- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Different behavior in running tests

by Norbert Hartl

The problem seems to be Phexample. There is an implementation of Given>>#consumeReturnValueAt: testMethod ^(cache includesKey: testMethod) ifTrue: [ cache removeKey: testMethod ] ifFalse: [ |ret| testMethod setUp. ret := testMethod perform: testMethod selector. testMethod tearDown. ret ] that caches test executions. I don't know why it does this and I don't know why an existing key is removed. But in my case I have tests with side effects while storing objects in MongoDB. The DB is properly reseted at the beginning of a test and if a cached test is not executed the necessary state is not produced. I'm not sure what exactly triggers the problem. It is probably the way how I use Phexample. I do a lot of MyTestCase new given: #anotherStateProducingTest because I cannot align the inheritance chain in parallel to the state needs to be there (and Given is used as a singleton). This is a Phexample problem IMHO. I really like Phexample because state preparation is often something that is annoying while writing tests. On the other hand your tests become a database of test execution state. The management of that state is not properly solved and only works in the simple cases as expected. Norbert Am 23.06.2014 um 14:32 schrieb Norbert Hartl <norbert(a)hartl.name>: > I use Sunit + Phexample in my project. The model uses a dynamic variable. In order to ease the writing of tests I implemented in my TestCase > > IAMModelTests>>#runCase > > self createContext do: [ super runCase ] > > with > > IAMContext>>#do: aBlock > ^ IAMCurrentContext > value: self > during: aBlock > > It works basically. The method #runCase is only execute once. The #given: calls inside a test method do not trigger further runCase executions. The strange thing is that all tests succeed when I run them alone e.g. in Nautilus. Whenever I run the tests via the testrunner 2 tests fail. Those 2 tests are not any more special than the other tests. And they fail reliably. > > Does anyone has a hint what is different when being run in testrunner. I cannot debug the problem easily because then I would run the test alone and it will succeed. > > thanks, > > Norbert > >

June 23, 2014

40Inbox

by Yuriy Tymchuk

Hi guys, do I need to be added as a contributor to Pharo40Inbox in order to commit there? Because I managed to commit slice somehow, but not the package itself. (My SmalltalkHub user is YuriyTymchuk) Uko

June 23, 2014

Re: [Pharo-dev] How to debug code using DynamicVariable?

by Esteban A. Maringolo

Norbert, I share the same feelings you have. And couldn't find any other way of dealing with them, other than doing the same thing Sven suggests. Regards! Esteban A. Maringolo 2014-06-23 9:03 GMT-03:00 Sven Van Caekenberghe <sven(a)stfx.eu>: > Your analysis is correct: DynamicVariables are hard to impossible to debug, the debugger always runs in another process. I don't think there is a solution here (maybe let the debugger process copy/share the binding ?) apart from what you suggest: make sure you enter the debugger after the lookup and assignment of the value to a local variable. > > server := ZnCurrentServer value. > > Still, I find DynamicVariables very useful, especially architecture/design-wise. > > On 23 Jun 2014, at 13:50, Norbert Hartl <norbert(a)hartl.name> wrote: > >> In my code I'm using a DynamicVariable to request a context object when needed. Until now I knew the name DynamicVariable only from seaside. There it is called WADynamicVariable and it is an exception. So I blindly assumed the pharo DynamicVariable works the same. >> I thought this might be a good optimization not to travel the stack all the time but put in the process. >> Now that I am using it I can see the difference. I find it real hard using it because I don't know how to debug/step in code. DynamicVariable is a process specific variable but as soon as a debugger opens it is very likely to be in another process. This makes stepping in method using the DynamicVariable impossible. The only way round is to set break points after the dynamic lookup and step from there. But this feels just wrong. >> What would be the best way to have DynamicVariable and be able to debug anything? Or is there a variant that uses the stack instead of the "active" process? >> >> thanks, >> >> Norbert > >

June 23, 2014

Different behavior in running tests

by Norbert Hartl

I use Sunit + Phexample in my project. The model uses a dynamic variable. In order to ease the writing of tests I implemented in my TestCase IAMModelTests>>#runCase self createContext do: [ super runCase ] with IAMContext>>#do: aBlock ^ IAMCurrentContext value: self during: aBlock It works basically. The method #runCase is only execute once. The #given: calls inside a test method do not trigger further runCase executions. The strange thing is that all tests succeed when I run them alone e.g. in Nautilus. Whenever I run the tests via the testrunner 2 tests fail. Those 2 tests are not any more special than the other tests. And they fail reliably. Does anyone has a hint what is different when being run in testrunner. I cannot debug the problem easily because then I would run the test alone and it will succeed. thanks, Norbert

June 23, 2014

Re: [Pharo-dev] How to debug code using DynamicVariable?

by Norbert Hartl

Am 23.06.2014 um 14:03 schrieb Sven Van Caekenberghe <sven(a)stfx.eu>: > Your analysis is correct: DynamicVariables are hard to impossible to debug, the debugger always runs in another process. I don't think there is a solution here (maybe let the debugger process copy/share the binding ?) apart from what you suggest: make sure you enter the debugger after the lookup and assignment of the value to a local variable. > > server := ZnCurrentServer value. > > Still, I find DynamicVariables very useful, especially architecture/design-wise. Agreed, I find context sensitive programming important because I don't like global state and you can solve this particular problem very easy. But I like the exception based dynamic variable much more. There the dynamic value is stored in a MethodContext which is a proper place for it. A process being only a runtime decorator for a stack is asking for trouble. Everything that copies/references the stack is suspect to produce errors. Of course the process variables are a lot faster. Don't know if a hybrid solution could work here. Norbert > > On 23 Jun 2014, at 13:50, Norbert Hartl <norbert(a)hartl.name> wrote: > >> In my code I'm using a DynamicVariable to request a context object when needed. Until now I knew the name DynamicVariable only from seaside. There it is called WADynamicVariable and it is an exception. So I blindly assumed the pharo DynamicVariable works the same. >> I thought this might be a good optimization not to travel the stack all the time but put in the process. >> Now that I am using it I can see the difference. I find it real hard using it because I don't know how to debug/step in code. DynamicVariable is a process specific variable but as soon as a debugger opens it is very likely to be in another process. This makes stepping in method using the DynamicVariable impossible. The only way round is to set break points after the dynamic lookup and step from there. But this feels just wrong. >> What would be the best way to have DynamicVariable and be able to debug anything? Or is there a variant that uses the stack instead of the "active" process? >> >> thanks, >> >> Norbert > >

June 23, 2014

Re: [Pharo-dev] How to debug code using DynamicVariable?

by Sven Van Caekenberghe

Your analysis is correct: DynamicVariables are hard to impossible to debug, the debugger always runs in another process. I don't think there is a solution here (maybe let the debugger process copy/share the binding ?) apart from what you suggest: make sure you enter the debugger after the lookup and assignment of the value to a local variable. server := ZnCurrentServer value. Still, I find DynamicVariables very useful, especially architecture/design-wise. On 23 Jun 2014, at 13:50, Norbert Hartl <norbert(a)hartl.name> wrote: > In my code I'm using a DynamicVariable to request a context object when needed. Until now I knew the name DynamicVariable only from seaside. There it is called WADynamicVariable and it is an exception. So I blindly assumed the pharo DynamicVariable works the same. > I thought this might be a good optimization not to travel the stack all the time but put in the process. > Now that I am using it I can see the difference. I find it real hard using it because I don't know how to debug/step in code. DynamicVariable is a process specific variable but as soon as a debugger opens it is very likely to be in another process. This makes stepping in method using the DynamicVariable impossible. The only way round is to set break points after the dynamic lookup and step from there. But this feels just wrong. > What would be the best way to have DynamicVariable and be able to debug anything? Or is there a variant that uses the stack instead of the "active" process? > > thanks, > > Norbert

June 23, 2014

How to debug code using DynamicVariable?

by Norbert Hartl

In my code I'm using a DynamicVariable to request a context object when needed. Until now I knew the name DynamicVariable only from seaside. There it is called WADynamicVariable and it is an exception. So I blindly assumed the pharo DynamicVariable works the same. I thought this might be a good optimization not to travel the stack all the time but put in the process. Now that I am using it I can see the difference. I find it real hard using it because I don't know how to debug/step in code. DynamicVariable is a process specific variable but as soon as a debugger opens it is very likely to be in another process. This makes stepping in method using the DynamicVariable impossible. The only way round is to set break points after the dynamic lookup and step from there. But this feels just wrong. What would be the best way to have DynamicVariable and be able to debug anything? Or is there a variant that uses the stack instead of the "active" process? thanks, Norbert

June 23, 2014

Re: [Pharo-dev] Pillar and tables

by kilon alios

done about the table conversion is it also hard to output just the text of the table , with no formatting, inside /table tags or something like that ? I know emacs lisp but I used it only to setup my init.el for extremely simple stuff. On Mon, Jun 23, 2014 at 12:22 PM, Damien Cassou <damien.cassou(a)gmail.com> wrote: > On Sun, Jun 22, 2014 at 9:16 PM, kilon alios <kilon.alios(a)gmail.com> > wrote: > > I have just ported 4.1 section of PBE to pillar. The automatic conversion > > tool of Damien does not work with tables so I did that by hand. > > > Thanks Kilon for these bug reports. Could you please report the > latex2pillar-related ones on > https://github.com/pillar-markup/pillar-mode ? For the table > conversion, there is no way I will implement it, it is just too > difficult. > > -- > Damien Cassou > http://damiencassou.seasidehosting.st > > "Success is the ability to go from one failure to another without > losing enthusiasm." > Winston Churchill > >

June 23, 2014

Re: [Pharo-dev] naming convention for dangerous methods

by Damien Pollet

>From what I've seen of that convention in Ruby, it gets misused. They also have the convention of ending selectors with a question mark for predicates, which works quite well (besides the fact that it's hardcoded in the languageâ¦ you can't do this, for instance, because only one of ? or ! can end the selector def wtf??! return 42 end As for the ! indicating side-effects, there are methods with side effects that don't end in !, and some that don't really that have it. In the end, the meaning gets diluted, IMHO. In Smalltalk, we have method categories, I'd like to exploit them more than just see them in the browser. If we could use them as tags instead of as a partition, that would be more flexible too On 22 June 2014 19:23, stepharo <stepharo(a)free.fr> wrote: > > >> There was a previous thread [1] that involved discussion of the naming >> convention for methods doing "dangerous" in-place mutatation - e.g. >> sort/sorted, reverse/reversed - and even though I now know there is such a >> naming convention, its hard to keep track of which is which. > > >> Now today I happened to be looking for the first time at some Ruby code >> and came across their naming convention for "dangerous" methods of having a >> trailing exclamation mark - e.g. sort/sort!, reverse/reverse! - which is >> REALLY intuitive. How extreme is it to wonder if Pharo might some day be >> able to support such a convention? > > You know we cannot change the world in one day. So I like this convention > but did you measure how much work it would mean > to follow it. > > What would be good is > to have systematically method comments > to build tools to give some hints and to slowly make sure that we follow > our convention. > analyse what is reasonable to do. > > Stef > > >> >> [1] >> http://forum.world.st/11635-Race-condition-in-SequenceableCollection-gt-gt-… >> >> >> > > -- Damien Pollet type less, do more [ | ] http://people.untyped.org/damien.pollet

June 23, 2014

prepared inclusion of OSWindow

by stepharo

Hi Igor and Ronie following igor advice I'm preparing to integration OSWindow in Pharo. I copied all the packages to Pharo/OSWindow. Now to fully work I imagine that SDL2 should be linked to the VM so that it can work. I need to know which package make sense to be loaded I imagine that this is OSWindow-Core OSWindow-SDL2 https://pharo.fogbugz.com/default.asp?12743#103040 indicates package: 'NBXLib-Core'; and I wonder why Why do we need also the OSWindow-X11 if we use SDL Stef

June 23, 2014