W3C home > Mailing lists > Public > public-html@w3.org > May 2007

Re: Support Existing Content

From: David Dailey <david.dailey@sru.edu>
Date: Tue, 01 May 2007 11:31:46 -0400
Message-Id: <6.2.5.6.1.20070501103230.01dbba20@sru.edu>
To: Murray Maloney <murray@muzmo.com>,HTML WG <public-html@w3.org>

At 10:14 AM 5/1/2007, Murray Maloney wrote:

>Sorry. This doesn't help me at all. It still speaks to user agents.
>
>         HTML 5 will be a superset of all previous incarnations of HTML,
>         whether as W3C specification, user agent instantiations, or
>         instances of HTML documents on the web and on intranets.
>         (As a result, user agents which can accommodate HTML 5 should
>         be able to accommodate earlier versions of HTML.)
>
>Does that say what needs to be said?

Hi Murray. It is back to a place now that I can understand it again. 
So that's good. (I also understand now that merely changing wording 
from browser to UA was not enough.) But now I think it may have two 
problems (#1 is more serious but, alas, rather longwinded. #2 may 
just reflect my poor understanding of what it means to be a superset).

1. This wording may be a little too strong for most folks.

Example: In year 2000 I built a web app. It was an animation studio. 
Users could incorporate (using input type=file) images into a work 
space. The app then allowed those images to be resequenced (with what 
was a fairly clever conservation of screen real estate). The files 
could be bulk uploaded. By that I mean, once the user identified a 
first filename (say x1.jpg), then JavaScript interrogated subsequent 
filenames (starting with x2.jpg, and working upward) using onload and 
onerror. I was a bit surprised that the browsers let me get away with 
it, but I got the thing working in IE and in Netscape. (The 
resolution of pathnames was different, so it was rather painful.) The 
user could adjust timing and the nature of the animation (looping vs 
oscillating) It created, as output, the source code of an HTML 
document with JavaScript, that would run such an animation wrote that 
code into a textarea and encouraged users to copy and paste that code 
into an HTML editor. It allowed non-programming users to create 
animations that ran in the browser. (I did not, [however and gasp] 
use style sheets -- CSS didn't seem to work very well across browsers 
from the little bit of reading and experimenting I had done -- so the 
UI used tables to layout widgets.)

Within a year it became the first or second response to a Google 
search for the words "javascript" and "animation". The page got lots 
of hits. I got email about it from countries I had never heard of. 
(The UN has gotten very big you know). That lasted for a couple of 
years. I got busy with other stuff. A couple of years ago I noticed 
that it was still working in IE, but it wasn't anymore in Netscape. I 
had rather started to give up on Netscape anyway.  My app's Google 
ranking deteriorated over the years and I did not try to maintain the 
code -- as I say I got busy with other stuff.

Recently I opened the page again and discovered to my dismay (though 
without surprise) that it now works in none of {IE, FF, Opera}. I 
don't know why. Undoubtedly I did something really dumb, according to 
some after-the-big-game quarterbacking. (Heck I used to change loop 
indexes in Fortran to simulate recursion to investigate 
non-associative universal algebras, and had seen such naive and 
"unstructured" code blown up by Fortran 77. And I was new to 
JavaScript at the time -- having thought it was a toy language prior 
to being assigned to teach it. Oh the textbooks in those days! No 
wonder there is so much questionnable content nowadays.)

I would like it if HTML5 would resurrect that page from the graveyard 
(http://srufaculty.sru.edu/david.dailey/javascript/animator20.html
  in case anybody is really ambitious), but, realistically I don't 
think it will. The page belongs in a place for curiosities that used 
to work, or a museum of quixotic endeavors. From what I gather, the 
security risks associated with allowing me to probe pathnames in your 
filespace were enough to convince somebody somewhere that I should 
not be able to do what I was doing. It's okay. I didn't view it as 
the fault of the standards community.

2. If HTML5/W3C is to be a superset of all previous incarnations of 
HTML, would that not include HTML5/WHATWG which has already promised 
that it will be superset of HTML5/W3C? Consider the set of all sets 
which do not contain themselves. Perhaps I'm just confused again.

cheers,
David 
Received on Tuesday, 1 May 2007 15:31:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:15:58 GMT