search

haroldthimbleby / improving-science

Paper, supplementary material and all data for a paper "Improving science that uses code"
2 stars 0 forks source link

"make check-versions" fails with case-sensitive file systems #1

Closed khinsen closed 1 year ago

khinsen commented 1 year ago

On a standard Linux system, make check-versions fails when testing bibtex due to a typo in makefile.

The line

    @$(call checkVersion,Bibtex,BibTeX 0.99d (TeX Live 2021),)

should read

    @$(call checkVersion,bibtex,BibTeX 0.99d (TeX Live 2021),)

i.e. bibtex in all lower-case.

haroldthimbleby commented 1 year ago

That's very helpful; thanks. On Darwin (which I'm using) the disk format can be set, as I have done, so case doesn't matter -- and so you don't notice the problem :-)

Anyways, now fixed.

best,

Harold

Prof. Harold Thimbleby PhD FLSW FRCPE HonFRSA HonFRCP See Change Fellow in Digital Health Fix IT - see https://tinyurl.com/3hkrazdp Web site - http://harold.thimbleby.net

On Mon, 20 Feb 2023 at 17:43, Konrad Hinsen @.***> wrote:

On a standard Linux system, make check-versions fails when testing bibtex due to a typo in makefile.

The line

@$(call checkVersion,Bibtex,BibTeX 0.99d (TeX Live 2021),)

should read

@$(call checkVersion,bibtex,BibTeX 0.99d (TeX Live 2021),)

i.e. bibtex in all lower-case.

— Reply to this email directly, view it on GitHub https://github.com/haroldthimbleby/Software-Enginering-Boards/issues/1, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEMNBTH5HZBXY7HB4E5DUOLWYOULZANCNFSM6AAAAAAVCESPEM . You are receiving this because you are subscribed to this thread.Message ID: @.***>

haroldthimbleby commented 1 year ago

Hi Konrad

Thanks so much for those observations/suggestions; they've led to many minor improvements in the scripts as well. I really appreciate your attention to detail. (I put your name in the acknowledgements in the previous revision.)

(The text of the paper and supplementary material hasn't changed, I've only made mods to the makefile, etc, to make things smoother - well, more honestly, as you noticed, to make things work!)

best,

Harold

Prof. Harold Thimbleby PhD FLSW FRCPE HonFRSA HonFRCP See Change Fellow in Digital Health Fix IT - see https://tinyurl.com/3hkrazdp Web site - http://harold.thimbleby.net

On Mon, 20 Feb 2023 at 17:43, Konrad Hinsen @.***> wrote:

On a standard Linux system, make check-versions fails when testing bibtex due to a typo in makefile.

The line

@$(call checkVersion,Bibtex,BibTeX 0.99d (TeX Live 2021),)

should read

@$(call checkVersion,bibtex,BibTeX 0.99d (TeX Live 2021),)

i.e. bibtex in all lower-case.

— Reply to this email directly, view it on GitHub https://github.com/haroldthimbleby/Software-Enginering-Boards/issues/1, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEMNBTH5HZBXY7HB4E5DUOLWYOULZANCNFSM6AAAAAAVCESPEM . You are receiving this because you are subscribed to this thread.Message ID: @.***>

khinsen commented 1 year ago

Harold,

thanks for fixing these issues! My goal is to get this running to the end.

I am particularly interested in exploring your code because I am currently working on a MOOC that explains the techniques for reproducible computations. Full automation is one important point, and make is one of the two tools we use to illustrate how this can be done (the other one is snakemake).

BTW, you might be interested in Maneage, a framework for automation that is similar to your approach in many ways. The main difference is that it doesn't merely record versions of all software packages, but actually installs a specific version fixed in the project's definition. Better reproducibility, at the price of lengthy compilations before the first run.

Best, Konrad.

khinsen commented 1 year ago

Also interesting: https://tinyurl.com/tuto-techno

Seen yesterday in a thesis defense. Fully automated high-performance computing!

haroldthimbleby commented 1 year ago

Fascinating ideas, ripe for generalisingI had two background thoughts:We are bad at describing new ideas abstractly - a lot of it is concrete “how it works” from which reader has to be the scientist objectifying and inventing the categories / principles Have you seen my relit paper - https://www.sciencedirect.com/science/article/pii/S0167642317302952From Harold's iPhoneThe Patron Saint of copying people into emails is St Francis of a CC.   Too brief? Here's why: https://medium.com/virtual-teams-for-systemic-change/the-email-charter-an-idea-worth-spreading-688b8940f892On 23 Feb 2023, at 15:20, Konrad Hinsen @.***> wrote: Also interesting: https://tinyurl.com/tuto-techno Seen yesterday in a thesis defense. Fully automated high-performance computing!

—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you modified the open/close state.Message ID: @.***>

khinsen commented 1 year ago

relit looks like an interesting variant on the many similar tools already out there! Looks like a good choice for those who are proficient at LaTeX already.