Microsoft jscript / mshtml coding standards

[C. (http://symcbean.blogspot.com/)]

Hi all,

There's a bit of a gap but what works in MSIE version x and what works
in other browsers. But for various reasons I'm currently only worrying
about forward compatability in MSIE.

Does anyone know if MS publish any coding standards for jscript /
mshtml (Microsoft version of ECMAScript and the DOM resp)?
Particularly wrt derefencing MSHTML elements within jscript?

TIA

C.

 

[RobG]

Hi all,

There's a bit of a gap but what works in MSIE version x and what works
in other browsers.

I think you meant *between* what works in MSIE and other browsers.

But for various reasons I'm currently only worrying
about forward compatability in MSIE.

I hope they are good reasons given that the use of IE on the web is
now as low as 50% to 60%, depending on whose statistics you believe.
It's not that hard to make web applications cross-browser from the
start, but likely very hard to make one cross-browser if written for a
particular client (depending on the complexity of course).

Does anyone know if MS publish any coding standards for jscript /
mshtml (Microsoft version of ECMAScript and the DOM resp)?
Particularly wrt derefencing MSHTML elements within jscript?

I don't know if you'd call them standards, but their documentation is
here:

<URL: http://msdn.microsoft.com/en-gb/ie/default.aspx >

Follow the links for "HTML & DHTML" and "JScript". Be very careful,
some of their documentation seems to be quite old, some examples won't
work in the latest IE and some *only* in IE. Consider sticking to
standards and making allowances for IE only where absolutely
necessary.

To save you a bit of grief trying to find stuff, the "HTML and DHTML
Reference" is here:

<URL: http://msdn.microsoft.com/en-gb/library/ms533050(vs.85).aspx >

Their are links to very handy Netscape and Mozilla resources here:

<URL: http://www.mozilla.org/docs/web-developer/ >

 

[Thomas 'PointedEars' Lahn]

I hope they are good reasons given that the use of IE on the web is
now as low as 50% to 60%, depending on whose statistics you believe.

I don't want to start a "stats war" here, but I think you are seriously
mistaken. First of all, it should not merely be a matter of belief but
the credibility of a numerical analysis such as statistics should be
based on the well-founded assessment that the numbers can be considered
representative. That is, the greater the data base, the more diverse
the sources that are used and the less biased the interpretation, the
more representative and thus more credible the numbers are.

And secondly, as for that, due to the methods they use (as described in
"About Our Market Share Statistics"), I find the analysis of Net
Applications, which says the *total* market share of Internet Explorer
6.0 and 7.0, although going down the last months, would still have been
about 69% (ca. 47% + 22%) in November 2008, a bit more representative
and thus more credible than your numbers. YMMV.¹

It's not that hard to make web applications cross-browser from the
start, but likely very hard to make one cross-browser if written for a
particular client (depending on the complexity of course).

Full ACK. And JFTR: 20% of 100 million users are still 20 million
potential customers.


PointedEars
___________
¹ It has often been argued here rightfully that the User-Agent header
as-is is not a reliable means to determine what the user agent is.
However, what does not appear to have been considered to date is how
likely a false positive would be. Insofar, for Web statistics that
have a large data base only, the possibility of a false positive
might turn out not to be relevant.

 

[C. (http://symcbean.blogspot.com/)]

I think you meant *between* what works in MSIE and other browsers.

Yes - sorry - brain/finger link overflow.

<URL:http://msdn.microsoft.com/en-gb/ie/default.aspx>

Follow the links for "HTML & DHTML" and "JScript".  Be very careful,
some of their documentation seems to be quite old, some examples won't
work in the latest IE and some *only* in IE.  Consider sticking to
standards and making allowances for IE only where absolutely
necessary.

Yes - found this - and the DHTML stuff. Unfortunately my problem is
not with writing my own code but a political battle with a non-
technical manager regarding a codebase which only works in the
current* version of MSIE. I'm more looking for ammunition to convince
them that I'm right regarding forward compatability and they're not.

(*actually its only tested against a rather old version of MSIE)

C.

 

[Thomas 'PointedEars' Lahn]

[...] Unfortunately my problem is not with writing my own code but a
political battle with a non- technical manager regarding a codebase
which only works in the current* version of MSIE. I'm more looking
for ammunition to convince them that I'm right regarding forward
compatability and they're not.

Assuming your metaphor, you have not yet provided a clear target
to shoot at.


PointedEars

 

[Jorge]

I don't want to start a "stats war" here, but I think you are seriously
mistaken.  First of all, it should not merely be a matter of belief but
the credibility of a numerical analysis such as statistics should be
based on the well-founded assessment that the numbers can be considered
representative.  That is, the greater the data base, the more diverse
the sources that are used and the less biased the interpretation, the
more representative and thus more credible the numbers are.

And secondly, as for that, due to the methods they use (as described in
"About Our Market Share Statistics"), I find the analysis of Net
Applications, which says the *total* market share of Internet Explorer
6.0 and 7.0, although going down the last months, would still have been
about 69% (ca. 47% + 22%) in November 2008, a bit more representative
and thus more credible than your numbers.  YMMV.¹

<http://marketshare.hitslink.com/>

Take a look at the "Operating System Market Share": they claim 89.6%
for Windows.
OTOH, they claim as well that IE's market share is 69.7%. Let's round
those numbers to 90 and 70%:

As IE runs only in Windows, that would mean that of every 9 PC's
running Windows, 7 navigate with IE (77.7%). I just need to look
around to see that the reality is very different. The counters of my
own pages also show a very different situation: IE (6+7) varies
between 28+12%= 40%, and 29+19%= 48%. And that's only slightly worse
than, for example, w3schools's observed IE's market share (46.6%).

And there's this: <http://www.bazon.net/mishoo/articles.epl?
art_id=958>