jpg/whatever from the edj? Yes. edje_decc e17-tree.edj will give you the png :) This isnt the right way to do it. It works here because the bg has 1 image (dunno if this one even does have only 1, havent looked at it). The proper way to this (which will work with multi image edjes and animated ones) would be to use edje_thumb. -- Hisham Mardam Bey MSc (Computer Science) http://hisham.cc/ +9613609386 Codito Ergo Sum (I Code Therefore I Am) --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_idt77alloc_id492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] Background for E16
Thanks all for the help... :D edje_decc also worked since it has only one image. Love this background! :D Cheers, Paulo Matos Hisham Mardam Bey said: Is there a way to get the png/jpg/whatever from the edj? Yes. edje_decc e17-tree.edj will give you the png :) This isnt the right way to do it. It works here because the bg has 1 image (dunno if this one even does have only 1, havent looked at it). The proper way to this (which will work with multi image edjes and animated ones) would be to use edje_thumb. -- Hisham Mardam Bey MSc (Computer Science) http://hisham.cc/ +9613609386 Codito Ergo Sum (I Code Therefore I Am) --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_idt77alloc_id492opÌk ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users -- Paulo Jorge Matos - pocm at sat inesc-id pt Web: http://sat.inesc-id.pt/~pocm Computer and Software Engineering INESC-ID - SAT Group --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_idt77alloc_id492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
[e-users] compile errors
Hello, I am trying to compile from CVS today but i always get thise error for e-modules and examine: dpkg-buildpackage: cannot determine name of current package i am running out of ideas on how to fix this. Thank you Steve -- Steve Ong Torrefranca http://www.on-tvmall.com.tw --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] Entrance and .xinitrc
phriedrich wrote: On Fri, 29 Jul 2005 03:22:59 +0100 Paulo Matos [EMAIL PROTECTED] wrote: I'd like to start some programs with E16, using Entrance (latest CVS) as login manager. Well, I've tried .xinitrc lines but it seems Entrance ignores it. How can I start xscreensaver and engage for example with my E16 release? You could add any program by editing ~/.e16/e_config--0.0.snapshots and add a new entry for a program. Because thats not very comfortable and only could be done with e _not_ running (otherwise the changes will be overwritten by e) there's a second solution. Once I added - NEW: run NAME: run CLASS: run CMD: /home/phriedrich/.e16/run.sh -- as the four first lines of the aboved named config file and created the run.sh-script on the described place. Now I can easily add new programs to autostart just by inserting a line. From my run.sh: --- #! /bin/bash xmodmap .xmodmap xscreensaver macosd -n Eterm --theme irssi --name=Irssi --title=Irssi -e irssi --- Hope that helps you, best regards, Friedrich Post Scriptum: Maybe we could have a file like '~/.e16/e_config--0.0.autostart' in which we can easily add programs to start without the loop way with calling a run.sh. Or maybe even better a '~/.e16/menus/autostart' which also had the advance to be more intuitive and could also be edited with the menu-editor. Just an idea With e16.8 I think this should do what you want: $ eesh set misc.session.cmd_init /home/phriedrich/.e16/run.sh Beware! The script must exit, as your run.sh does. /Kim --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
[e-users] Problem using boost::bind and ecore_ipc from Enlightenment
Hi all,
I have a Problem with C++ Member Functions and Pointers to this Member
Functions so that they fit into a Function with take a normal Function.
I guess i'm a Newbie, otherwise i should have already solved it, i think. ;)
I want to use IPC Stuff from Ecore. But before you can make use of ecore_ipc
you must initialize it, as far as i can see.
I atached 4 Files, where i build a little Scenario:
1. Example.cpp // The Main Program
2. Testclass.hpp
3. Testclass.cpp
4. Makefile // Perhaps you must adjust some of the Variables
I commented the important Points in the Code again.
So i must first call
ecore_ipc_init() // Testclass Line 68
and
must supply 3 Functions to ecore for handling Events
1. A Function that is called, when a new IPC Server is added
2. A Function that is called, when a IPC Server is deleted
3. A Function that is called, when some Message is sent
Namely:
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_ADD, ipc_server_add, NULL );
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DEL, ipc_server_del, NULL );
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DATA, ipc_server_data, NULL );
ipc_server_add, ipc_server_del and ipc_server_data are the functions that
are available outside the Class Testclass
The Example Program as it is compiles and i can start it, the ecore_pic_init()
succeds the 3 Functions get registered but this 3 submitted Functions are
standing outside my Testclass,
but my Target is to have they inside my Testclass.
That's my Problem, i didn't get the Clue to give these MemberFunctions to these
3 Functions:
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_ADD, ipc_server_add, NULL );
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DEL, ipc_server_del, NULL );
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DATA, ipc_server_data, NULL );
I tried Things like that, but that didn't helped me:
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DEL, boost::bind(
boost::typeint(), Testclass::ipc_server_add ), NULL );
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DEL, boost::bind(
boost::typeint(), Testclass::ipc_server_add, _1, _2, _3 ), NULL );
So i want to ask to the Mailinglist in Hope somebody knows how to do this Task
in the right Way. ;)
Thanks for your help and Greets to all,
Christian#include Testclass.hpp
#include iostream
using namespace std;
int
main( int argc, char *argv[] ) {
cout Creating an Instance of the Testclass \n;
Testclass * testclass = new Testclass();
if ( !testclass-Init() ) {
cout Initialization of ecore_ipc failed.\n;
} else {
cout Initialization of ecore_ipc succeded.\n;
}
cout Delete the Instance of the Testclass\n;
delete testclass;
return 0;
}
Makefile
Description: Binary data
#include Testclass.hpp
#include iostream
// Boost
#include boost/format.hpp
using boost::format;
using namespace std;
/**
* ipc_server_add - when we connect to the ipc daemon
* @data -
* @type -
* @event - the Ecore_Ipc_Event_Server_Add that triggered us
*/
int
ipc_server_add( void *data, int type, void *event ) {
Ecore_Ipc_Event_Server_Add * e;
e = ( Ecore_Ipc_Event_Server_Add * ) event;
cout format( Server add \n );
return 1;
}
/**
* ipc_server_del - when we disconnect from the ipc daemon
* @data -
* @type -
* @event - the Ecore_Ipc_Event_Server_Del that triggered us
*/
int
ipc_server_del( void *data, int type, void *event ) {
Ecore_Ipc_Event_Server_Del * e;
e = ( Ecore_Ipc_Event_Server_Del * ) event;
cout format( Server del \n );
return 1;
}
/**
* ipc_server_data -
* @data -
* @type -
* @event - the Ecore_Ipc_Event_Server_Data that triggered us
*/
int
ipc_server_data( void *data, int type, void *event ) {
Ecore_Ipc_Event_Server_Data * e;
e = ( Ecore_Ipc_Event_Server_Data * ) event;
cout format( Server sent [%i] [%i] (%i) \%s\\n ) % e-major % e-minor % e-size % e-data;
return 1;
}
Testclass::Testclass()
: server( NULL ) {}
Testclass::~Testclass() {
if ( server != NULL ) {
// ecore_ipc_server_del( server );
}
}
bool
Testclass::Init() {
if ( ecore_ipc_init() 1 ) {
return 0;
}
// Create a IPC Server
server = ecore_ipc_server_add( ECORE_IPC_LOCAL_USER, Testclass IPC Server, 0, NULL );
if ( server == NULL ) {
cout Creating new IPC server failed\n;
return 0;
}
// Here is my Problem, i didn't get a valid Pointer to one of my Memberfunctions
// My Tries looks like these:
// ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DEL, boost::bind( boost::typeint(), Testclass::ipc_server_add ), NULL );
// ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DEL, boost::bind( boost::typeint(), Testclass::ipc_server_add, _1, _2, _3 ), NULL );
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_ADD, ipc_server_add, NULL );
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DEL, ipc_server_del, NULL );
ecore_event_handler_add( ECORE_IPC_EVENT_SERVER_DATA, ipc_server_data, NULL );
char buf[ 512 ];
char *msg = Hello World;
int
Re: [e-users] Problem using boost::bind and ecore_ipc from Enlightenment
ecore_* takes a pointer to a function whereas boost::bind returns an object (that can be used as a function by operator overloading). to use a member function you need both a pointer to an object instance and a pointer to the function, which boost::bind wraps up in an object, but which the efl can't use. it would be nice to use stuff like boost::bind, but that's a c++ thing and the efl is c --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] Problem using boost::bind and ecore_ipc from Enlightenment
ecore_* takes a pointer to a function whereas boost::bind returns an object (that can be used as a function by operator overloading). to use a member function you need both a pointer to an object instance and a pointer to the function, which boost::bind wraps up in an object, but which the efl can't use. it would be nice to use stuff like boost::bind, but that's a c++ thing and the efl is c Ok so i must use something else then i guess. Thanks for this good Explanation! I didn't realize this. So i guess i must code a Mini IPC Thing or think about restructuring my Framework in another Fashion. ;) BtW: Enlightenment rocks! Greets, Christian --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] Problem using boost::bind and ecore_ipc from Enlightenment
On Sun, 31 Jul 2005 22:08:37 +0200 Der Lockruf des Kaos [EMAIL PROTECTED] babbled: ecore_* takes a pointer to a function whereas boost::bind returns an object (that can be used as a function by operator overloading). to use a member function you need both a pointer to an object instance and a pointer to the function, which boost::bind wraps up in an object, but which the efl can't use. it would be nice to use stuff like boost::bind, but that's a c++ thing and the efl is c Ok so i must use something else then i guess. Thanks for this good Explanation! I didn't realize this. yeah - basically you will need to write a c callback to c++ boost::bind style calling to make it work. one side of this must me a standard C function call of the prototype ecore expects as handler callbacks. it is possible as you can create a generic callback handler that uses the data pointer as the pointer to the boots:bind: object and then proceeds to call down further from there using c++ as the object you gety is both an object AND a callback (from boost::bind). btw - just theorising here - no practical examples for you. remember efl is c and done in a c way making it easy for c. c++ is often very similar and it is possible to bridge the 2 BUT when you jump into heavily OO c++ construct systems like boost::bind the separation between c and it gets larger and harder to bridge. So i guess i must code a Mini IPC Thing or think about restructuring my Framework in another Fashion. ;) BtW: Enlightenment rocks! Greets, Christian --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users -- - Codito, ergo sum - I code, therefore I am -- The Rasterman (Carsten Haitzler)[EMAIL PROTECTED] 裸好多 [EMAIL PROTECTED] Tokyo, Japan (東京 日本) --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
Re: [e-users] Problem using boost::bind and ecore_ipc from Enlightenment
i'm not sure how you mean to use a smart_ptr - maybe we're talking about different things here? a smart_ptr is also an object and can't be passed to a function expecting a function pointer. as for the example, i'm guessing you mean the SignalHandler class? i don't really see the point of that template at all; have you tried calling that constructor with anything other than a function for the CallBack type? Vincent Torri wrote: Can't you wrap ecore objects with a boost::smart_ptr ? I use the smart_ptr to wrap objects of the Gstreamer library without any problem. After that, I think it is not difficult to use the C++ stuff of boost. For examples, maybe you can have a look there: http://cvs.sourceforge.net/viewcvs.py/avisynth2/avisynth/src/gstreamer/?only_with_tag=avisynth_3_0 Hope this helps Vincent On Sun, 31 Jul 2005, xnowfall wrote: ecore_* takes a pointer to a function whereas boost::bind returns an object (that can be used as a function by operator overloading). to use a member function you need both a pointer to an object instance and a pointer to the function, which boost::bind wraps up in an object, but which the efl can't use. it would be nice to use stuff like boost::bind, but that's a c++ thing and the efl is c --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
[e-users] Funny bug with iconbar?
Hey ho, A very happy user of the All New Enlightenment, I noticed some odd things going on with the iconbar. I'm using the Debian repository at gefechtsdienst.de. Yesterday, I started rearranging iconbar and favorite .orders, installing another theme than the default while I was at it. Now most launchers work fine from the favorites menu, i.e. I haven't noticed the phenomenon there, but starting applications from the iconbar often results in funny error messages. The gedit icon, for example, half of the time yields an error saying Enlightenment was unable to run the program:The command was not found. The program it ought to be trying to run, is just an empty string, it seems. Without changing anything though, every once in a few clicks, it suddenly mentions a program execution line such as images/10 or a certain font name or whatever. At first, I had it with the other icons too, but the problem has persisted only with this specific application; in some way, for other icons, it just disappeared without me doing anything at all. I made sure to -set-exe of gedit.eap several times, but it doesn't change a bit. I wonder, to conclude this all too verbose mail, about two things: * Why on earth can I start the program just fine from the favorites menu, and not from the iconbar, seeing they both use the same eap? * Where on earth does it get an empty string or other gibberish as an -exe line? Has anyone seen this before? Cheers, Tom -- keys: http://tmp.verbreyt.be/files/ (abwaerts.asc verbreyt.asc) --- np: Pink Floyd - Have a Cigar --- SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477alloc_id=16492op=click ___ enlightenment-users mailing list enlightenment-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-users
