[longisland6]

over the last year or so i've really become an adherent of web standards. i'm in the process of designing my first site that is (hopefully) 100% valid for xhtml transitional, css and section 508 complience.

i've been a web designer/developer for about 8 years now and i've produced quite a few websites and applications. these were all functional, and in some cases, beautiful pieces of work that functioned perfectly for the audience they were intended for. unfortunately, most of them were tag soup behind the scenes and did not validate.

the site(s)/applications i've designed, developed and maintain for my primary 9 to 5 job is a collection of pages and apps that include graphic design, html, flash, javascript, server-side scripting (asp) and database development (mysql). and...all these pages are styled with externally linked css based on the browser someone is using (calm down and take a deep breath...). i will hopefully soon be able to re-work the site to make it more standards compliant. in the meantime i am using server-side scripting (asp, http user agent) to determine the browser type/version. i've been reading about how easy it is to spoof this string and throw off these results. i've got 2 questions about this:

1. how many of my site users actually do this, or even know how or what it is? my guess is zero. and...

2. what is the best server-side method of determining the browser type/version of a site user? (no client side detection; i've got to consider someone who, for whatever reason has javascript turned off or some other limitation in place that, in imho, severely limits the experience of the web)

so... thanks in advance for any suggestions.

[pgo]

1. Not many, probably.

2. It's a bad idea. I've never had any real need to do it, myself, outside of using conditional comments to target IE (proprietary code, but still standards compliant). All the other browsers work like they're supposed to.

[longisland6]

well yes, ie is the problem. but since it's currently the primary browser in use today, by a wide margin, this becomes an issue.

i know it's a bad idea...

[longisland6]

what suggestions does everyone have for writing flash that is standards complient? i've been reading about sifr(?) but it appears this is for writing text only, is this correct, or can it display any flash file regardless of content?

[achtungbaby]

Originally Posted by longisland6

what suggestions does everyone have for writing flash that is standards complient? i've been reading about sifr(?) but it appears this is for writing text only, is this correct, or can it display any flash file regardless of content?

If you're talking about general flash embedding, Geoff Stearns' swfobject is the way to go. I've used it on nearly every flash project for the last three years. Aside from it just being plain better than most of the other techniques and methods out there, Stearns puts a lot of time supporting it and improving it. Basically swfobject will search through your page for an element you specify, and when it finds it, will clear out whatever's inside and replace it with your swf.

sIFR uses the same sort of technique but for typography only. It also rocks.

[longisland6]

i currently use the latest version of swfobject in a number of my sites and i agree, it works great. i love the support and the swfobject forum. i haven't yet tried to validate a site (xhtml transitional) with swfobject but i've heard it doesn't validate. any thoughts?

[pgo]

It has to validate, it's just JavaScript. It should be ignored by the validator (if it's done properly), I would think...

[hawken]

conditional comments often help me get projects out the door a lot quicker.

[longisland6]

conditional comments just might be the trick. thank you.