October 2015 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] FastTable question

by Cyril Ferlicot

We will at least keep it for Synectique in order to have a widget with an action button for now. I don't think you want an action button to be integrated inside the FTTableMorph ? When we will integrate the next configuration of FT into Pharo I will remove it from the FastTable package. On Thu, Oct 22, 2015 at 8:55 AM, Esteban Lorenzano <estebanlm(a)gmail.com> wrote: > > > I thought we were removing FTFastTableComposer, as is not such a great idea. > > Esteban > > -- Cheers Cyril Ferlicot

Oct. 22, 2015

Re: [Pharo-dev] FastTable question

by stepharo

Excellent How did you do it? Le 20/10/15 18:47, Ferlicot D. Cyril a Ã©crit : > Le 20/10/2015 18:42, stepharo a Ã©crit : >> for the mooc videos I must use larger fonts. >> So I ended up hacking and changing the hardcoded number in default >> >> defaultRowHeight >> ^ 24 >> >> rowHeight >> "This is the row height your rows will have. Cells answered in >> dataSource will be forced to have >> this height number... We force it instead allowing lists to have >> any height because the logic to >> calculate rows becomes complicated. Possible, but complicated :)" >> >> ^ rowHeight ifNil: [ self class defaultRowHeight ] >> >> >> Esteban two questions: >> >> - when I read the code of FTmorph I was thinking that a lot of it has >> not much to do with UI but looks like a model :). >> >> - if I were about to manage the height of a row based on its actual >> contents where should I put it. >> FTTableContainerMorph is containing what is visible but I wonder why it >> was asking this to the FastTable and to the elements displayed? >> >> Why the ContainerMorph does not ask the FTTableRowMorph for its height >> and it could be computed based on the font. >> I still did not get where the strings built. >> >> >> may be here >> >> updateExposedRows >> >> ... >> >> | cell | >> cell := (self owner dataSource >> cellColumn: (columns at: columnIndex) >> row: rowIndex). >> cell width: (columnWidths at: columnIndex). >> >> We should add more comments to methods. >> >> >> Stef >> >> >> >> drawOn: canvas >> | x y cellWidth cellHeight rowsToDisplay rowSubviews >> highligtedRowIndexes primarySelectionIndex | >> >> self bounds ifNil: [ ^ self ]. "Nothing to show yet" >> self owner ifNil: [ ^ self ]. >> >> x := self left + self class rowLeftMargin. >> y := self top. >> cellWidth := self width - self class rowLeftMargin. >> cellHeight := self owner rowHeight. >> ^^^^^^^^^^^^^^^^^^^^^^^ >> >> highligtedRowIndexes := >> self owner selectedRowIndexes, >> self owner highlightedRowIndexes. >> primarySelectionIndex := self owner selectedRowIndex. >> >> "For some superweird reason, calling #calculateExposedRows here >> instead in #changed (where >> it should be called) is 10x faster. Since the whole purpose of this >> component is speed, for >> now I'm calling it here and adding the #setNeedRecalculateRows >> mechanism. >> History, please forgive me." >> self updateAllRows. >> >> rowsToDisplay := self exposedRows. >> rowSubviews := OrderedCollection new: rowsToDisplay size + 1. >> headerRow ifNotNil: [ >> headerRow bounds: (x@y extent: cellWidth@cellHeight). >> y := y + cellHeight + self owner intercellSpacing. >> rowSubviews add: headerRow ]. >> >> rowsToDisplay keysAndValuesDo: [ :rowIndex :row | | visibleHeight | >> visibleHeight := cellHeight min: (self bottom - y). >> row bounds: (x@y extent: cellWidth@visibleHeight). >> y := y + visibleHeight + self owner intercellSpacing. >> rowSubviews add: ((highligtedRowIndexes includes: rowIndex) >> ifTrue: [ >> "IMPORTANT: I need to set owner to nil because otherwise >> it will trigger an >> invalidation of the owner when adding morph to >> selectionMorph, causing an >> infinite loop" >> self >> toSelectionRow: (row privateOwner: nil) >> primary: primarySelectionIndex = rowIndex ] >> ifFalse: [ row ]) ]. >> >> submorphs := rowSubviews asArray. >> super drawOn: canvas. >> needsRefreshExposedRows := false >> > If you load the development version of FT it will be fine normally :) > Maybe not perfect for now but usable with large fonts. (You have to > close and reopen the existing FT if you change font size because the > change is hard coded and not announce). > > I adapted also the FTFastTableComposer so that it manage the font size. >

Oct. 22, 2015

Re: [Pharo-dev] FastTable question

by Esteban Lorenzano

> On 20 Oct 2015, at 18:47, Ferlicot D. Cyril <cyril.ferlicot(a)gmail.com> wrote: > > Le 20/10/2015 18:42, stepharo a Ã©crit : >> for the mooc videos I must use larger fonts. >> So I ended up hacking and changing the hardcoded number in default >> >> defaultRowHeight >> ^ 24 >> >> rowHeight >> "This is the row height your rows will have. Cells answered in >> dataSource will be forced to have >> this height number... We force it instead allowing lists to have >> any height because the logic to >> calculate rows becomes complicated. Possible, but complicated :)" >> >> ^ rowHeight ifNil: [ self class defaultRowHeight ] >> >> >> Esteban two questions: >> >> - when I read the code of FTmorph I was thinking that a lot of it has >> not much to do with UI but looks like a model :). >> >> - if I were about to manage the height of a row based on its actual >> contents where should I put it. >> FTTableContainerMorph is containing what is visible but I wonder why it >> was asking this to the FastTable and to the elements displayed? >> >> Why the ContainerMorph does not ask the FTTableRowMorph for its height >> and it could be computed based on the font. >> I still did not get where the strings built. >> >> >> may be here >> >> updateExposedRows >> >> ... >> >> | cell | >> cell := (self owner dataSource >> cellColumn: (columns at: columnIndex) >> row: rowIndex). >> cell width: (columnWidths at: columnIndex). >> >> We should add more comments to methods. >> >> >> Stef >> >> >> >> drawOn: canvas >> | x y cellWidth cellHeight rowsToDisplay rowSubviews >> highligtedRowIndexes primarySelectionIndex | >> >> self bounds ifNil: [ ^ self ]. "Nothing to show yet" >> self owner ifNil: [ ^ self ]. >> >> x := self left + self class rowLeftMargin. >> y := self top. >> cellWidth := self width - self class rowLeftMargin. >> cellHeight := self owner rowHeight. >> ^^^^^^^^^^^^^^^^^^^^^^^ >> >> highligtedRowIndexes := >> self owner selectedRowIndexes, >> self owner highlightedRowIndexes. >> primarySelectionIndex := self owner selectedRowIndex. >> >> "For some superweird reason, calling #calculateExposedRows here >> instead in #changed (where >> it should be called) is 10x faster. Since the whole purpose of this >> component is speed, for >> now I'm calling it here and adding the #setNeedRecalculateRows >> mechanism. >> History, please forgive me." >> self updateAllRows. >> >> rowsToDisplay := self exposedRows. >> rowSubviews := OrderedCollection new: rowsToDisplay size + 1. >> headerRow ifNotNil: [ >> headerRow bounds: (x@y extent: cellWidth@cellHeight). >> y := y + cellHeight + self owner intercellSpacing. >> rowSubviews add: headerRow ]. >> >> rowsToDisplay keysAndValuesDo: [ :rowIndex :row | | visibleHeight | >> visibleHeight := cellHeight min: (self bottom - y). >> row bounds: (x@y extent: cellWidth@visibleHeight). >> y := y + visibleHeight + self owner intercellSpacing. >> rowSubviews add: ((highligtedRowIndexes includes: rowIndex) >> ifTrue: [ >> "IMPORTANT: I need to set owner to nil because otherwise >> it will trigger an >> invalidation of the owner when adding morph to >> selectionMorph, causing an >> infinite loop" >> self >> toSelectionRow: (row privateOwner: nil) >> primary: primarySelectionIndex = rowIndex ] >> ifFalse: [ row ]) ]. >> >> submorphs := rowSubviews asArray. >> super drawOn: canvas. >> needsRefreshExposedRows := false >> > If you load the development version of FT it will be fine normally :) > Maybe not perfect for now but usable with large fonts. (You have to > close and reopen the existing FT if you change font size because the > change is hard coded and not announce). > > I adapted also the FTFastTableComposer so that it manage the font size. I thought we were removing FTFastTableComposer, as is not such a great idea. Esteban > > -- > Cyril Ferlicot > > http://www.synectique.eu > > 165 Avenue Bretagne > Lille 59000 France >

Oct. 22, 2015

Re: [Pharo-dev] FastTable question

by Esteban Lorenzano

Hi, Sorry late response but yesterday I was not around. > On 20 Oct 2015, at 18:42, stepharo <stepharo(a)free.fr> wrote: > > for the mooc videos I must use larger fonts. > So I ended up hacking and changing the hardcoded number in default > > defaultRowHeight > ^ 24 > > rowHeight > "This is the row height your rows will have. Cells answered in dataSource will be forced to have > this height number... We force it instead allowing lists to have any height because the logic to > calculate rows becomes complicated. Possible, but complicated :)" > > ^ rowHeight ifNil: [ self class defaultRowHeight ] > > > Esteban two questions: > > - when I read the code of FTmorph I was thinking that a lot of it has not much to do with UI but looks like a model :). this is not a question :) in the case of rowHeight, it determines how the row looks (and not what it shows) then is UI (IMO)â¦ anyway Iâm aware this is not 100% MVC but knowing the height of rows is very important to calculate in a fast way how many rows fit and get the rows from datasource (an operation that happens a lot so is convenient to have it as fast as possible (at least in a âfast tableâ). this do not supports variable height but that can be achieved by extending the table without compromising the performance of the parent(just very few use-cases need a variable height table). > > - if I were about to manage the height of a row based on its actual contents where should I put it. as described, you need to extend the functionality, this is not supported in current version (reason already explained). For me, to support this you need to extend: - rowHeight should be rowHeightFor: index - visibleRows and around need to take first visible row height, then iterate one by one asking for the each row size. - datasource needs to be extended with ârowHeight: indexâ > FTTableContainerMorph is containing what is visible but I wonder why it was asking this to the FastTable and to the elements displayed? explained above: speed needs. > > Why the ContainerMorph does not ask the FTTableRowMorph for its height and it could be computed based on the font. > I still did not get where the strings built. I didnât thought about, thatâs why is not calculating font size :) Alsoâ¦ I cannot know your table will contain textâ¦ thatâs a task for the datasource, then (to answer appropriate row height)â¦ but it needs to be fast! :) Esteban > > > may be here > > updateExposedRows > > ... > > | cell | > cell := (self owner dataSource > cellColumn: (columns at: columnIndex) > row: rowIndex). > cell width: (columnWidths at: columnIndex). > > We should add more comments to methods. > > > Stef > > > > drawOn: canvas > | x y cellWidth cellHeight rowsToDisplay rowSubviews highligtedRowIndexes primarySelectionIndex | > > self bounds ifNil: [ ^ self ]. "Nothing to show yet" > self owner ifNil: [ ^ self ]. > > x := self left + self class rowLeftMargin. > y := self top. > cellWidth := self width - self class rowLeftMargin. > cellHeight := self owner rowHeight. > ^^^^^^^^^^^^^^^^^^^^^^^ > > highligtedRowIndexes := > self owner selectedRowIndexes, > self owner highlightedRowIndexes. > primarySelectionIndex := self owner selectedRowIndex. > > "For some superweird reason, calling #calculateExposedRows here instead in #changed (where > it should be called) is 10x faster. Since the whole purpose of this component is speed, for > now I'm calling it here and adding the #setNeedRecalculateRows mechanism. > History, please forgive me." > self updateAllRows. > > rowsToDisplay := self exposedRows. > rowSubviews := OrderedCollection new: rowsToDisplay size + 1. > headerRow ifNotNil: [ > headerRow bounds: (x@y extent: cellWidth@cellHeight). > y := y + cellHeight + self owner intercellSpacing. > rowSubviews add: headerRow ]. > > rowsToDisplay keysAndValuesDo: [ :rowIndex :row | | visibleHeight | > visibleHeight := cellHeight min: (self bottom - y). > row bounds: (x@y extent: cellWidth@visibleHeight). > y := y + visibleHeight + self owner intercellSpacing. > rowSubviews add: ((highligtedRowIndexes includes: rowIndex) > ifTrue: [ > "IMPORTANT: I need to set owner to nil because otherwise it will trigger an > invalidation of the owner when adding morph to selectionMorph, causing an > infinite loop" > self > toSelectionRow: (row privateOwner: nil) > primary: primarySelectionIndex = rowIndex ] > ifFalse: [ row ]) ]. > > submorphs := rowSubviews asArray. > super drawOn: canvas. > needsRefreshExposedRows := false >

Oct. 22, 2015

Re: [Pharo-dev] [Pharo-users] [ANN] The next Pharo Sprints at Lille

by Juraj Kubelka

Hi, we are pleased to announce that University of Chile (PLEIAD laboratory) is organizing Pharo Sprint at Oct 23, from 10:00 to 13:00 CLP (15:00-18:00 Lille time). We will join Pharo Sprint at Lille using IRC channel #pharo. If you want to come, we will stay in the new building at Beauchef 851, Poniente, office 320. Everyone is welcome. Object Profile <http://objectprofile.com/> company sponsors the chilean participationâparticipants will enjoy delicious snacks :-) Cheers, Juraj > On Sep 8, 2015, at 12:35, Marcus Denker <marcus.denker(a)inria.fr> wrote: > > > Hello, > > We looked at the calendar and fixed the dates for all public sprints that we will do this > year here locally. These are all Fridays: > > Sept 11 > Oct 23 > Nov 27 > Dec 18 > > Here is the invitation for this Friday: > âââââââââââââââââââââââ > > We will organize a Pharo sprint / Moose dojo Friday, 11th September, starting at > 10:00am. (Local Time Lille). > > It will be at the Inria Lille, Building B, third floor (RMoD offices). > Remotely, you can join us on the official IRC channel #pharo on > irc.freenode.net server. During the sprint, we will try to synchronize > local and remote Pharo sprinters. > > As the building is not open to the public, please contact us before if > you plan to come.

Oct. 21, 2015

Re: [Pharo-dev] Is "Enterprise Pharo" available in PDF?

by Gabriel Cotelli

https://ci.inria.fr/pharo-contribution/view/Books/job/EnterprisePharoBook/l… On Wed, Oct 21, 2015 at 1:06 PM, Richard Sargent < richard.sargent(a)gemtalksystems.com> wrote: > In looking at http://files.pharo.org/books/enterprisepharo/, I see a link > to > get the book in PDF form. ("The full book is available as a free printable > PDF download.") The link takes me to > https://gforge.inria.fr/scm/?group_id=3552, which seems to be the source > code repository. > > Is the PDF version available, even as a running CI artefact? > > > > > -- > View this message in context: > http://forum.world.st/Is-Enterprise-Pharo-available-in-PDF-tp4857121.html > Sent from the Pharo Smalltalk Developers mailing list archive at > Nabble.com. > >

Oct. 21, 2015

Re: [Pharo-dev] ConfigurationOf and Git

by Dale Henrichs

On 10/21/2015 05:55 AM, Ben Coman wrote: > > > On Wed, Oct 21, 2015 at 2:49 AM, Dale Henrichs > <dale.henrichs(a)gemtalksystems.com > <mailto:dale.henrichs@gemtalksystems.com>> wrote: > > Here's the documentation for the Metacello github:// repository > description[1]. ... > > In recent versions of Metacello I have made it possible for you to > use pattern matching in the <version identifier> to provide for > symbolic-version-like facility for github references. > > Instead of stableForPharo4 you would associate a semantic version > with the code that is "stableForPharo4" for example `4.0.0` and > use the tag `v4.0.0` to mark the commit that is "stableForPharo4". > > Similarly you'd use the tag `v5.0.0` to mark the commit that is > "stableForPharo5". > > Then you'd use the following in your baseline: method (not in a > symbolicVersion method). > > spec for: #'pharo4.0.x' > do: [spec > baseline: 'Project' > with: [ spec repository: 'github://username/Project:v4.?/']]. > spec for: #'pharo5.0.x' > do: [spec > baseline: 'Project' > with: [ spec repository: 'github://username/Project:v5.?/']]. > > If you end up with a new commit that patches a Pharo4 problem, > you'd tag that commit as `v4.0.1` and so on .... The above `v4.?` > pattern will match `v4.0.1` and you'll pick up that tag the next > time you refresh your build (i.e., do a `get` on the Project > baseline ... which causes a new download from Github) ... > > > Intuitively I would have thought that #'pharo4.0.x' > would correspond to repository: 'github://username/Project:v4.0.?/' > I seem to be missing something. > I guess it's not intuitive to me, because I'm not sure what point you are making:) In general terms I'm not very intuitive, so this is my personality flaw:) Dale

Oct. 21, 2015

Is "Enterprise Pharo" available in PDF?

by Richard Sargent

In looking at http://files.pharo.org/books/enterprisepharo/, I see a link to get the book in PDF form. ("The full book is available as a free printable PDF download.") The link takes me to https://gforge.inria.fr/scm/?group_id=3552, which seems to be the source code repository. Is the PDF version available, even as a running CI artefact? -- View this message in context: http://forum.world.st/Is-Enterprise-Pharo-available-in-PDF-tp4857121.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

Oct. 21, 2015

Re: [Pharo-dev] ConfigurationOf and Git

by Ben Coman

On Wed, Oct 21, 2015 at 2:49 AM, Dale Henrichs < dale.henrichs(a)gemtalksystems.com> wrote: > Here's the documentation for the Metacello github:// repository > description[1]. ... > > In recent versions of Metacello I have made it possible for you to use > pattern matching in the <version identifier> to provide for > symbolic-version-like facility for github references. > > Instead of stableForPharo4 you would associate a semantic version with the > code that is "stableForPharo4" for example `4.0.0` and use the tag `v4.0.0` > to mark the commit that is "stableForPharo4". > > Similarly you'd use the tag `v5.0.0` to mark the commit that is > "stableForPharo5". > > Then you'd use the following in your baseline: method (not in a > symbolicVersion method). > > spec for: #'pharo4.0.x' > do: [spec > baseline: 'Project' > with: [ spec repository: 'github://username/Project:v4.?/']]. > spec for: #'pharo5.0.x' > do: [spec > baseline: 'Project' > with: [ spec repository: 'github://username/Project:v5.?/']]. > > If you end up with a new commit that patches a Pharo4 problem, you'd tag > that commit as `v4.0.1` and so on .... The above `v4.?` pattern will match > `v4.0.1` and you'll pick up that tag the next time you refresh your build > (i.e., do a `get` on the Project baseline ... which causes a new download > from Github) ... > Intuitively I would have thought that #'pharo4.0.x' would correspond to repository: 'github://username/Project:v4.0.?/' I seem to be missing something. cheers -ben > > HTH, > > Dale > > [1] > https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloScr… > [2] > https://github.com/dalehenrich/metacello-work/issues/277#issuecomment-58970… > > > On 10/19/2015 08:05 PM, Gabriel Cotelli wrote: > > I was wondering if something like this can work: > > Given a ConfigurationOfProject (a subclass of ConfigurationOf) defining: > > stable: spec <symbolicVersion: #stable> spec for: #'pharo4.0.x' do: [spec > baseline: 'Project' with: [ spec repository: 'github://username/Project:stableForPharo4/']]. > spec for: #'pharo5.0.x' do: [spec baseline: 'Project' with: [ spec > repository: 'github://username/Project:stableForPharo5/']]. > where stableForPharo4 and stableForPharo5 are tags in the git repo (and > assuming the BaselineOfProject is defined in this commits). Is this > supposed to work : > Metacello new > configuration: 'Project'; > repository: 'github://username/Project:master'; > version: #stable; > load. ?? > I've tried using a non-symbolic version and it works but I can't make it > work for #stable. I'd like to use this configuration for the Configuration > Browser/Catalog Browser. > Any help is appreciated. > > >

Oct. 21, 2015

[pharo-project/pharo-core]

by GitHub

Branch: refs/tags/50397 Home: https://github.com/pharo-project/pharo-core

Oct. 21, 2015