(old)Puppy Linux Discussion Forum

sorry folks, I have been up above my ears (literally) in paint. !I hate painting ceilings!

However I have downloaded joe's latest v762 and will build today. If it passes my basic set it should be up tonight/tomorrow for your wider testing.
EDIT: Done, see first post for the 32bit test pet, x86_64 will appear this weekend.

@Oscar, I suspect it's google earth but when I have built the next version can you try again. If it's still happening I shall pass it on to joe for a check.

If the ceiling's fairly plain, you can use a specialized roller cover for ceilings from Sears (hardly any dripping!) and go a bit slower with it..

more jwm-ish (dotpet of 762 and etc)

Handles well in racy 5491 and slacko55. One of the differences I looked for was the handling of 'serial translucency' with kcompmgr -n enabled - it does indeed seem more simplified, in essence the 'stepping' of the translucency effect according to layer of individual windows. The overall visual effect is similar to pre-750 versions but slightly different in use (actually seems more robust in keeping with Joe's overall plan). Moving windows by r-click-drag in the pager feels quite smooth and even the 'pop' as desktop is changed when moving over halfway past the boundary is hardly noticeable.

Custom-compile in p431. Oh well, still has the drawing-remnant issue for max, max-h and max-v when using titlebar button for that in p431 (but fine with window or tasklist button menu) with kcompmgr -n enabled. Noticeably less smooth in p431 for moving windows by r-click-drag in the pager also. The stepping of serial translucency is simplified for sure, much more obvious when the xpad windows I use for test have the same bg. Comparing against vgit-731, sloppy focus looks a bit better integrated now too.

All very quick and responsive, looks like a winner to me.

thanks 'f00, but the house is a pre-1st world war timber and iron roofed building, nothing is plain, straight, square,etc.,..., but I like it as it's 'small' and there is only me (and four cats) living in it nowadays. And before anyone yells about the cats, they are all desexed, were strays, and up here in the wheat belt, a necessity to keep mice out of the house and workshops for at least 9 months of the year.

However, back on topic.

I've just downloaded a new snapshot v768, so that should come the test way uploaded tonight, one thing it has is some updated po files that need checking please by our non-english speakers please, and a few more supplied if people have time. from an initial build, it seems a bit faster with window creation as joe has apparently done some changes and removed some redundant code.

I have uploaded onto the first post JWM Version 768.

I have tested it as ok and it's time to have extensive testing. You have seven days!

If this version is good it will be rebuilt as the next Release version.

Hi,

I use Precise 5.6 with a 1024 x 768 pixel screen.

After installing jwm 768 and Restarting X both abiword
and dillo open their 936 x 702 windows in a new position
(some 25 pixels right of their pre-768-installation point).

My regards

PS I used 936 x 702 because if I made my dillo or abiword
less than full screen windows with other than a 4 x 3 ratio
they only opened in the exact position I wanted on exactly
every second opening.

@jasper, hi

Is it only those two apps or do all apps start to the right?

Also, if you shrink them a little bit more, do they still move right?

before I put a problem to joe, I had best get some further info as I cannot duplicate your problem on any of my boxes, although i don't have precise 5.6 loaded anywhere, downloading one now. By the way there is a 5.6.1 bugfix out.

Hi again,

I reduced abiword to 900 x 675 and dillo to 800 x 600 and
checked the pre and post v 768 positions with a pixel ruler
and saw no movement, Since you have almost certainly hit the
nail on head and as they are the only apps I want to and can
open in a specific position I did not try to test more apps.

I did post in BarryK's Precise 5.6/5.6.1 at the very bottom
of page 5 about the superb speed increase in JWM menus
thanks to technosaurus - I expect Joe and yourself already
know about that.

I will not be suggesting to BarryK that there may be a
minor bug in Precise 5.6/5.6.1 (though I do hope npierce
will update Barry on his abiword non-US spell-checker
problem fixes).

To be honest I have never used the layer optioms in earnest
and I gather JWM v768 Test has some speed increases and
a nice right click shade option. I only use JWM and ROX so
I'm very pleased Joe is making improvements and although
I was and am happy to try to test v768 I have no idea of
any other changes.

My regards

I fear that my request for easier window minimization (issue #45) may make it easier to break nested windows. Ideally, when all of the windows in a stack have individual minimize buttons, you'd expect to be able to minimize all of them using the taskbar entry as well. But when you're dealing with windows that cannot be closed until the child window is closed... things are easier to break.

This can be duplicated in QtWeb. Options -> Advanced -> Search Providers -> Add New. Minimize and restore the main window several times, and one of the nested windows will break. Granted, you had the ability to minimize multiple nested windows before issue #45 was resolved, you just couldn't do it from the taskbar. But I can't duplicate this in v574.

@jasper, I think your problem has been fixed by joe with the latest snapshot.

@session, Yours may also be ok now, joe has done some fixes that may have sorted it all out.

I shall build, test and make available a V772 test build tomorrow, I have to go out tonight to a CFA meeting.

ChangeLog<---(from current all the way back to 2005!) says

No more custom cursors for close/minimize/maximize.

as of Fri May 31 18:42:46 2013

If you like that feature, it can be swapped back in with the src/cursor.c from 762 .. doesn't seem to break anything in 772 as far as I can tell. Of course whether the custom hover cursors look 'right' depends on your cursor set

I have uploaded a test version 775, have fun!

Hi,

Thank you for v775 (Test). but the movement persists.

My regards

...meanwhile, my problem is fixed.

Resizing while maximized is now disabled... I've used that quite often to set window size, but it became redundant after recent versions made it so easy to maximize X or Y; this is certainly more streamlined.

@jasper: window placement is a little different after v768 in some apps... you can always use the --geometry option, say 936x702+0+0 (curiously, dillo won't respect 0+0, you have to do 0+1 to get as close to the corner as possible)

@jasper, I have added a problem to joe's issue list for you.

EDIT 14june: I have just downloaded v777, I shall build and test tonight before uploading tomorrow as i'm working today elsewhere.

@ scsijon - my thanks

@ session - I tried +0+1 and +1+1, but no joy - perhaps as
I'm using dillo v 2.2 from Tman.

My regards

@jasper

Joe answered with:-

The window placement has changed somewhat, so I would expect that some windows might open in different locations. Are the applications being started with a "-geometry" option?

I must admit that I haden't seen this mentioned in the changelog, else I would have mentioned it.

Can you try to work it out with sessions help?

I have put the problem on joe's system on hold until needed.

regards

@jasper: It seems that the change has happend to fix the problems relating to windows dissapearing off screen.
Details from joe are:

The relevant commits that would have changed the behavior are 3ee8ddc and 4dd7c84. The problem these commits solve relates to the placement of windows that would go off the screen if placed where JWM would normally place a window.

Commit 3ee8ddc makes it so JWM correctly accounts for the window border size.

Commit 4dd7c84 makes it so JWM doesn't revert to the upper left-most position on the screen if the placement algorithm fails (instead, only the direction of the overflow is reset). In addition, 4dd7c84 constrains newly mapped windows to the area of the screen that JWM is willing to use even if a program specifies its own position.

So it looks like we are stuck with them as it's fixing other problems.

---------------------------------

@all: I have uploaded a test build of v777 for your testing.

If all is ok after some user testing, I intend using this for the next release version in about seven days so can I have responses in the thread with the puppy's you have used or tested it against listed please.

I shall also build a x86_64 as previously promised now I have found which box the Fatdog x86 partition is on!

regards

EDIT: 785 will NOT be uploaded tonight. Barryk had found a problem relating to geany which joe had fixed, however it's broken again.

Hi scsijon,

Thank you for all the trouble you went to in following up my
observation and advising why Joe made the change for our
greater general good.

My regards

Although joe is up to version 790 and I have built a few since v777, problems have been found that don't make me confident enough to be willing to release new test versions.

We shall have to be a little patient.

Well, we have a new test version V805.

It contains a few more changes and fixups.

have fun