On 29 Jun 2018, at 23:32, Ben Coman <btc@openinworld.com> wrote:

Per (https://github.com/exercism/meta/issues/91)
it looks like the syntax highlighter is "prism" (https://prismjs.com/).

It briefly crossed my mind to wonder if using ".st" for both traditional fileout format and tonel format might cause complications,
but that might be jumping at shadows.�� Ultimately the both comprise Smalltalk syntax.

cheers -ben

On 30 June 2018 at 01:55, Tim Mackinnon <tim@testit.works> wrote:
Hi Ben - it looks completely do-able.

Just submitted my hello world for python - and it wasn���t anything particularly special. I thought your suggestion of using Tonel is perfect - I���m just not sure how well the exercism site will render tonel format (ideally with correct colour coding).

I���ve noticed there are lots of Textmate themes for Smalltalk floating around (and indeed I took one and put it on my iPad for a text app who���s name I���ve forgotten and it seemed to render tonal fine). So if exercism uses that - then it really shouldn���t be a ton of work to make the submit/pull calls like you suggested? It seems that exercism is just providing conversations and a push pull mechanism - so really its git with training wheels? Or have I missed something?

Tim

On 29 Jun 2018, at 17:14, Ben Coman <btc@openinworld.com> wrote:



On 29 June 2018 at 22:53, Tim Mackinnon <tim@testit.works> wrote:
Presumably someone is working on Pharo Exercism (did I notice Ben or Hilaire asking about the tonel format for it?)

Is it possible for someone to put a comment in the repo for it - so we can coordinate effort? I���m assuming that someone can at least get hello world working as an example?

It would also be handy if someone who understands how it works, can give a quick explanation here - as I would offer to suggest it as something for the next UK Smalltalk gathering next month. I���m assuming the cli submits files to some travis runner and you get results back - and we just have to fill something in that will interpret those files and launch a command line pharo to get the results (presumably handling walkbacks in a way that the results get fed back). And then I���m guess we have a pharo add-in that would let you do this from the image some way too?

Tim

I don't know much or even how I ended up tuning in on it - but of several of these sorts of initiatives I've heard about, Exercism looks like a good one.
I haven't put any thought into exercises, but have been thinking about infrastructure.����

Here is some discussion about making a code exporter to be able to use the command line tool to interact with the server.

I've been looking at how the command line tool communicates with the server��
to determine how hard it would be too hard to write a Pharo direct interface to their server.
I wasn't planning to share my experiments yet, but since it may be a while before I get a chance to extend it further,
perhaps it interests someone...

On 30 June 2018 at 06:38, Tim Mackinnon��<tim@testit.works>��wrote:
Nice one Ben - I think a few quick iterations could get something out the door and then we can refine it and make it really cool!

Btw, for on lookers, I'm more and more liking what I see in Exercism,��
with templates to help generate problem descriptions and test data for exercises.
For example...
* https://github.com/exercism/problem-specifications/blob/master/exercises/dominoes/description.md
* https://github.com/exercism/problem-specifications/blob/master/exercises/dominoes/canonical-data.json
*��https://github.com/exercism/docs/blob/master/language-tracks/exercises/anatomy/readmes.md

cheers -ben