> > which is exactly what i mean: you have to shout from the
> > outside. so if i get a file by mail and process it i wouldn't know
> > what to do with it.
> If you didn't tell me anything, I would run it with the (hopefully
> some day to be default) input encoding UTF8. If your file was clean
> ASCII, it would work in any case. Only if you had used some limited 8
> bit encoding you would need to tell me about the encoding explicitly.
my prediction is that for a good number of years to come you have to live with
documents being in current type of 8bit encodings. anyway, that's again the
input encoding part which I think is not that interesting. right now I don't
think anything other than
- default ascii (one day perhaps in fact default utf8)
- special input encoding used requires explicit mentioning in the document,
eg via some sort of inputenc declaration
is suitable for the near future (and also in my opinion) sufficient
> > the switch to 2e was a very valuable and sad lession in this
> > respect.
> I am not sure what you are referring to. From my limited experience I
> can see two problems with 2e:
my guess is your experience is related to Germany which from the outset was 2e
positive for various reasons, one of it being a strong base for NFSS1 and
i was referring to the fact that "nearly" 2e died a death like say lollipop
(which is quite a nice format but never got used); ie that people didn't even
install it and if you don't have the base then you don't get new packages or
> - Many users weren't using the new features and running in
> compatibility mode. This may be sad, but it has nothing to do with
> the underlying engine, and everything to do with changes (to the
> better) in the LaTeX user interface.
that would have been an acceptable (though not nice) start of acceptance, far
more problematical would have been people sticking with 2.09 and refusing to
use it. it took a very long time for american (and other) journals, say, to
accept 2e sources.
> > but more importantly and that is my second point or the point i
> > tried to make: LaTeX should "work" on TeX as a platform. so if i
> > would build the kernel in a way that this is not possible then
> > people would not go with LaTeX not for a long time.
> It is not for me to judge if TeX as an engine is sufficient or
> desirable, I don't know the real issues well enough.
in many respects it is not but so far it is
a) the most widely available one
b) the most stable one (everywhere the same)
c) it doesn't have the problem of maintenance (even if no bugs get fixed
d) it doesn't have the problem of featurism, ie it is stable and reliable
most or even all of the successors of TeX have been under strong development
which is fine as such but LaTeX also requires a strong compatibility between
different sites and that is not provided if the underlying program changes a
lot. pdftex is a good example, in itself a success it was of no much use to
parse source files around using anything other than the basic "don't output
dvi output pdf" until recently where it finally got more or less
stable. problem with pdftex now is that it lacks a suitable maintenance
future given that Than might not be able to support it much longer.
if you look at it from the point of available successors:
- etex is stable but the features are far too internal to make a convincing
argument for people to switch to a version running only on that program
we tried to get further features into etex hoping to get a version which
would allow to produce things really selling but unfortunately that never
got over a number of specification sessions (see project web site articles
again http://www.latex-project.org/papers/etex-meeting-notes.pdf )
- pdftex has a (potential) maintenance problem (and also lacks the features i
would like to see in a successor of TeX)
- omega is or was up to now too unstable in its features, ie far too much
evolving to base LaTeX on it and also lacks above features to some
extend. i guess it will get stable now (if not already being so), but it is
a pity (though very understandable) that it mainly tried to solve one part
of the picture.
so if there is a new latex which replaces 2e then it needs a basis fast or
else too many people will stick with 2e; then the professional side (eg the
journals) will stick 2e (if they don't even use 209:-) since they will say
there is no need for putting money or resources into updating their classes
and files; and if so even people who would like to go to a successor find that
they have nobody to exchange their files with or send them to to get published
the problem why something like the linux model/takeover can't work with
something like LaTeX is that you can't simply add or upgrade without becoming
incompatible with the rest of the world and that is a big need for LaTeX.
with linux itdoesn't matter that i run suse 6.2 (i think) on the box i'm
writing from while my old laptop has perhaps even 5.x and you might have 7.1
--- it only concerns the individual box and the need for a certain driver for
it and some goodies on the X etc. but the boxes can talk to each other without
any problems. not so with LaTeX.
> Anyway, Frank, I just got your last mail in my inbox (need to read the
> details more carefully), and I think we agree that it's worth
> exploring if there would be a substantial advantage for having some
> engine with Unicode internal reprentation.
it surely is, though i'm not convinced that the time has come, given that the
current LICR actually is as powerful (or more powerful in fact) than unicode
ever can be.
> Anyway, I think that
> Knuthian TeX should not be treated as a holy cow (if sound arguments
i'm not treating it as such. but i want LaTeX <whatever> to succeed. and that
means keeping a close eye on how the user base will react and how to influence
> can be made), and a move to LaTeX3 which will by definition break the
> "stability" of LaTeX may be a good point to make a move beyond TeX if
> such a more is deemed eventually necessary.
oh yes, but it would be far easier to do so if it comes with a lot of "sexy"
items that people will jump on (whether they need it or not) and that is not
easy to provide. and you need a critical mass