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

[pharo-project/pharo-core]

by GitHub

Branch: refs/tags/50059 Home: https://github.com/pharo-project/pharo-core

May 22, 2015

[pharo-project/pharo-core] 93ae2a: 50059

by GitHub

Branch: refs/heads/5.0 Home: https://github.com/pharo-project/pharo-core Commit: 93ae2a76e4cae1e04906a07c81771312d9ba39e1 https://github.com/pharo-project/pharo-core/commit/93ae2a76e4cae1e04906a07c… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2015-05-22 (Fri, 22 May 2015) Changed paths: M Collections-Unordered.package/Set.class/instance/enumerating/collect_.st M CollectionsTests.package/SetTest.class/instance/tests - iterating/testCollect.st M NautilusRefactoring.package/NautilusRefactoring.class/class/shortcuts/buildSourceCodeShortcutsOn_.st A OpalCompiler-Core.package/OCKeyedSet.class/instance/copying/collect_.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50058.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50059.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50058.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50059.st M ScriptLoader50.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st Log Message: ----------- 50059 14535 IdentitySet>>collect: answers a Set, not an IdentitySet. https://pharo.fogbugz.com/f/cases/14535 15584 TClassDescription>>#sourceCodeRefactoringMenu: and NautilusRefactoring>>#buildSourceCodeShortcutsOn: need review https://pharo.fogbugz.com/f/cases/15584 http://files.pharo.org/image/50/50059.zip

May 22, 2015

Re: [Pharo-dev] [Reflectivity] Progress for Breakpoints

by Marcus Denker

>> >> https://pharo.fogbugz.com/f/cases/15516/add-suggestion-for-breakpoint <https://pharo.fogbugz.com/f/cases/15516/add-suggestion-for-breakpoint> >> >> this is of course completely untestedâ¦ it works only when there is no unaccepted >> text in the editor and maybe only within Nautilus. >> >> And of course existing breakpoints are not yet in any form visualised in the editor. > > This is in 50041. Please use with careâ¦ still needs work to be really usable. > I now finished to add everything from Claraâs original SmartBreakpoints project. There are still lots of things that need to be done -> Something goes wrong when re-adding a one-shot breakpoint again. -> Adding should be disabled when the editor has non-submitted code. -> The experience in the debugger is not that nice (need to think about using the <debuggerCompleteToSender> trick and explore the filtering capabilities of the debugger) I am now continuing to flesh out the Reflectivity features: -> hook AST generation is now done by HookGenerator -> correct hooks are already generated for the case that a link passes arguments to the meta object (see the RFReification hierarchy, for now we have only #object *and* actual compiling does not yet work due to missing semantic analysis of the AST snippets for the hooks) -> We have support for âvirtualâ meta objects. Any valid reification can be the metaObject (this is fun for #node). TODO - add all reifications for all nodes (#arguments, #node, #context, #value #name, #arg1â¦) - do semantic analysis correctly and add test for links with arguments - add conditions with parameters (e.g. for object specific links: [:object | object == myObject]) - handle primitives: when putting a link on a primitive method, wrap the compiledMethod instead of AST manipulation. - â¦. I will continue to flesh it out and send examples to the list as these things start to work for real. Marcus

May 22, 2015

Re: [Pharo-dev] [Pre-ANN][Job] We are looking for a "Engineer R&D"

by Marcus Denker

Hi, No job posting yet as we are seeing what exact kind of person we can hire. There is a certain chance that this will not be constrained to âtwo years after universityâ. Job posting will be finalised next week, if you are interested (or now someone who might) ==> please contact me already now. Marcus > On 19 May 2015, at 17:10, Marcus Denker <marcus.denker(a)inria.fr> wrote: > > Hi, > > We will put the official announce online later this weekâ¦ we are looking to fill a a âyoung engineer positionâ. > This is for recent graduates who finished their degree 2013/2014/2015. > > The topic is to work for 2 years on improving âremoteâ debugging for cases like server deployment, typical > cloud scenarios and when using embedded Pharo images (e.g. for robotics). > > If you are interested (or know someone who might), please contact me. > (official announcement will be online later this week). > > Marcus >

May 22, 2015

Re: [Pharo-dev] Opportunity to steal good ideas from dad ;)

by phil＠highoctane.be

Thinking of the <=> operator, this can be useful for things like this: dataset := #(1 3 4 5 7 8 5 2 3 0). mean := dataset average. dataset collect: [ :each | mean <=> each ] --> #(1 1 -1 -1 -1 -1 -1 1 1 1) This we can also do with coll := #(1 3 4 5 7 8 5 2 3 0). mean := coll average. coll collect: [ :each | (mean - each) sign ] but still. Other use cases for <=> ? Phil On Fri, May 22, 2015 at 11:43 AM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > > On 21 May 2015, at 16:50, Esteban A. Maringolo <emaringolo(a)gmail.com> > wrote: > > > > 2015-05-21 10:01 GMT-03:00 phil(a)highoctane.be <phil(a)highoctane.be>: > >> > >> > >> On Thu, May 21, 2015 at 2:19 PM, Sean P. DeNigris < > sean(a)clipperadams.com> > >> wrote: > >>> > >>> Squeak 4.6 "Spur" Release Notes > >>> http://wiki.squeak.org/squeak/6192 > >> > >> > >> Tag sort functions looks great to have around. > >> > >> About the spaceship operator (<=>), is this something we'd like to have > >> around? (I like it). http://en.wikipedia.org/wiki/Three-way_comparison > > > > http://www.smalltalkhub.com/#!/~emaringolo/SortFunctions > > > > I use them extensively. > > > > Esteban A. Maringolo > > I had a look. > > Apart from the fact that for some reason, the extensions on MAElement are > in the Core package, hence loading just Core gave errors, things looked OK. > > This is generally useful, small and a very good example of super cool OO > design. > > I would vote to integrate this (<=> and all). > > Sven > > >

May 22, 2015

Re: [Pharo-dev] New Collections for Pharo

by stepharo

Hi did you write tests? did your implementation propose APIs that are compatible with the Collection API? Stef

May 22, 2015

Re: [Pharo-dev] Opportunity to steal good ideas from dad ;)

by stepharo

If people wants to harvest some changes from Squeak, please do it. Collections Speed up testing different sized strings. Interval >> #sum is now O(1) instead of O(n). Introduction of Set>>#addNewElement: optimizes graph-traversal. Slightly faster SequenceableCollection >> #copyUpT (for ByteString receivers). Stricter and faster implementation of String >> #format:. Optimized Interval >> #sum, OrderedCollection >> #at: and OrderedCollection >> #at:put:. Simpler and faster Character >> #tokenish. Speed-up #endsWith: for strings. Morphic Sped up #valueOfProperty:ifAbsent: and friends. Symbol table improvements such as thread-safety and consistent enumeration. Threadsafe random. > Squeak 4.6 "Spur" Release Notes > http://wiki.squeak.org/squeak/6192 > > > > ----- > Cheers, > Sean > -- > View this message in context: http://forum.world.st/Opportunity-to-steal-good-ideas-from-dad-tp4827836.ht… > Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com. > >

May 22, 2015

Re: [Pharo-dev] Morphic menus

by stepharo

Hi mathieu can you open a bug entry? and publish your solution. Tx Le 21/5/15 14:50, Matthieu Lacaton a Ã©crit : > Hello everyone, > > I had to understand how menus work in Morphic lately and I noticed > some things : > > 1) When a menu has strictly more than 1 submenu active it cannot > regain the focus if the mouse doesn't pass over the submenus first. It > is a bit hard for me to explain so I will give you an example : > Open a Nautilus browser on any package and right click on a class. Now > browse "Refactoring => Code Rewriting => Rewrite Code but don't click > on any of these items. You will have 3 menus. Now if you exit the last > menu without returning to the previous one (make a loop from the right > for example if you have enough place) and go back directly to the > first one you'll see that it doesn't respond (it is like frozen). > > I don't know if it is a bug or if it is intended but I find it strange > as it is not how it is working on Windows nor Ubuntu for example. > > So if it is indeed a bug, what I propose as a "fix" would be this: > > Replace *: > > * > > *MenuItemMorph >>* /activateOwner Menu: evt/ > "Activate our owner menu; e.g., pass control to it" > owner ifNil:[^false]. "not applicable" > (owner fullContainsPoint: evt position) ifFalse:[^false]. > owner activate: evt. > ^true > > > by: > > *MenuItemMorph >>* /activateOwner Menu: evt/ > > / > / > > / | popUpOwner | > > / > / owner ifNil: [^false]. > / > / (owner fullContainsPoint: evt position) ifTrue: [ > / > / owner activate: evt. > / > / ^true. > ]. > > / > / popUpOwner := owner popUpOwner. > / > / [ popUpOwner ] whileNotNil: [ > / > / (popUpOwner owner fullContainsPoint: evt position) ifTrue: [ > / > / popUpOwner owner activate: evt. > / > / ^true. > ] > / > / ifFalse: [ > / > / popUpOwner := popUpOwner owner popUpOwner. > ]. > ]. > / > / ^false. > / > > > It should work with any number of submenus but i couldn't find > anywhere in Pharo with more than 3 menus so i couldn't test it further. > > 2) When you leave a submenu and quickly come back to its popUpOwner, > the item in the submenu won't be deselected. So if you navigate a bit > in the menu and then want to come back to this submenu, you won't be > able to select the item by passing your mouse over it. It is not a big > deal in 99.9% of the time as you will still be able to click on it but > when this item is supposed to open a submenu it won't. You need to go > to another item in the submenu and to come back. > > I noticed that it happened because if you are leaving the submenu too > quickly there is no "mouseLeave" event generated so the itemMenu is > not deselected. Thus, as it is considered as selected, if you come > back to it later a "mouseEnter" event will not trigger anything. > > It is quite easy to reproduce with the same menu than for the 1) : > Right click on a class in a Nautilus browser then browse Refactoring > => Code Rewriting. > Now if you quickly come back to Refactoring sometimes you'll see that > Code Rewriting won't be deselected (it will still be blue after the > 500ms time out). If it is the case you can go back on "Rename..." and > then if you quickly come back to Refactoring => Code Rewriting, you > won't be able to get the 3rd submenu. > > I can reproduce this very easily but it is not really a problem > because it only happens if you are fast with your mouse. > > I did not try to fix it but I guess that reinitializing all the > itemMenus to unselected when poping a menu could do the trick. > > 3) This one is mostly insignificant because it is quite hard to > reproduce and doesn't trigger often at all. But sometimes when you > quickly navigate between 3 menus (I have not been able to reproduce it > with only 2 menus) when you want to go back to the first one it > freezes and it doesn't handle the mouseOver events anymore (you can > still click though). > This is not the same thing than the first point because here you > correctly pass through the previous submenus to go to the first one. > I don't really know why this happens sometimes but I think it may be > because the focus was not passed quickly enough between windows and > one on the submenus gets deactivated and removed from the world wheras > it still had the focus on him. > > Anyway these 3 things are not really important but I just wanted to > share what i found with you in case you think it is valid. > > Cheers, > > Matthieu > / > /

May 22, 2015

Re: [Pharo-dev] Opportunity to steal good ideas from dad ;)

by Sven Van Caekenberghe

> On 21 May 2015, at 16:50, Esteban A. Maringolo <emaringolo(a)gmail.com> wrote: > > 2015-05-21 10:01 GMT-03:00 phil(a)highoctane.be <phil(a)highoctane.be>: >> >> >> On Thu, May 21, 2015 at 2:19 PM, Sean P. DeNigris <sean(a)clipperadams.com> >> wrote: >>> >>> Squeak 4.6 "Spur" Release Notes >>> http://wiki.squeak.org/squeak/6192 >> >> >> Tag sort functions looks great to have around. >> >> About the spaceship operator (<=>), is this something we'd like to have >> around? (I like it). http://en.wikipedia.org/wiki/Three-way_comparison > > http://www.smalltalkhub.com/#!/~emaringolo/SortFunctions > > I use them extensively. > > Esteban A. Maringolo I had a look. Apart from the fact that for some reason, the extensions on MAElement are in the Core package, hence loading just Core gave errors, things looked OK. This is generally useful, small and a very good example of super cool OO design. I would vote to integrate this (<=> and all). Sven

May 22, 2015

Re: [Pharo-dev] Best practice for overwriting methods in other classes

by Marcus Denker

> On 22 May 2015, at 07:08, Kasper Osterbye <kasper(a)itu.dk> wrote: > > I have now twice in two different situations felt a need to "steal" a method > in Nautilus and make that method an extension method in my own package. > > In situation A, I wanted to steal the #buildCommentPane to use a different > Morph for class comments, > In situation B, Sergio and I needed to redefine #getComments and > #addComments:notifying: to fill out behaviour for package comments. > > What is the kosher way to work and contribute in such situations? I guess it > will also pop up in situations outside of Nautilus. > Semantically, overrides are never âsaveâ as they do not compose: two packages overriding, which version is the correct one? In the past people tried to e.g. add support to monticello to install the original method when an override gets unloaded. But this all breaks down as soon as two packages override the same method. Overrides therefore should be used just for experimental codeâ¦ what i do is to put it on *mypackage-override so I do not forget to find a solution later that does not need it. Marcus

May 22, 2015