- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Discussing FileSystem >> #workingDirectory and VM process getcwd(3)

by Dale Henrichs

this pinged my memory about OSX ... the funkiness is that you had to arrange for PWD to be set "correctly" when double clicking an application ... when running from the command line PWD was set "correctly" ... This was affecting GemStone because I was using the PWD of the client in certain circumstances (and years ago) ... Sounds like this quirk probably doesn't apply, but I still think it is something to be aware of ... Dale On 3/21/16 11:13 AM, Mariano Martinez Peck wrote: > Hi guys, > > In a couple of recent threads, there were discussions > about FileSystem >> #workingDirectory VS VM process getcwd(3) and > the fact that they do not match. > > The current state is as follows: > > 1) *If Pharo was launched from double click* then we do not care about > its location and I think it's correct to default to image location: > > 1.1) FileSystem >> #workingDirectory refers to the directory > containing the .image file > 1.2) The VM process getcmd() is the default one of the particular OS > ($HOME in my OSX). > > So...there is a mismatch between #workingDirectory and VM's process cmd. > > *I think (as Damien said), that we should do a chdir() for the VM > process to match the .image location. * > > 2) *If Pharo was launched from command line* then we likely are > interested in the current CMD from where we launched the VM > process (imagine using Pharo for scripting). The current status is: > > 2.1) FileSystem >> #workingDirectory still refers to the directory > containing the .image file which I think it is incorrect. > 2.2) The CMD of the VM process is correct. But again, there is a > mismatch with #workingDirectory. > > *For this case, I think we should set FileSystem >> #workingDirectory > to CMD. * > > Before start discussing how to implement this, do you agree with the idea? > > Thanks in advance, > > -- > Mariano > http://marianopeck.wordpress.com

March 21, 2016

Discussing FileSystem >> #workingDirectory and VM process getcwd(3)

by Mariano Martinez Peck

Hi guys, In a couple of recent threads, there were discussions about FileSystem >> #workingDirectory VS VM process getcwd(3) and the fact that they do not match. The current state is as follows: 1) *If Pharo was launched from double click* then we do not care about its location and I think it's correct to default to image location: 1.1) FileSystem >> #workingDirectory refers to the directory containing the .image file 1.2) The VM process getcmd() is the default one of the particular OS ($HOME in my OSX). So...there is a mismatch between #workingDirectory and VM's process cmd. *I think (as Damien said), that we should do a chdir() for the VM process to match the .image location. * 2) *If Pharo was launched from command line* then we likely are interested in the current CMD from where we launched the VM process (imagine using Pharo for scripting). The current status is: 2.1) FileSystem >> #workingDirectory still refers to the directory containing the .image file which I think it is incorrect. 2.2) The CMD of the VM process is correct. But again, there is a mismatch with #workingDirectory. *For this case, I think we should set FileSystem >> #workingDirectory to CMD. * Before start discussing how to implement this, do you agree with the idea? Thanks in advance, -- Mariano http://marianopeck.wordpress.com

March 21, 2016

Re: [Pharo-dev] feedback from yet another student presentation

by Dale Henrichs

Tobias Pape has been improving SS3 over the years and the latest versions are in continual use at HPI ... I can find out if there is serious interest ... Dale On 3/10/16 11:15 PM, Serge Stinckwich wrote: > On Fri, Mar 11, 2016 at 8:08 AM, stepharo <stepharo(a)free.fr> wrote: >> >> Le 10/3/16 22:14, Alexandre Bergel a Ã©crit : >>>> - it was reallllly hard to register a project on smalltalkhub just after >>>> we created an account. >>>> This already happened to me during the open-devs. >>>> So this is a really big show stopper for the mooc. Because people will >>>> gave up just after the first lecture :( >>> The situation will be much easier I hope once git is well supportedâ¦ >> >> But will we have to answer 100 of problems during the mooc? >> And if people drop after just because they cannot save their code as we say >> it :( > Yes this is real annoying problem ... Maybe you can ask Gemstone to > support the MOOC > with a dedicated ss3 instance (or use ss3). >

March 21, 2016

Re: [Pharo-dev] [Pharo-users] DeprecationAssistant

by Yuriy Tymchuk

Oh, thatâs an interesting use case. Itâs not easy no make predictions further in stack so by default it is assumed that the change should happen in the method that has called the deprecated one. In fact usually this should help with the method deprecation, and in the particular case we are dealing with a class deprecation hacked as a method deprecation :) But itâs good that now I know more use cases ;) Uko > On 21 Mar 2016, at 14:56, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > Ah okay! > But how does the system resolve it ? I mean, the class RTLabelled does not refer to RTLabeled. So, what is the magic behind your tool? > > I found an interesting situation. Look at this: > > <Screen Shot 2016-03-21 at 10.53.09 AM.png> > Alexandre > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu <http://www.bergel.eu/> > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > >> On Mar 21, 2016, at 10:19 AM, Yuriy Tymchuk <yuriy.tymchuk(a)me.com <mailto:yuriy.tymchuk@me.com>> wrote: >> >> Hi Alex, >> >> this is not about criticsâ¦ >> >> If you have a method like >> >> labeledObject >> ^ RTLabelled new >> >> in the latest roassal image and you will execute it, you will get a debugger deprecation warning with an option to fix that. >> >> Cheers. >> Uko >> >>> On 21 Mar 2016, at 14:07, Alexandre Bergel <alexandre.bergel(a)me.com <mailto:alexandre.bergel@me.com>> wrote: >>> >>> Hi! >>> >>> I cannot figure out what I should expect after loading DeprecationAssistant. I have loaded it and created a method: >>> -=-=-=-=-=-=-=-=-= >>> foo >>> self deprecated: 'use bar insteadâ >>> -=-=-=-=-=-=-=-=-= >>> >>> Is there anything that should be displayed in the critic list? >>> >>> Alexandre >>> -- >>> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >>> Alexandre Bergel http://www.bergel.eu <http://www.bergel.eu/> >>> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >>> >>> >>> >>>> On Mar 21, 2016, at 9:59 AM, Yuriy Tymchuk <yuriy.tymchuk(a)me.com <mailto:yuriy.tymchuk@me.com>> wrote: >>>> >>>> HI everyone, >>>> >>>> last week Iâve created a little tool that automates deprecation resolution. The goal is to provide a help based on the existing code (no need for extra actions). You can read more here: http://blog.yuriy.tymch.uk/2016/03/deprecationassistant-aiding-developers.h… <http://blog.yuriy.tymch.uk/2016/03/deprecationassistant-aiding-developers.h…> >>>> >>>> You can install and try it out with: >>>> Metacello new >>>> repository: 'github://Uko/DeprecationAssistant' <github://Uko/DeprecationAssistant'>; >>>> baseline: 'DeprecationAssistant'; >>>> load >>>> >>>> This will add additional actions to the debugger once you will get a deprecation warning. >>>> >>>> Have fun! >>>> Cheers. >>>> Uko >>> >> >

March 21, 2016

Re: [Pharo-dev] feedback from yet another student presentation

by Sean P. DeNigris

abergel wrote >> So this is a really big show stopper for the mooc. Because people will >> gave up just after the first lecture :( > The situation will be much easier I hope once git is well supportedâ¦ IFAICT, git is well-supported right now. Having our entire infrastructure written in Pharo is definitely still a great dream for me, but realistically, a Pharo-backed repo is so far down on the list of priorities for our limited manpower that I wouldn't waste a moment on it. I moved all my public projects to GitHub (and private ones to BitBucket for free) because I was horrified once it dawned on me that all my project loading and saving (including private projects) was unencrypted on all our standard repos. When I combine that with BitBucket/GitHub's built-in issue tracker, wiki, and collaborative features, it was a no brainer. Although I resisted the switch to git for quite a while, I haven't looked back. With FileTree, I rarely use any more git than `git add -A`, `git commit`, and `git push origin master`, and with GitFileTree, even that basic knowledge of git is unnecessary. Top it all off with git being considered sexy and cool right now, why not just start them off with GitFileTree and be done with it? ----- Cheers, Sean -- View this message in context: http://forum.world.st/feedback-from-yet-another-student-presentation-tp4883… Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

March 21, 2016

Re: [Pharo-dev] [Pharo-users] DeprecationAssistant

by Alexandre Bergel

Ah okay! But how does the system resolve it ? I mean, the class RTLabelled does not refer to RTLabeled. So, what is the magic behind your tool? I found an interesting situation. Look at this: Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > On Mar 21, 2016, at 10:19 AM, Yuriy Tymchuk <yuriy.tymchuk(a)me.com> wrote: > > Hi Alex, > > this is not about criticsâ¦ > > If you have a method like > > labeledObject > ^ RTLabelled new > > in the latest roassal image and you will execute it, you will get a debugger deprecation warning with an option to fix that. > > Cheers. > Uko > >> On 21 Mar 2016, at 14:07, Alexandre Bergel <alexandre.bergel(a)me.com <mailto:alexandre.bergel@me.com>> wrote: >> >> Hi! >> >> I cannot figure out what I should expect after loading DeprecationAssistant. I have loaded it and created a method: >> -=-=-=-=-=-=-=-=-= >> foo >> self deprecated: 'use bar insteadâ >> -=-=-=-=-=-=-=-=-= >> >> Is there anything that should be displayed in the critic list? >> >> Alexandre >> -- >> _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: >> Alexandre Bergel http://www.bergel.eu <http://www.bergel.eu/> >> ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. >> >> >> >>> On Mar 21, 2016, at 9:59 AM, Yuriy Tymchuk <yuriy.tymchuk(a)me.com <mailto:yuriy.tymchuk@me.com>> wrote: >>> >>> HI everyone, >>> >>> last week Iâve created a little tool that automates deprecation resolution. The goal is to provide a help based on the existing code (no need for extra actions). You can read more here: http://blog.yuriy.tymch.uk/2016/03/deprecationassistant-aiding-developers.h… <http://blog.yuriy.tymch.uk/2016/03/deprecationassistant-aiding-developers.h…> >>> >>> You can install and try it out with: >>> Metacello new >>> repository: 'github://Uko/DeprecationAssistant' <github://Uko/DeprecationAssistant'>; >>> baseline: 'DeprecationAssistant'; >>> load >>> >>> This will add additional actions to the debugger once you will get a deprecation warning. >>> >>> Have fun! >>> Cheers. >>> Uko >> >

March 21, 2016

[pharo-project/pharo-core]

by GitHub

Branch: refs/tags/50656 Home: https://github.com/pharo-project/pharo-core

March 21, 2016

[pharo-project/pharo-core] 188c4e: 50656

by GitHub

Branch: refs/heads/5.0 Home: https://github.com/pharo-project/pharo-core Commit: 188c4e76bb2357beb8e4a08fc7ed89af682f9139 https://github.com/pharo-project/pharo-core/commit/188c4e76bb2357beb8e4a08f… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2016-03-21 (Mon, 21 Mar 2016) Changed paths: M ConfigurationOfSessionManager.package/ConfigurationOfSessionManager.class/instance/symbolic versions/stable_.st A ConfigurationOfSessionManager.package/ConfigurationOfSessionManager.class/instance/versions/version15_.st A HudsonBuildTools20.package/HDTestReport.class/instance/testing/hasErrors.st A HudsonBuildTools20.package/HDTestReport.class/instance/testing/hasFailures.st M HudsonBuildTools20.package/HDTestReport.class/instance/testing/hasFailuresOrErrors.st M HudsonBuildTools20.package/TestCommandLineHandler.class/README.md M HudsonBuildTools20.package/TestCommandLineHandler.class/instance/private/runPackages.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50655.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - scripts/script50656.st R ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50655.st A ScriptLoader50.package/ScriptLoader.class/instance/pharo - updates/update50656.st M ScriptLoader50.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st M System-SessionManager.package/SessionManager.class/class/compatibility - startupList/initializedSessionManager.st A System-SessionManager.package/SessionManager.class/instance/accessing/maxPriority.st A System-SessionManager.package/SessionManager.class/instance/registration/registerLast_inCategory_.st Log Message: ----------- 50656 17844 make SystemSettingsPersistence the last startup executed https://pharo.fogbugz.com/f/cases/17844 17852 command line test runner should be able to exit without error if there are failures https://pharo.fogbugz.com/f/cases/17852 http://files.pharo.org/image/50/50656.zip

March 21, 2016

Re: [Pharo-dev] Fwd: Really usability fix for Versionner

by Sean P. DeNigris

stepharo wrote > it is IMPOSSIBLE to use Versionner for managing projects like > SciSmalltalk. Yes, also it is impossible to use with Metacello's BaselineOf (created in 2012). This means that all git-hosted projects using Metacello in the recommended way (e.g. storing most info in a BaselineOf committed in the repo alongside the project code) are back to writing dependency info by hand - obviously a huge drag now that we've been spoiled by Versionner! ----- Cheers, Sean -- View this message in context: http://forum.world.st/Fwd-Really-usability-fix-for-Versionner-tp4884177p488… Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

March 21, 2016

Re: [Pharo-dev] [Pharo-users] DeprecationAssistant

by Yuriy Tymchuk

Hi Alex, this is not about criticsâ¦ If you have a method like labeledObject ^ RTLabelled new in the latest roassal image and you will execute it, you will get a debugger deprecation warning with an option to fix that. Cheers. Uko > On 21 Mar 2016, at 14:07, Alexandre Bergel <alexandre.bergel(a)me.com> wrote: > > Hi! > > I cannot figure out what I should expect after loading DeprecationAssistant. I have loaded it and created a method: > -=-=-=-=-=-=-=-=-= > foo > self deprecated: 'use bar insteadâ > -=-=-=-=-=-=-=-=-= > > Is there anything that should be displayed in the critic list? > > Alexandre > -- > _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: > Alexandre Bergel http://www.bergel.eu <http://www.bergel.eu/> > ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;. > > > >> On Mar 21, 2016, at 9:59 AM, Yuriy Tymchuk <yuriy.tymchuk(a)me.com <mailto:yuriy.tymchuk@me.com>> wrote: >> >> HI everyone, >> >> last week Iâve created a little tool that automates deprecation resolution. The goal is to provide a help based on the existing code (no need for extra actions). You can read more here: http://blog.yuriy.tymch.uk/2016/03/deprecationassistant-aiding-developers.h… <http://blog.yuriy.tymch.uk/2016/03/deprecationassistant-aiding-developers.h…> >> >> You can install and try it out with: >> Metacello new >> repository: 'github://Uko/DeprecationAssistant' <github://Uko/DeprecationAssistant'>; >> baseline: 'DeprecationAssistant'; >> load >> >> This will add additional actions to the debugger once you will get a deprecation warning. >> >> Have fun! >> Cheers. >> Uko >

March 21, 2016