[Pharo-project] Progress Update
I promised to fix my partial refactoring for the ProgressBar. Finally this evening I hacked on the whole progress mess :), now I am almost done with one little bug left. - the progress bar is always responsive - all processes showing a progress with priority <60 can be interrupted - refreshes periodically 16 times per second - only forces a manual refresh when blocking the UI - no more out of bounds progress bars - no more ugly display buffer pattern when resizing the screen during a running progress this is a massive improvement for me :) For instance now you can simply interrupt the test runner with a click on the progress bar. I have one bug left, that is when I run the test runner, the UI event handling gets somehow disturbed. Windows no longer receive events, but everything is directly forwarded to the World. Opening the World menu is the only thing you can do :/ Anyone has a clue? The slice is in the inbox, see issue http://code.google.com/p/pharo/issues/detail?id=7103 best cami
Thanks a lot for all the hard work, Camillo. You are improving Pharo a lot - often by just doing the hard work on your own. All these UX improvements are the icing on the internal cleanups. On 06 Dec 2012, at 13:05, Camillo Bruni <camillobruni@gmail.com> wrote:
I promised to fix my partial refactoring for the ProgressBar. Finally this evening I hacked on the whole progress mess :), now I am almost done with one little bug left.
- the progress bar is always responsive - all processes showing a progress with priority <60 can be interrupted - refreshes periodically 16 times per second - only forces a manual refresh when blocking the UI - no more out of bounds progress bars - no more ugly display buffer pattern when resizing the screen during a running progress
this is a massive improvement for me :) For instance now you can simply interrupt the test runner with a click on the progress bar.
I have one bug left, that is when I run the test runner, the UI event handling gets somehow disturbed. Windows no longer receive events, but everything is directly forwarded to the World. Opening the World menu is the only thing you can do :/
Anyone has a clue? The slice is in the inbox, see issue
http://code.google.com/p/pharo/issues/detail?id=7103
best cami
Camillo I do not know. Now I will have a look at your change because I was also fixing a bit the progressBar class. Removing some useless methods and I grouped all the job class under kernel-job and progressbar under morphic progress bar because I was fed up to see them in process and system support mixed with other.
I promised to fix my partial refactoring for the ProgressBar. Finally this evening I hacked on the whole progress mess :), now I am almost done with one little bug left.\\
Cool you are addicted!
- the progress bar is always responsive - all processes showing a progress with priority <60 can be interrupted - refreshes periodically 16 times per second - only forces a manual refresh when blocking the UI - no more out of bounds progress bars - no more ugly display buffer pattern when resizing the screen during a running progress
this is a massive improvement for me :) For instance now you can simply interrupt the test runner with a click on the progress bar.
Did you check that it works with 1 to: 1 because in the past I got a zeroDivide
I have one bug left, that is when I run the test runner, the UI event handling gets somehow disturbed. Windows no longer receive events, but everything is directly forwarded to the World. Opening the World menu is the only thing you can do :/
Anyone has a clue? The slice is in the inbox, see issue
http://code.google.com/p/pharo/issues/detail?id=7103
best cami
On 2012-12-06, at 18:14, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote:
Camillo I do not know. Now I will have a look at your change because I was also fixing a bit the progressBar class. Removing some useless methods and I grouped all the job class under kernel-job and progressbar under morphic progress bar because I was fed up to see them in process and system support mixed with other.
ok, that looks fine :) separate packages should clarifiy things a bit :) Maybe I have time to look into it, today. Otherwise I won't touch the computer for 7 days :P and enjoy the georgious argentinian country side!
I promised to fix my partial refactoring for the ProgressBar. Finally this evening I hacked on the whole progress mess :), now I am almost done with one little bug left.\\
Cool you are addicted!
- the progress bar is always responsive - all processes showing a progress with priority <60 can be interrupted - refreshes periodically 16 times per second - only forces a manual refresh when blocking the UI - no more out of bounds progress bars - no more ugly display buffer pattern when resizing the screen during a running progress
this is a massive improvement for me :) For instance now you can simply interrupt the test runner with a click on the progress bar.
Did you check that it works with 1 to: 1 because in the past I got a zeroDivide
there should be no more min-max, hence no more 0 divide possible. But yeah, I got a bit lost in fixing the UI part ;) so I still have to do the proper transformations on the rest of the image.
On Dec 7, 2012, at 7:17 PM, Camillo Bruni wrote:
On 2012-12-06, at 18:14, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote:
Camillo I do not know. Now I will have a look at your change because I was also fixing a bit the progressBar class. Removing some useless methods and I grouped all the job class under kernel-job and progressbar under morphic progress bar because I was fed up to see them in process and system support mixed with other.
ok, that looks fine :) separate packages should clarifiy things a bit :)
I could not separate the packages because we found a bug in rpackage integration.
Maybe I have time to look into it, today. Otherwise I won't touch the computer for 7 days :P and enjoy the georgious argentinian country side!
there should be no more min-max, hence no more 0 divide possible. But yeah, I got a bit lost in fixing the UI part ;) so I still have to do the proper transformations on the rest of the image.
I see what you mean because there are plenty of places where it is based on size. I will try to integrate my changes with progressBar (I try to fix the size problem but it often depends on the users). Stef
please fix the progress bars. i downloaded latest image today and it blocks all existing windows from being accessible. the only cure i found is tell world to remove all windows. new windows working normally. -- Best regards, Igor Stasenko.
On Dec 7, 2012, at 9:10 PM, Igor Stasenko wrote:
please fix the progress bars. i downloaded latest image today and it blocks all existing windows from being accessible. the only cure i found is tell world to remove all windows. new windows working normally.
Or I will rollback the change. Sorry my fault.
-- Best regards, Igor Stasenko.
participants (4)
-
Camillo Bruni -
Igor Stasenko -
Stéphane Ducasse -
Sven Van Caekenberghe