May 2012 - Pharo-users - lists.pharo.org

Re: [Pharo-users] how to test forked morph rendering

by Hilaire Fernandes

Hi, Your problem is Morph operation taking place only in the UI thread, therefore blocking the user, is it that? Hilaire On 30/05/2012 07:43, Tudor Girba wrote: > Hi, > > I have a piece of code that builds a morph in a parallel process, and I would like to test that it builds the right thing. > > The building of the morph can take long. My current solution is to add a delay, but this is ugly and unreliable: > > "setup and trigger the morph building" > (Delay forMilliseconds: 1000) wait. > "assert that the morph is correct" > > The question is if there is a better way. > > Cheers, > Doru > > > -- > www.tudorgirba.com > > "We are all great at making mistakes." > > > > > > > > > -- Dr. Geo -- http://www.drgeo.eu

May 30, 2012

Freetype tips with loading fonts

by Hilaire Fernandes

Hello In case, it is useful to someone. It is how I clean up the known fonts in the image and load an external one located along the image. FreeTypeFontProvider unload. FreeTypeFontProvider current updateFromFileEntry: (FileDirectory default entryAt: 'UnDotum.ttf') directory: FileDirectory default locationType: #imageRelative. FreeTypeFontProvider current buildFamilies. FreeTypeFontSelectorDialogWindow new open. "to see the loaded font" StandardFonts balloonFont: (LogicalFont familyName: 'UnDotum' pointSize: 10). StandardFonts defaultFont: (LogicalFont familyName: 'UnDotum' pointSize: 11). Hilaire -- Dr. Geo -- http://www.drgeo.eu

May 30, 2012

how to test forked morph rendering

by Tudor Girba

Hi, I have a piece of code that builds a morph in a parallel process, and I would like to test that it builds the right thing. The building of the morph can take long. My current solution is to add a delay, but this is ugly and unreliable: "setup and trigger the morph building" (Delay forMilliseconds: 1000) wait. "assert that the morph is correct" The question is if there is a better way. Cheers, Doru -- www.tudorgirba.com "We are all great at making mistakes."

May 30, 2012

Re: [Pharo-users] How does the UI update during this block?

by Sean P. DeNigris

Benjamin Van Ryseghem-2 wrote > > Probably because the fork create a new process at priority 40, which is > also the priority of the ui process. > So the processor dispatch between them I guess :) > I see where Morphic may give an opportunity to switch back to my block in #interCyclePause:, but when does the processor get a chance to switch out of my block? I tried debugging, but couldn't figure out a way to hook in before Morphic got control back (e.g. I put a haltOnce in Process>>resume, which was never called apparently; and putting a halt in the block itself just loops through the block repeatedly). Here's the code accessing a global variable for logging: m := MyMorph new openInWorld. "MyMorph>>#step does - Messages add: 'step'." Messages := OrderedCollection new. trues := 0. [ (1 to: 250000000) do: [ :e | "(Delay forMilliseconds: 100) wait." e \\ 5000000 == 0 ifTrue: [ Messages add: 'block - true'. trues := trues + 1 ] ]. m delete. WorldState addDeferredUIMessage: [Messages explore] ] forkAt: 30. p.s. If I fork at 30, there is approximately a 1:1 interleaving of the world stepping and the ifTrue: part being executed. If I fork at 40 (UI priority), there is (very roughly) a 5:1 between the ifTrue: and world stepping, but it's very inconsistent. I did not protect access to the global Messages, because I was worried the lock would change the process scheduling. -- View this message in context: http://forum.world.st/How-does-the-UI-update-during-this-block-tp4632333p46… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

May 29, 2012

[FUN] Pharo Zen Workspace

by Sven Van Caekenberghe

Inspired by a Twitter conversation between Esteban Maringolo & Esteban Lorenzano, I quickly made this little snippet. This is a cool list, it should be featured more prominently: | zenValues | zenValues := #( 'Easy to understand, easy to learn from, easy to change.' 'Objects all the way down.' 'Examples to learn from.' 'Fully dynamic and malleable.' 'Beauty in the code, beauty in the comments.' 'Simplicity is the ultimate elegance.' 'Better a set of small polymorphic classes than a large ugly one.' 'Classes structure our vocabulary.' 'Messages are our vocabulary.' 'Polymorphism is our esperanto.' 'Abstraction and composition are our friends.' 'Tests are important but can be changed.' 'Explicit is better than implicit.' 'Magic only at the right place.' 'One step at a time.' 'There is no unimportant fix.' 'Learning from mistakes.' 'Perfection can kill movement.' 'Quality is an emerging property.' 'Simple processes to support progress.' 'Communication is key.' 'A system with robust abstractions that a single person can understand.' ). Smalltalk tools openWorkspace label: 'Pharo Zen'; toggleStylingActive; contents: (String streamContents: [ :stream | stream << 'Our values and convictions are condensed in this simple list:'; cr; cr. zenValues do: [ :each | stream << each ] separatedBy: [ stream cr ] ]). Sven -- Sven Van Caekenberghe http://stfx.eu Smalltalk is the Red Pill

May 29, 2012

Re: [Pharo-users] Code Completion Table with dead entries

by StÃ©phane Ducasse

> >> that's because it does not has anything to do with the code completion, but with the fact that symbols are still there (they are not removed when you remove the method). OCompletion just shows you a bunch of *symbols* you used lately (sorted with some -not much- intelligence), not methods or classes or anything... > Understand. So it shows the expected behaviour, even if it not nice one. > >> Of course that can be fixed (and should be, eventually)... but who can take it? The "in house pharo team" now is really busy stabilizing 2.0 to take another feature action right now. > I have no expectations here that this should fixed/enhanced with the upcoming 2.0. In my eyes it > was not clear, if it was a bug. Anyway i can open a feature request. ;-) > >> But I think I don't need to remember you that the "full pharo team" are not just us, it includes all of you too. Any help (implementing cool stuff, fixing bugs or just reporting them) is always welcome (and talking by myself, I welcome much more the first, "implementing cool stuff" ;) > I just started with Pharo, so give me a few weeks (to implement cool stuff ;-)) excellent! > > Appreciate your work. tx > > >> best, >> Esteban >> >> On May 29, 2012, at 5:07 PM, Volkert Barr wrote: >> >>> >>>>> Hello, >>>>> >>>>> i have a question. Why does the code completion offers methods that were removed? >>>> >>>> probably because it works at the symbol level and not method dictionary. >>>> May be removing a method could unintern its selector from the symbol cache. >>>> >>>>> I have some >>>>> cases where deleted methods still listed as "symbol (no implementors)". >>>> >>>> where? >>>> In nautilus? >>> In workspace, nautilus, ... the removed methods still listed in the code completion popup. >>> >>> I removed the messages with nautilus and also with the default system browser. Same effect. >>> >>> if this is an bug, i can open a ticket. >>> >>> Volkert >>> >>> >>> >> >> > > -- > www.nivoba.de > www.software-wartung.net > www.erlang-dach.org > > PGP-Key on request / PGP-SchlÃ¼ssel auf Anfrage > >

May 29, 2012

Re: [Pharo-users] How does the UI update during this block?

by Benjamin

Probably because the fork create a new process at priority 40, which is also the priority of the ui process. So the processor dispatch between them I guess :) Ben On May 29, 2012, at 2:04 PM, Sean P. DeNigris wrote: > trues := 0. > [ > (1 to: 200000000) do: [ :e | > " (Delay forMilliseconds: 100) wait." > e \\ 1000000 == 0 ifTrue: [ trues := trues + 1 ] ]. > m delete ] fork. > > I thought I would have to uncomment the #wait to give the UI a chance to > update, but it works as is... > > -- > View this message in context: http://forum.world.st/How-does-the-UI-update-during-this-block-tp4632333.ht… > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. >

May 29, 2012

Re: [Pharo-users] [Pharo-project] GemSource downtime - May 29 2pm PDT

by Dale Henrichs

REMINDER of GemSource downtime scheduled for 2pm PDT today! ----- Original Message ----- | From: "Dale Henrichs" <dhenrich(a)vmware.com> | To: Pharo-project(a)lists.gforge.inria.fr | Sent: Tuesday, May 22, 2012 10:08:42 AM | Subject: [Pharo-project] GemSource downtime - May 29 2pm PDT | | GemSource (http://seaside.gemstone.com) will be down from 2-3pm PST | on May 29 to add additional storage. | | Please let me know if that will be a problem. | | Dale | |

May 29, 2012

How does the UI update during this block?

by Sean P. DeNigris

trues := 0. [ (1 to: 200000000) do: [ :e | " (Delay forMilliseconds: 100) wait." e \\ 1000000 == 0 ifTrue: [ trues := trues + 1 ] ]. m delete ] fork. I thought I would have to uncomment the #wait to give the UI a chance to update, but it works as is... -- View this message in context: http://forum.world.st/How-does-the-UI-update-during-this-block-tp4632333.ht… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

May 29, 2012

Re: [Pharo-users] Code Completion Table with dead entries

by Sven Van Caekenberghe

On 29 May 2012, at 18:05, Volkert Barr wrote: > No effect. But this is more complicated: OK, the method was removed, code completion shows the method based on the fact that the symbol exists. Now, compacting the table could remove the symbol, or not, maybe it is still used somewhere else (recent changes, changesets, monticello, â¦). IMHO, this feels like interfering with GC. Completions is a hard problem: it would be dead slow if it was 100% accurate. Sven

May 29, 2012