Sorting issues for consideration for siunitx v2.2

I’ve been leaving siunitx alone for a while, concentrating on bug fixes in the v2.1 branch. The  list of issues has continued to grow, and I’m now getting some organisation done before starting on items for v2.2. Some of these are more likely to get tackled, some rather less likely, but it’s best to get everything logged! If you look at the closed issues, you’ll see that part of getting organised is closing some bugs where I don’t feel that it is appropriate to take action. What I would say is that if you’ve got something you’d like considering, put it into the database. I do try to keep up with ideas by e-mail or from forums, but do forget some.

On particular thing I want to think about is the naming is some options and macros. I’ve had some discussion with Marcus Foster from CSIRO Information Management & Technology about siunitx, and he’s pointed out various errors on my part. Some of those, in the documentation, have been fixed. At the code level, he pointed out that what \SI prints are properly called quantities, and that units are separated by products. So I’m thinking of some reasonably radical renaming of macros and options (with the old ones retained, of course!). Feedback on these ideas would be welcome. At the same time, he’s not at all keen on the ‘qualifiers’ concept, but on that I think users would not be happy if I removed it!

3 thoughts on “Sorting issues for consideration for siunitx v2.2

  1. Out of interest what is the objection to the qualifiers? Is it that they are not part of the SI system so should not be in a package dedicated to producing SI units (quantities)?

  2. [latexpage] More or less, yes. The particular worry was that ( U = 1000,mathrm{V}_{mathrm{max}} ) is incorrect and should be written as ( U_{mathrm{max}} = 1000,mathrm{V} ). That is really not what qualifiers are for: the case I’m more worried about is something like ( mathrm{kg}_{mathrm{polymer}},mathrm{mol}_{mathrm{cat}}^{-1} ), which can’t really be applied to a variable.

    My take here is that: (1) realism sometimes is more useful than strictly following rules and (2) I did put ‘SI’ in brackets in the package description – there are lots of non-SI uses for siunitx!

Leave a Reply