- Pharo-dev - lists.pharo.org

Re: AI systems & Smalltalk

by Kasper Osterbye

I have been trying my stuff on Ollama, Mistral, Gemini, OpenAI, Claude and Grok, and some others. And for each of these I have tried four of their models. All are different, and all comparable. I hope to get some questions I cannot answer - such are mostly the most interesting. Best, Kasper > On 23 Jan 2026, at 15.05, omar abedelkader <omar.abedelkader(a)inria.fr> wrote: > > Hi > > Iâm working on ChatPharo, which uses a different Agent in the backend, Ollama, Gemini, and Mistral > > https://github.com/omarabedelkader/ChatPharo > > -- > OA > >> Le 23 janv. 2026 Ã 14:46, Tim Mackinnon <tim(a)testit.works> a Ã©crit : >> >> Hi John - this sounds very interesting and the next UKSTUG meeting is on the 28th cover AI in Smalltalk, so I will mention this and see how comparable/complementary that work is - ideally leading to some folks making your solution run in Pharo as well. (of course - might be interesting for you to join us on the 28th too if you have time) >> >> Tim >> >> On Fri, 23 Jan 2026, at 3:24 AM, John M McIntosh wrote: >>> This evening I added ChatGPT support to our work on AI systems talking to Squeak and Cuis systems. >>> Likely 10 minutes to enable Pharo to join the party, any takers? >>> https://github.com/CorporateSmalltalkConsultingLtd/ClaudeSmalltalk >>> >>> .... >>> John M. McIntosh. Corporate Smalltalk Consulting Ltd https://www.linkedin.com/in/smalltalk >>> >> >

Jan. 23, 2026

Re: AI systems & Smalltalk

by omar abedelkader

Hi Iâm working on ChatPharo, which uses a different Agent in the backend, Ollama, Gemini, and Mistral https://github.com/omarabedelkader/ChatPharo -- OA > Le 23 janv. 2026 Ã 14:46, Tim Mackinnon <tim(a)testit.works> a Ã©crit : > > Hi John - this sounds very interesting and the next UKSTUG meeting is on the 28th cover AI in Smalltalk, so I will mention this and see how comparable/complementary that work is - ideally leading to some folks making your solution run in Pharo as well. (of course - might be interesting for you to join us on the 28th too if you have time) > > Tim > > On Fri, 23 Jan 2026, at 3:24 AM, John M McIntosh wrote: >> This evening I added ChatGPT support to our work on AI systems talking to Squeak and Cuis systems. >> Likely 10 minutes to enable Pharo to join the party, any takers? >> https://github.com/CorporateSmalltalkConsultingLtd/ClaudeSmalltalk >> >> .... >> John M. McIntosh. Corporate Smalltalk Consulting Ltd https://www.linkedin.com/in/smalltalk >> >

Jan. 23, 2026

Re: AI systems & Smalltalk

by Tim Mackinnon

Hi John - this sounds very interesting and the next UKSTUG meeting is on the 28th cover AI in Smalltalk, so I will mention this and see how comparable/complementary that work is - ideally leading to some folks making your solution run in Pharo as well. (of course - might be interesting for you to join us on the 28th too if you have time) Tim On Fri, 23 Jan 2026, at 3:24 AM, John M McIntosh wrote: > This evening I added ChatGPT support to our work on AI systems talking to Squeak and Cuis systems. > Likely 10 minutes to enable Pharo to join the party, any takers? > https://github.com/CorporateSmalltalkConsultingLtd/ClaudeSmalltalk > > .... > John M. McIntosh. Corporate Smalltalk Consulting Ltd https://www.linkedin.com/in/smalltalk >

Jan. 23, 2026

AI systems & Smalltalk

by John M McIntosh

This evening I added ChatGPT support to our work on AI systems talking to Squeak and Cuis systems. Likely 10 minutes to enable Pharo to join the party, any takers?https://github.com/CorporateSmalltalkConsultingLtd/ClaudeSmalltalk .... John M. McIntosh. Corporate Smalltalk Consulting Ltd https://www.linkedin.com/in/smalltalk

Jan. 23, 2026

Claude & Smalltalk

by John M McIntosh

Hi, if you've been watch Claude Code taking over the world, missing interacting with Smalltalk, then I've build out an interface so you can talk to Cuis Smalltalk. However the Python MCP bridge interfaces to a Smalltalk solution via MQTT so a bit of work could be done to enable one to talk to a Pharo image. https://github.com/CorporateSmalltalkConsultingLtd/SmalltalkCuisMCP .... John M. McIntosh. Corporate Smalltalk Consulting Ltd https://www.linkedin.com/in/smalltalk

Jan. 21, 2026

Re: Looking for Metacello beta testers

by stephane ducasse

Hi Gabriel Happy new year > Nice, Metacello needs some love. Indeed :) And we are making progress. > I will try with our repos and report back if any problems arise. Excellent. > > Are there any docs or details on the strict mode? Not that I remember. Guille can tell us. > Is it enabled by default, No because nothing would work :) > or do we need to enable it per repo? I can give it a try too. I will ask Guille to tell us. > > Regards, > Gabriel > > On Mon, Dec 29, 2025 at 5:55â¯AM stephane ducasse <stephane.ducasse(a)inria.fr <mailto:stephane.ducasse@inria.fr>> wrote: >> Hello >> >> Under the radar, we are actively continiously cleaning Pharo deep core so that we can work >> on new features. >> >> For metacello, in Pharo 13 we introduced a strict mode (so that we can detect hidden dependencies), >> and other loading features to support module experimentations. It will make Pharo code more robust >> in the long term. >> >> Since a couple of months we are actively simplifying/cleaning Metacello (as the version available in Pharo) because we want to introduce new features such as the possibility to force the reload of baselines and others. >> >> Now the unit test coverage of metacello is quite low and we do not use/know all the features >> that are available and that you may rely on. >> >> So we would love to have beta testers that can just try to load their code in Pharo 14 alpha >> and report potential bugs or features drop we would have introduced. >> >> So do not hesitate and this is a nice way to help pharo >> >> S.

Jan. 4, 2026

Re: Looking for Metacello beta testers

by Gabriel Cotelli

Nice, Metacello needs some love. I will try with our repos and report back if any problems arise. Are there any docs or details on the strict mode? Is it enabled by default, or do we need to enable it per repo? I can give it a try too. Regards, Gabriel On Mon, Dec 29, 2025 at 5:55â¯AM stephane ducasse <stephane.ducasse(a)inria.fr> wrote: > Hello > > Under the radar, we are actively continiously cleaning Pharo deep core so > that we can work > on new features. > > For metacello, in Pharo 13 we introduced a strict mode (so that we can > detect hidden dependencies), > and other loading features to support module experimentations. It will > make Pharo code more robust > in the long term. > > Since a couple of months we are actively simplifying/cleaning Metacello > (as the version available in Pharo) because we want to introduce new > features such as the possibility to force the reload of baselines and > others. > > Now the unit test coverage of metacello is quite low and we do not > use/know all the features > that are available and that you may rely on. > > So we would love to have beta testers that can just try to load their code > in Pharo 14 alpha > and report potential bugs or features drop we would have introduced. > > So do not hesitate and this is a nice way to help pharo > > S.

Jan. 2, 2026

"Hera 1.1 â A Shipload of New Functionality" â New blog post on all: objects all: theTime

by Koen De Hondt

Dear Pharo users and developers, I presented Hera, a BDD tool for Pharo, at ESUGâ25. That was version 1.0. I am happy to announce that version 1.1 is available. You can read about all new and improved functionality on https://all-objects-all-the-time.st/#/blog/posts/18. As always, all feedback is welcome. Happy reading! Ciao, Koen

Dec. 31, 2025

Looking for Metacello beta testers

by stephane ducasse

Hello Under the radar, we are actively continiously cleaning Pharo deep core so that we can work on new features. For metacello, in Pharo 13 we introduced a strict mode (so that we can detect hidden dependencies), and other loading features to support module experimentations. It will make Pharo code more robust in the long term. Since a couple of months we are actively simplifying/cleaning Metacello (as the version available in Pharo) because we want to introduce new features such as the possibility to force the reload of baselines and others. Now the unit test coverage of metacello is quite low and we do not use/know all the features that are available and that you may rely on. So we would love to have beta testers that can just try to load their code in Pharo 14 alpha and report potential bugs or features drop we would have introduced. So do not hesitate and this is a nice way to help pharo S.

Dec. 29, 2025

Re: [DEBT 2026] Future Debugging Techniques Call for Papers, submission 7th of March 2026

by Kasper Osterbye

Go for it, Go for it. One of the many wonders of Pharo is how well the IDE is in all aspects, including debugging. â Kasper > On 24 Dec 2025, at 18.54, Steven Costiou <steven.costiou(a)inria.fr> wrote: > > > DEBT 2026: Call for Papers > > ================================================================================ > DEBT 2026, Workshop co-located with ECOOP 2026, Brussels, Belgium > Fourth Workshop on Future Debugging Techniques > (https://2026.ecoop.org/home/debt-2026#Call-for-Papers) > ================================================================================ > > Important Dates > ----------------------- > * Abstract submission deadline: Sat 28 Feb 2026 > * Paper submission deadline: Sat 7 Mar 2026 > * Author notification: Wed 15 Apr 2026 > * Workshop: Monday 29 June 2026 > > ================================================================================ > > DEBTâ26 is looking to advance state-of-the-art to debug modern software. We welcome researchers from all related areas aimed at helping with the hard task of diagnosing the root cause of bugs, including dynamic and static debugging techniques, online and postmortem debuggers, delta debugging, automatic bug finding, novel visualization techniques for debugging programs, etc. > > The workshop aims to gather the community and foster discussion from different perspectives. That is why we seek submissions in the form of papers as well as talks and tool demonstrations. > > The workshop is a venue for all approaches to debugging. A non-exclusive list of topics of interest is: > > Debugging techniques, from static to dynamic techniques. > Innovative visualisation techniques. > Techniques targeted specific programming models and execution models (e.g., concurrent and parallel programming, microservices, distributed ledgers, web, etc. ) or hardware (e.g., debugging micro-controllers, mobile devices, Big Data applications, etc.). > Case studies and evaluation of such techniques, e.g., user studies on visualisation tools, debuggers, etc. > Surveys, taxonomies of bugs and bug patterns, and current practices/uses of debugging approaches. > > > Papers must be formatted according to the guidelines for ACM sigplan papers (\documentclass[sigplan,screen]{acmart}), see https://www.sigplan.org/Resources/Author/ <https://www.sigplan.org/Resources/Author/> for details. > > Submissions should be made via Easychair using the following link: > https://easychair.org/conferences/?conf=debt2026 <https://easychair.org/conferences/?conf=debt2026> > The accepted papers will be published in the ACM Digital Library, though authors will be able to opt out of this publication, if desired. At least one author of an accepted paper must register for the workshop and attend the event to present the work, and participate in the discussions. > > > AUTHORS TAKE NOTE: > The official publication date is the date the proceedings are made available in the ACM Digital Library. This date may be up to two weeks prior to the first day of your conference. The official publication date affects the deadline for any patent filings related to published work. > > Workshop Organizers > --------------------------- > > * Burcu Kulahcioglu Ozkan, Delft University of Technology, Netherlands > * Steven Costiou, Centre Inria de lâUniversitÃ© de Lille, France > > For more information see: https://2026.ecoop.org/home/debt-2026 > > > > > >

Dec. 27, 2025