> I disagree. Any program that can't copy with illegal input is a
> very bad program. Just think of all the exploits with web applications
> that allow worms and viruses to spread. That's just because there
> seems to be a mindset, especially with many Microsoft engineers, that
> handling bad input is not what they're responsible for.
Hi Thomas. I agree with your statement in general(minus the straw man
aspect), but I don't think that is the case here. My original comment
may have been misunderstood. My intent was to counter the comment that
the RB XML classes are "fragile". In my experience with these classes,
they don't usually willingly allow you to use bad data, they just might
not give you the most helpful error message when it doesn't work. But
out of the hundreds of functions in the XML classes, taking the fact
that this one aspect isn't as polished and helpful as it should be (as
Andrew unfortunately learned) and extending it to the statement that
"shame the XML stuff is still so fragile though" and "Just think of all
the exploits with web applications that allow worms and viruses to
spread." are both incorrect and unnecessarily dramatic.
As for your other message about encodings, you are correct that you can
do other encodings, but UTF-8 is assumed if you don't tell the library
differently.
-stephen
_______________________________________________
Search the archives of this list here:
<http://www.realsoftware.com/listarchives/lists.html>
|