Warning: This thread is seriously off-topic now.
In fact, I'll change the subject to reflect this.
Furthermore, I cannot imagine a "good, anticiapting programmer"
writing < instead of '<' inside quotes for the "difficulties"
someone might have parsing HTML, and then leaving off the quotes.
Perhaps, but the parser has to cope with HTML from both experts
*and* amateurs. To cope with missing quotes from amateurs,
it might choose to treat the first > as the tag terminator
then look to see if it can make sense of the tag contents.
It occurs to me that you may be confusing this hypothetical
parser with one that I would write, but that's not my point.
You may also be confusing the poor HTML that the parser is
trying to defend against as something I might write, but
that isn't my point either.
My point is that anyone writing HTML for public consumption
must be mindful of the fact that it will be interpreted by
all manner of parsers; some good, some bad, some good but old,
some new but poor, and some that for pragmatic reasons are
operating in quirks mode to defend against poor HTML. So,
as long as the HTML author doesn't produce invalid HTML,
they should make sensible choices to guard against likely
faults in HTML parsers (unless they don't give a damn about
the web in the first place, of course).
I really hate that attitude. Instead of blaming the people
writing buggy browsers, or other bad parsers, you blame the
people writing HTML!
That's because web developers should know better and anyway
can make corrections easily, whereas on many minority
computing platforms there is no choice of web browser. The
user has to use what's available, or write their own which
is usually not practical or even possible.
Given that the standards allow a free choice of whether to
write angle brackets in attribute values either raw or as
character entities, it doesn't make sense to mindlessly
choose the technique that will cause *more* problems rather
than less for those receiving your HTML. This is just one
example of where a socially conscious choice can improve the
web for everyone. There are plenty of other areas where
careful selection of the correct technique can make a big
difference to accessibility and usability, but it requires
the HTML author to have reasonable clue about the chronology
of the introduction of new technologies, how widespread
support for them is, and how well older browsers degrade.
Which of the following techniques would *you* choose:
1. Encode smart quotes as:
a) unicode entities like ’ etc.
b) Windows codepage 1252 entities
c) convert them to ascii ' and "
2. Site navigation:
a) superkewl Flash app the user has to learn first
b) javascripted rollover images with no alt text
c) static images with alt text and plain href links
d) normal textual links
3. Links that you want the user to open in another window:
a) using a javascript: scheme so you can position the
window and strip it of the normal controls
b) using <a href="whatever" target="_blank">
c) normal href links so the user can make their
own choice about whether they want a new window
4. Glossy company branding to impress people:
a) a Flash splash page with wizzy animations (after all nobody
still uses dial-up, and who wants to get indexed anyway)
b) an animated GIF that constantly draws the eye
c) a static image of the company logo in one corner
5. Named anchors within a page:
a) use <div id="name"> and to hell with older browsers
b) use <a name="name"> which works everywhere
6. Headings, subheadings, and bold:
a) use <div class="head">, <div class="subhead">
and <span class="emphasis">
b) use <h1 class="head">, <h2 class="head">
and <b class="emphasis">
7. Multimap.com style application where the positional
relationship between page elements is important:
a) use positional CSS for everything
b) use HTML tables
8. An HTML form:
a) with a javascript button to submit it
b) with a real submit button
I could probably go on to fill a book with a list of these
sort of choices, but then I'm an experienced web developer
with sufficient clue. The vast majority of kiddie web
deeziners out there would be completely oblivious to the
existence of a choice, and anyway would pick (a) from every
selection just because it's the newest wizzy technology that
gives them the maximum scope for creativity, "so it *has* to
be the right choice doesn't it". It would never even occur
to them that using minimal new technology to achieve their
goals is better than using the maximally new and fragile
technique. Such deeziners have little understanding of what
they're doing (beyond the use of whatever wysiwyg editor
they're using) and they care even less about social niceties,
such as ensuring accessibility to the widest audience,
keeping their HTML and graphics small, neat and efficient
for the benefit dial-up modem users, web caches, etc, or
allowing people to view the site on any browser, at any font
size, or in any window size. They don't care about allowing
people (or bots) to automatically crawl and scrape the site,
in fact they probably think that's a *bad* thing and would
prefer everyone to enter their site only from the front page
so they can throw the right combination of popup advertising
at the hapless suckers!
Not only does the appallingly fragile construction of most
websites reduce the general quality of the web, but it also
imposes a pressure of extinction on minority browsers and
platforms that don't have sufficient market share and
financial muscle to keep up with the grubby complexity that
results from this. Furthermore, it raises the barrier for
entry to anyone wishing to write their own browser, crawler,
or other web client. Despite being a competent Linux user
and fan, I still do the majority of my work on an alternative
platform (RISC OS) because its GUI usability *completely*
outclasses anything available on Linux. (Of course, I have
it networked to my Linux box for the best of both worlds.)
Alternative platforms have much to offer and, just as we
should look after the bio-diversity of the rainforest, we
should avoid needlessly killing off computing platforms in a
mindless lust for the latest kewl thing, otherwise we'll
look back and wonder why we didn't see it coming when some
megacorp owns the world and there are no freedoms left.
So you see, you might think it's a small thing, but when I
see someone advocating the use of fragile markup (needlessly)
in the full knowledge that some browsers won't cope with it
and suggesting that browsers should just get fixed and
upgraded, I hope you now see why I oppose this socially
harmful and myopic attitude as a matter of utmost principle.
