oy boris, i'm once again in over my head, independent of whatever brehnin is asking (he wants to make the push himself?), there are at least three issues 1) am i right that you're gonna leave me to push releases to the julia package manager? that's ok by me and given your previous help i can probably manage, but of course it's fine by me if you intend to continue to do it. 2) i made a new release (7.1.9) mainly because i had screwed up the LINUX_makefile so that the precompiled "release asset" was no longer statically linked. because the other corrections were minor, and because i thought the push to the package manager was not automatic, i didn't bother with trying to make the push myself. 3) MC_fit is in development and not documented, though it does do some useful things (fitting data, thermobarometry). Given the lack of documentation I don't think it's all that useful to push it to the package manager, but if its presence somewhere (the makefile?) makes that automatic or is causing other problems then tell me and i'll modify whatever needs to be modified. Since I use MC_fit it's a easier for me if i keep it as part of Perple_X, so that's my preference.

________________________________________ From: Boris Kaus ***@***.***> Sent: Thursday, October 31, 2024 4:56 PM To: JuliaPackaging/Yggdrasil Cc: Connolly James; Mention Subject: Re: [JuliaPackaging/Yggdrasil] Update for Perple_X v7.1.9 (PR #9719) 1. am i right that you're gonna leave me to push releases to the julia package manager? that's ok by me and given your previous help i can probably manage, but of course it's fine by me if you intend to continue to do it. I would actually prefer if you maintain this, as I'm involved in many other open-source projects and can't always keep up with this. Ofcourse, if @brenhinkeller<https://github.com/brenhinkeller> is happy to help that is fine as well. of course, that's fine by me as well. 1. i made a new release (7.1.9) mainly because i had screwed up the LINUX_makefile so that the precompiled "release asset" was no longer statically linked. because the other corrections were minor, and because i thought the push to the package manager was not automatic, i didn't bother with trying to make the push myself. I guess the issue is that as soon as you make a new release of Perple_X, some users would like to incorporate this immediately and compile new binaries. As long make sure that make all works, it should be fine. i'll update the makefile. it's odd that it worked on the previous push, but perhaps i had modified the makefile and then overwrote in 7.1.9. In any case, if you (brehnin) want to push 7.1.9, then tell me and i'll delete and remake the release with the corrected makefile. 1. MC_fit is in development and not documented, though it does do some useful things (fitting data, thermobarometry). Given the lack of documentation I don't think it's all that useful to push it to the package manager, but if its presence somewhere (the makefile?) makes that automatic or is causing other problems then tell me and i'll modify whatever needs to be modified. Since I use MC_fit it's an easier for me if i keep it as part of Perple_X, so that's my preference. In the current build_tarballs.jl file, MC_fit is listed<https://github.com/JuliaPackaging/Yggdrasil/blob/f114f44d4b53223c74d250bf4e0aea2e06658d2d/P/Perple_X/build_tarballs.jl#L17> as a file to be compiled. If you just take this out for now, and also comment the library, it won't be included in the BinaryBuilder version. — Reply to this email directly, view it on GitHub<#9719 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AL3BGODR2PU3YIM2J3FAOBTZ6JHKZAVCNFSM6AAAAABQ6MIVTCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDINJQGIZTMNBRHA>. You are receiving this because you were mentioned.Message ID: ***@***.***>

the correction for gridding is only relevant for analyzing 1d gridded minimization done by vertex. if that's the ....ah, i bet it was your student who told me of the problem, so yes it is relevant.

? i just looked at the 7.1.9 "makefile" and it does have the MC_fit target and dependencies. it's not easy for me to do a make on WINDOWS, but i see no reason why "make all" should fail (and that's consistent with the push working for 7.1.8).

sorry! i see it now, MC_fit in not in "MYOBJ", i'll push it now.

you're very welcome, that bug was a bad one albeit for a special case.