RE: Why Doesn't IE5 use the DTD to Validate?

Subject: RE: Why Doesn't IE5 use the DTD to Validate?
From: "Didier PH Martin" <martind@xxxxxxxxxxxxx>
Date: Thu, 1 Apr 1999 16:06:05 -0500
Hi Ed,

<YourComment>
However,  I came into this when we were talking about the proper behaviour
of a validating parser and, by implication, the claims that a vendor can
make in this regard and, further, what users -- programmers or browsers --
can expect from all of this. Talking about 'practical' scenarios, past
history (which has proven notoriously inadequate for predicting the future)
is beside the point. Does the parser behave according the specification or
not. In this instance it seems pretty clear that it doesn't and that the
implementation, given time pressures and/or business directions or whatever,
has/is/will cause confusion and inconvenience to everyone -- as individuals
and as groups or organizations trying to sort out what to do about adoption,
product evaluation, etc., etc.

The simplest route to take as a developer -- I'm saying this as a person
whose professional responsibility is to evaluate these technologies and
choose low-cost, easily maintained and consistent/compatible products -- is
to follow the spec or, if you can't, say so and why. The conclusion I come
to professionally is that, as far as XML, etc. is concerned, IE5 as it
stands now is not a viable solution for my organization. This, because of
bugs sure but mostly because of the 'extensions' and 'incompleteness' and
'idiosyncrasies' in implementation of the standard.
</YourComment>

<Reply>
This is precisely the point. According to the actual specs nothing is said
about if a XML interpreter should or should not validate a document. To the
question: does this browser follow the specs? Actually the reference
documenst are: XML specifications version 1.0 and CSS1 (we can also say
CSS2, but the browser got code freeze before the specs where finalized and
got the feedback round). In these specs nothing is said if an interpreter
should or not validate documents. It seems that the implementation is not
totally bug free. I expect the same thing from Mozilla and getting full 100%
specs compliance is the ideal situation rarely attained ( but desirable).
Now about choice. We live in a free place and you can choose whatever you
want. Now, taking all philosophical positions apart. Is IE 5.x so bugged? I
don't think so. I personally found some bugs concerning encoding and CSS
processing. Is these bugs so bad that the product is not usable? No. Will
Mozilla be a good alternative? I expect that yes. Is mozilla better than IE.
Lets see the final release before and some lab test to get an objective
opinion.

This said, this is your strict right to choose whatever you want. We also
have, as professionals, the right to search for truth. This is what I am
trying to do by not taking into consideration political issues and consider
both IE and Mozilla as viable alternatives (sorry I do not have time to
examine Opera). What is interesting me is to know what are real facts, and
evaluate the product stability and accordance to actual and not future specs
(remember that XSL is still work in progress). So my own obervation is that
CSS implementation is still not complete but improved since the last time. I
expect that Mozilla will also mostly but not completely (except if we accept
delays) implement CSS.

By the way, it is today the first Mozilla's birthday. So happy birthday
Mozilla :)
</reply>

Regards
Didier PH Martin
mailto:martind@xxxxxxxxxxxxx
http://www.netfolder.com

<<attachment: winmail.dat>>

Current Thread