because of that, I have an idea, how if the window get half size of screen if we drag to the right or left border of screen? like kde or wdz8 did.
one more, how if we add script:
Code: Select all
vbox scrollable=\"true\"[CLOSED]JWM Desktop Updates to V755(Release) and V910(Test)
I downloaded the new jwm. I think this is good, change the cursor at minimize, maximize, and close button. but now I can't drag my window to another desktop,just stack at border of screen, but at the previous version I can.
because of that, I have an idea, how if the window get half size of screen if we drag to the right or left border of screen? like kde or wdz8 did.
one more, how if we add script:
in /usr/local/jwmconfig2/theme_switcher, so that if there are a lot of jwm theme, the window not too high. like pemasu did
because of that, I have an idea, how if the window get half size of screen if we drag to the right or left border of screen? like kde or wdz8 did.
one more, how if we add script:
@joewing
Thanks for the example(clarification)/tips! Agreed that your wm is best served lean but with just the right 'fat' content to make it tasty :9
Good idea to make a personal cursor theme although perhaps a bit involved (elegant and useful for the industrious, however). I 'cheated' by modifying src/cursors.c (before compile) to show left_ptr for the titlebar button hovers as some users would like.
___more testing - puppy 4.31
A vintage pup, to be sure (jwm vsvn-457). The stock 704 just wouldn't fly (or even make it to desk), good thing I got other wms
. Too much different to mess with bandaids, so a fresh compile in p431 did the trick properly. A few minor adjustments and version-switching is do-able.
Something I've noticed in the two compiles I just did .. memory footprint is noticeably less (from a ballpark 1.3% of RAM on my machine to 0.8%). Possibly the configure options I used..
CFLAGS="-mtune=generic -O2 -fomit-frame-pointer" CXXFLAGS="${CFLAGS}" --prefix=/usr --localstatedir=/var --disable-confirm --disable-cairo --disable-fribidi --disable-nls --disable-debug
..since cairo was disabled, rsvg was also (implicit).
@members
Pm me if you'd like an archive (I rarely do dotpets) of my 'special' 704 ~and/or~ 704-for-p431
If scsijon indicates, I could attach them if that's okay (otherwise not)
___possible request(s)
1) border/outline-less tray/tasklist buttons for a 'flat' look or an example of what/how to edit for such
2) very minor .. rounded popup/tooltips?
Thanks for the example(clarification)/tips! Agreed that your wm is best served lean but with just the right 'fat' content to make it tasty :9
Good idea to make a personal cursor theme although perhaps a bit involved (elegant and useful for the industrious, however). I 'cheated' by modifying src/cursors.c (before compile) to show left_ptr for the titlebar button hovers as some users would like.
___more testing - puppy 4.31
A vintage pup, to be sure (jwm vsvn-457). The stock 704 just wouldn't fly (or even make it to desk), good thing I got other wms
. Too much different to mess with bandaids, so a fresh compile in p431 did the trick properly. A few minor adjustments and version-switching is do-able.Something I've noticed in the two compiles I just did .. memory footprint is noticeably less (from a ballpark 1.3% of RAM on my machine to 0.8%). Possibly the configure options I used..
CFLAGS="-mtune=generic -O2 -fomit-frame-pointer" CXXFLAGS="${CFLAGS}" --prefix=/usr --localstatedir=/var --disable-confirm --disable-cairo --disable-fribidi --disable-nls --disable-debug
..since cairo was disabled, rsvg was also (implicit).
@members
Pm me if you'd like an archive (I rarely do dotpets) of my 'special' 704 ~and/or~ 704-for-p431
If scsijon indicates, I could attach them if that's okay (otherwise not)
___possible request(s)
1) border/outline-less tray/tasklist buttons for a 'flat' look or an example of what/how to edit for such
2) very minor .. rounded popup/tooltips?
- Attachments
-
- 704special.png
- illustration
- (14.21 KiB) Downloaded 744 times
'f00, the stock jwm I built is for puppy 5 level's, i'm not suprised it didn't work with 431.
If you want to attach the archive to a pm i'll pet it up and make available on the first post.
I did consider trying to build a puppy 214, but decided not, unless I have a few requests.
I no longer have a working 431 since that machine had a power supply malfunction and put 240v on the 5v power rail, quite a light and smoke show at the time.
If you want to attach the archive to a pm i'll pet it up and make available on the first post.
I did consider trying to build a puppy 214, but decided not, unless I have a few requests.
I no longer have a working 431 since that machine had a power supply malfunction and put 240v on the 5v power rail, quite a light and smoke show at the time.
I'm having the same problem, can't find the setting to allow windows to be dragged off screen edge to another virtual desktop.recobayu wrote:I downloaded the new jwm. I think this is good, change the cursor at minimize, maximize, and close button. but now I can't drag my window to another desktop,just stack at border of screen, but at the previous version I can....
Other than that 704 seems good, tried with qjackctl (qt) and audacity (wxWidgets) - both appear working fine.
@`foo:
You can make tray buttons border-less/flat by setting border="false". This wasn't supported in the past for task lists, but as of snapshot 707, you can do this for task lists as well.
I'm kind of divided on rounded popups. They were there in the past and then I removed them when redoing some of the code...
@recobayu, @darkcity:
To drag to another virtual desktop, drag the window to the side of the screen and hold it there. The default timeout is 1000 milliseconds, but you can change that using the undocumented "delay" attribute to MoveMode (I guess I should document that).
As far as Inkscape is concerned, that's really annoying. Mouse bindings might be more important than I thought.
You can make tray buttons border-less/flat by setting border="false". This wasn't supported in the past for task lists, but as of snapshot 707, you can do this for task lists as well.
I'm kind of divided on rounded popups. They were there in the past and then I removed them when redoing some of the code...
@recobayu, @darkcity:
To drag to another virtual desktop, drag the window to the side of the screen and hold it there. The default timeout is 1000 milliseconds, but you can change that using the undocumented "delay" attribute to MoveMode (I guess I should document that).
As far as Inkscape is concerned, that's really annoying. Mouse bindings might be more important than I thought.
I thought Joe disabled this (hi Joedarkcity wrote:I'm having the same problem, can't find the setting to allow windows to be dragged off screen edge to another virtual desktop.recobayu wrote:I downloaded the new jwm. I think this is good, change the cursor at minimize, maximize, and close button. but now I can't drag my window to another desktop,just stack at border of screen, but at the previous version I can....
Other than that 704 seems good, tried with qjackctl (qt) and audacity (wxWidgets) - both appear working fine.
)To send a window to another desktop click the icon in the top left of the title bar and choose "Send to" or right click the window's entry in the pager, same procedure.
Puppy Linux Blog - contact me for access
Okay I got it working now just wait for the 1000mil delay. Its good to have some delay. I'd probably lower it about 500mil.joewing wrote:
@recobayu, @darkcity:
To drag to another virtual desktop, drag the window to the side of the screen and hold it there. The default timeout is 1000 milliseconds, but you can change that using the undocumented "delay" attribute to MoveMode (I guess I should document that).
As far as Inkscape is concerned, that's really annoying. Mouse bindings might be more important than I thought.
The Inkscape thing should be configurable in Inkscape - unless JWM is 'grabbing' the drag action in some way.
I've found 50 milliseconds to be about right for me
I find it counter intuitive that the window appears in the same place when the next virtual desktop to switch to. It would be cool if the window appears on the opposite edge (ie. drag off far left -> appears on far right).
Code: Select all
<!-- The move mode (outline or opaque), virtual desktop switch delay -->
<MoveMode delay="50">outline</MoveMode>@darkcity - mmm, bearing in mind that jwm uses discreet desktops (rather than viewport mode, etc) there's a few issues
1) layout of desktops - serial horizontal/vertical or grid, it's an exercise for the mapping to determine where/how a window appears on the 'next' desktop. Seems like when dragging in the pager, a window will flip to the dragto desk when it reaches > 50% of the window - there's some geometry math involved in the positioning but it does flip intuitively according to my layout borders in 704
2) ring or linear - ring connects the last desk to the first in circular fashion
, believe jwm uses linear
as for grab, that's pretty much in the man
@scsijon - I should run some compatibility check for jwmconfig2 items in p431, wrapper should work for the vdesk difference but I'm so used to setting it up for manual switching via personal and binary.. that way I can keep the buttonmask deal in the old version(s)
Had some fun cobbling together a custom cursor set (no more all-black + pointer!)
Now to move on to 707 …
1) layout of desktops - serial horizontal/vertical or grid, it's an exercise for the mapping to determine where/how a window appears on the 'next' desktop. Seems like when dragging in the pager, a window will flip to the dragto desk when it reaches > 50% of the window - there's some geometry math involved in the positioning but it does flip intuitively according to my layout borders in 704
2) ring or linear - ring connects the last desk to the first in circular fashion
, believe jwm uses linearas for grab, that's pretty much in the man
@scsijon - I should run some compatibility check for jwmconfig2 items in p431, wrapper should work for the vdesk difference but I'm so used to setting it up for manual switching via personal and binary.. that way I can keep the buttonmask deal in the old version(s)
Had some fun cobbling together a custom cursor set (no more all-black + pointer!)
Now to move on to 707 …
That's it?!`f00 wrote:I 'cheated' by modifying src/cursors.c (before compile) to show left_ptr for the titlebar button hovers as some users would like.
And the issue becomes moot? If it's this simple, I'd like to do a new compile. Just need to decide on the right, latest, "good" version. I've noticed, Slacko has version 692, btw.
Thank you for sharing this trick.
Of course then you just get a 'standard' pointer for the hovers. Some cursor sets like Neutral_Plus_Inverted (aka npi) actually look really good for the hovers - the appearance is fairly intuitive, anyhow .. even a few surprises in that set (a proper right_ptr for one)
On the testing side, did a quick-o compile of 707 in p431 - it crunked out on src/fonts with default options enabled, but fine with my usual disables .. glad scsijon does do his with fribidi and so on for the wider base of users. For a lark, I tried that binary in racy 5.4.91 .. seems fine on all the major functions.
Minor - tried the new tasklist attribute border="0" (or "false") in tray file, border/outline persists on buttons (windowbuttons in tasklist and traybuttons as well)
It's a cosmetic preference to be sure, full windows or shaded titlebars take a border/outline well but it can look a bit overdone in the confines of a tray (with some themes it's more noticeable than others). Maybe something odd on my side (disables or using the p431 binary in racy), will test again when scsijon does a 707..
heh, the real magic is having good source code to work with
.. even a few surprises in that set (a proper right_ptr for one)On the testing side, did a quick-o compile of 707 in p431 - it crunked out on src/fonts with default options enabled, but fine with my usual disables .. glad scsijon does do his with fribidi and so on for the wider base of users. For a lark, I tried that binary in racy 5.4.91 .. seems fine on all the major functions.
Minor - tried the new tasklist attribute border="0" (or "false") in tray file, border/outline persists on buttons (windowbuttons in tasklist and traybuttons as well)
It's a cosmetic preference to be sure, full windows or shaded titlebars take a border/outline well but it can look a bit overdone in the confines of a tray (with some themes it's more noticeable than others). Maybe something odd on my side (disables or using the p431 binary in racy), will test again when scsijon does a 707..heh, the real magic is having good source code to work with
I shall be building jwm's thursday night,
!I need a change from working on mage2 as I have been running around in tighter circles the last few days, I finally realized the problem last night. Nearly three months work unnecessary, ouch!
@'f00, maybe instead of playing with mick's jwmconfig2 with 431 you should use the 691 config pet on the first page as it's barryk's puppy standard', just updated for the new jwm format.
EDIT: WED
@ all joe is up to 715, I shall build tomorrow pm and upload thur night as i've had a sudden 'burst of brain power' and I think that I have sorted out most of my remaining mage2 problems!
!I need a change from working on mage2 as I have been running around in tighter circles the last few days, I finally realized the problem last night. Nearly three months work unnecessary, ouch!
@'f00, maybe instead of playing with mick's jwmconfig2 with 431 you should use the 691 config pet on the first page as it's barryk's puppy standard', just updated for the new jwm format.
EDIT: WED
@ all joe is up to 715, I shall build tomorrow pm and upload thur night as i've had a sudden 'burst of brain power' and I think that I have sorted out most of my remaining mage2 problems!
JWM V704 release & 716 test
I have updated both the Release Version to 704, and Test Version to 716.
Attached is the changelog for those interested. Rename it without the fake .gz and open with geany to view it.
Will testers make sure they state what puppy they are testing on please.
As usual, the Test Version will be rebuilt to Release Version Quality approximately 14 days after no problems have been found unless a further test version is required.
Have fun folks.
scsijon
Attached is the changelog for those interested. Rename it without the fake .gz and open with geany to view it.
Will testers make sure they state what puppy they are testing on please.
As usual, the Test Version will be rebuilt to Release Version Quality approximately 14 days after no problems have been found unless a further test version is required.
Have fun folks.
scsijon
- Attachments
-
- ChangeLog_002.asc.gz
- (178.58 KiB) Downloaded 441 times
tested on racy 5.4.91 (kernel 3.8.0)
Everything looks good in normal mode.
With kcompmgr -n enabled .. toggling max-full, max-v and max-h leaves remnants on the display. If compiled with Shape disabled, there are no remnants (but windows lose the rounded corners ).
___a few pldf-links for those interested in kcompmgr and related___
Patriot's 2.0.1 stuff
'related' somewhat
___jwm2-716-w5c in other pups___
precise 5.3.96 (kernel 3.2.29) - ok
quirky 1.4.2 (kernel 2.6.32.28 ) - good¹
wary 5.1.4 (kernel 2.6.32-smp) - sub-par with xcompmgr, otherwise fine
wary 0.2.0 (kernel 2.6.33.2) - ok²
¹ - unable to test with compmgr (odd xorg doesn't recognize extension)
² - needs newer syntax for virtual desk layout (was older 'count' style)
___721 compiled on p431___addenda 130404___
.. really seems to be getting a handle on the compmgr-max-toggling thing ..
puppy 4.3.1 - still has remnants (max-* toggled from titlebar buttons)
racy 5.4.91 - no remnants at all, very stable with all control methods
wary 5.1.4 - if toggled from winmenu, all is fine .. toggling from titlebar buttons gives a variety of repeatable results (differing dependent on if k or xcompmgr used)
___722 compiled on wary 5.1.4 ~and~ the w5c binary___addenda 130407___
precise 5.3.96 - w/kcompmgr all control methods good
.. w/xcompmgr, no unshade drawn from titlebar button nor kybd (I use Alt+s to toggle shade)
Everything looks good in normal mode.
With kcompmgr -n enabled .. toggling max-full, max-v and max-h leaves remnants on the display. If compiled with Shape disabled, there are no remnants (but windows lose the rounded corners
).___a few pldf-links for those interested in kcompmgr and related___
Patriot's 2.0.1 stuff
'related' somewhat
___jwm2-716-w5c in other pups___
precise 5.3.96 (kernel 3.2.29) - ok
quirky 1.4.2 (kernel 2.6.32.28 ) - good¹
wary 5.1.4 (kernel 2.6.32-smp) - sub-par with xcompmgr, otherwise fine
wary 0.2.0 (kernel 2.6.33.2) - ok²
¹ - unable to test with compmgr (odd xorg doesn't recognize extension)
² - needs newer syntax for virtual desk layout (was older 'count' style)
___721 compiled on p431___addenda 130404___
.. really seems to be getting a handle on the compmgr-max-toggling thing ..
puppy 4.3.1 - still has remnants (max-* toggled from titlebar buttons)
racy 5.4.91 - no remnants at all, very stable with all control methods
wary 5.1.4 - if toggled from winmenu, all is fine .. toggling from titlebar buttons gives a variety of repeatable results (differing dependent on if k or xcompmgr used)
___722 compiled on wary 5.1.4 ~and~ the w5c binary___addenda 130407___
precise 5.3.96 - w/kcompmgr all control methods good
.. w/xcompmgr, no unshade drawn from titlebar button nor kybd (I use Alt+s to toggle shade)
- Attachments
-
- shade-tb-xcompmgr.png
- 722 (w/xcompmgr) in precise 5.3.96
- (11.09 KiB) Downloaded 519 times
-
- remnants.png
- screencap from p431
- (6.74 KiB) Downloaded 552 times
-
- 716all.png
- some differences
- (128 KiB) Downloaded 659 times
Last edited by `f00 on Mon 08 Apr 2013, 18:05, edited 6 times in total.
thank you 'f00,
yes, the systemconf line is misleeding as we don't use it, but it has to be there in the build, it won't build without it. I remove the file before I pet it up to both save space and make sure the jwm default one isn't used as it's nothing like puppy's (it's really an example). JWM looks for a $HOME/.jwmrc first anyway, which for us is /root/.jwmrc.
I shall report the remnants problem. Personally I haven't seen it in testing, but I shall re-check and see what I find.
EDIT: a quickcheck and it's not happening with rox-filer, thunderbird, firefox or a console screen. However I have reported it anyway, maybe it's a quick find and fix. Maybe it's worth adding a link for kcompmgr in your message above in case he wants to download it to test with.
thanks
scsijon
yes, the systemconf line is misleeding as we don't use it, but it has to be there in the build, it won't build without it. I remove the file before I pet it up to both save space and make sure the jwm default one isn't used as it's nothing like puppy's (it's really an example). JWM looks for a $HOME/.jwmrc first anyway, which for us is /root/.jwmrc.
I shall report the remnants problem. Personally I haven't seen it in testing, but I shall re-check and see what I find.
EDIT: a quickcheck and it's not happening with rox-filer, thunderbird, firefox or a console screen. However I have reported it anyway, maybe it's a quick find and fix. Maybe it's worth adding a link for kcompmgr in your message above in case he wants to download it to test with.
thanks
scsijon
Round windows are madness anyway (pixels and screens aren't round). I still find it surprising that people want that feature in JWM...
(pixels and screens aren't round). I still find it surprising that people want that feature in JWM...Do you know a good gtkdialog program? Please post a link here
Classic Puppy quotes
ROOT FOREVER
GTK2 FOREVER
Classic Puppy quotes
ROOT FOREVER
GTK2 FOREVER
v720
Joe has released a new test version 720, I shall build, Test it myself, and if it's ok I will upload as a test version tomorrow, it's suppose to fix a few things like 'f00's.
regards
EDIT: We ended up with version 722, it's on test and available for download in the first post.
regards
EDIT: We ended up with version 722, it's on test and available for download in the first post.
-
OscarTalks
- Posts: 2196
- Joined: Mon 06 Feb 2012, 00:58
- Location: London, England
Just to report that the problem of SlimBoat Browser maximized window opening displaced is now FIXED in JWM version 726 (which it wasn't in 704).
Many thanks to Joe for all the hard work. It is very much appreciated.
By the way, personally I quite like the little changing cursor shapes. At least they give a visible indication that you are hovering in position over the button.
Many thanks to Joe for all the hard work. It is very much appreciated.
By the way, personally I quite like the little changing cursor shapes. At least they give a visible indication that you are hovering in position over the button.
Oscar in England
