In another post I described a page which Netscape 6 displays correctly and K-Meleon 0.6 does not.

This situation will recur as other non-IE browsers adopt Gecko, and in some cases K-Meleon will be OK while browser X won't, or K-Meleon version X will be OK and K-Meleon version Y won't. And of course the situation gets messier as the number of technologies used in web pages increases (e.g. last time I looked Opera had great CSS and lousy / non-existent DHTML).

So K-Meleon needs precise browser identification:
* identify as K-Meleon
* identify version and build
* in a form which makes it easy to parse out of the useragent string, e.g. "K-Meleon/version/build". Properties of the navigator object are not very useful because you have to know which browser you're dealing with in order to know which properties are available and what they tell you.

If the script needed for accurate browser identification threatens to become as large as scripts which perform functions the developer wants to deliver to the user, many developers (including me) will concentrate on detecting the big 2 and give other browsers 2nd best pages, possibly with "get a decent browser" messages.

As I said in other post, do object detection and code for DOM. That'll cure 90% of your problems (yes, I do know it looks impossible at first but it is not, really). If you need to detect browsers for remaining 10%, then thats bearable amount.

And no, you don't need to know what browser you're dealing with. You need to know what methods or properties it supports. It really doesn't matter which browser it is when you know it supports, for example, getElementsByTagName or setAttribute.

Philcha's original request for proper UA identification is reasonable, sans the whole concern for [client-side] script parsing.

There is no reason (in my mind, however small it is) to not identify K-Meleon as such. Not including some K-Meleon-specific UA substring confuses UA manipulation for Web-types interested in log parsing (wrongly inflating traffic numbers when K is masquerading) and wholly obscures the existence of K-Meleon to those unaware of the agent.

The folks at Opera maintain a unique substring (though I cannot recall whether that was always the case) and their lead in this regard seems to make sense.

Of the UA strings available under "Preferences - Privacy" I would think it best to re-focus the selections to:

* "empty" UA string
* K-Meleon specific UA string
* NN 4.x masquerade
* NN 6.x masquerade
* IE 6.x masquerade (is it problematic not to include a lower IE masq?)

Possibly joined by:

* "custom" UA entry field

Ordering and initial selection of the default UA string is debatable. I like the idea of a free-form UA entry field, but I fear that it just leads to UA naming pollution. The inclusion of an empty-string option (along with well-chosen alternative masques) might mitigate the need for a free-form field altogether. Maybe not?

That being said, it seems somewhat redundant to offer a free-form field for the UA string *and* the ability to set new drop-down options under "Preferences - Configs - Prefs". Maybe in this case, having more than one way to accomplish the same goal is a bit too much?

Oh, well. Just wanted to bump this topic up a bit.

I'd replhrase the question other way: are there any reasons to identify KM other than Mozilla release it's based on?

I'm not certain how to read "are there any reasons to identify [K-Meleon] other than [the] Mozilla release it's based on."

Should I read that as "apart from the Mozilla release upon which it is based", asking whether the K-Meleon UA string should or should not be identical to the supporting Mozilla core, to the exclusion of masquerade UA strings?

Or, should I read that as "beyond the Mozilla release upon which it is based", asking whether K-Meleon-specific substring(s) should or should not be added to distinguish it (further) from the supporting Mozilla core, specific to the single, non-masquerade UA string for K-Meleon?

I suspect you mean the latter.

I prefer to just use witty strings that site admins may or may not find offensive

Just some more food for thought on the matter, in terms of guiding principles:

Official Mozilla user-agent string specification

Which could be adopted for K-Meleon to produce a K-specific UA string like the following (using version 0.6 as a base):

K-Meleon/0.6 (Windows; U; Win98; en-US; build:547) Gecko/20011011
Of course, I have no clear notion of whether K-Meleon appropriates more of Mozilla than just the Gecko rendering component (in which case a 'rv:' substring patterned after Moz usage might be useful), etc. :p

*bump*

K-Meleon Already allows complete customization in this area...

I can think of a thousand good reason's why I'd want to mascarade as another browser. I like the post above about a catch phrase for webmasters to enjoy...

This also makes it possible for a webmaster to have a different view than someone else out there...

That said, I see no reason why K-Meleon should hide it's identity by default. However, as pointed out above what we really want to point out is the Gecko Engine behind the K-Meleon not K itself!!!

Gecko is what is gonna tell you exactly what display engine is viewing your page. If you have problems coding for K it's Gecko not K! If your code works on other versions of Gecko check the version you're having problems with and refer to Gecko NOT K...

I think the K team has the right idea by identifying K as Gecko. If, however, the K team decides to extend the Gecko engine outside of the Gecko development framework then K should note the differences from it's browser identification...

Yours Truly,
~TheRodster

a.k.a. Rodney Giles
Chief Engineering Architect
Toydrum Web & Media

I think as long as some devoplers code just for IE not dom then K should stay the way it is so the ones that want to go to a site with out having to use IE can go there and I also like it because I can go to a web site and see what they designed it for and see if I would like to support that site or not if its designed only for IE then I will not support that web site like I would if its not designed for just IE So I think that is a nice of K

What's about the statistics? It'd be nice if you could see how many users of x website use K-Meleon.

Best regards,
Josef.