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

Re: [Pharo-users] Jupyter notebooks / Dynabook / JupyterTalk

by H. Hirzel

Hello JesÃºs MarÃ The JavaScript kernel has custom output functions [1] such as $$.html(htmlString) $$.svg(svgString) $$.png(base64String) $$.jpeg(base64String) $$.mime(mimeBundle) What is the equivalent for $$.svg(svgString) for example? Regards Hannes [1] http://n-riesco.github.io/ijavascript/doc/custom.ipynb.html On 8/27/18, H. Hirzel <hannes.hirzel(a)gmail.com> wrote: > Actually there is a probably a configuration file, but not in the > place where I checked, I assume. > > I have Ubuntu or Ubuntu based test installations in partitions > sda6,7,8,9,10. I can not recall in which partition in actually > happened. > > However I found out about the command [1] > > jupyter --paths > > And I did this in some of the different test installations [2] > > In sda8 I have > > user8@Latitude2:~$ ls /usr/share/jupyter > kernels nbextensions > > user8@Latitude2:~$ ls /usr/share/jupyter/kernels > python3 > > user8@Latitude2:~$ ls /usr/local/share/jupyter/kernels > pharo > > > So the probable explanation is that there was actually a configuration > file but not in > > /usr/local/share/jupyter/kernels/pharo > > as you indicate in > https://github.com/jmari/JupyterTalk/blob/master/README.md > > but in > > /usr/share/jupyter/kernels > > > So my guess is that a download *nteract* (https://www.nteract.io/) > alone and pharo6.1 with Juyptertalk (thus no python notebook) should > work in a new Linux installation if there is a > > a kernel.json file with Pharo Smalltalk [3] > > in either > > /usr/local/share/jupyter/kernels/pharo > OR > /usr/share/jupyter/kernels/pharo > > I will test that and report to the list. > > --Hannes > > > > > [1] jupyter help > ================= > > user10@Latitude2:~$ jupyter --help > usage: jupyter [-h] [--version] [--config-dir] [--data-dir] [--runtime-dir] > [--paths] [--json] > [subcommand] > > Jupyter: Interactive Computing > > positional arguments: > subcommand the subcommand to launch > > optional arguments: > -h, --help show this help message and exit > --version show the jupyter command's version and exit > --config-dir show Jupyter config dir > --data-dir show Jupyter data dir > --runtime-dir show Jupyter runtime dir > --paths show all Jupyter paths. Add --json for machine-readable > format. > --json output paths as machine-readable json > > Available subcommands: bundlerextension console kernel kernelspec migrate > nbconvert nbextension notebook qtconsole run serverextension troubleshoot > trust > > > > [2] jupyter paths in different test installations > ================================================= > > > sda7 > ----- > jupyter --paths > config: > /home/user25/.jupyter > /home/user25/anaconda3/etc/jupyter > /usr/local/etc/jupyter > /etc/jupyter > data: > /home/user25/.local/share/jupyter > /home/user25/anaconda3/share/jupyter > /usr/local/share/jupyter > /usr/share/jupyter > runtime: > /run/user/1000/jupyter > > > sda8 > ------- > > jupyter --paths > config: > /home/user8/.jupyter > /usr/etc/jupyter > /usr/local/etc/jupyter > /etc/jupyter > data: > /home/user8/.local/share/jupyter > /usr/local/share/jupyter > /usr/share/jupyter > runtime: > /run/user/1000/jupyter > > > > user8@Latitude2:~$ ls /usr/share/jupyter > kernels nbextensions > user8@Latitude2:~$ ls /usr/share/jupyter/kernels > python3 > user8@Latitude2:~$ ls /usr/local/share/jupyter/kernels > pharo > user8@Latitude2:~$ > > > > sda10 > ----- > > > > user10@Latitude2:~$ jupyter --paths > config: > /home/user10/.jupyter > /usr/etc/jupyter > /usr/local/etc/jupyter > /etc/jupyter > data: > /home/user10/.local/share/jupyter > /usr/local/share/jupyter > /usr/share/jupyter > runtime: > /run/user/1000/jupyter > > > > > [3] Example kernel.json file for Pharo > ====================================== > > ls /usr/local/share/jupyter/kernels/pharo > kernel.json > user8@Latitude2:~$ cat /usr/local/share/jupyter/kernels/pharo/kernel.json > { > "argv": [ > "/home/user8/pharo-vm/bin/pharo", > "/home/user8/Pharo.image", > "ipharo", > "{connection_file}" > ], > "display_name": "Pharo Smalltalk", > "language": "smalltalk" > } > > > > > On 8/24/18, JesÃºs MarÃ <jmariaguirre(a)gmail.com> wrote: >> Last update works also on *nteract* >> >> In the other hand, I could not understand how Jupyter knows about Pharo >> kernel without the configuration file!!! Unbelievable! >> >> >> >> >> -- >> Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >> >> >

Aug. 29, 2018

Re: [Pharo-users] How do I get ZeroMQ working?

by H. Hirzel

To summarize this thread: Make sure Jupyter (http://jupyter.org/) is installed properly. Then ZeroMQ will work with the Pharo Smalltalk kernel ( https://github.com/jmari/JupyterTalk) HH. On 8/24/18, H. Hirzel <hannes.hirzel(a)gmail.com> wrote: > user25@Latitude2:~$ find . -name libzmq* > ./anaconda3/lib/pkgconfig/libzmq.pc > ./anaconda3/lib/libzmq.so.5 > ./anaconda3/lib/python3.6/site-packages/zmq/backend/cython/libzmq.pxd > ./anaconda3/lib/libzmq.so > ./anaconda3/lib/libzmq.so.5.1.5 > ./anaconda3/lib/libzmq.la > ./anaconda3/lib/libzmq.a > ./anaconda3/pkgs/zeromq-4.2.5-h439df22_0/lib/pkgconfig/libzmq.pc > ./anaconda3/pkgs/zeromq-4.2.5-h439df22_0/lib/libzmq.so.5 > ./anaconda3/pkgs/zeromq-4.2.5-h439df22_0/lib/libzmq.so > ./anaconda3/pkgs/zeromq-4.2.5-h439df22_0/lib/libzmq.so.5.1.5 > ./anaconda3/pkgs/zeromq-4.2.5-h439df22_0/lib/libzmq.la > ./anaconda3/pkgs/zeromq-4.2.5-h439df22_0/lib/libzmq.a > ./anaconda3/pkgs/pyzmq-17.0.0-py36h14c3975_0/lib/python3.6/site-packages/zmq/backend/cython/libzmq.pxd > user25@Latitude2:~$ > > > On 8/24/18, H. Hirzel <hannes.hirzel(a)gmail.com> wrote: >> Yes, I re-evaluated in the playground >> >> Metacello new >> baseline: 'JupyterTalk'; >> repository: 'github://jmari/JupyterTalk:master/repository'; >> load:'all' >> >> Then save and quit. >> I hope that this was sufficient. >> >> >> jupyter notebook >> >> create new Pharo Smalltalk notebook >> >> --> the notebook was created. >> >> Evaluate an expression >> >> --> the attached error comes up. The library is still not found. It is >> looking for version 3 wheras the anaconda installation has 5. >> >> >> At the moment for me anaconda support (600MB download) is nice to have >> but not necessary. >> >> The other ways of installing jupyter notebooks (pip3 and nteract, 50MB >> download) need less space and start giving good results. >> >> I suggest to put this as an issue at >> https://github.com/jmari/JupyterTalk for the moment. >> >> 'Retest anaconda support and fix if necessary' >> >> --Hannes >> >> On 8/24/18, JesÃºs MarÃ <jmariaguirre(a)gmail.com> wrote: >>> it's done. >>> please try again >>> Thanks >>> >>> >>> >>> -- >>> Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >>> >>> >> >

Aug. 29, 2018

Re: [Pharo-users] Dynabook and authentic learning pedagogy

by H. Hirzel

On 8/26/18, Sean P. DeNigris <sean(a)clipperadams.com> wrote: > Offray Vladimir Luna CÃ¡rdenas-2 wrote >> Smalltalk keeps its curse of being so concern with the future that was >> disconnected from most of the present and so, giving the majority of us >> no >> bridge between our present >> and the Dynabook vision. > > The idea of a bridge is an interesting one. This idea needs to be further elaborated and then the gaps should be identified and so that bridges may be built. This should make it possible to use existing material from the curriculum. > > One of the core Dynabook/Smalltalk principles is that the whole computer > should be available for inspection and experimentation by the user. As Dan > Ingalls said about operating systems "there shouldn't be one". > > However, the rest of the world is stuck in an application/stovepipe model. > While that may cripple and pigeonhole users' creativity, it's a reality > that > to be practically and generally useful the Dynabook would now have to > deal/interact effectively with that whole mess while we're waiting for the > rest of the world to catch up to our better way of doing things. > > > > ----- > Cheers, > Sean > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html > >

Aug. 29, 2018

Re: [Pharo-users] ssh client for Pharo

by Denis Kudriashov

Hi Andrei. Look at this: https://github.com/dionisiydk/libssh2-pharo-bindings It is very initial support but it was working in my tests ÑÑ, 29 Ð°Ð²Ð³. 2018 Ð³., 2:19 Andrei Stebakov <lispercat(a)gmail.com>: > Is there an ssh client for Pharo just like net-ssh for Ruby? > > Thanks! >

Aug. 29, 2018

Re: [Pharo-users] How best to cope with 2 editor/browser frameworks - Calypso and RubSmalltalkEditor

by Tim Mackinnon

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 >>>>> >>>> >>> >> >>

Aug. 29, 2018

Re: [Pharo-users] [Moose-dev] Re: Centering edges over composites

by HernÃ¡n Morales Durand

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 >

Aug. 29, 2018

ssh client for Pharo

by Andrei Stebakov

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

Aug. 29, 2018

Re: [Pharo-users] Centering edges over composites

by HernÃ¡n Morales Durand

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

Aug. 28, 2018

Looking for a pharos/seaside developer based in Turkey (Istanbul)

by Pierre Durand

Hello, I am looking for a pharo/seaside developer based in Turkey (Istanbul). Feel free to contact me Pierre

Aug. 28, 2018

Re: [Pharo-users] Using ZnEasy getPng: with an octet-stream

by Sven Van Caekenberghe

> On 28 Aug 2018, at 00:41, Andy Burnett <andy.burnett(a)knowinnovation.com> wrote: > > Next step of my Asana adventure. > > In a browser > > https://s3.amazonaws.com/profile_photos/206785263586755.0mU1QMOxVZLp3WiRWhW… > > downloads a png. > > When I try ZnEasy getPng: with the same address, I get a walkback saying expected image/png actual application/octet-stream. In AWS S3 you have to set the correct mime-type when you upload a file, this can also be changed afterwards. If not, it remains binary (application/octet-stream). It is possible to derive mime-types from extensions, though that is not perfect. See ZnMimeType class>>#forFilenameExtension: > Is it possible to handle this with ZnEasy, Not as it is today. > or do I have to use the ZnClient and handle the stream first? Yes, get the bytes and create the Form yourself. > Cheers > Andy >

Aug. 28, 2018