Note about stepping away: status?

[JeffreyBroll]

Has anyone stepped forward to take over development/maintenance? Is this still needed?

[goedman]

Thanks for asking. No one has stepped up yet.

Yes, if there is interest to keep these packages around, someone needs to take development/maintenance over.

Rob

[JeffreyBroll]

I apologize for the delay - last few months were a bit hectic. I'm able to pick this up if there aren't any better offers.

[goedman]

Hi Jeffrey, thanks for willing to step in!

Please let me know if there is anyway I can help you. Just now I did send you an invitation to become a co-owner.

Just as background info, over the last number of years I had about 4 objectives:

1. Use DataFrames to present the inference draws,
2. Use Pluto as my working platform,
3. Try to integrate features of CausalInference.jl in the Pluto notebook examples, and, since I started to work with Stan,
4. Separate the Stan methods (StanSample, StanOptimize, etc.)

Of course, you might have very different objectives and I think that would be very healthy!

Best,
Rob

[bparbhu]

I would like to be a part of the transition as well. I do have some projects that will be attached to this mainly a Bayesplot.jl package and I would like to help maintain this ecosystem.

[bparbhu]

I would like to help maintain this as well.

[goedman]

Hi Brian, just send you an invite.

Best, Rob

[bparbhu]

Awesome, thanks Rob, much appreciated!

[JeffreyBroll]

Great! I'm still playing catchup, thinking about the Pluto-related aspects and integrating with DataFrames (+/- maybe making sure Polars can also fit without reworking too much), and it looks like you're well ahead of me.

[goedman]

Polars.jl looks an interesting query approach. Unfortunately I could not install it in Julia or run https://pangoraw.github.io/Polars.jl/docs on Pluto. Not sure if it is supported on MacOS or I'm missing a step.

[bparbhu]

So I'll be updating a few packages that I will transfer ownership to the org. This will be BayesPlot.jl and starting TidyBayesPlot.jl to make a tidyverse friendly version of that package. Also trying to update Loo.jl as well and other stan support packages in addition to maintenance and development of the core stan packages within this org.

[bparbhu]

Also @goedman I noticed that many of the repositories that Stan.jl depends on are archived. Can we undo that? We would like to maintain them as well as they directly feed Stan.jl

[bparbhu]

Ah okay no prob, I'll do it then. Sorry about the confusion on my end.

[bparbhu]

Also, @JeffreyBroll would you like to meet up to discuss future plans for the org and also what development and maintenance cadence will look like. I'm also down to meet up once a week and we can go through things we want to do and ideas for future development. Right now I'm looking to add some supplemental packages like Loo.jl, Bayesplot.jl, SBC.jl, and a few others to help build out the ecosystem for the org. Also I want to make sure we can keep all of this updated as Rob and others have done a tremendous amount of work over the years and we want to make sure that still is happening. Also, we can touch base with the larger Stan community as well.

[bparbhu]

My email is brian.parbhu@gmail.com and my number is 203-543-4672

[JeffreyBroll]

That is a good idea; I've wanted to put together a draft roadmap and see if there's community input (from Stan or Julia discourse, ideally both). It might take me a little while before I can schedule a call but we'll get that sorted.

[bparbhu]

Hey @goedman , I'm also interested in taking over the orgs for RegressionAndOtherStoriesJulia and StatisticalRethinkingJulia. We're doing newer releases for Stan.jl so I want to make sure we'll update those examples as well as the Turing.jl examples as well.

[goedman]

Hi Brian, just send 2 invitations. ROS.jl is pretty bare, I found the R code rather R specific. Hope this works for you, best, Rob Rob J Goedman ***@***.***

…

On Nov 17, 2025, at 11:16, Brian Parbhu ***@***.***> wrote: Hey @goedman <https://github.com/goedman> , I'm also interested in taking over the orgs for RegressionAndOtherStoriesJulia and StatisticalRethinkingJulia. We're doing newer releases for Stan.jl so I want to make sure we'll update those examples as well as the Turing.jl examples as well. — Reply to this email directly, view it on GitHub <#139 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAMX2HDSGJAZRJROLZIFVT35IGJXAVCNFSM6AAAAABV25KUSWVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTIOJZGI4TMOA>. You are receiving this because you were mentioned.

[bparbhu]

Thanks Rob, I got the invitations. This definitely works I'll try and get them updated shortly after the release we'll do this week.
Understood, I'll take a look at ROS.jl over the weekend.

[bparbhu]

Hey @goedman I'm trying to register the package using Registrator.jl what was you way of going about it. Right now I'm doing a comment after the merge into master and it doesn't look like that's working. If it's possible could we go through what you did to do the release to the Julia registrar of packages.

[goedman]

Hi Brian, just noted your email, I’m traveling in Europe right now, so I don’t have my normal Julia work machine with me. From memory: 1. I always kept Julia on the latest version (towards the end `juliaup up`). 2. Once julia is up to date, I updated julia itself (`julia up`). These are only a few packages, mainly Pluto.jl. 3. Then I updated and tested all JuliaStan sub-packages JuliaBase, StanJulia, StanOptimize, etc. 4. Merged and registrated (see below) these. 5. Then for Stan.j: In my case I activated the Stan.jl project, entered ‘]’ and `up`. So at this point all package versions are the most recent available. If a package was held back, I figured out the reason and if possible fixed this. 6. Merge Stan.jl and once the merge was successful used JuliaHub to register the package. Registration through JuliaHub should work. 7. Go to juliajub.com, login and select packages in the left hand column, then select “register package”. You can’t put comments in the merged master, this will block further actions. I think there is something like a <NOBLOCK> marker to prevent this, but I don’t remember all the details. Hope this helps a bit, best, Rob Rob J Goedman ***@***.***

…

On Dec 19, 2025, at 20:40, Brian Parbhu ***@***.***> wrote: Hey @goedman <https://github.com/goedman> I'm trying to register the package using Registrator.jl what was you way of going about it. Right now I'm doing a comment after the merge into master and it doesn't look like that's working. If it's possible could we go through what you did to do the release to the Julia registrar of packages. — Reply to this email directly, view it on GitHub <#139 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAMX2EQTN3B2LS56Z66LJ34CRIEXAVCNFSM6AAAAABV25KUSWVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTKMZQGE3TONY>. You are receiving this because you were mentioned.

[bparbhu]

Understood, We'll follow this protocol from now on. Thanks again Rob and very appreciative of this!

[goedman]

Before merging any package, I also updated its version number.

[bparbhu]

understood.

[goedman]

Hi Brian, recently I needed to use Stan (cmdstan) from within Swift and again evaluated whether to create a single binary or multiple binaries (as I did in Julia a number of years ago). For Swift I decided to go back to a single binary (would be equivalent to a single Julia package handling all Stan methods). Since roughly 2020 I sometimes considered a similar update for Julia to lower the maintenance effort.Sent from my iPhoneOn Jan 24, 2026, at 08:15, Brian Parbhu ***@***.***> wrote: understood. —Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: ***@***.***>

[bparbhu]

Hi Rob, understood, we can take that approach I think at some point after this initial push to get StanJulia to work with 2.27 and 2.28. Right now I just want to make sure we get back to leveraging the work you already did and making sure we can safely expand the interface with packages. Still it makes sense, it would be less work to maintain once it's in that state, so I agree with the approach.

[goedman]

Brian, with "2.27 and 2.28", you mean cmdstan releases? I use cmdstan v2.38 currently for both Stan.jl and on Swift and R. Or is this about another package?

[bparbhu]

Yup, I'm referring to CmdStan releases.

[goedman]

Hi Brian, right now I’m struggling to get CmdStan to compile on my work laptop running MacOs 26.4 and Xcode 26.4 (both betas). While trying to figure this out, I also took a quick look at Stan.jl and today StanSample.jl merge requests. I wonder what is causing the GitHub token issues. I keep en getting GitHub messages that my tokens are expiring. I usually regenerate them but I’m now worried that causes above issues? I wonder if I should remove myself as owner/member? Are you owner of all these repositories?

[bparbhu]

I can step up as being owner for the repositories. We are owners for the org so I just have to make sure we use an appropriate token.

[bparbhu]

Also what version of CmdStan are you using?

[bparbhu]

Also, how have you been Rob? It's been a while.

[goedman]

Hi Brian, hope all is well on your side too! I'm doing fine. Trying to get some stuff finished before spring and summer really kick in and I usually spend less time behind my computer. On my back-up machine I installed Stan 2.38.0 with no problems. On my work machine, using the latest beta versions of MacOS and Xcode as I'm investigating what is possible with Apple's on-board LLM, I'll try to figure out why Stan 2.38.0 doesn't compile.

[bparbhu]

I'm doing well Rob, can't complain. Glad to hear you're doing well! Cool, understood, yeah I need to get a mac laptop just to test things out now. Especially for StanJulia repos. I didn't know apple had an on-board LLM that sounds pretty cool. Yeah I haven't tried 2.38 just yet so anything odd definitely needs to get reported.

[goedman]

It turned out the issues were caused by spotty GitHub.com access to the (Cmd)Stan downloads. Initially it complained about not being able to find some files required for TBB. But occasionally it also failed much later on, when building stansummary and/or checking for the latest version of the stanc compiler. Always complaining it couldn't resolve github.com. But late yesterday evening the problems (I expect nameserver issues) were gone and it cloned and compiled CmdStan 2.38.0 without problems. Testing StanSample.jl and Stan.jl worked fine. On MacOS this is with the very latest (couple of days ago released by Apple) clang++ v21 part of the 26.4 beta versions of MacOS and Xcode toolset. So that is good news!