August 2018 - Pharo-users - lists.pharo.org

When we create a missing method in debugger - why does it end with a dot?
by Tim Mackinnon Aug. 30, 2018

Aug. 30, 2018

I am looking at porting my missing class creation code into Pharo 7, and I notice there is a DynamicMessageImplementor that does this for methods - but Iâm wondering why when we create a missing method with âself souldBeImplemented.â with a trailing â.â? The Smalltalk style guide normally says the last statement shouldnât have a dot - and I notice some recent cleanups removed some dots on methods (presumably the formatter). This is a bit of a special case as the intent is to replace that method - but shouldnât we create without the dot to set a good example anyway? If so, I can include that change in my pr. Tim

3 2

Latest Pharo 7 image doesn't give normal debugger on a method with undefined class method - get System Error
by Tim Mackinnon Aug. 30, 2018

Aug. 30, 2018

I was going to look at testing my Create class in the debugger fixes however the latest image behaves in a manner Iâve not seen before. In a fresh P7 image - build 1198 64bit osx (with latest vm downloaded by launcher) - I have created a new class in a new package TimTest:TimTest>>testMethod "self new testMethod" FooObject new someMethod Where FooObject and someMethod are undefined and when I saved the method I left them as undefined. I then evaluated - TimeTest new testMethod in a playground and I get a white screen with *** System error handling failed *** (screenshot below) In an earlier image (build 1167) this wasnât the case? Has something been broken? Tim

1 1

Tracing Method Calls
by Arturo Zambrano Aug. 29, 2018

Aug. 29, 2018

Hi all, I have a piece of undocumented software that I need understand ... One thing I would like to do is to have a complete trace of method calls (receivers, parameters) for certain scenarios. As an "aspects"guy, my first idea was to use Phantom, but it seems not active anymore. Could someone please point me to some tools that I could use for this? (using Pharo 6, 6.1 or 7) Thanks in advance Cheers Arturo

3 4

Jupyter notebooks / Dynabook / JupyterTalk
by H. Hirzel Aug. 29, 2018

Aug. 29, 2018

Hello What is the status of this project? https://github.com/jmari/JupyterTalk Hannes

9 48

ssh client for Pharo
by Andrei Stebakov Aug. 29, 2018

Aug. 29, 2018

Is there an ssh client for Pharo just like net-ssh for Ruby? Thanks!

2 3

Project based learning and Dynabook
by Hilaire Aug. 29, 2018

Aug. 29, 2018

Hi, Some though on the topic http://blog.drgeo.eu/post/2018/Project-based-learning Hilaire -- Dr. Geo http://drgeo.eu

1 0

How do I get ZeroMQ working?
by H. Hirzel Aug. 29, 2018

Aug. 29, 2018

Hello I installed JupyterTalk [1]. With it comes a ZeroMQ implementation [2]. This was done on Ubuntu 18.04.1 (Bionic Beaver, 64bit). However the ZeroMQ tests all fail. I assume there is something missing in the installation. How can I test that? Any other suggestions? Kind regards Hannes [1] JupyterTalk (https://github.com/jmari/JupyterTalk) is a Smalltalk kernel for the Jupyter notebook system: http://jupyter.org/ [2] ZeroMQ is a high-performance asynchronous messaging library, aimed at use in distributed or concurrent applications. It provides a message queue, but unlike message-oriented middleware, a ZeroMQ system can run without a dedicated message broker. https://en.wikipedia.org/wiki/ZeroMQ

4 12

Dynabook and authentic learning pedagogy
by Hilaire Aug. 29, 2018

Aug. 29, 2018

Hi, A few notes on how Dynabook and authentic learning methodology could be related. http://blog.drgeo.eu/post/2018/Authentic-learning Hilaire -- Dr. Geo http://drgeo.eu

4 7

Re: [Pharo-users] How best to cope with 2 editor/browser frameworks - Calypso and RubSmalltalkEditor
by Tim Mackinnon Aug. 29, 2018

Aug. 29, 2018

Denis - Iâve hopefully corruptly issues a PR on SystemCommands that shows how I approached this. I think its come out quite neat - but it will need the lower level Rubric based PR to be approved first (and then Calypso will have some useful jump to next keyword and intelligently widen selection commands). Tim > On 21 Aug 2018, at 20:12, Tim Mackinnon <tim(a)mackinnon.email> wrote: > > Thanks Denis - I figured that might be the case. As it turns out - the text selection protocol for the 2 editors is different so I can delegate that to a code block and put it in the > Code mode. > > Anyway Iâll figure it out and show you guys in a or (Iâm almost there) > > Tim > > Sent from my iPhone > > On 21 Aug 2018, at 17:16, Denis Kudriashov <dionisiydk(a)gmail.com <mailto:dionisiydk@gmail.com>> wrote: > >> Hi Tim. >> >> Currently commands are loaded after rubric. So RubSmalltalkEditor can't use them directly. >> >> 2018-08-21 13:03 GMT+01:00 Tim Mackinnon <tim(a)testit.works <mailto:tim@testit.works>>: >> AH - this makes sense then that SystemCommands is in its own Repo then (although all these tiny repos get quite difficult the track so you can understand where/how to contribute). >> >> This leads to another question - if I stick my commands for extend selection and jump to keyword in SystemCommands - is the load order such that RubSmalltalkEditor can safely access those commands (I know that for Calypso its ok, but can I reference those commands in the older RubSmalltalkEditor). Otherwise I will need to put the guts of my commands in the RubSmalltalkCodeMode which is shared by al editors (but it feels a bit too generic place for textMorph navigation stuff). >> >> Tim >> >> >> >>> On 21 Aug 2018, at 05:48, Esteban Lorenzano <estebanlm(a)gmail.com <mailto:estebanlm@gmail.com>> wrote: >>> >>> Hi, >>> >>>> On 20 Aug 2018, at 00:39, Denis Kudriashov <dionisiydk(a)gmail.com <mailto:dionisiydk@gmail.com>> wrote: >>>> >>>> >>>> 2018-08-19 23:12 GMT+01:00 Tim Mackinnon <tim(a)testit.works <mailto:tim@testit.works>>: >>>> Thanks Denis - I guess for now I can put the methods in RubSmalltalkEditor so that the playground can operate and then my Calypso plugins can also reference that code too (if class methods). >>>> >>>> Is the idea that playground will become part of Calypso too? >>>> >>>> No. >>>> What we can try to achieve is to reused Commander everywhere. So these SystemCommands will be reused by different tools. >>> >>> Yes, this is what we are targeting, to have a centralised command library that manages all system commands (so we do not need to duplicate code all around). >>> With enough time to implement, all tools will use same command mechanism. >>> >>> Esteban >>> >>>> >>>> If so, I can add the methods I need as an extension (although I guess this wonât work as the keyboard mapping for playground in one method and not extensible like Calypso which uses pragmas). >>>> >>>> Iâll do something as a pr and we can figure it out. >>>> >>>> Itâs worth saying, that for my Calypso plugins Iâve had to abuse your mechanism a bit as the simple #execute mechanism doesnât give me access to the text morph to control the cursor. I was able to do it though via the prompting mechanism (I just donât show any ui prompt). >>>> >>>> Tim >>>> >>>> Sent from my iPhone >>>> >>>> On 19 Aug 2018, at 14:28, Denis Kudriashov <dionisiydk(a)gmail.com <mailto:dionisiydk@gmail.com>> wrote: >>>> >>>>> Hi. >>>>> >>>>> I dont't know answer to your question. But for the note: >>>>> Calypso was needed to override existing way to spawn implementors/senders and so on. So I added subclass of RubSmalltalkEditor - ClyTextEditor which overrides required methods. >>>>> So if you will put new methods into the RubSmalltalkEditor then Calypso will be able to use them. >>>>> >>>>> 2018-08-19 16:55 GMT+01:00 Tim Mackinnon <tim(a)testit.works <mailto:tim@testit.works>>: >>>>> Hi Iâm trying to work out the best way to cope with the fact that we are in a hybrid place where the Playground is using an RubSmalltalkEditor but Calypso has its own mechanisms for editing code. >>>>> >>>>> I have commands that should/can work in both the playground as well as code method editor (in Calypso). >>>>> >>>>> Previously my code lived in RubSmalltalkEditor and Nautilus was using that too - but now Iâm not sure where to put some reusable methods that use RBParser and the AST to find the best place to position your cursor. >>>>> >>>>> I can make a class method on RubSmalltalkEditor and Calypso commands could reference that (but that feels wrong somehow). I could make some extension methods on RBParser and both places reference that - but then who should own the extension method (as I think RubSmalltalkEditor is lower level and Calypso is then loaded on top). >>>>> >>>>> Anyone have some good suggestions? >>>>> >>>>> Tim >>>>> >>>> >>> >> >>

1 0

Re: [Pharo-users] [Moose-dev] Re: Centering edges over composites
by HernÃ¡n Morales Durand Aug. 29, 2018

Aug. 29, 2018

Hi Chris, I attached a screenshot of the desired result. It seems to me the problem is not where the arrow is connected, but the Tree Layout, which is positioning the "Test" label as if it would be the root of a tree with the svg figures as their leaves. I still don't get how to apply a vertical layout (not tree) to a RTCompositeShape, assuming that could be a reasonable solution. Cheers, HernÃ¡n 2018-08-28 18:25 GMT-03:00 Chris Cunningham <cunningham.cb(a)gmail.com>: > Hi, > I looks like the arrow is still centered on the shapes; it is just that the > 'shapes' are now composite shapes consisting of the text 'Test' and the > drawn/filled shape. > > What I think you want is a way to connect the arrows between the SVG > sub-shapes, correct? I am not sure (at all) how to do that. > > -cbc > > On Tue, Aug 28, 2018 at 12:04 PM HernÃ¡n Morales Durand > <hernan.morales(a)gmail.com> wrote: >> >> Attached screenshots just in case someone has a clue. >> Cheers, >> >> HernÃ¡n >> >> >> 2018-08-23 14:31 GMT-03:00 HernÃ¡n Morales Durand >> <hernan.morales(a)gmail.com>: >> > Hi, >> > >> > When connecting two composite shapes with RTEdgeBuilder, the connector >> > position loose its center over the connecting shapes. >> > >> > It's easier to explain with two scripts: >> > >> > This one looks good, the connector (RTArrowedLine) is centered: >> > >> > | view shapes myElems | >> > view := RTView new. >> > myElems := 1 to: 2. >> > shapes := (RTSVGPath new >> > path: 'm 3.96875,9.2604167 h 31.75 V 25.135417 C 22.489583,23.8125 >> > 17.197917,34.395833 3.96875,27.78125 Z'; >> > fillColor: Color red; >> > borderColor: Color black; >> > borderWidth: 1.2; >> > scale: 1.4) elementsOn: myElems. >> > view addAll: shapes. >> > RTEdgeBuilder new >> > view: view; >> > shape: (RTArrowedLine new >> > color: Color white; >> > yourself); >> > elements: shapes; >> > connectFrom: 1 to: 2. >> > RTTreeLayout new >> > verticalGap: 30; >> > horizontalGap: 30; >> > applyOn: shapes. >> > view inspect. >> > >> > However when adding a label in a composite, both label and arrow looks >> > shifted: >> > >> > | view shape shapes myElems | >> > view := RTView new. >> > myElems := 1 to: 2. >> > shape := RTCompositeShape new >> > add: (RTLabel new text: 'Test'); >> > add: (RTSVGPath new >> > path: 'm 3.96875,9.2604167 h 31.75 V 25.135417 C >> > 22.489583,23.8125 17.197917,34.395833 3.96875,27.78125 Z'; >> > fillColor: Color red; >> > borderColor: Color black; >> > borderWidth: 1.2; >> > scale: 1.4); >> > vertical; >> > yourself. >> > shapes := shape elementsOn: myElems. >> > view addAll: shapes. >> > RTEdgeBuilder new >> > view: view; >> > shape: (RTArrowedLine new >> > color: Color white; >> > yourself); >> > elements: shapes; >> > connectFrom: 1 to: 2. >> > RTTreeLayout new >> > verticalGap: 30; >> > horizontalGap: 30; >> > applyOn: shapes. >> > view inspect. >> > >> > Any idea how to align these shapes? >> > >> > Cheers, >> > >> > HernÃ¡n >> _______________________________________________ >> Moose-dev mailing list >> Moose-dev(a)list.inf.unibe.ch >> https://www.list.inf.unibe.ch/listinfo/moose-dev > > > _______________________________________________ > Moose-dev mailing list > Moose-dev(a)list.inf.unibe.ch > https://www.list.inf.unibe.ch/listinfo/moose-dev >

1 0