- Pharo-users - lists.pharo.org

Re: [Pharo-users] Pharo 3 beta impressions : SystemBrowser on first visit to new install

by Marcus Denker

On 10 Jan 2014, at 14:46, Marcus Denker <marcus.denker(a)inria.fr> wrote: > > On 10 Jan 2014, at 14:33, Nicolai Hess <nicolaihess(a)web.de> wrote: > >> >> 2014/1/10 Robert Shiplett <grshiplett(a)gmail.com> >> 1) on first open after install on Windows, the SystemBrowser shows a list of class names in LastModified classes - but I have modified no classes in this image >> >> This list is from Recent Messages List (World Menu / Tools / Recent Messages). Not only your changes. >> > This is because I disabled image cleanUp last week for checking if that was the source for the crashâ¦ It is on my TODO to add this again. > I committed a new ScriptLoader that does itâ¦ it will be picked up next some an update is prepared. Marcus

Jan. 10, 2014

Re: [Pharo-users] Pharo 3 beta impressions : SystemBrowser on first visit to new install

by Marcus Denker

> > > 2) on Quit no Save, even after selecting not to Save a second time, the pointer flips to show WRITING TO DISK which was surprising and also seemed not quite as one might expect > > This disturbed me too, but I think this was always there, all Pharo versions and even on Squeak. I fixed itâ¦ https://pharo.fogbugz.com/f/cases/12612/Do-not-show-image-save-mouse-icon-w… The reason was that it has Cursor write showWhile: [ Smalltalk snapshot: save andQuit: true. ] so it uses the write hand even when save is false. Marcus

Jan. 10, 2014

Re: [Pharo-users] Pharo 3 beta impressions : SystemBrowser on first visit to new install

by Marcus Denker

On 10 Jan 2014, at 14:33, Nicolai Hess <nicolaihess(a)web.de> wrote: > > 2014/1/10 Robert Shiplett <grshiplett(a)gmail.com> > 1) on first open after install on Windows, the SystemBrowser shows a list of class names in LastModified classes - but I have modified no classes in this image > > This list is from Recent Messages List (World Menu / Tools / Recent Messages). Not only your changes. > This is because I disabled image cleanUp last week for checking if that was the source for the crashâ¦ It is on my TODO to add this again. Marcus

Jan. 10, 2014

Re: [Pharo-users] Pharo 3 beta impressions : SystemBrowser on first visit to new install

by Nicolai Hess

2014/1/10 Robert Shiplett <grshiplett(a)gmail.com> > 1) on first open after install on Windows, the SystemBrowser shows a list > of class names in LastModified classes - but I have modified no classes in > this image > This list is from Recent Messages List (World Menu / Tools / Recent Messages). Not only your changes. > > 2) on Quit no Save, even after selecting not to Save a second time, the > pointer flips to show WRITING TO DISK which was surprising and also seemed > not quite as one might expect > This disturbed me too, but I think this was always there, all Pharo versions and even on Squeak. > > 3) the title bar changes to SystemBrowser seemed more like bread-crumbs > ... and displayed previous package name even after a single select in the > list even if that previous package had not been entered. Perhaps this is > optional behavior for the browser title bar ? Loosing one's place in a > browser has long been a fact of life, but is this intended to track where I > was last ? But I was not THERE ( I had not viewed a single item in that > package ) > > Yes it is a bug in the SystemBrowser (PackageTreeNautilus a subclass of Nautilus). It works in Nautilus ( Try with "Nauilus open" in a workspace). There are still many little bugs in the SystemBrowser. Thank you for testing and reporting back. Nicolai

Jan. 10, 2014

Pharo 3 beta impressions : SystemBrowser on first visit to new install

by Robert Shiplett

1) on first open after install on Windows, the SystemBrowser shows a list of class names in LastModified classes - but I have modified no classes in this image 2) on Quit no Save, even after selecting not to Save a second time, the pointer flips to show WRITING TO DISK which was surprising and also seemed not quite as one might expect 3) the title bar changes to SystemBrowser seemed more like bread-crumbs ... and displayed previous package name even after a single select in the list even if that previous package had not been entered. Perhaps this is optional behavior for the browser title bar ? Loosing one's place in a browser has long been a fact of life, but is this intended to track where I was last ? But I was not THERE ( I had not viewed a single item in that package ) NOTE : to see the browser title changes 1) select classes last changed ( top item left list ) 2) select a class ( I selected Komiter ) 3) single click elsewhere in the left list ( I selected as AST package ) 4) single click on your "where I intend to view a class package" - but the browser title bar still shows you as in that AST package that you never entered ... Or I have been on Android and in workbench-ed eclipse-fer-Java too much of late ;-) robert Canada

Jan. 10, 2014

Re: [Pharo-users] [Pharo-dev] Two videos to explain what we do when integrating a fix.

by Sven Van Caekenberghe

On 07 Jan 2014, at 10:57, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > Thanks, Stef, I will have a look in the coming days. > Now I wonât have many excuses not to try ;-) And I did, and it worked ! #30683 Thanks for all the help, the screencasts really helped to increase my confidence, the git and CI integration is really seriously impressive. > On 06 Jan 2014, at 16:18, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > >> Hi guys >> >> I produced two videos to show how simple is an integration now (It could be better see below). >> This can influence some of you to help us: sven, ben, igor :)? >> >> First one: install some scripts >> http://stephane.ducasse.free.fr/Videos/PharoIntegration/DownloadingScripts.… >> >> Second one >> http://stephane.ducasse.free.fr/Videos/PharoIntegration/integratingAChange.… >> >> ./prepare >> start menu >> merge 9look at the fixâ¦.) >> freeze menu >> publish menu >> ./publish >> >> I do not think that Monticello is the best tool for managing our changes and code (and I asked money in the past to get an engineer but it was refused and I spent the holiday building a new case to ask money to pay an engineer to >> get a real integration with Git) but for now this is what we have and it is working. I do not want that end users to get expose to >> git on the command-line with different platforms. We should keep the simplicity of Pharo. >> >> Stef >> >> >

Jan. 10, 2014

Re: [Pharo-users] Possible bug with Voyage-Mongo date

by raphonic

Mongo-BSON was the package overriding duration. It over-rode duration's accessor with duration ^Duration seconds: self asSeconds nanoSeconds: 0 -- View this message in context: http://forum.world.st/Possible-bug-with-Voyage-Mongo-date-tp4735597p4735638… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Jan. 10, 2014

Re: [Pharo-users] Possible bug with Voyage-Mongo date

by raphonic

For Date today duration, I got '41281:18:30:00' and for (Date readFrom: '01-23-2004' readStream) duration: '37641:18:30:00'. I've loaded only a few packages over the default image Seaside3(3.1) VoyageMongo, JSON I'm currently looking at the Date class to see if some package overrides any methods. Thanks. -- View this message in context: http://forum.world.st/Possible-bug-with-Voyage-Mongo-date-tp4735597p4735632… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Jan. 10, 2014

Re: [Pharo-users] Possible bug with Voyage-Mongo date

by Sven Van Caekenberghe

On 10 Jan 2014, at 08:18, raphonic <raphonic(a)outlook.com> wrote: > I just ran the entire test suite for Pharo and it seems that the problem > isn't with voyage or the mongo test driver. The > DateTest>#testDuration test failed with this message 'Got 37641:18:30:00 > instead of 1:00:00:00'. I'm running the latest pharo3.0 vm and image on > Ubuntu 12 x86. Anyone know what could be the problem? That seems very weird: it basically says that the duration of a Date, which should be one day, 24 hours, is totally wrong ! What do you get if you do Date today duration. (Date readFrom: '01-23-2004' readStream) duration. I think one of the packages you loaded did some override in Chronology and broke stuff. Sven

Jan. 10, 2014

Re: [Pharo-users] Possible bug with Voyage-Mongo date

by raphonic

I just ran the entire test suite for Pharo and it seems that the problem isn't with voyage or the mongo test driver. The DateTest>#testDuration test failed with this message 'Got 37641:18:30:00 instead of 1:00:00:00'. I'm running the latest pharo3.0 vm and image on Ubuntu 12 x86. Anyone know what could be the problem? -- View this message in context: http://forum.world.st/Possible-bug-with-Voyage-Mongo-date-tp4735597p4735598… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Jan. 10, 2014