From: Ed Brey (brey_at_[hidden])
Date: 2000-10-11 11:31:42
From: "David Abrahams" <abrahams_at_[hidden]>
> > HTML authors have gotten away with trouble for a while due to the
> > prevalence of a small number of very popular browsers.
> What kind of trouble? I am not being resistant here, I just hardly know
> _anything_ about HTML at a deep level.
As an example, take the page on amortized complexity put by my professor for
my data structures class last semester:
Under IE5.5, the last two-thirds of the text is all blue. Chances are my
prof tested it with Netscape, under which it likely displays fine. However,
there are several errors in the HTML; it just happens to look OK Netscape.
The trouble is that you can get lucky and have it work for a subset of
browsers, but will fail beyond what you test. Many authors test on Netscape
and IE, which gives a fuzzy feeling because they make up such a large
percentage of browsers, but still leaves the door open to trouble with
future or less popular browsers. An analogy is the trouble with code like
int* i = v.begin();
It works a lot of the time, but eventually it will catch up to you and bite
you. Fortuantely, verificiation is easier with HTML than C++ in this
> > To me it seems only fitting that a group committed to writing
> > standard-conforming C++ would produce standard-conforming HTML.
> I totally agree with the intention. But it certainly seems like it could
> a serious additional barrier to entry at boost, an idea which I don't like
> at all. I hear that most of the automated tools generate garbage, by an
> HTML-expert's standards, but I'm not sure I do any better by hand. I sure
> wish someone else would weigh in on this; I don't know enough to speak
I don't think that hand-generated, super-clean HTML should be a requirement,
but rather just a desire. We should take it where we can get it. However,
we must recognize that although reading the code is pretty easy, the
learning curve to write it is harder, and so especially when starting new
documents, having generated HTML is a better than a large time investement.
> what-the-heck-is-css-ly y'rs,
CSS stands for Cascading Style Sheets. The general intent is to provide a
mechanism to separate content from formatting. The idea is that by
performing a few tweaks on a single file, we could totally change the look
and feel of the Boost site (although the content would remain the same). In
the style sheet, you assign formatting semantics to tags like H1 and CODE,
and also create style classes like Copyright, which you assign formatting
to. The author is freed from concern over formatting, and just needs to
properly categorize his content (like how UNIX man pages are written, I
think). The "Cascading" part of CSS is that you can have any number of
levels of refinement of formatting as you go from the outer to inner scopes.
For example, a library could override or augment a general boost style, and
a page could override a library style. The result is a lot of power and
flexibility of appearance with minimal coding and maximal reuse.
For the full scoop, check out:
Now that I've hopefully wet your appetite, the bad news. Support is shoddy
to the point where any more than minimal use is just not practical at this
Boost list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk