Hi Joe, Yes, and thanks very much for all the work on the MacPorts package for Demeter. I appreciate it very much, and I know Bruce does too. The challenges are very real, and it is not any failing on the part of MacPorts or you or any of the other maintainers of needed packages. wxPerl has always been challenging, and especially on macOS. Keeping very old configuration code for the Ifeffit library (with now very old autoconf scripts, readline, and an ancient plotting library, and the added hassle of using gfortran) is also very challenging. The combination of these very old packages is just too much to be stable. To be clear, we knew almost immediately (so, 20+ years ago) that Ifeffit had a lot of real limitations. An interactive embedded macro language written in Fortran77 (ie, with static, global memory) is going to be a performance and maintenance nightmare - and it was. Having wrappers (Perl, Python) that call a dynamic library worked great, and Demeter is awesome. Transitioning to use RPC to a Python runtime engine for Larch works well, but support for Demeter stalled at about the time Larch became complete enough for that to be the obvious default. As a result, the Demeter installers still use the now very old ifeffit, and no package bundles a Larch installation with the Demeter GUIs. The installer works on Windows because they have very stable executables. But macOS is very hard, and even Linux is not easy (not too long ago, I did get Ifeffit to compile and run on Linux for a friend, but I am not advertising how I did that). I have tried to convey all of these issues here many times. This is all open-source software, and anyone is free to use whatever versions of these codes they like. The counterpoint to that is that no one is obligated to support these. The Larix GUI is an attempt to recover and maybe expand the features of Demeter for XAS analysis in a package that can be supported and sustained. It is a work in progress, but we have made real progress with help from lots of people. Some of that help is feedback and bug reports, asking why something is different from Demeter (it might be intentional, it might not be), or complaining about stuff that doesn't work well. These are all helpful and welcome. Anyway, thanks again Joe! On Thu, Aug 7, 2025 at 10:22 AM Fowler, Joseph W. (Fed) via Ifeffit < ifeffit@millenia.cars.aps.anl.gov> wrote:

Dear Matt, Many thanks for opining on Mac installations of x-ray code. Much more, many thanks for building a really nice suite of useful tools for x-ray physics. (Congratulations on the new major version of Larch!) I am the official MacPorts ZjQcmQRYFpfptBannerStart This Message Is From an External Sender This message came from outside your organization.

ZjQcmQRYFpfptBannerEnd

Dear Matt,

Many thanks for opining on Mac installations of x-ray code. Much more, many thanks for building a really nice suite of useful tools for x-ray physics. (Congratulations on the new major version of Larch!)

I am the official MacPorts “maintainer” of Ifeffit and the Demeter suite and have been for roughly a decade. For several years, I was able to keep the port alive, albeit barely. For roughly the last 4-5 years, I have been unsuccessful building Ifeffit under the MacPorts project; without it, Demeter can’t be used.

Given that the package is unmaintained and pulled from GitHub, I will plan to work with MacPorts to remove the demeter and ifeffit packages, or mark them as permanently moribund, or whatever they favor. Perhaps that will help clarify for people that installing these tools on Macs is no longer possible in 2025.

Hi Kyle, Yep, it has been well-established that trying to build and run Demeter on macOS is very hard, and the results will be fragile. If you got advice in the past 5 years that was something other than "don't bother", then I would

ZjQcmQRYFpfptBannerStart

Best wishes,

Joe Fowler

NIST Quantum Sensors Division, Boulder Colorado

ZjQcmQRYFpfptBannerEnd

*From: *Matthew Newville on behalf of Matt Newville *Date: *Wednesday, August 6, 2025 at 2:56 PM *To: *XAFS Analysis using Ifeffit *Cc: *Kluherz, Tk *Subject: *[Ifeffit] Re: Help installing Demeter on a Mac

Hi Kyle,

Yep, it has been well-established that trying to build and run Demeter on macOS is very hard, and the results will be fragile. If you got advice in the past 5 years that was something other than "don't bother", then I would say you were misled.

Larch installs without any trouble, and Larix runs just fine on macOS. Yep, there are a few things that Larix does not do that Demeter does, and there are many things that Larix does that Demeter does not. Larch and Larix are actively developed. If some feature is not working well for you, we have plenty of mechanisms for asking for help, getting feedback, and contributing. There have been 5 releases of Larch in the past 12 months.

Most installations of Demeter use only the Fortran Ifeffit library. This has not been supported for more than 10 years. It has many known bugs that will not ever be fixed. I recently removed the repository from GitHub, as some people were still trying to use and build that code. This needs to stop.

These are open-source software, written by volunteers. You are free to use any version or modification of these you want – that is one of the points of open-source software. But you cannot expect support for very old software, especially when the developers are clearly stating that it is not supported.

Please stop asking about installing Demeter on macOS. It is not going to be supported.

--Matt ifeffit mailing list: https://millenia.cars.aps.anl.gov/mailman3/lists/ifeffit.millenia.cars.aps.a... to unsubscribe, send mail to ifeffit-leave@millenia.cars.aps.anl.gov

-- --Matt Newville <newville at cars.uchicago.edu> 630-327-7411