Hi Kim,
 
> > (the http authentication scheme is a good example).
> 
> Actually, this is a debated issue...

I know, which is why I tried to steer clear. I was part of an
ongoing debate on Full-Disclosure about this for quite a while. My
position remains that the host portion of a URL contain the
<authentication> portion, as specified in the URL schema, even if
it's "not recommended" to do this (and the reasoning behind this is
purely *social* in the RFC's, not functional), it still remains as
part of the URL schema and, IMO, should continue to be supported by
default. MS provides a means to revert functionality via a registry
hack, so I'm not exactly furious, but it's not something I'm happy
with them about.

My point in bringing this up is *not* to debate the RFC's or whether
MS was right or wrong. My point is that regardless of whether
they're right or wrong, they *did* apply a change that really
screwed up applications (like some of mine) that relied on this
handling for remote locations to authenticate (I never included the
password, only the branch location as
http://[EMAIL PROTECTED]/webservice/ ).

Another good example would be how MS changed the default behavior of
htm files under SP2. For years, literally, people could expect to
double click on an HTM file and have it open in a browser. Users
install an OS update and suddenly DQSD opens Word or Notepad when
the user has their settings to "pagetemplate" variable set.

When I develop applications that rely on MS components to function,
I remain very wary of what 'exceptions' I trust to remain
consistent. Due to their ongoing battles with security and
bug-management, MS is the most common offender at changing how
things work. I would feel much better about this if we did not rely
on their XML parsing engine to process the content.

Just because the RFC or the W3 says one thing, doesn't mean MS is
necessarily going to implement it that way. Proprietary browser
extensions/filters, funky basic HTML handling, that they were
already *using* an authentication scheme in the URL to begin with
(one way or the other, they either "were" or "are" wrong)... I don't
trust that the handling of XML will remain consistent.


> My whole rationale for moving the config files to XML is to 
> make them easier to machine-edit, so we can build a powerful
> configuration UI. If we don't have a UI to offset the
> confusion a bit, I don't think we should do this at all.

I commend that and for the most part I agree; I'm just voicing my
concerns. As a hard-core text-based geek I'm still gunning for the
ability to be able to use a command-line approach to making
configuration changes. I'd really like to see "config var=value" in
the next version.


> Actually... Come to think of it, we can support both.

That's got my vote, as long as it doesn't noticeably degrade the
load time.

Regards,

Shawn K. Hall
http://12PointDesign.com/
http://ReliableAnswers.com/

'// ========================================================
    Old soldiers never die, they just fade away. 
        -- General Douglas MacArthur (1880-1964)




-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
_______________________________________________
Archive: https://lists.sourceforge.net/lists/listinfo/dqsd-devel

Reply via email to