Re: [Jmol-users] Re: Jmol prerelease 10.00.37 - Jmol.js JavaScript Library updates

[Miguel <mi...@jm...>]

>> 1. Using browser sniffing to write tags using js makes problems
=5Bsnip=5D
> I agree. It would be better, for instance, for the code to detect the
> presence of window.ActiveXObject to understand to create the IE-type
> object instead of checking to see that navigator.appVersion includes
> =22IE=22 or not. I know for some aspects you can't do this, especially
> with the strange Mac configurations that are out there, but as much as
> possible, it is the functionality, not the name, that should determine
> what choices are made for applet/object

Correct.

If you folks come up with some better and reliable ways to distinguish
between browsers then I will be glad to incorporate the changes into
Jmol.js.

>> 2. This sort of thing may allow one to generate partial code that
>> satisfies HTML4.01 or XHTML1.0 strict in appropriate browsers, but
>> unless you also use javascript to generate the DTD line, that's is bou=
nd
>> to be incorrect on some browsers. Not that it really matters, as
>> browsers only take limited notice of the DTD (see my next posting).
>
> Right. And you have to have your SERVER sending out the page as
> =22application/xhtml+xml=22 type.

Q: Can't you put the content-type in the header as a meta tag and have it=

override what the server returns?

(I am not even sure that this would be useful)

> AND, you have to require no use of
> Internet Explorer by your users (because IE does not and probably will
> not ever support XML).

Q: Has MSFT made any official statements to this effect?


Miguel