Re: A prediction about the Presidential campaign

On 3 Feb 2008 23:40:32 -0500, Keith F. Lynch <kfl@xxxxxxxxxxxxxx> wrote:
Daniel Silevitch <dmsilev@xxxxxxxxxxxx> wrote:
Keith F. Lynch <kfl@xxxxxxxxxxxxxx> wrote:
Daniel Silevitch <dmsilev@xxxxxxxxxxxx> wrote:
Even at the level of basic readibility, things like kerning and
ligatures make a difference, and that's not something that's
easily done in straight HTML.

Those are a function of the font. Don't most fonts contain those?
Presumably the reader will choose whichever font he finds most
readable.

Some fonts do, some don't. And those that do, you generally have
to explicitly invoke them for 'fl' or whatever (ie it's a separate
character code rather than a normal f and l).

There aren't any fonts that automatically use that ligature whenever
an "f" is immediately followed by an "l"? Why not?

I'm not a font designer. I have no idea what the technical issues are. I
just know that 'fi' is usually a different glyph entirely from 'f' + 'i'
crammed close together. For instance, the 'i' in an 'fi' ligature
doesn't have a dot. I suspect that you might be able to add ligature
support at the level of the rendering engine (ie at the web-browser
level), but I don't think any of the major browsers have done so, so
it's either difficult or sufficiently not in demand to be worth the
bother.

Good point. One of the WSFA Journals I put online had some math that
had been printed from a long-lost LaTeX file. I hand-coded HTML to be
readable everywhere, including character-cell terminals. I'll let you
judge the results, which are at http://wsfa.org/journal/j96/b/index.htm#p

Looks fine. Now add some integrals, including sub and superscript
limits in the right places. Some compound fractions. A matrix,
with large deliminators. Etc. Suddenly, not so trivial.

You've found the one of the few weak spots in character cell
terminals. I actually miss the ability to see hairy equations more
than I miss the ability to see pictures.

I suspect that most people wouldn't care about equations, but might find
the lack of picture capability annoying. Looking at the popularity of
sites like youtube, the lack of audio and video capability would
probably also be a deal-breaker for many.

The system used in Wikipedia works pretty well, though it does take
some getting used to. Equations look like, for instance:

\zeta(s) = \sum_{n=1}^\infty \frac{1}{n^s}

Presumably in a graphical browser this displays as it would in a
textbook. But the text alternative is a lot more readable than
"[imgmap]" or "[eqn23.jpg]," which are what I would see on some
math-oriented websites.

In a graphical browser, Wikipedia displays equations by using a LaTeX
engine to generate (low-resolution) bitmaps which are then embedded in
the page via IMG tags. That works, sort of. Looks OK on the screen,
looks like crap if you want to zoom in or print it out. That's about the
only real general-purpose solution for math in HTML (MathML exists, but
browser support for it is far from universal).

I prefer that things be readable to everyone rather than look
really pretty to some and totally unreadable to others.

I prefer not to have to spend my time hand-coding HTML.

Me too, but if it's worth creating a web page, it's worth doing it
right. I'd rather create ten good pages than a hundred crappy ones,
even if I can do the latter in less time.

You miss the point. I'm not creating a web page. I'm creating a document
which happens to be accessible via the web. And PDF takes precisely zero
effort, since that's the format I have the output in anyway. And is
better suited for the content I'm working with.

Oh, another problem with HTML: When embedding plots in a document, it's
greatly preferred to use a vector format, since a common practice is for
people to zoom in to look at some fine bit of detail that's attracted
their interest. Easy in PDF: I just tell my graphing program to spit out
an EPS file, and I'm all set. Hard in HTML: I need to find a converter
from EPS or PDF to SVG, and then hope that the target audience has a
browser that actually understands complicated SVG files. Oh, and does
SVG support font embedding? If not, it's no good for what I need.

And the vast vast majority of my target audience can read PDF.

The majority, maybe, but I doubt it's a vast majority.

Between Windows, OS X, and Linux w/ some windowing system, and looking
around in my dpeartment, I feel quite comfortable claiming 'vast
majority'.[1] Also, you can still read PDFs if you have a dumb terminal,
just as long as there's a Postscript printer somewhere on your local
network. Further, since the actual published journals always supply
their online content as PDF and only occassionally as HTML, anyone
without the capability to read PDFs and without the desire to walk down
the hall to one of the public computers or public printers is probably
reading the literature in its dead-tree version and hence *really*
doesn't care about file format arguments.


For the tiny fraction that can't, the arXiv preprint server can
supply the source .tex and .eps files and they can figure things
out for themselves.

I suspect that most people who can't do PDF can't do TeX or
EPS either.

You're probably right, but given the prevalence of PDFs all through the
science community, the intersection of the sets of "interested in a
particular paper" and "can't read PDFs" is going to be pretty small. And
if that set is non-empty and if it's a paper which hasn't appeared in
paper form yet, there's always the option of emailing the author and
asking to have a copy faxed to them. I've never heard of that happening,
but I suppose it could.

-dms

[1] I can't recall seeing a single dumb terminal on anyone's desk.
Assuming that our department is vaguely typical, that suggests that the
incident of said terminals in physics departments (the audience I'm
targeting) is on the order of 1% at most.
.

Relevant Pages

  • Re: How to fix ligatures problem in pdf while converting it into xml
    ... The problem is while I convertthe document some of the characters(fi ... Is there any way to extract text from a PDF so it doesn't carry the ... always -- in the font used in the original PDF. ... a "Th" ligature. ...
    (comp.text.pdf)
  • Re: web page accessibility :: PDF and print versions
    ... But allow me to set aside the matter of PDF and inquire ... regarding HTML, inasmuch as I am creating a HTML version ... When I click the PRINT button on the browser to print out a copy of a ... And some web pages print with a very small font, ...
    (Debian-User)
  • Re: String Length in Pixels
    ... I have a table displayed on the screen in HTML and I wish to ... Unfortunately in PDF I need to specify the column width ... over how the font is displayed, what font is displayed and what size it is. ... Web apps have a wider range of ...
    (microsoft.public.dotnet.framework.aspnet)
  • Re: web page accessibility :: PDF and print versions
    ... But allow me to set aside the matter of PDF and inquire ... regarding HTML, inasmuch as I am creating a HTML version ... When I click the PRINT button on the browser to print out a copy of a ... And some web pages print with a very small font, ...
    (Debian-User)
  • Re: web page accessibility :: PDF and print versions
    ... visitor to enlarge the type via the browser. ... I do not know to specify the font size used in the ... print version of a HTML document. ... Every PDF reader I'm aware of can zoom in. ...
    (Debian-User)
Loading