In https://supersitenewmss.superpartners.com.au/Login.aspx?FundCode=CBUS (it's OK, you don't need to actually log in!!!) at line 168 we find an element closing tag </p--> within a <div>. There is no start tag <p-->, and given these two factors KM 1.6.0b2 spits the dummy, chucks a wobbly and throws an immediate fatal error as required by the XHTML standards:

Quote

When a conforming XML processor detects a fatal error it, in standards terminology, “MUST NOT” continue processing in the normal way — i.e. there is an absolute prohibition on error recovery if a well–formedness violation is detected.

from http://www.dev-archive.net/articles/xhtml.html

(Extensible Markup Language (XML) 1.0 (Fifth Edition), 1.2 Terminology, fatal error.)

The problem I (we all) have is that KM is the only browser I know of (I don't and will never use FF) that applies this strictness: Opera 11.51 doesn't. It affects us all, because while web pages should be written to standards and well-formedness should be checked, these things often slip past us. Result? We can't use KM for these pages even though the site otherwise doesn't care what browser we have.

I understand that KM is one of the "standards-compliant " browsers, which is one reason I use it. However, sometimes some rather silly edicts must be disregarded, like this case.

Is there any way future versions can fail to spit the dummy when detecting "fatal errors" in XHTML? Or should this be reported as a bug?

Gordon.

____________________
Understanding the scope of the problem is the first step on the path to true panic. [Florence Ambrose, "Freefall" 01372 January 22, 2007 http://freefall.purrsia.com/ff1400/fv01372.htm]



Edited 1 time(s). Last edit at 10/17/2011 07:11AM by gordon451.

@ Gordon

It took me a while to get the error up. I did not have a valid certificate for the server. I had to use a temporary exception.

I see your point about the error stopping KM with description of the error. I don't know how to answer your question. For myself, I like the strict rule. I do not know enough about XML XHTML etc to know if the error is trivial or major. I do know the KM 1.6 is more strict than is KM 1.5.4. I could not get Policy Manager to run with 1.6 until there was a new version made available. I could see the character which stopped KM but I did not know how to fix it.

If the devs can work through the current problem of connecting to the ever changing Gecko code, maybe you could ask for the ability to override the error. I have no idea what this would mean for the workload of the devs.

Hanlon’s razor is an eponymous adage named after Robert J. Hanlon that states: “Never attribute to malice that which is adequately explained by stupidity.”

JamesD

UA change to FF fixes it.

Quote
JamesD
I could not get Policy Manager to run with 1.6 until there was a new version made available.

1.6 and policy-manager worked on XP.

Thanks everyone -

@jsnj -- setting the UA to FF (3.5) works like a charm. But now I am really confused :s

@JamesD -- I know I asked if there's any way future versions of KM could be made or persuaded to not throw an error in these cases, but it seems this is part of the Gecko engine, and probably beyond the reach of our devs. And believe me, the error in question -- actually all well-formedness errors -- is more cosmetic than functional.

I spent some considerable time (the pun is unintentional ) thinking about a workaround, but all I can come up with is the old Proxomitron (http://www.proxomitron.info/), set to remove all XHTML headers and replace with HTML headers. This would bypass the mandatory error-checking. I think. Of course, this also means I have to set up the proxy, etc... A right royal PITA in fact.

It seems we may have to live with what I personally consider a Mozilla Gecko bug given that the problem is obviously uncommon. Or perhaps we'll all get lucky and they will respond to my request that they fix the page.

{edit: remove dribble}

Gordon.

____________________
Understanding the scope of the problem is the first step on the path to true panic. [Florence Ambrose, "Freefall" 01372 January 22, 2007 http://freefall.purrsia.com/ff1400/fv01372.htm]



Edited 1 time(s). Last edit at 10/18/2011 12:29PM by gordon451.

Ummmm... This is really weird :s

I have just spoofed as Opera 11.51 which did not trigger the XHTML error; and as IE6, which didn't throw any error but wouldn't display the page, however I could view the source -- and it was the requested login page :O

So, is KM not merely spoofing an id, but also trying very hard to be the chosen browser?

I cannot see anything looking like browser-sniffing or feature-sniffing.

Gordon.

____________________
Understanding the scope of the problem is the first step on the path to true panic. [Florence Ambrose, "Freefall" 01372 January 22, 2007 http://freefall.purrsia.com/ff1400/fv01372.htm]