- Pharo-dev - lists.pharo.org

Re: [Pharo-dev] Spec Understandability

by HernÃ¡n Morales Durand

2014-10-21 15:20 GMT-03:00 kilon alios <kilon.alios(a)gmail.com>: > Why are you saying that. You can get halos on any morph, just halos do not >> show up when you do not want them. >> > > I am saying because of my experience with halos , I have not seen anything > new added to them or improved the last 2 years I am using Pharo. > > Did you ever look at the code of Halos? Halos are not defined per Morph >> but in a global tables. So we started to clean them. >> > > No why would I need to looks at the code of halos ? I am its user not its > developer. It great you clean the code of Halos but what that means to me > as user ? Not much I am afraid. > > Do you think that spec is nothing in terms of dev. Give me 1/1000 of >> the ressources put on one of this project and we will have a real UIBuilder. >> >> Stef >> >> > Thats the problem I see many very good projects like Spec , PharoLauncher, > Pillar, AsmJit, Nativeboost , Athens , Opal and the list goes on and on. > But I see no interest in having a UIBuilder. Apparently GUI designers are > very important for most IDEs out there and judging from the developers > using them I would say also most developers but is not for Pharo. Why ? > > You say you want the resources of 1/1000 do you really say that you need > more than 10 developers to make a UIBuilder ? Here examples of very small > development teams that have produced capable IDEs with GUI designers > included > > Building a good GUI designer is one of those problems where you cannot "solve" with 1-2 unexperienced programmers (and I am not talking about of confident programmers, I mean real experience). Is not about just painting widgets, is about layered APIs, application integration, extending specialized widgets, how to integrate second party widgets, portability, usability, etc. It is a domain where you cannot support the whole learning curve of a post-doc, and expect to see successful results. In short: You have to hire experienced developers. > http://gambas.sourceforge.net/en/main.html > > http://www.geany.org/Main/Authors > > http://www.lazarus.freepascal.org/ > > and the list goes on... I dont think that any of them had more than 1 -2 > and certainly they were not full time devs. Gambas alone is one developer > effort for the entire IDE, language, libraries, GUI designer etc etc. > > I didn't check but I wonder how they perform with resizing layouts? Do they use spacers? Do they use constraints? > Why such small development teams do dare to implement features that huge > IDEs do have ? Mainly because it usually comes down to small teams, take a > look at Visual Basic for example, I was reading its history recently and > found out that VB was no Microsoft invention (surprise - surprise) but > there was actually a very small company that create it as a product that > then they were hired from Microsoft. Delphi has a small development team as > well, actually it was single developer that leaded the effort and was > mainly responsible for its creation. > > So add this to one of my ranting and one of my complains. I love spec , I > love nativeboost , I love Playground and new GT tools but none of these > will ever start to compete with the usefulness of having a GUI designer > when up to 50% of a modern GUI based apps code is creating and managing > that GUI. > > The problem with Pharo's GUI designer is not that it sucks, its > undocumented , it severely lacks in features. But that it simple does not > exist. And without good GUI designer and an undocumented GUI API my remark > that on the GUI front Pharo is not in a good state is as honest as it can > be. > > Yes, and I prefer not having a GUI builder than having a bad one. > On the other hand on other fronts, like web interfacing, databases, > calling c libraries , vector graphics, unicode , true type fonts, JIT VM > even 64 bit Pharo has made a lot more significant progress. So apparently > Pharo resources wise seems to do great. Actually I am impressed how fast it > progress forward , its amazing. >

Oct. 21, 2014

Re: [Pharo-dev] Configurations and Github for a specialised configuration browswe

by kilon alios

Thank you Dale, this is more simple than I expected. >From what you linked it looks like by [:comitish] you mean the git branch .Which in my case is master so it should be Metacello new baseline: 'BaselineOfXXX'; repository: 'github://kilon/phEphestos:master/' get; load. https://github.com/kilon/phEphestos On Tue, Oct 21, 2014 at 9:52 PM, Dale Henrichs < dale.henrichs(a)gemtalksystems.com> wrote: > Kilon, > > You can use an expression like the following to load projects directly > from github using Metacello 1.0.0-beta.32.16: > > Metacello new > baseline: 'BaselineOfXXX'; > repository: 'github://<user name>/<project name>[:comitish]/<path to > filetree repo>' > get; > load. > > there are some examples of Seaside load expressions here[1]. Whenever you > use the `get` command, the github repo is refreshed from github. The `load` > loads directly from the local github cache for the project. > > Dale > > [1] https://github.com/GsDevKit/Seaside31#seaside31 > > > > On Tue, Oct 21, 2014 at 11:33 AM, kilon alios <kilon.alios(a)gmail.com> > wrote: > >> So I am so happy with how Pharo works with Github, that I decided to even >> port my configurations there and I also decided to create a small >> configuration browser which will subclass the existing configuration >> browser for installing tools for my project Ephestos so I dont have to >> pollute the existing configuration projects with all my sub-projects. >> >> In sort I want to host configurations to Github similarly to how >> MetaRepos are host in Smalltalkhub. I want the user to be able to install >> those configurations without a need to install git >> >> I asked this I think before and I was told how to, but it was part of >> another thread and it got lost inside my email abyss. >> >> So How I do that ? >> > >

Oct. 21, 2014

Re: [Pharo-dev] Configurations and Github for a specialised configuration browswe

by Dale Henrichs

Kilon, You can use an expression like the following to load projects directly from github using Metacello 1.0.0-beta.32.16: Metacello new baseline: 'BaselineOfXXX'; repository: 'github://<user name>/<project name>[:comitish]/<path to filetree repo>' get; load. there are some examples of Seaside load expressions here[1]. Whenever you use the `get` command, the github repo is refreshed from github. The `load` loads directly from the local github cache for the project. Dale [1] https://github.com/GsDevKit/Seaside31#seaside31 On Tue, Oct 21, 2014 at 11:33 AM, kilon alios <kilon.alios(a)gmail.com> wrote: > So I am so happy with how Pharo works with Github, that I decided to even > port my configurations there and I also decided to create a small > configuration browser which will subclass the existing configuration > browser for installing tools for my project Ephestos so I dont have to > pollute the existing configuration projects with all my sub-projects. > > In sort I want to host configurations to Github similarly to how MetaRepos > are host in Smalltalkhub. I want the user to be able to install those > configurations without a need to install git > > I asked this I think before and I was told how to, but it was part of > another thread and it got lost inside my email abyss. > > So How I do that ? >

Oct. 21, 2014

Re: [Pharo-dev] Inspector for Opal IR -> source mapping

by kilon alios

great work Marcus , look like GT tools have a huge potential . Would that work with a debugger too ? On Tue, Oct 21, 2014 at 2:53 PM, Marcus Denker <marcus.denker(a)inria.fr> wrote: > This is now in update 317 > > > > > > > On Tue, Oct 21, 2014 at 11:19 AM, Marcus Denker <marcus.denker(a)inria.fr> > wrote: > >> Hi, >> >> There is now special inspector view for the Opal IR: >> >> >> <IRInspector.png> >> >> TODO: show how it maps to Bytecode and AST. >> >> https://pharo.fogbugz.com/f/cases/14241/add-a-inspector-view-for-IRMethods >> >> >> >

Oct. 21, 2014

Configurations and Github for a specialised configuration browswe

by kilon alios

So I am so happy with how Pharo works with Github, that I decided to even port my configurations there and I also decided to create a small configuration browser which will subclass the existing configuration browser for installing tools for my project Ephestos so I dont have to pollute the existing configuration projects with all my sub-projects. In sort I want to host configurations to Github similarly to how MetaRepos are host in Smalltalkhub. I want the user to be able to install those configurations without a need to install git I asked this I think before and I was told how to, but it was part of another thread and it got lost inside my email abyss. So How I do that ?

Oct. 21, 2014

Re: [Pharo-dev] Spec Understandability

by kilon alios

> > Why are you saying that. You can get halos on any morph, just halos do not > show up when you do not want them. > I am saying because of my experience with halos , I have not seen anything new added to them or improved the last 2 years I am using Pharo. Did you ever look at the code of Halos? Halos are not defined per Morph but > in a global tables. So we started to clean them. > No why would I need to looks at the code of halos ? I am its user not its developer. It great you clean the code of Halos but what that means to me as user ? Not much I am afraid. Do you think that spec is nothing in terms of dev. Give me 1/1000 of the > ressources put on one of this project and we will have a real UIBuilder. > > Stef > > Thats the problem I see many very good projects like Spec , PharoLauncher, Pillar, AsmJit, Nativeboost , Athens , Opal and the list goes on and on. But I see no interest in having a UIBuilder. Apparently GUI designers are very important for most IDEs out there and judging from the developers using them I would say also most developers but is not for Pharo. Why ? You say you want the resources of 1/1000 do you really say that you need more than 10 developers to make a UIBuilder ? Here examples of very small development teams that have produced capable IDEs with GUI designers included http://gambas.sourceforge.net/en/main.html http://www.geany.org/Main/Authors http://www.lazarus.freepascal.org/ and the list goes on... I dont think that any of them had more than 1 -2 and certainly they were not full time devs. Gambas alone is one developer effort for the entire IDE, language, libraries, GUI designer etc etc. Why such small development teams do dare to implement features that huge IDEs do have ? Mainly because it usually comes down to small teams, take a look at Visual Basic for example, I was reading its history recently and found out that VB was no Microsoft invention (surprise - surprise) but there was actually a very small company that create it as a product that then they were hired from Microsoft. Delphi has a small development team as well, actually it was single developer that leaded the effort and was mainly responsible for its creation. So add this to one of my ranting and one of my complains. I love spec , I love nativeboost , I love Playground and new GT tools but none of these will ever start to compete with the usefulness of having a GUI designer when up to 50% of a modern GUI based apps code is creating and managing that GUI. The problem with Pharo's GUI designer is not that it sucks, its undocumented , it severely lacks in features. But that it simple does not exist. And without good GUI designer and an undocumented GUI API my remark that on the GUI front Pharo is not in a good state is as honest as it can be. On the other hand on other fronts, like web interfacing, databases, calling c libraries , vector graphics, unicode , true type fonts, JIT VM even 64 bit Pharo has made a lot more significant progress. So apparently Pharo resources wise seems to do great. Actually I am impressed how fast it progress forward , its amazing.

Oct. 21, 2014

Re: [Pharo-dev] Spec Understandability

by Ben Coman

Ralph Boland wrote: > > Besides error intractability, there is also a cost to understanding the > > system via exploration - a glorious benefit of Morphic. In the past, one > > only had to bring up halos on and inspect a button to see how to > duplicate > > its action. Now, to see what happens when one presses Versionner's > "Commit > > the project" button, you have to drill down in the inspector from > > aButtonXyzMorph -> model -> actionHolder -> value, a tedious and opaque > > process. > > > Of course, I'm not in any way against Spec or its goals, just sharing my > > unfolding experience using it. > > I'm running Squeak 4.0 but I assume my question still applies. If not, > other > Pharo users might want to know. > > I am not in the habit of using halos, not really sure what they are > for. But your description implies they are quite useful. > What does "bring up halos on and inspect a button to see how to duplicate > > its action" mean? Could you please describe in step by step detail. > > Ralph Boland Say you wanted your application to add a new package to the system. So you wonder how would to do that. Well... the system already does it! Right click in the first pane of the System Browser and you get a menu it <Add package...>. So why not just copy how that does it? But how do you look under the hood to see how it works you ask? Halos to the rescue! Hold the key combination to invoke halos and click on the menu item (varies between platforms, mac is <shift-option>, win is <shift-alt>) so that "ToggleMenuItemMorph" shows as the object type the halos are related to. On the right click the spanner "Debug" icon, and <inspect morph>. Scroll down looking for something that looks like a relevant method call (in this case 'target' contains [ target addPackge ]. #addPackage looks a likely candidate, so browse implementors and put a "self halt" at the top of the methods. Then back in the System Browser, invoke the <Add package...> menu item, and the debugger appears. Now you can trace through the execution of the system code that creates package. As an example, nothing beats real-life code. ----- Another use I have is sometimes to go World > System, use the Duplicate Halo on the <Enable halt/inspect once> menu item to drag it onto the background and using it as a button. HTH -ben

Oct. 21, 2014

Re: [Pharo-dev] Spec Understandability

by stepharo

On 21/10/14 14:20, kilon alios wrote: > I may have but then halos have been abandoned in pharo at least > functionality wise , Why are you saying that. You can get halos on any morph, just halos do not show up when you do not want them. > I have not seen any improvement to them from when I started using > Squeak 2 years ago and if I remember there were once a time you had to > enable them in pharo settings. Did you ever look at the code of Halos? Halos are not defined per Morph but in a global tables. So we started to clean them. > Also both a gui builder for squeak and another for pharo have been > abandonware . So it looks pharo is not doing well on GUI front > especially when it has to compete with gui designers in visual basic, > visual studio, QT, GTK and many more. Do you think that spec is nothing in terms of dev. Give me 1/1000 of the ressources put on one of this project and we will have a real UIBuilder. Stef > > On Tue, Oct 21, 2014 at 2:01 PM, Ben Coman <btc(a)openinworld.com > <mailto:btc@openinworld.com>> wrote: > > kilon alios wrote: > > For me there is nothing that sucks more than having to type > code to create a GUI in year 2014.... This a strong reason > why I miss my days coding in Delphi so much. > at least halos try to compensate for that partly.I love halos > please dont remove them, good thing they suck for you, they > rock for me. > > > I think you misread it. The key word is the "implementation" sucks. > It seems Bloc is cleaning the implementation, not removing the > functionality. > > -ben > > The fact that I can close a problematic morph that refuses to > exit, by bringing its halo and clicking x , is enough reason > to keep them around for me. > I love the idea of halos the idea that you can hack any gui > with simple actions, with no code invloved, and resynthesised > it is definetly something I find immensely useful though the > implementation of halos is far from ideal. > On Tue, Oct 21, 2014 at 9:18 AM, stepharo <stepharo(a)free.fr > <mailto:stepharo@free.fr> <mailto:stepharo@free.fr > <mailto:stepharo@free.fr>>> wrote: > > Note that halo implementation sucks! > In Bloc alain clean it and he is able to plug halos and > package them > cleanly. > > In addition in Bloc he took the decision not to implement > for now > the direct creation of certain widget because > this kind of feature leads to spaghetti dependencies. > > Stef > > > > On 20/10/14 20:35, Sean P. DeNigris wrote: > > Ralph Boland wrote > > I am not in the habit of using halos, not really > sure what > they are for. > But your description implies they are quite useful. > What does "bring up halos on and inspect a button > to see > how to duplicate > its action" mean? Could you please describe in > step by step > detail. > > Hi Ralph! I think the best way to understand why > Morphic is > revolutionary is > to read the papers of its creators (e.g. > http://web.media.mit.edu/~__jmaloney/papers/__DirectnessAndLivenessInMorphi… > <http://web.media.mit.edu/%7E__jmaloney/papers/__DirectnessAndLivenessInMorp…> > > <http://web.media.mit.edu/~jmaloney/papers/DirectnessAndLivenessInMorphic.pdf > <http://web.media.mit.edu/%7Ejmaloney/papers/DirectnessAndLivenessInMorphic.…>>). > I also wrote http://stackoverflow.com/a/__4410189/424245 > <http://stackoverflow.com/a/4410189/424245> which > should give a > taste and also answers your question more directly. HTH. > > > > ----- > Cheers, > Sean > -- > View this message in context: > http://forum.world.st/Spec-__Understandability-__tp4778370p4785597.html > > <http://forum.world.st/Spec-Understandability-tp4778370p4785597.html> > Sent from the Pharo Smalltalk Developers mailing list > archive at > Nabble.com. > > > > > > > >

Oct. 21, 2014

Re: [Pharo-dev] Spec Understandability

by kilon alios

I may have but then halos have been abandoned in pharo at least functionality wise , I have not seen any improvement to them from when I started using Squeak 2 years ago and if I remember there were once a time you had to enable them in pharo settings. Also both a gui builder for squeak and another for pharo have been abandonware . So it looks pharo is not doing well on GUI front especially when it has to compete with gui designers in visual basic, visual studio, QT, GTK and many more. On Tue, Oct 21, 2014 at 2:01 PM, Ben Coman <btc(a)openinworld.com> wrote: > kilon alios wrote: > >> For me there is nothing that sucks more than having to type code to >> create a GUI in year 2014.... This a strong reason why I miss my days >> coding in Delphi so much. >> at least halos try to compensate for that partly.I love halos please dont >> remove them, good thing they suck for you, they rock for me. >> > > I think you misread it. The key word is the "implementation" sucks. > It seems Bloc is cleaning the implementation, not removing the > functionality. > > -ben > > The fact that I can close a problematic morph that refuses to exit, by >> bringing its halo and clicking x , is enough reason to keep them around for >> me. >> I love the idea of halos the idea that you can hack any gui with simple >> actions, with no code invloved, and resynthesised it is definetly >> something I find immensely useful though the implementation of halos is far >> from ideal. >> On Tue, Oct 21, 2014 at 9:18 AM, stepharo <stepharo(a)free.fr <mailto: >> stepharo(a)free.fr>> wrote: >> >> Note that halo implementation sucks! >> In Bloc alain clean it and he is able to plug halos and package them >> cleanly. >> >> In addition in Bloc he took the decision not to implement for now >> the direct creation of certain widget because >> this kind of feature leads to spaghetti dependencies. >> >> Stef >> >> >> >> On 20/10/14 20:35, Sean P. DeNigris wrote: >> >> Ralph Boland wrote >> >> I am not in the habit of using halos, not really sure what >> they are for. >> But your description implies they are quite useful. >> What does "bring up halos on and inspect a button to see >> how to duplicate >> its action" mean? Could you please describe in step by step >> detail. >> >> Hi Ralph! I think the best way to understand why Morphic is >> revolutionary is >> to read the papers of its creators (e.g. >> http://web.media.mit.edu/~__jmaloney/papers/__ >> DirectnessAndLivenessInMorphic__.pdf >> <http://web.media.mit.edu/~jmaloney/papers/ >> DirectnessAndLivenessInMorphic.pdf>). >> I also wrote http://stackoverflow.com/a/__4410189/424245 >> <http://stackoverflow.com/a/4410189/424245> which should give a >> taste and also answers your question more directly. HTH. >> >> >> >> ----- >> Cheers, >> Sean >> -- >> View this message in context: >> http://forum.world.st/Spec-__Understandability-__ >> tp4778370p4785597.html >> <http://forum.world.st/Spec-Understandability- >> tp4778370p4785597.html> >> Sent from the Pharo Smalltalk Developers mailing list archive at >> Nabble.com. >> >> >> >> >> >> > >

Oct. 21, 2014

[pharo-project/pharo-core] 695fbf: 40317

by GitHub

Branch: refs/heads/4.0 Home: https://github.com/pharo-project/pharo-core Commit: 695fbfc73e087100e61aeb64753b897c920570a7 https://github.com/pharo-project/pharo-core/commit/695fbfc73e087100e61aeb64… Author: Jenkins Build Server <board(a)pharo-project.org> Date: 2014-10-21 (Tue, 21 Oct 2014) Changed paths: M AST-Core.package/RBIdentifierToken.class/definition.st A AST-Core.package/RBIdentifierToken.class/instance/accessing/stop.st A AST-Core.package/RBIdentifierToken.class/instance/accessing/stop_.st M AST-Core.package/RBKeywordToken.class/definition.st A AST-Core.package/RBKeywordToken.class/instance/accessing/stop.st A AST-Core.package/RBKeywordToken.class/instance/accessing/stop_.st A AST-Core.package/RBLiteralArrayNode.class/instance/accessing/selectorParts.st A AST-Core.package/RBLiteralArrayNode.class/instance/accessing/sourcePointer_.st A AST-Core.package/RBLiteralArrayNode.class/instance/accessing/stop_.st A AST-Core.package/RBMessageNode.class/instance/accessing/sourcePointer_.st A AST-Core.package/RBMessageNode.class/instance/accessing/stop_.st M AST-Core.package/RBMethodNode.class/instance/accessing/start.st M AST-Core.package/RBMethodNode.class/instance/accessing/stop.st A AST-Core.package/RBProgramNode.class/instance/accessing/selectorParts.st M AST-Core.package/RBProgramNode.class/instance/accessing/sourceInterval.st A AST-Core.package/RBProgramNode.class/instance/accessing/sourcePointer_.st A AST-Core.package/RBProgramNode.class/instance/accessing/stop_.st A AST-Core.package/RBProgramNode.class/instance/accessing/token.st A AST-Core.package/RBProgramNode.class/instance/replacing/calculateInnerCommentsSize_.st A AST-Core.package/RBProgramNode.class/instance/replacing/calculateSourceInterval_.st M AST-Core.package/RBProgramNode.class/instance/replacing/replaceWith_.st A AST-Core.package/RBToken.class/instance/accessing/sourcePointer_.st M AST-Tests-Core.package/RBProgramNodeTest.class/instance/testing-replacing/testReplaceLiteral.st M AST-Tests-Core.package/RBProgramNodeTest.class/instance/testing-replacing/testReplaceLiteralArray.st M AST-Tests-Core.package/RBProgramNodeTest.class/instance/testing-replacing/testReplaceMessage.st M AST-Tests-Core.package/RBProgramNodeTest.class/instance/testing-replacing/testReplaceMessageArgument.st M AST-Tests-Core.package/RBProgramNodeTest.class/instance/testing-replacing/testReplaceMessageReceiver.st M AST-Tests-Core.package/RBProgramNodeTest.class/instance/testing-replacing/testReplaceVariable.st A OpalCompiler-Core.package/IRInstruction.class/instance/inspector/children.st A OpalCompiler-Core.package/IRInstruction.class/instance/inspector/sourceInterval.st A OpalCompiler-Core.package/IRInstruction.class/instance/printing/printOn_.st A OpalCompiler-Core.package/IRMethod.class/instance/inspector/children.st R OpalCompiler-Core.package/IRMethod.class/instance/inspector/gtInspectorIrIn_.st M OpalCompiler-Core.package/IRPrinterV2.class/instance/visiting/visitPushClosureCopy_.st A OpalCompiler-Core.package/IRSequence.class/instance/inspector/children.st A OpalCompiler-Core.package/IRSequence.class/instance/inspector/sourceInterval.st A OpalTools.package/GTIRBrowser.class/README.md A OpalTools.package/GTIRBrowser.class/definition.st A OpalTools.package/GTIRBrowser.class/instance/building/compose.st A OpalTools.package/GTIRBrowser.class/instance/building/sourceIn_.st A OpalTools.package/GTIRBrowser.class/instance/building/treeIn_.st A OpalTools.package/extension/IRInstruction/instance/gtInspectorSourceCodeIn_.st A OpalTools.package/extension/IRMethod/instance/gtInspectorBytecodeIn_.st A OpalTools.package/extension/IRMethod/instance/gtInspectorIrIn_.st A OpalTools.package/extension/IRMethod/instance/gtInspectorSourceCodeIn_.st A OpalTools.package/extension/IRMethod/instance/gtInspectorTreeIn_.st A OpalTools.package/extension/IRSequence/instance/gtInspectorSourceCodeIn_.st A ScriptLoader40.package/ScriptLoader.class/instance/pharo - scripts/script317.st A ScriptLoader40.package/ScriptLoader.class/instance/pharo - updates/update40317.st M ScriptLoader40.package/ScriptLoader.class/instance/public/commentForCurrentUpdate.st Log Message: ----------- 40317 14241 add a inspector view for IRMethods https://pharo.fogbugz.com/f/cases/14241 14254 AST method replaceWith: does not change source interval https://pharo.fogbugz.com/f/cases/14254 http://files.pharo.org/image/40/40317.zip

Oct. 21, 2014