[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Browser user-agent considered harmful (was Re: wine and IE)



on Wed, Feb 26, 2003 at 08:52:33AM -0500, Brian Stults (bstults@soc.ufl.edu) wrote:
> Karsten M. Self wrote:
> >on Tue, Mar 04, 2003 at 10:28:43PM -0600, Kent West (westk@acu.edu) wrote:
> >
> >>Karsten M. Self wrote:
> >>
> >>
> >>>Moreover:  even in the corporate world, if two business needs make
> >>>conflicting browser demands, you may not be able to accomodate them
> >>>(particularly for specific versions of MSIE after 5.x).
> >>>
> >>>For this and other reasons, I consider the browser user-agent string to
> >>>be harmful.  You can join the protest:
> >>>
> >>> http://twiki.iwethey.org/twiki/bin/view/Main/UserAgentString
> >>>
> >
> >
> 
> I also wholeheartedly support strict adherence to W3C standards.  That 
> is why I was surprised to find that the site referenced above, which you 
> maintain, does not validate as "XHTML 1.0 Transitional" according to W3C 
> standards.  I'm just curious if there is a reason why, when you so 
> clearly advocate W3C standards.

It appears you've neglected to include the d-u list on this response.
Reply redirected to list.

Truth to tell, I haven't checked the complaince myself. 

>From a pragmatic standpoint, the page (and site) renders in a wide range
of browsers -- MSIE, Mozilla, Phoenix, Konqueror, lynx, w3m, and dillo,
among others.  Part of the IETF concensus is "working code", and this
does work, by a fairly broad measure.

You're not disclosing what specific tools you're using to validate the
site.  Running 'tidy -e' against a downloaded page, I get the following
error output (note that these are all warnings, not errors):

    HTML Tidy for Linux (vers 1st March 2002; built on Mar  8 2002, at 22:42:03)
    Parsing "UserAgentString"
    line 2 column 1 - Warning: <html> has XML attribute "xml:lang"
    line 11 column 59 - Warning: unknown attribute "hight"
    line 11 column 155 - Warning: <img> element not empty or not closed
    line 12 column 56 - Warning: <br> element not empty or not closed
    line 14 column 1 - Warning: <table> lacks "summary" attribute
    line 27 column 1 - Warning: <table> lacks "summary" attribute
    line 39 column 1 - Warning: <table> lacks "summary" attribute
    line 44 column 1 - Warning: <br> element not empty or not closed
    line 50 column 1 - Warning: <table> lacks "summary" attribute
    line 173 column 1 - Warning: <table> lacks "summary" attribute
    line 186 column 1 - Warning: <table> lacks "summary" attribute
    line 194 column 98 - Warning: <a> escaping malformed URI reference

    UserAgentString: Doctype given is "-//W3C//DTD XHTML 1.0 Transitional//EN"
    UserAgentString: Document content looks like XHTML 1.0 Transitional
    12 warnings, 0 errors were found!

Given that the page is a mix of hand-edited and generated TWiki HTML,
I'd have to attend to it closely to identify the specific issues.  And
given the publicly accessible nature of the site, short of a validation
tool built into TWiki (certainly a possibility), it's likely that total
compliance with the XHTML standard is impossible.  Keeping the
noncompliance to a handful of warnings seems acceptable to me.

For further discussion of this issue at the TWiki development site, see:

    http://twiki.org/cgi-bin/view/Codev/XhtmlConsideredHarmful

Peace.

-- 
Karsten M. Self <kmself@ix.netcom.com>        http://kmself.home.netcom.com/
 What Part of "Gestalt" don't you understand?
   If spam is the question, Spamassassin is the answer.
     http://spamassassin.taint.org/



Reply to: