mailing-list for TeXmacs Users

[David Allouche <address@hidden>]

Well, I see the LWN comment spawned much more discussion here than on LWN.

Obviously, TeXmacs has little hope of succeeding as a replacement for 
TeX, simply because people who use TeX do not want to replace it, and 
people who do not use TeX still find TeXmacs too difficult to use.

I agree that the existing user documentation *is* hard to read and 
understand, but it looks like the people that Joris know find it easy. So 
there is little hope there unless someone else make a complete rewrite.

Some people want a document for people who know little or nothing about 
computing. There currently exists such a document (though it is a work in 
progress). It is the tutorial, and you can contribute to it on the Wiki.

Now, I think that LaTeX compatibility is not a priority. It is important, 
but is also a real Sysiphus stone. LaTeX is *very* big, full of 
irregularities, and has poor semantics. It is important to improve 
compatibility on punctual problems (like designing a stylesheet needed by 
a specific kind of public) though I will probably not do it unless I am 
paid for it.

What is important is to do "more and other" than LaTeX instead of "new 
and improved". So I am going to keep focus on my current improvements of 
structrured editing abilities. Incremental rewriting and validation, 
support for literate programming and tigher integration with other 
interactive tools are the real features which are going to make TeXmacs 
successful. And since we are going to do all of that with good XML 
support, we will be what today's LaTeX users will want when they will 
have to adopt a interchange format which is less braindead than LaTeX; 
but that is actually only a side effect, they are not our real target. 
Again, the point is not wether XML is good or bad, but it is what people 
use for structured documents nowadays, and they need WYSIWYG tool to work 
for it.

Repeat after me: let LaTeX user use LaTeX, let Word users use StarOffice, 
we can do what they do better than them (well, almost), but we will 
prevail because we will do what no other free software is doing 
satisfyingly by any standard: WYSIWYG edition of structured documents, 
not WYSIWYG edition of TeX-like documents or high-quality word 
processing. So XML is THE must.

About the spacing, I agree that TeXmacs clearly still has problem with 
that. For example, try to make a document title over several lines. For 
the center alignement to be correct you need to create two "title" 
environments, but then, there is no spacing between their ink boxes. That 
can be fixed by adding a "small space after". Idem for the eqnarray* 
problem described by Phil. That is easy to fix, but that should not even 
appear in the first place.

Moreover, TeXmacs have default margins which noticeably narrower than the 
margin in, say, article documentclass in LaTeX.

One other thing where spacing sucks is with the positioning of section 
titles. LaTeX place chapter titles (in book style) at the right postion, 
which is NOT the top of the page body, but much lower. The correct 
position is determined by a complicated geometric method (shown in any 
good typesetting book) which takes in account the page size, and the end 
result is that the title is placed where the eyes naturally goes *at 
first*. I have not noticed such a sophistication in the TeXmacs styles 
yet.

Michael John Downes says whitespacing is a compromise between readability 
and cost of paper. I absolutely do not agree. Whitespace is here to 
improve readability. Full stop. If you want to save on paper, okay, but 
then do not pretend you are doing high quality typography. The point is 
not that too much whitespace costs more paper, but that too much 
whitespace *reduces* readability.

Also Joris seems to say there is a way to customize the minimal 
horizontal distance between ascents and descent to allow skyline fitting. 
What is it?
-- 

                                  -- David --