The Trunk: Morphic-nice.683.mcz

Nicolas Cellier uploaded a new version of Morphic to project The Trunk:
http://source.squeak.org/trunk/Morphic-nice.683.mcz

==================== Summary ====================

Name: Morphic-nice.683
Author: nice
Time: 20 September 2013, 10:47:32.05 pm
UUID: 3c0205a8-73ce-406f-a9d3-71c03be3335f
Ancestors: Morphic-cmm.682

Use displayWorldSafely rather than doOneCycle in PluggableListMorph.
No idea if it really does the intended preselection job, but at least it avoids all these red morphs of death to yellow cross my workflow.

=============== Diff against Morphic-cmm.682 ===============

Item was changed:
  ----- Method: PluggableListMorph>>changeModelSelection: (in category 'model access') -----
  changeModelSelection: anInteger
        " Change the model's selected item index to be anInteger. Enable the pre selection highlight. Step the World forward to let the pre selection highlight take effect. "

        self rowAboutToBecomeSelected: (self uiIndexFor: anInteger).
+       World displayWorldSafely.
-       World doOneCycle.
        setIndexSelector ifNotNil: [
                model perform: setIndexSelector with: anInteger ].!

On Tue, 24 Sep 2013, Nicolas Cellier wrote:

Ah my bad, I have overriden a package with same name !!! (This was from source.squeak.org web interface, and I got no warning...)
How can I undo? Not sure whether I have a copy of the original or not...

No idea. I haven't found it on my pc, nor on the CI server. Maybe Randal's backups have them. I wonder how hard is it to recreate a .mcz file from a previous .mcz and a .mcd.


Levente

2013/9/24 <[hidden email]>
      Nicolas Cellier uploaded a new version of Morphic to project The Trunk:
      http://source.squeak.org/trunk/Morphic-nice.683.mcz

      ==================== Summary ====================

      Name: Morphic-nice.683
      Author: nice
      Time: 20 September 2013, 10:47:32.05 pm
      UUID: 3c0205a8-73ce-406f-a9d3-71c03be3335f
      Ancestors: Morphic-cmm.682

      Use displayWorldSafely rather than doOneCycle in PluggableListMorph.
      No idea if it really does the intended preselection job, but at least it avoids all these red morphs of death to yellow cross my workflow.

      =============== Diff against Morphic-cmm.682 ===============

      Item was changed:
        ----- Method: PluggableListMorph>>changeModelSelection: (in category 'model access') -----
        changeModelSelection: anInteger
              " Change the model's selected item index to be anInteger. Enable the pre selection highlight. Step the World forward to let the pre selection highlight take effect. "

              self rowAboutToBecomeSelected: (self uiIndexFor: anInteger).
      +       World displayWorldSafely.
      -       World doOneCycle.
              setIndexSelector ifNotNil: [
                      model perform: setIndexSelector with: anInteger ].!

On 23-09-2013, at 5:30 PM, Nicolas Cellier <[hidden email]> wrote:

> I think I have it on another machine, I'll try later in the morning...
>
>
> 2013/9/24 Levente Uzonyi <[hidden email]>
> On Tue, 24 Sep 2013, Nicolas Cellier wrote:
>
> Ah my bad, I have overriden a package with same name !!! (This was from source.squeak.org web interface, and I got no warning...)

> How can I undo? Not sure whether I have a copy of the original or not…

What has actually happened here? I see a nice orderly progression of Morphic 678 (currently in my image) to 679, 680, …683 with no nasty looking changes at any step. In fact there isa single change from cmm.682 to your nice.683 and that is swapping World displayWorldSafely for World doOneCycle.

I'd quite like to be able to save my small changes to Morphic, so hopefully there's nothing weird going on here...


tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
How many of you believe in telekinesis? Raise my hands....

I produced two mcz with exact same name on two different machines.

The first one (20 september) went in inbox.

The second one (21 september) went in trunk.

Today (24 september) I moved the inbox version into the trunk overriding the one of 21st...

I checked update-nice.246.mcm, and it points to Morphic-cmm.682 fortunately... Otherwise, this could break the update stream (because mcm retain the UUID). So one problem less.

However, what will happen on every package-cache with conflicting version is something I can't predict, and that I'd rather would not want to learn, so I'll try to republish the original trunk version.

2013/9/24 tim Rowledge <[hidden email]>

On 23-09-2013, at 5:30 PM, Nicolas Cellier <[hidden email]> wrote:

> I think I have it on another machine, I'll try later in the morning...
>
>
> 2013/9/24 Levente Uzonyi <[hidden email]>
> On Tue, 24 Sep 2013, Nicolas Cellier wrote:
>
> Ah my bad, I have overriden a package with same name !!! (This was from source.squeak.org web interface, and I got no warning...)

> How can I undo? Not sure whether I have a copy of the original or not…

What has actually happened here? I see a nice orderly progression of Morphic 678 (currently in my image) to 679, 680, …683 with no nasty looking changes at any step. In fact there isa single change from cmm.682 to your nice.683 and that is swapping World displayWorldSafely for World doOneCycle.

I'd quite like to be able to save my small changes to Morphic, so hopefully there's nothing weird going on here...


tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
How many of you believe in telekinesis? Raise my hands....

I have republished Morphic-nice.683 from 20th -> Morphic-nice.684.

So please, tim, restart from that version if you want to publish this evening.

I'll then upload Morphic-nice.683 (21st) later and we'll do the merge later.

2013/9/24 Nicolas Cellier <[hidden email]>

I produced two mcz with exact same name on two different machines.

The first one (20 september) went in inbox.

The second one (21 september) went in trunk.

Today (24 september) I moved the inbox version into the trunk overriding the one of 21st...

I checked update-nice.246.mcm, and it points to Morphic-cmm.682 fortunately... Otherwise, this could break the update stream (because mcm retain the UUID). So one problem less.

However, what will happen on every package-cache with conflicting version is something I can't predict, and that I'd rather would not want to learn, so I'll try to republish the original trunk version.

2013/9/24 tim Rowledge <[hidden email]>

On 23-09-2013, at 5:30 PM, Nicolas Cellier <[hidden email]> wrote:

> I think I have it on another machine, I'll try later in the morning...
>
>
> 2013/9/24 Levente Uzonyi <[hidden email]>
> On Tue, 24 Sep 2013, Nicolas Cellier wrote:
>
> Ah my bad, I have overriden a package with same name !!! (This was from source.squeak.org web interface, and I got no warning...)

> How can I undo? Not sure whether I have a copy of the original or not…

What has actually happened here? I see a nice orderly progression of Morphic 678 (currently in my image) to 679, 680, …683 with no nasty looking changes at any step. In fact there isa single change from cmm.682 to your nice.683 and that is swapping World displayWorldSafely for World doOneCycle.

I'd quite like to be able to save my small changes to Morphic, so hopefully there's nothing weird going on here...


tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
How many of you believe in telekinesis? Raise my hands....

Squeaksource has no UI for retrieving previous versions, but it does preserve them:

squeaksource@box2:~/ss$ ls -l trunk/Morphic-nice.683.*

-rw-r--r--  1 squeaksource squeaksource 1651876 Sep 23 23:55 trunk/Morphic-nice.683.2013-09-23.23-54-59

-rw-r--r--  1 squeaksource squeaksource 1651857 Sep 23 23:55 trunk/Morphic-nice.683.mcz

- Bert -

On 2013-09-24, at 14:33, Nicolas Cellier <[hidden email]> wrote:

Absolutely no trace of original file, neither on my disks, nor in my Time Machine...

I wonder how I could ever publish in these conditions...

Name: Morphic-nice.683
Author: nice
Time: 21 September 2013, 11:19:28.071 pm
UUID: 881a99fa-1788-410b-b676-

c648e1b662f2
Ancestors: Morphic-cmm.682

2013/9/24 Nicolas Cellier <[hidden email]>

I have republished Morphic-nice.683 from 20th -> Morphic-nice.684.

So please, tim, restart from that version if you want to publish this evening.

I'll then upload Morphic-nice.683 (21st) later and we'll do the merge later.

2013/9/24 Nicolas Cellier <[hidden email]>

I produced two mcz with exact same name on two different machines.

The first one (20 september) went in inbox.

The second one (21 september) went in trunk.

Today (24 september) I moved the inbox version into the trunk overriding the one of 21st...

I checked update-nice.246.mcm, and it points to Morphic-cmm.682 fortunately... Otherwise, this could break the update stream (because mcm retain the UUID). So one problem less.

However, what will happen on every package-cache with conflicting version is something I can't predict, and that I'd rather would not want to learn, so I'll try to republish the original trunk version.

2013/9/24 tim Rowledge <[hidden email]>

On 23-09-2013, at 5:30 PM, Nicolas Cellier <[hidden email]> wrote:

> I think I have it on another machine, I'll try later in the morning...
>
>
> 2013/9/24 Levente Uzonyi <[hidden email]>
> On Tue, 24 Sep 2013, Nicolas Cellier wrote:
>
> Ah my bad, I have overriden a package with same name !!! (This was from source.squeak.org web interface, and I got no warning...)

> How can I undo? Not sure whether I have a copy of the original or not…

What has actually happened here? I see a nice orderly progression of Morphic 678 (currently in my image) to 679, 680, …683 with no nasty looking changes at any step. In fact there isa single change from cmm.682 to your nice.683 and that is swapping World displayWorldSafely for World doOneCycle.

I'd quite like to be able to save my small changes to Morphic, so hopefully there's nothing weird going on here...


tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
How many of you believe in telekinesis? Raise my hands....

Hi Nicolas,

wouldn't be "self changed" enough to trigger a redraw of the
PluggableListMorph?

This fix is correct in the sense that #doOneCycle starts processing
events'n'stuff again, which should clearly not be intended here. However,
#displayWorldSafely forces an immediate redraw. That's why I think that it
could be sufficient to just call #changed or "fullBounds ifNil:[self
invalidRect: self outerBounds] ifNotNil:[self invalidRect: fullBounds]" to
avoid conflicts with other change-update stuff in the list morph. :)

Good luck for localizing the missing version! :)

Best,
Marcel



--
View this message in context: http://forum.world.st/The-Trunk-Morphic-nice-683-mcz-tp4709903p4710055.html
Sent from the Squeak - Dev mailing list archive at Nabble.com.

Hi Marcel,
that looks better to my eyes indeed.

IMHO, Morph>>changed should be renamed redraw, changed should just call redraw, and changeModelSelection: too.

Should I submit that?

2013/9/24 Marcel Taeumel <[hidden email]>

Hi Nicolas,

wouldn't be "self changed" enough to trigger a redraw of the
PluggableListMorph?

This fix is correct in the sense that #doOneCycle starts processing
events'n'stuff again, which should clearly not be intended here. However,
#displayWorldSafely forces an immediate redraw. That's why I think that it
could be sufficient to just call #changed or "fullBounds ifNil:[self
invalidRect: self outerBounds] ifNotNil:[self invalidRect: fullBounds]" to
avoid conflicts with other change-update stuff in the list morph. :)

Good luck for localizing the missing version! :)

Best,
Marcel



--
View this message in context: http://forum.world.st/The-Trunk-Morphic-nice-683-mcz-tp4709903p4710055.html
Sent from the Squeak - Dev mailing list archive at Nabble.com.

Or maybe invalidate instead of redraw...

2013/9/24 Nicolas Cellier <[hidden email]>

Hi Marcel,
that looks better to my eyes indeed.

IMHO, Morph>>changed should be renamed redraw, changed should just call redraw, and changeModelSelection: too.

Should I submit that?

2013/9/24 Marcel Taeumel <[hidden email]>

Hi Nicolas,

wouldn't be "self changed" enough to trigger a redraw of the
PluggableListMorph?

This fix is correct in the sense that #doOneCycle starts processing
events'n'stuff again, which should clearly not be intended here. However,
#displayWorldSafely forces an immediate redraw. That's why I think that it
could be sufficient to just call #changed or "fullBounds ifNil:[self
invalidRect: self outerBounds] ifNotNil:[self invalidRect: fullBounds]" to
avoid conflicts with other change-update stuff in the list morph. :)

Good luck for localizing the missing version! :)

Best,
Marcel



--
View this message in context: http://forum.world.st/The-Trunk-Morphic-nice-683-mcz-tp4709903p4710055.html
Sent from the Squeak - Dev mailing list archive at Nabble.com.

Though, I don't see why this would be necessary, just calling changed seems OK.

2013/9/24 Nicolas Cellier <[hidden email]>

Or maybe invalidate instead of redraw...

2013/9/24 Nicolas Cellier <[hidden email]>

Hi Marcel,
that looks better to my eyes indeed.

IMHO, Morph>>changed should be renamed redraw, changed should just call redraw, and changeModelSelection: too.

Should I submit that?

2013/9/24 Marcel Taeumel <[hidden email]>

Hi Nicolas,

wouldn't be "self changed" enough to trigger a redraw of the
PluggableListMorph?

This fix is correct in the sense that #doOneCycle starts processing
events'n'stuff again, which should clearly not be intended here. However,
#displayWorldSafely forces an immediate redraw. That's why I think that it
could be sufficient to just call #changed or "fullBounds ifNil:[self
invalidRect: self outerBounds] ifNotNil:[self invalidRect: fullBounds]" to
avoid conflicts with other change-update stuff in the list morph. :)

Good luck for localizing the missing version! :)

Best,
Marcel



--
View this message in context: http://forum.world.st/The-Trunk-Morphic-nice-683-mcz-tp4709903p4710055.html
Sent from the Squeak - Dev mailing list archive at Nabble.com.

Done.

squeaksource@box2:~/ss/trunk$ ll Morphic-*683*

-rw-r--r--  1 squeaksource squeaksource   86474 Sep 23 23:56 Morphic-nice.683(cmm.682).2013-09-23.23-56-22

-rw-r--r--  1 squeaksource squeaksource   83388 Sep 23 23:56 Morphic-nice.683(cmm.682).mcd

-rw-r--r--  2 squeaksource squeaksource 1651876 Sep 23 23:55 Morphic-nice.683.2013-09-23.23-54-59

-rw-r--r--  1 squeaksource squeaksource 1651857 Sep 23 23:55 Morphic-nice.683.mcz

squeaksource@box2:~/ss/trunk$ mv Morphic-nice.683.mcz Morphic-nice.683.mcz.bak  

squeaksource@box2:~/ss/trunk$ mv Morphic-nice.683.2013-09-23.23-54-59 Morphic-nice.683.mcz

squeaksource@box2:~/ss/trunk$ mv 'Morphic-nice.683(cmm.682).mcd' 'Morphic-nice.683(cmm.682).mcd.bak'

squeaksource@box2:~/ss/trunk$ mv 'Morphic-nice.683(cmm.682).2013-09-23.23-56-22' 'Morphic-nice.683(cmm.682).mcd'

squeaksource@box2:~/ss/trunk$ ll Morphic-*683*

-rw-r--r--  1 squeaksource squeaksource   86474 Sep 23 23:56 Morphic-nice.683(cmm.682).mcd

-rw-r--r--  1 squeaksource squeaksource   83388 Sep 23 23:56 Morphic-nice.683(cmm.682).mcd.bak

-rw-r--r--  2 squeaksource squeaksource 1651876 Sep 23 23:55 Morphic-nice.683.mcz

-rw-r--r--  1 squeaksource squeaksource 1651857 Sep 23 23:55 Morphic-nice.683.mcz.bak

I think that should suffice ...

- Bert -

On 2013-09-24, at 19:37, Nicolas Cellier <[hidden email]> wrote:

So please Bert, could you:
rm  trunk/Morphic-nice.683.mcz;
mv  trunk/Morphic-nice.683.2013-09-23.23-54-59  trunk/Morphic-nice.683.mcz

2013/9/24 Bert Freudenberg <[hidden email]>

Squeaksource has no UI for retrieving previous versions, but it does preserve them:

squeaksource@box2:~/ss$ ls -l trunk/Morphic-nice.683.*

-rw-r--r--  1 squeaksource squeaksource 1651876 Sep 23 23:55 trunk/Morphic-nice.683.2013-09-23.23-54-59

-rw-r--r--  1 squeaksource squeaksource 1651857 Sep 23 23:55 trunk/Morphic-nice.683.mcz

- Bert -

On 2013-09-24, at 14:33, Nicolas Cellier <[hidden email]> wrote:

Absolutely no trace of original file, neither on my disks, nor in my Time Machine...

I wonder how I could ever publish in these conditions...

Name: Morphic-nice.683
Author: nice
Time: 21 September 2013, 11:19:28.071 pm
UUID: 881a99fa-1788-410b-b676-

c648e1b662f2
Ancestors: Morphic-cmm.682

2013/9/24 Nicolas Cellier <[hidden email]>

I have republished Morphic-nice.683 from 20th -> Morphic-nice.684.

So please, tim, restart from that version if you want to publish this evening.

I'll then upload Morphic-nice.683 (21st) later and we'll do the merge later.

2013/9/24 Nicolas Cellier <[hidden email]>

I produced two mcz with exact same name on two different machines.

The first one (20 september) went in inbox.

The second one (21 september) went in trunk.

Today (24 september) I moved the inbox version into the trunk overriding the one of 21st...

I checked update-nice.246.mcm, and it points to Morphic-cmm.682 fortunately... Otherwise, this could break the update stream (because mcm retain the UUID). So one problem less.

However, what will happen on every package-cache with conflicting version is something I can't predict, and that I'd rather would not want to learn, so I'll try to republish the original trunk version.

2013/9/24 tim Rowledge <[hidden email]>

On 23-09-2013, at 5:30 PM, Nicolas Cellier <[hidden email]> wrote:

> I think I have it on another machine, I'll try later in the morning...
>
>
> 2013/9/24 Levente Uzonyi <[hidden email]>
> On Tue, 24 Sep 2013, Nicolas Cellier wrote:
>
> Ah my bad, I have overriden a package with same name !!! (This was from source.squeak.org web interface, and I got no warning...)

> How can I undo? Not sure whether I have a copy of the original or not…

What has actually happened here? I see a nice orderly progression of Morphic 678 (currently in my image) to 679, 680, …683 with no nasty looking changes at any step. In fact there isa single change from cmm.682 to your nice.683 and that is swapping World displayWorldSafely for World doOneCycle.

I'd quite like to be able to save my small changes to Morphic, so hopefully there's nothing weird going on here...


tim
--
tim Rowledge; [hidden email]; http://www.rowledge.org/tim
How many of you believe in telekinesis? Raise my hands....