Re: [jats-list] Computer algorithms

Subject: Re: [jats-list] Computer algorithms
From: "Christina Von Raesfeld tvonraesfeld@xxxxxxxx" <jats-list-service@xxxxxxxxxxxxxxxxxxxxxx>
Date: Tue, 13 Mar 2018 17:14:31 -0000
I agree that some particularly complex algorithms may be better suited as art,
but most are not, and at a minimum it seems like an indentation solution
should be available for algorithms in JATS (looks like @indent-level is
currently only supported for <verse-line>). The problem with using fixed-font
spaces to achieve indentation is that evolving rendering platforms make it
impossible to predict where long lines will wrap and additional spaces would
be needed to maintain alignment. JATS should be able to support indent-levels
for algorithm lines, which may or may not be preceded by non-indented line
numbers, and which may include inline formulas (as mathML and/or images).

Tina von Raesfeld
PLOS Production Technical Lead

From: Mirko Janc mjanc@xxxxxxxxxxxxx
Sent: Tuesday, March 13, 2018 9:42 AM
To: jats-list@xxxxxxxxxxxxxxxxxxxxxx
Subject: Re: [jats-list] Computer algorithms

I fully agree with Debbie's comment that algorithms are almost "pieces of
art". However, even in such case, they are highly structured, and the
impressionist and postmodern strokes of color serve easier navigation through
hard-to-understand code. Presence of heavy math is unavoidable in almost all
of the pseudo-code. JATS was never envisioned as an art-capturing XML scheme,
but I think that it is worth doing some additional development and also
"teach" authors what is OK to have in the code, and what will invariably be
deferred to a figure or some supplemental material not in the main body of
their article.

I may collect some cases and send to the list, but I am not sure how to
transmit it as attachments are not allowed. We are by no means champions of
algorithmship, but still have significant needs. What about ACM journals, or


>Shading and color in <code> are handled the same way
>as shading and color in <table>, using <styled-content>.
><code> does not allow inline math.
><preformat> does allow inline math.
>None of this means I advocate you change your practice,
>I just wanted to make it clear what JATS CAN do and to
>understand your requirements a little better. When I wrote
>algorithms, they were pseudo-code, no more, no less, using
>shading and color definitely. But I was not in the OSA field
>of study, and I was a programmer rather a long time ago in
>a limited number of programming languages.
>Mirkos post convinces me that I do not begin to understand
>the range of typography inside an algorithm, but they
>sound more and more like art, with keywords for searching
>rather than searchable text.

Current Thread