How broadly do we define "friends" of TeX?

Question

The original title of this site was, as I recall, "TeX, LaTeX, and friends". That seems to have now been trimmed down to just "TeX - LaTeX", but other variants such as ConTeXt are still on topic. How broad does it get?

To be specific, I'm trying to generate PDF documents from PHP. There will be a standard template, but different actual content each time. Escaping values from PHP to produce valid LaTeX looks very very tricky. It would be easy to introduce all sorts of subtle security bugs by missing some escape sequences.

SILE has a similar aim to LaTeX, but has two syntaxes, one intended to be human-editable (and similar to LaTeX), the other based on XML. XML is a mature technology easy to manipulate with a programming language. Using SILE instead of LaTeX may be an excellent idea. On the other hand, if I run into difficulties with SILE, can I ask about them here? Is it on topic?

(A search on Meta for SILE finds nothing.)

Answer 1

Well, to me it seems that there are four categories of software:

1. TeX-based software. To see whether your software is TeX-based, there are IMHO two basic questions:

   • Are vast majority of TeX primitives relevant to the software in any way?
   • Is the software based on an expansion language?

   The answer to both questions is certainly "yes" for LaTeX, ConTeXt, LaTeX3, e-TeX, Lua(La)TeX etc.

2. TeX-inspired software. There is certainly MathJaX and LilyPond here, for instance.

3. Software with LaTeX output. Think GnuPlot, PanDoc or even SageMath and Maple.

4. Auxiliary software. These are stuff like BibTeX, MakeIndex, PDFCrop etc.

Now, (at least in my opinion) Categories 1 and 4 are clearly covered, Category 2 is clearly not covered (unless you in some way compare such software to something from category 1, this could be on-topic), and Category 3 is covered partially (based on whether it's the output code that's discussed or not).

There are two main reasons for the strong distinction between Categories 1 and 2. First, pragmatically, people here specialize in TeX-based software. Second, conceptually, we tend to strongly distinguish the way how the software works inside.

SILE seems to be in Category 2 and whence is not covered.

Answer 2

I think we need to start here by thinking about the broader picture. The aim of TeX-sx is to provide answers to questions on topics 'somewhat' related to TeX. In particular, the site needs to offer a realistic chance of the questions being answered well, and for there to be some logic to looking here rather than elsewhere. This fits into the general StackOverflow model: provide useful answers to both the original questioner and to others with the same issues/queries.

The site has always been 'TeX, LaTeX and friends' to avoid limiting just to LaTeX and to have 'LaTeX' in the title so 'TeX' doesn't put people off! That's a bit wordy hence the 'TeX-sx' or similar abbreviation. At the same time, we've never aimed to cover 'Computer typography' or 'Programmatic typography' as a whole. So how do we define the 'friends' to fit in with the above idea of wanting answers? There needs to be some reasonable probably of a good answer (or answers) from a set of people with relevant expertise (both to answer and to review/vote). (People may well have expertise on non-TeX topics too, but those questions can still belong elsewhere.)

The answer by yo' has provided four categories of material that we might divide questions into, and he suggests that SILE falls outside of our scope. I'd look at this slightly differently (though not massively). Where a question is about SILE ('How do I do X in SILE?') it seems to me to be off-topic: it's not something 'we' have expertise on and is likely directly at a single person (the SILE author). On the other hand, where there is some link to TeX code we might regard it as on-topic. For example, a 'compare and contrast' for the paragraph-breaking algorithm, which might contrast SILE with TeX90 and ask what can be done in LuaTeX to match SILE might well be on-topic.