July 2012 - Pharo-dev - lists.pharo.org

[Pharo-project] a little pearl from SystemProgressMorph

by Tudor Girba

Just for fun read the below piece of code :) SystemProgressMorph>>label: shortDescription min: startMinValue max: startMaxValue "Answer the block that updates the progress bar." "some fun stuff added (by kph) - bar value: #label. - tell me my current label. - bar value: 'newLabel'. - enable changing the label from within the workBlock - bar value: #increment. - enable progress by one without keeping a counter - bar value: #decrement. - go backwards (if you really have to, useful for an abort, or rollback)! - bar value: newBigNum. - change the max on the fly - when you find there is more to do. - bar value: (bar value: #setMax) + 20 - change the max on the fly - when you find there is more/less to do. - bar value: (bar value: #setMin) - 20 - change the min on the fly - not sure why you would want to. " | slot range barSize lastRefresh maxValue minValue bar| maxValue := startMaxValue. minValue := startMinValue. ((range := maxValue - minValue) <= 0 or: [(slot := self nextSlotFor: shortDescription) = 0]) ifTrue: [^[:barVal| 0 ]]. self recenter. self openInWorld. barSize := -1. "Enforces a inital draw of the morph" lastRefresh := Time millisecondClockValue. bar := bars at: slot. ^[:barValArg | | barVal return newBarSize | barVal := barValArg. return := nil. bar := bars at: slot. "new fun stuff here" barVal == #current ifTrue: [ return := barSize ]. barVal == #label ifTrue:[ return := (labels at: slot) contents ]. barVal == #setMax ifTrue: [ return := maxValue. maxValue := minValue ]. barVal == #setMin ifTrue: [ return := minValue. minValue := maxValue ]. barVal == #increment ifTrue: [return := barVal := barSize + 1]. barVal == #decrement ifTrue: [ return := barVal := barSize - 1]. (barVal isString and: [barVal isSymbol not]) ifTrue: [ (labels at: slot) contents: barVal. barVal := #refresh]. barVal == #refresh ifTrue: [self currentWorld displayWorld. return := true]. (barVal == SmallInteger maxVal or: [ barVal == #finished ]) ifTrue: [return := slot]. return ifNil: [ barVal > maxValue ifTrue: [return := maxValue := barVal]. barVal < minValue ifTrue: [return := minValue := barVal]. (barVal between: minValue and: maxValue) ifTrue: [newBarSize := (barVal - minValue / range * BarWidth) truncated. newBarSize = barSize ifFalse: [barSize := newBarSize. (Time millisecondsSince: lastRefresh) > 25 ifTrue: [barVal := #refresh ]]]. barVal == #refresh ifTrue: [ ((bar valueOfProperty: #useStripes) ifNil: [false]) ifTrue: [bar fillStyle origin: bar position - ((Time millisecondClockValue // 50 \\ 30) @ 0)]. bar barSize: barSize. self currentWorld displayWorld. lastRefresh := Time millisecondClockValue]]. return] Cheers, Doru -- www.tudorgirba.com "Every thing has its own flow"

July 25, 2012

Re: [Pharo-project] How read the Pharo release version

by Dario Trussardi

Thanks Camillo, i forgot : how i can read the file name of the image ? The: Pharo-1.4-xxxxxx.image Ciao, grazie Dario > SystemVersion current highestUpdate > > seems to work in Pharo (but I have no clue what will happen in 1.0 as > it is already quite old and unmaintained) > > On 2012-07-25, at 16:09, Dario Trussardi wrote: > >> Ciao, >> >> i'm interested to understand the Pharo version of the image. >> >> The System menu About report : >> >> Pharo-1.0 Latest update: #10517 >> >> >> How i can read it with method ? >> >> >> Thank, >> >> Dario >> > >

July 25, 2012

Re: [Pharo-project] How read the Pharo release version

by Camillo Bruni

SystemVersion current highestUpdate seems to work in Pharo (but I have no clue what will happen in 1.0 as it is already quite old and unmaintained) On 2012-07-25, at 16:09, Dario Trussardi wrote: > Ciao, > > i'm interested to understand the Pharo version of the image. > > The System menu About report : > > Pharo-1.0 Latest update: #10517 > > > How i can read it with method ? > > > Thank, > > Dario >

July 25, 2012

[Pharo-project] How read the Pharo release version

by Dario Trussardi

Ciao, i'm interested to understand the Pharo version of the image. The System menu About report : Pharo-1.0 Latest update: #10517 How i can read it with method ? Thank, Dario

July 25, 2012

Re: [Pharo-project] How to filter out method extensions in Ring?

by Veronica Isabel Uquillas Gomez

Hi Hernan, Sorry, I have been totally disconnected because of my thesis. Ok I quickly reply... You need to instantiate methods with #realClass:selector: to manipulate objects from the image.. otherwise, they are static (within a ring source model). The category is not assigned then is not possible to know if they are from a class extension. ArrayedCollection selectors asSortedCollection collect: [ : each | RGMethodDefinition realClass: ArrayedCollection selector: each ] then you can use: aRGMethod isExtension Cheers, Veronica On 23 Jul 2012, at 13:45, HernÃ¡n Morales Durand wrote: > Hello all, > You may help me on this one. I want to filter out method selectors which are extensions of a particular class. For example, in Pharo 1.4 the ArrayedCollection class has these selectors: > ArrayedCollection selectors asSortedCollection > > I would want all of them except #writeOnGZIPByteStream: which is in "Compression" package. I did a try to Ring > > ArrayedCollection selectors asSortedCollection collect: [ : each | > RGMethodDefinition > class: ArrayedCollection > selector: each ] > > with MethodReference I did : > > ... ( each category beginsWith: '*' ) not ... > > but it seems not supported in Ring? Any hints? > > Cheers, > > HernÃ¡n >

July 25, 2012

Re: [Pharo-project] Way to see diffs between different builds?

by Patrik Sundberg

On Wed, Jul 25, 2012 at 1:25 PM, Camillo Bruni <camillobruni(a)gmail.com>wrote: > > On 2012-07-25, at 14:10, Patrik Sundberg wrote: > > > Hi, > > > > Is there a way to see the diffs (in terms of changes to image) between > one > > 2.0 build number and another? I don't see any obvious link on the CI > server > > to it, and the github export seems to be done in a way that makes the > diffs > > enourmous. > > you can use the git repos and filter out non-st methods: > git diff $VERSION_A $VERSION_B -- $(git diff --name-only $VERSION_A > $VERSION_B | grep .st) > > Ok, I'll try that. I haven't looked at it more closely, but how come the diffs get so HUGE? is there any easy wins to get more "proper" diffs? > If not, is there a way to do it in image? I can only see changes for > > individual MC packages, not a more system level diff to see exactly what > > changed from one build to another. > > We should have a global Pharo MC Package which has all other packages as > dependencies, that would simplify quite a few things here... > > Otherwise you're stuck with downloading the image for a specific version > which might be corrupted, and thus leaves you without any chance of > inspecting the changes... > > Right, a fully packetized system with meta-packages would seem ideal.

July 25, 2012

Re: [Pharo-project] Way to see diffs between different builds?

by Camillo Bruni

On 2012-07-25, at 14:10, Patrik Sundberg wrote: > Hi, > > Is there a way to see the diffs (in terms of changes to image) between one > 2.0 build number and another? I don't see any obvious link on the CI server > to it, and the github export seems to be done in a way that makes the diffs > enourmous. you can use the git repos and filter out non-st methods: git diff $VERSION_A $VERSION_B -- $(git diff --name-only $VERSION_A $VERSION_B | grep .st) Git even has a utility to find the version that introduced a certain bug :) http://incitecode.com/2009/10/13/finding-bugs-with-git-bisect/ > If not, is there a way to do it in image? I can only see changes for > individual MC packages, not a more system level diff to see exactly what > changed from one build to another. We should have a global Pharo MC Package which has all other packages as dependencies, that would simplify quite a few things here... Otherwise you're stuck with downloading the image for a specific version which might be corrupted, and thus leaves you without any chance of inspecting the changes... > I was starting to dig into the metacello bootstrapping problem in 2.0 and > became clear to me it'd help me a lot if I could more easily follow the > diffs. > > Thanks, > Patrik

July 25, 2012

Re: [Pharo-project] ContextPart >> copyTo: -> copyUpTo: ?

by Frank Shearar

On 25 July 2012 13:19, Mariano Martinez Peck <marianopeck(a)gmail.com> wrote: > > > On Wed, Jul 25, 2012 at 1:30 PM, Frank Shearar <frank.shearar(a)gmail.com> > wrote: >> >> On 25 July 2012 11:42, Mariano Martinez Peck <marianopeck(a)gmail.com> >> wrote: >> > Guys. Did I miss something or ContextPart>>copyTo: should be renamed to >> > #copyUpTo: ? >> >> You feel that #copyTo: implies an inclusive bound? >> >> There's only one >> #copyTo: implementor (ContextPart) and only one #copyUpTo: >> (SequenceableCollection) and both are "up to and including" copies. In >> other words there isn't a clear precedence one way or the other for >> having #copyTo: meaning either "and including" or "not including". >> Your argument is, I guess, that there should be? (The method comments >> are perfectly clear on what both method do, at least.) >> > > With #copyTo: I feel that I am copying from one place (source) to a target, > when what it actually does (if I understood correctly) is to copy the source > (receiver) up to the parameter (stop there). So #copyTo: is totally > misleading. By that argument the correct name would be #copyFromSelfTo: or #copyFromHereTo:, neither of which look appetising. (The names would semi-parallel SequenceableCollection's #copyFrom:to:, only a ContextPart is already an element in a collection.) frank >> frank >> >> > Thanks, >> > >> > -- >> > Mariano >> > http://marianopeck.wordpress.com >> > >> > > > > -- > Mariano > http://marianopeck.wordpress.com >

July 25, 2012

Re: [Pharo-project] ContextPart >> copyTo: -> copyUpTo: ?

by Mariano Martinez Peck

On Wed, Jul 25, 2012 at 1:30 PM, Frank Shearar <frank.shearar(a)gmail.com>wrote: > On 25 July 2012 11:42, Mariano Martinez Peck <marianopeck(a)gmail.com> > wrote: > > Guys. Did I miss something or ContextPart>>copyTo: should be renamed to > > #copyUpTo: ? > > You feel that #copyTo: implies an inclusive bound? There's only one > #copyTo: implementor (ContextPart) and only one #copyUpTo: > (SequenceableCollection) and both are "up to and including" copies. In > other words there isn't a clear precedence one way or the other for > having #copyTo: meaning either "and including" or "not including". > Your argument is, I guess, that there should be? (The method comments > are perfectly clear on what both method do, at least.) > > With #copyTo: I feel that I am copying from one place (source) to a target, when what it actually does (if I understood correctly) is to copy the source (receiver) up to the parameter (stop there). So #copyTo: is totally misleading. > frank > > > Thanks, > > > > -- > > Mariano > > http://marianopeck.wordpress.com > > > > -- Mariano http://marianopeck.wordpress.com

July 25, 2012

[Pharo-project] Way to see diffs between different builds?

by Patrik Sundberg

Hi, Is there a way to see the diffs (in terms of changes to image) between one 2.0 build number and another? I don't see any obvious link on the CI server to it, and the github export seems to be done in a way that makes the diffs enourmous. If not, is there a way to do it in image? I can only see changes for individual MC packages, not a more system level diff to see exactly what changed from one build to another. I was starting to dig into the metacello bootstrapping problem in 2.0 and became clear to me it'd help me a lot if I could more easily follow the diffs. Thanks, Patrik

July 25, 2012