February 2013 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] Wow... we've closed a lot of issues

by Sean P. DeNigris

Nicolas Cellier wrote > Though i bet some of the 500 were closed for lack of activity rather > than for resolution ;) While I agree that closing to prevent rot would be okay too, 1275 were definitely integrated (and tagged as such). Of the 134 that are marked closed, if you go through those, some were integrated too (e.g. fix included in a slice for another issue). So at least 82% were actually fixed, which continues to amaze me, even as I rehash it now :) -- View this message in context: http://forum.world.st/Wow-we-ve-closed-a-lot-of-issues-tp4669901p4669963.ht… Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

Feb. 14, 2013

Re: [Pharo-project] Wow... we've closed a lot of issues

by Nicolas Cellier

Though i bet some of the 500 were closed for lack of activity rather than for resolution ;) It's not a bad thing, letting them rot has no much value and prevent progress (it's stupid to repeatedly loose time in analyzing issues that nobody cares of). After all, what counts is that Pharo has enough task forces to throw into making a better system (simpler, less convoluted, more powerful etc...). This is a nice thing and this is what the number of closed issues reflects. Bravo Nicolas 2013/2/14 Esteban Lorenzano <estebanlm(a)gmail.com>: > yeah we opened a lot. IMO, as you say, that's a good thing... you need to break things to create new things > but there were at least 500 issues opened since the beginning of times that we have closed in this release. > > so, 500 "historical issues"plus all the issues we created (most of them non bugs,just modifications, enhancements, additions) > not a bad number, I think :) > > cheers, > Esteban > > On Feb 14, 2013, at 5:55 PM, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: > >> You must also ask how many issues you have created since 1.4, because >> I doubt there are 1500 issues applying to 1.4. >> It's not a critic. >> It simply reflect the fact that it is necessary to temporarily break >> things in order to change underlying system (events, text, browser >> etc...) >> Especially when the system has evolved up to a point where it is >> fragile and hardly maintainable. >> >> Nicolas >> >> 2013/2/14 Sean P. DeNigris <sean(a)clipperadams.com>: >>> I was just doing some housecleaning, deleting some old bug fix images and I >>> was struck by how many important things we've accomplished. I'm still >>> dumbfounded that a community of our size has resolved 1548 issues since 1.4. >>> >>> Cheers, >>> Sean >>> >>> >>> >>> -- >>> View this message in context: http://forum.world.st/Wow-we-ve-closed-a-lot-of-issues-tp4669901.html >>> Sent from the Pharo Smalltalk mailing list archive at Nabble.com. >>> >> > >

Feb. 14, 2013

Re: [Pharo-project] PetitParser in 2.0

by Esteban Lorenzano

hi, something like: Gofer it renggli: 'petit'; package: 'ConfigurationOfPetitParser'; load. (ConfigurationOfPetitParser project version: '1.5') load: #(Core PetitXml PetitJson PetitXPath PetitCSV) UI will not load in 2.0 right now because it uses glamour who is being migrated. It will throw a warning because BlockContext is not in the image anymore. You can safely ignore it. I copied the configuration into the "untrusted" directory. Some benevolent soul could prepare a reliable configuration for 2.0 and copy it into MetaRepoForPharo20... :) Esteban On Feb 14, 2013, at 6:46 PM, "H. Hirzel" <hannes.hirzel(a)gmail.com> wrote: > Hello > > How do I load > > PetitParser > > into Pharo-2.0? > > There is no entry in the configuration browser. > > Regards > > Hannes >

Feb. 14, 2013

Re: [Pharo-project] Wow... we've closed a lot of issues

by Esteban Lorenzano

yeah we opened a lot. IMO, as you say, that's a good thing... you need to break things to create new things but there were at least 500 issues opened since the beginning of times that we have closed in this release. so, 500 "historical issues"plus all the issues we created (most of them non bugs,just modifications, enhancements, additions) not a bad number, I think :) cheers, Esteban On Feb 14, 2013, at 5:55 PM, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: > You must also ask how many issues you have created since 1.4, because > I doubt there are 1500 issues applying to 1.4. > It's not a critic. > It simply reflect the fact that it is necessary to temporarily break > things in order to change underlying system (events, text, browser > etc...) > Especially when the system has evolved up to a point where it is > fragile and hardly maintainable. > > Nicolas > > 2013/2/14 Sean P. DeNigris <sean(a)clipperadams.com>: >> I was just doing some housecleaning, deleting some old bug fix images and I >> was struck by how many important things we've accomplished. I'm still >> dumbfounded that a community of our size has resolved 1548 issues since 1.4. >> >> Cheers, >> Sean >> >> >> >> -- >> View this message in context: http://forum.world.st/Wow-we-ve-closed-a-lot-of-issues-tp4669901.html >> Sent from the Pharo Smalltalk mailing list archive at Nabble.com. >> >

Feb. 14, 2013

Re: [Pharo-project] OneClick for 2.0 available again

by Marcus Denker

Thanks, I changed the template and script -> correct .exe mentioned -> 2.0 instead of 1.0 The name of the .exe will change once again to "Pharo.exe" before we release. (We should then remember to update the README) On Feb 14, 2013, at 6:41 PM, "H. Hirzel" <hannes.hirzel(a)gmail.com> wrote: > Yes, Marcus, it is Pharo 2.0 now starts fine in MSWindows when > double-clicking on NBCog.exe. > > What now remains to be updated is the README.txt [1] > Replace > CogVM.exe --> NBCog.exe > > And probably remove 'Pharo 1.0' > > --Hannes > > > > [1] Current version or README.txt > > Pharo 1.0 > > - Mac: launch Pharo2.0-one-click.app > - Linux: launch Pharo2.0-one-click.sh > - Windows: launch CogVM.exe > > This distribution was built February 14, 2013. > > Please Note: > ------------ > > Pharo does not yet natively run on a 64bits Linux. Just install the > ia32/i386 runtime libraries and you should be fine. On Ubuntu and > other > debian-based systems, these libraries are named 'ia32-libs'. > > > On 2/11/13, Marcus Denker <marcus.denker(a)inria.fr> wrote: >> On Mon, Feb 11, 2013 at 12:32 PM, Torsten Bergmann <astares(a)gmx.de> wrote: >>> Hannes wrote >>>> Thorsten, >>>> >>>> reading what you have written, am I right that we just need a >>>> NBCog.ini >>>> >>>> file for MSwindows? >>> >>> We already have an INI file in the Pharo 2.0 one-click. >>> This INI file contains the correct entry about where to find the image >>> file. >>> >>> But it is named "CogVM.ini" instead of "NBCog.ini" since according to the >>> naming convention it has to be named like the VM (which is currently >>> NBCog.exe). >>> >>> So as I wrote just rename CogVM.ini into NBCog.ini and it works. >>> >>> >> I did this now and the next build should have the right ini >> >> Marcus >> >> -- >> Marcus Denker -- denker(a)acm.org >> http://www.marcusdenker.de >> >> >

Feb. 14, 2013

[Pharo-project] PetitParser in 2.0

by H. Hirzel

Hello How do I load PetitParser into Pharo-2.0? There is no entry in the configuration browser. Regards Hannes

Feb. 14, 2013

Re: [Pharo-project] Rectangle>>intersects: change Rationale

by Nicolas Cellier

Sure, any other point on the circle defined by this diameter would be a valid vertex... Nicolas 2013/2/14 Frank Shearar <frank.shearar(a)gmail.com>: > On 14 February 2013 17:27, Nicolas Cellier > <nicolas.cellier.aka.nice(a)gmail.com> wrote: >> I don't want nothing, but you inherit from very specialized class >> historically used to draw on pixel boundaries... >> A rectangle that you cannot even rotate, common ;) > > You can't define a unique rectangle with only two points anyway, if > you permit them to be rotated - consider the flipped version / > reflected in the line between the two points. > > When you add the constraint that their sides must be > horizontal/vertical you remove a degree of freedom, and then you get > unique rectangles. > > frank > >> Nicolas >> >> 2013/2/14 Igor Stasenko <siguctua(a)gmail.com>: >>> On 14 February 2013 18:10, Nicolas Cellier >>> <nicolas.cellier.aka.nice(a)gmail.com> wrote: >>>> That means you probably need to change some graphics primitives. >>>> >>>> If you want to go this way (have mathematical geometric objects), then >>>> I think you're on the step off Juan Morphic 3... >>>> >>> well, how you can reason about intersecting rectangles without using geometry? >>> >>> i do not want a rectangle which is not a rectangle in common sense, >>> but rectangle in "our" sense, >>> just because it fits us better. >>> >>> >>>> Nicolas >>> >>> >>> -- >>> Best regards, >>> Igor Stasenko. >>> >> >

Feb. 14, 2013

Re: [Pharo-project] Rectangle>>intersects: change Rationale

by Nicolas Cellier

2013/2/14 Igor Stasenko <siguctua(a)gmail.com>: > On 14 February 2013 18:27, Nicolas Cellier > <nicolas.cellier.aka.nice(a)gmail.com> wrote: >> I don't want nothing, but you inherit from very specialized class >> historically used to draw on pixel boundaries... >> A rectangle that you cannot even rotate, common ;) >> > > Well, from that perspective, then we should not touch anything and let > it stink forever :) > Sure, they still serves low abstractions for low requirements... Despite they are present in Fun Squeak, not everybody needs a tilted menu or window... However, given the variety of pixel density these days, non scalable graphics are going to be a problem. Nicolas > P.S. I think that we should revert the change to #intersects: > > > -- > Best regards, > Igor Stasenko. >

Feb. 14, 2013

Re: [Pharo-project] Rectangle>>intersects: change Rationale

by Frank Shearar

On 14 February 2013 17:27, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: > I don't want nothing, but you inherit from very specialized class > historically used to draw on pixel boundaries... > A rectangle that you cannot even rotate, common ;) You can't define a unique rectangle with only two points anyway, if you permit them to be rotated - consider the flipped version / reflected in the line between the two points. When you add the constraint that their sides must be horizontal/vertical you remove a degree of freedom, and then you get unique rectangles. frank > Nicolas > > 2013/2/14 Igor Stasenko <siguctua(a)gmail.com>: >> On 14 February 2013 18:10, Nicolas Cellier >> <nicolas.cellier.aka.nice(a)gmail.com> wrote: >>> That means you probably need to change some graphics primitives. >>> >>> If you want to go this way (have mathematical geometric objects), then >>> I think you're on the step off Juan Morphic 3... >>> >> well, how you can reason about intersecting rectangles without using geometry? >> >> i do not want a rectangle which is not a rectangle in common sense, >> but rectangle in "our" sense, >> just because it fits us better. >> >> >>> Nicolas >> >> >> -- >> Best regards, >> Igor Stasenko. >> >

Feb. 14, 2013

Re: [Pharo-project] OneClick for 2.0 available again

by H. Hirzel

Yes, Marcus, it is Pharo 2.0 now starts fine in MSWindows when double-clicking on NBCog.exe. What now remains to be updated is the README.txt [1] Replace CogVM.exe --> NBCog.exe And probably remove 'Pharo 1.0' --Hannes [1] Current version or README.txt Pharo 1.0 - Mac: launch Pharo2.0-one-click.app - Linux: launch Pharo2.0-one-click.sh - Windows: launch CogVM.exe This distribution was built February 14, 2013. Please Note: ------------ Pharo does not yet natively run on a 64bits Linux. Just install the ia32/i386 runtime libraries and you should be fine. On Ubuntu and other debian-based systems, these libraries are named 'ia32-libs'. On 2/11/13, Marcus Denker <marcus.denker(a)inria.fr> wrote: > On Mon, Feb 11, 2013 at 12:32 PM, Torsten Bergmann <astares(a)gmx.de> wrote: >> Hannes wrote >>>Thorsten, >>> >>>reading what you have written, am I right that we just need a >>> NBCog.ini >>> >>>file for MSwindows? >> >> We already have an INI file in the Pharo 2.0 one-click. >> This INI file contains the correct entry about where to find the image >> file. >> >> But it is named "CogVM.ini" instead of "NBCog.ini" since according to the >> naming convention it has to be named like the VM (which is currently >> NBCog.exe). >> >> So as I wrote just rename CogVM.ini into NBCog.ini and it works. >> >> > I did this now and the next build should have the right ini > > Marcus > > -- > Marcus Denker -- denker(a)acm.org > http://www.marcusdenker.de > >

Feb. 14, 2013