August 2014 - Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Latch for Pharo (Opinion needed)

by garduino

Thanks Phil! Your opinion is very encouraging to me! Cheers. -- View this message in context: http://forum.world.st/Latch-for-Pharo-Opinion-needed-tp4775525p4775536.html Sent from the Pharo Smalltalk Developers mailing list archive at Nabble.com.

Aug. 31, 2014

Re: [Pharo-dev] SqNumberParser and PGConnection in Pharo 3

by Esteban A. Maringolo

Yes... it was intentional. Because there is no "official" maintainer of the GLORP package, I didn't want to add stuff to the main codebase. Same story as with the PostgresV2 package. Regards! Esteban A. Maringolo 2014-08-31 19:16 GMT-03:00 Sven Van Caekenberghe <sven(a)stfx.eu>: > > On 31 Aug 2014, at 12:27, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > >>> ps: There is also other versions of GLORP I didn't bless as #stable. >> >> That seems a bit more complicated: the Glorp ancestry is mixed up, we should merge this carefully: >> >> you made versions 86 & 87 while those already existed, the latest 88 skips your version. >> >> in any case, 88 now works for me, but I fear it skips your code > > Ah, I see, this was an intentional branch, right ? > > Anyway, your addition seems non-intrusive, since it only adds a type, which does not get used unless one specifically asks for it, correct ? > > If so, I can merge this, commit a new version and update the config. > > Sven

Aug. 31, 2014

Re: [Pharo-dev] SqNumberParser and PGConnection in Pharo 3

by Sven Van Caekenberghe

On 31 Aug 2014, at 12:27, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: >> ps: There is also other versions of GLORP I didn't bless as #stable. > > That seems a bit more complicated: the Glorp ancestry is mixed up, we should merge this carefully: > > you made versions 86 & 87 while those already existed, the latest 88 skips your version. > > in any case, 88 now works for me, but I fear it skips your code Ah, I see, this was an intentional branch, right ? Anyway, your addition seems non-intrusive, since it only adds a type, which does not get used unless one specifically asks for it, correct ? If so, I can merge this, commit a new version and update the config. Sven

Aug. 31, 2014

Re: [Pharo-dev] Experiments in Code Typography

by Alain Rastoul

About AST, we did something at work in dotnet for syntax checking when typing SQL requests in a tool (not only simple statements, some of the developers requests are real monsters, iirc one of the biggest was 42k), and we built the ast on the fly at each typing, it was incredibly fast. I was sceptic at the beginning, but given the power of current cpu, few milliseconds between key strokes is more than enough. Perhaps it could/should be done only on "special" characters like spaces ;[]() -> and so on, or only on a special keymapping (reformating at each keystroke could be very disturbing).

Aug. 31, 2014

Re: [Pharo-dev] Latch for Pharo (Opinion needed)

by phil＠highoctane.be

Looks very interesting and innovative. As a Seasider, I welcome that! Phil. Le 31 aoÃ»t 2014 23:32, "GermÃ¡n Arduino" <garduino(a)gmail.com> a Ã©crit : > Hi Guys: > > I'm working in a project to write a SDK for use Latch with Pharo. Latch [1] > is a product/service to add a degree of security to web applications and is > developed and marketed by 11Paths [2] , a company from TelefÃ³nica. > > It provides very interesting security services [3] but they do not have a > Smalltalk SDK (Only Ruby, PHP, Java, .NET, Python, C and PowerShell) then I > think that I could write a Pharo SDK and a reference (example) > implementation in Seaside, to bring to the Pharo community an interesting > tool to add more protection to the Pharo web applications. > > Do you think that this SDK could be of interest for the community? Are you > using other tools different of Latch to add security to your web > applications? > > Your opinion will be of great help. > > Thanks. > > [1] https://latch.elevenpaths.com/www/index.html > [2] https://www.elevenpaths.com/index.html > [3] https://latch.elevenpaths.com/www/why.html > > -- > Saludos / Regards, > GermÃ¡n Arduino > www.arduinosoftware.com >

Aug. 31, 2014

Latch for Pharo (Opinion needed)

by GermÃ¡n Arduino

Hi Guys: I'm working in a project to write a SDK for use Latch with Pharo. Latch [1] is a product/service to add a degree of security to web applications and is developed and marketed by 11Paths [2] , a company from TelefÃ³nica. It provides very interesting security services [3] but they do not have a Smalltalk SDK (Only Ruby, PHP, Java, .NET, Python, C and PowerShell) then I think that I could write a Pharo SDK and a reference (example) implementation in Seaside, to bring to the Pharo community an interesting tool to add more protection to the Pharo web applications. Do you think that this SDK could be of interest for the community? Are you using other tools different of Latch to add security to your web applications? Your opinion will be of great help. Thanks. [1] https://latch.elevenpaths.com/www/index.html [2] https://www.elevenpaths.com/index.html [3] https://latch.elevenpaths.com/www/why.html -- Saludos / Regards, GermÃ¡n Arduino www.arduinosoftware.com

Aug. 31, 2014

Re: [Pharo-dev] Contributions about Pharo and Moose

by Alain Rastoul

Thank you Stef for your answers. For Moose, I will search on code.google, and about loading it in Pharo4, i tried the development version (since Pharo4 is in development I assumed the latest Moose configuration was needed), if a ci job can load it , sure that the problem is me ... wouldn't be the first time btw :) And you are right, there is no confirmation email (I was stupidely waiting for it, as it was stated on the web page iirc) but I got connected without. Great! Regards, Alain

Aug. 31, 2014

Re: [Pharo-dev] Experiments in Code Typography

by stepharo

On 31/8/14 21:47, Thierry Goubier wrote: > Something like that, yes. > > Or instead of an heuristic to get an ast, have an editor which has the > logic of editing the ast (instead of typing text which will be rightly > or wrongly matched to a tentative ast). I do not think it will work because this is really constrained and typing code is not that constrained. What would be nice is to have a set of rules that help terminating a tree to the closest one. ie | -> | | x isNil ifTrue: => x isNil ifTrue: [] ... and to format only what has been typing but in presence of the complete tree. > I'm looking into doing some stuff about asts and editing for > SmaCC-generated asts. Some stuff about having core building blocks for > non-smalltalk IDE written in Pharo. > > Thierry > > Le 31/08/2014 21:14, stepharo a Ã©crit : >> You know I would like to have >> >> format as you type. >> >> if someone wants to work on that it would be great. >> The point is to have some heuristics to get as fast as possible a >> "complete" ast. >> And pretty print it for example each time a space is added. >> >> stef >> >> >> > > > >

Aug. 31, 2014

Re: [Pharo-dev] Experiments in Code Typography

by Thierry Goubier

Something like that, yes. Or instead of an heuristic to get an ast, have an editor which has the logic of editing the ast (instead of typing text which will be rightly or wrongly matched to a tentative ast). I'm looking into doing some stuff about asts and editing for SmaCC-generated asts. Some stuff about having core building blocks for non-smalltalk IDE written in Pharo. Thierry Le 31/08/2014 21:14, stepharo a Ã©crit : > You know I would like to have > > format as you type. > > if someone wants to work on that it would be great. > The point is to have some heuristics to get as fast as possible a > "complete" ast. > And pretty print it for example each time a space is added. > > stef > > >

Aug. 31, 2014

Re: [Pharo-dev] Having some trouble on the Pharo Contributions CI

by Sven Van Caekenberghe

Hey Stef, In the mean time I got my CI job working, there were some errors on my part in there as well (as usual ;-) The question remains how to detect that you are running on CI, because I see this: + ./pharo Pharo.image eval OSPlatform current environment at: #CI ifAbsent: [#none] #none + export CI=ci.inria.fr + ./pharo Pharo.image eval OSPlatform current environment at: #CI ifAbsent: [#none] 'ci.inria.fr' So, $CI is not set by default ... Sven On 31 Aug 2014, at 21:22, stepharo <stepharo(a)free.fr> wrote: > hi sevn > > christophe will certainly reply to you tomorrow morning. > > stef > On 31/8/14 17:36, Sven Van Caekenberghe wrote: >> Hi, >> >> I seem to be having some trouble on the Pharo Contributions CI. I successfully created some projects before, but this time something strange is happening. >> >> It seems as if a win7 slave is doing the build instead of the linux slaves that I am used to. >> >> Has something changed ? >> >> I want to detect that I am running on the CI so that I can skip some tests that require resources not available there, but I fear the following won't work on Windows: >> >> runningOnInriaCI >> | platform ci | >> ^ [ >> platform := OSPlatform current. >> ci := platform environment at: #CI ifAbsent: [ '' ]. >> ci = 'ci.inria.fr' ] >> on: MessageNotUnderstood >> do: [ false ] >> >> What is the recommended way to use linux-only slaves ? >> >> Any other ideas ? >> >> Thanks, >> >> Sven >> >> > >

Aug. 31, 2014