- Pharo-dev - lists.pharo.org

monkey timeout

by Nicolai Hess

I think we have to review and revert some changes for SmallLint (RB-Classes). issue 14096 and 14098 for instance, rules on package X are executed on all methods in class Y if package X extends "one" methode in class Y. -> issue 14228

Oct. 18, 2014

Re: [Pharo-dev] #openPharoZenWorkspace broken

by Sven Van Caekenberghe

Perfect! I'll make a slice. > On 18 Oct 2014, at 13:28, Marcus Denker <marcus.denker(a)gmail.com> wrote: > > replace it by > > UIManager default > edit: self pharoZenValuesContents > label: 'Pharo Zen'. > > > > > On Sat, Oct 18, 2014 at 1:20 PM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > So I noticed that #openPharoZenWorkspace is broken, because that code makes assumptions that GT-Playground does not offer. > > I know this has been discussed before, what is the strategy going forward ? > > Sven > >

Oct. 18, 2014

Re: [Pharo-dev] #openPharoZenWorkspace broken

by Marcus Denker

replace it by UIManager defaultÂ edit: self pharoZenValuesContents label: 'Pharo Zen'. On Sat, Oct 18, 2014 at 1:20 PM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > So I noticed that #openPharoZenWorkspace is broken, because that code makes assumptions that GT-Playground does not offer. > I know this has been discussed before, what is the strategy going forward ? > Sven

Oct. 18, 2014

#openPharoZenWorkspace broken

by Sven Van Caekenberghe

So I noticed that #openPharoZenWorkspace is broken, because that code makes assumptions that GT-Playground does not offer. I know this has been discussed before, what is the strategy going forward ? Sven

Oct. 18, 2014

Re: [Pharo-dev] Mysteries of the compiler

by Nicolai Hess

2014-10-17 13:28 GMT+02:00 Marcus Denker <marcus.denker(a)inria.fr>: > > -- > Marcus Denker > Sent with Airmail > > On 17 Oct 2014 at 12:46:20, Nicolai Hess (nicolaihess(a)web.de) wrote: > > I ran some testcases with your changes and I expect some test > are failing now, but actually they don't fail but others do. > > > > Yes, but I think these tests are wrong. > (needs to be carefully checked). > > Ok, some checks: OCASTCheckerTest testSemanticAnalysisOnNonMethodNode is this a testcase? what does it test? testNoRemoteBlockArgument testNoRemoteBlockTemp both look similiar. testsingleRemoteTempVarWhileWithTempNotInlined I would add assertions for the "temp" var the others look fine. OCASTClosureAnalyzerTest testOptimizedBlockWriteInNestedBlockCase2: the testmethos like (OCOpalExamples>>#optimizedBlockWriteInNestedBlockCase2) arent supposed to be runnable code, but I would change the "(true)whileTrue:" to -> "[true] whileTrue:" anyway. testOptimizedBlockWriteInNestedBlockCase3 look at the comment for : self assert: ast scope copiedVars size = 1. "Is this correct, I think that the copied vars should be empty." I think the number of copiedVars is right, because there is a outer block. the same for testOptimizedBlockWriteInNestedBlockCase4, the whileTrue is inside of a block. in testOptimizedBlockWrittenAfterClosedOverCase1 there is a comment "problem: as block is optimized, this var does not need to be remote" self assert: (scopes third tempVector at: 'temp') isRemote. But I think this is correct, the temp var is within a block the others look fine. Changing the method visitVariableNode: and lookupVariableForWrite: makes these tests failing: testOptimizedBlockWrittenAfterClosedOverCase1 it fails in self deny: (ast scope lookupVar: 'index') isEscaping. I think the test is right, the var index is only used within the whileTrue block. the same for testOptimizedBlockWrittenAfterClosedOverCase2 testExampleSimpleBlockLocalWhile fails, I think the test is right, but I am not sure. testOptimizedBlockWriteInBlock fails in self assert: ast scope tempVars size = 1. But I think the test is right, there is only one local var, and it is written only in the whileTrue:[block] the same in testOptimizedBlockWriteInNestedBlockCase2, allthough the whileTrue: is in another ifTrue:[] block the same for the other failing tests.

Oct. 18, 2014

Re: [Pharo-dev] Question on Package comments [SO]

by kilon alios

There this dark magic in internet called Google search, as a Jedi yourself you may not use but I a sith lord myself I use it all the time. I have used more time to find out about thing than I have posted here, you know why ? Because its quite rare for google search not to able to find something. Google search is some serious black magic. So here is the workflow Step 1- you google search "pharo documentation" , on of the top search results is the pharo website. Now the fact that Pharo has a centralised website should make you suspicious that if a centralised reference documentation is not linked there then you should be almost certain that it does not exist. Step 2- you have went through the links of the documentation and you are still convinced. No problemo, pick a library that is not documented in there and is part of the pharo official distribution, hey why not , pick 10 libraries / packages , add some DLLs to spice it up and use again google search. Do you find documentation links for them, are those links in any case centralised in a single web site ? or at least in a single pharo tool ? No. You can now safely assume that there is no such thing. This process is called deductive reasoning. Also it does not take much of browsing SO to realise that is not the right place to ask these questions. If you google pharo it even gives you the link to this mailing list. To be fair however this may not be entirely his fault and Pharo website is to be blamed here to for example in the community sub page it says "If you have any question regarding Pharo, use StackOverflow <http://stackoverflow.com/tags/pharo>." This is a mistake , a very big one. StackOverflow most definetly does not accept any question and should not be used as the main way to ask any kind questions. This should change to at least "If you have specific practical questions regarding Pharo , use StackOverflow, for generic questions use the mailing list". I sorry if I sound a bit too sarcastic but seriously we need as a internet community and I am talking about the whole internet and not just pharo , agree that we need to stress the mandatory use of google. It will save people time, especially those that ask questions and expose them in information that a forum or a QA website will never offer them. On Sat, Oct 18, 2014 at 1:50 AM, Nicolai Hess <nicolaihess(a)web.de> wrote: > > > 2014-10-18 0:38 GMT+02:00 kilon alios <kilon.alios(a)gmail.com>: > >> I disagree and I downvoted the question because its clear he did not even >> bother to google for documentation. >> > > He asked: "Does there exist any comprehensive reference for what each one > [Package] is" > The answer is "no". How can he found this out without asking? >

Oct. 18, 2014

Re: [Pharo-dev] Question on Package comments [SO]

by Alain Rastoul

I agree with Kilon. May be he is true about a lack of a package/class reference documentation/system like in java or dotnet but I find the question a little biased and SO policy is right on that : "questions to find (...) tools, libraries (...) are off-topic, they tend to attract opinionated answers and spam. Describe the problem (...)". If he already downloaded the VM, he could have asked in the newsgroup instead. May be building a tool (javadoc like, extracting class and method comments with additional tags) could be of some interest ? Not only for newcomers. Regards, Alain Le 18/10/2014 02:54, Sean P. DeNigris a Ã©crit : > Nicolai Hess wrote >>> I disagree and I downvoted the question because its clear he did not even >>> bother to google for documentation. >> He asked: "Does there exist any comprehensive reference for what each one >> [Package] is" >> The answer is "no". How can he found this out without asking? > > I have to agree with Nicolai here. He wasn't asking "how do I find the docs > for package Xyz?" in which case search would be obvious. What exactly would > he type into google to find out if Pharo had a package catalog, like Squeak > Map? I think it's a logical question that could be difficult to google. > > > > ----- > Cheers, > Sean > -- > View this message in context: http://forum.world.st/Question-on-Package-comments-SO-tp4785191p4785240.html > Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com. > >

Oct. 18, 2014

Re: [Pharo-dev] Question on Package comments [SO]

by Sean P. DeNigris

Nicolai Hess wrote >> I disagree and I downvoted the question because its clear he did not even >> bother to google for documentation. > He asked: "Does there exist any comprehensive reference for what each one > [Package] is" > The answer is "no". How can he found this out without asking? I have to agree with Nicolai here. He wasn't asking "how do I find the docs for package Xyz?" in which case search would be obvious. What exactly would he type into google to find out if Pharo had a package catalog, like Squeak Map? I think it's a logical question that could be difficult to google. ----- Cheers, Sean -- View this message in context: http://forum.world.st/Question-on-Package-comments-SO-tp4785191p4785240.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

Oct. 18, 2014

Re: [Pharo-dev] Question on Package comments [SO]

by Nicolai Hess

2014-10-18 0:38 GMT+02:00 kilon alios <kilon.alios(a)gmail.com>: > I disagree and I downvoted the question because its clear he did not even > bother to google for documentation. > He asked: "Does there exist any comprehensive reference for what each one [Package] is" The answer is "no". How can he found this out without asking?

Oct. 17, 2014

Re: [Pharo-dev] Question on Package comments [SO]

by kilon alios

I disagree and I downvoted the question because its clear he did not even bother to google for documentation. Maybe he forgot or maybe he meants something else than it is already available. I am mostly against SO strict policy but in this case I will have to agree with it, the question is too vague for it anyway. On Sat, Oct 18, 2014 at 1:27 AM, Nicolai Hess <nicolaihess(a)web.de> wrote: > 2014-10-17 23:43 GMT+02:00 Sean P. DeNigris <sean(a)clipperadams.com>: > >> Nicolai Hess wrote >> > It is a valid question >> >> I agree. And at the same time, after reading their reasoning, while >> seemingly overzealous, their decision seems logically sound based on SO >> rules. I added this comment: >> >> I think that the closest you have is >> ci.inria.fr/pharo-contribution/job/PharoProjectCatalog/â¦, which is an >> attempt-in-progress in making a package catalog. Also, most better-known >> Pharo projects seem to be hosted on SmalltalkHub which provides a Markdown >> description for each project (provided of course that someone has filled >> it >> out!) >> > > thank you! > > > >> >> >> >> ----- >> Cheers, >> Sean >> -- >> View this message in context: >> http://forum.world.st/Question-on-Package-comments-SO-tp4785191p4785223.html >> Sent from the Pharo Smalltalk Developers mailing list archive at >> Nabble.com. >> >> >

Oct. 17, 2014