comment #3:

> Thanks Boud, good (and useful!) point ;-)! It is now in Maneage under Commit 2fadf4ba6f.
>
> I just removed the '$$(pwd)' part because the '$$unchecked' file is not in project's top directory. It is in the same directory as the ultimate target ('$@'). After all, it was defined like this:
>

> unchecked=$@.unchecked

>
> So its in the same place that the ultimate target should be, but is just appended with an '.unchecked' suffix.

I see that my branch of maneage has already diverged a bit - in my case the '$$(pwd)' is necessary. :)  A merge or commit proposal will have to be left to later...

> By the way, it would also be great if instead of `, you could use a single-quote/apastrophe (') to mark code components in commit messages (or comments). This is recommended by the GNU Coding standards and would help in making the code more consistent ;-).

I was mostly using the backtick because github and bitbucket use something like markdown, in which the backtick is useful for strings of code. That's not an ethical reason, it's a practical reason. Savannah uses a different form of markup, so in principle it's a good idea to switch here. Also, I see that codeberg, running gitea , does not seem to pass git commit comments through any sort of font convertor like markdown or rich markup.

Following https://www.gnu.org/prep/standards/html_node/Quote-Characters.html#Quote-Characters I'm happy to try either apostrophes or double quotes.

I included a full quote of your previous comment, but replaced '-verbatm-' by '-verbatim-' in order to properly close the verbatim section. It turned out that your verbatim section was kept secret from the html rendered version because of the typo... :)

Here's a one-line commit to clarify to the user how s/he might want to solve a failed checksum for downloads:

https://codeberg.org/boud/maneage_dev/commit/9226955b70dd7476aa21c7b821dbd160ecd73791

I propose that this be like on gnuastro(?) - a permanent task for posting pull requests for small minor commits.

Here I propose commit 0529684b0c, with text changes in README-hacking.md . The most important change is to clarify to the user (who may not yet have used Zenodo) that reserving a Zenodo DOI and saving is an intermediate step that still allows changes. Only "publish" freezes the content of the research package.

https://codeberg.org/boud/maneage_dev/commit/0529684b0c3db7472bb257c690c143bddf8560e3