HTML - 56 errors, 41 warnings
CSS - 601 errors
Feeds - Valid but with Recommendations
Web Content Accessibility (WCAG) Checklist Report - FAILED
Section 508 Checklist Report - FAILED
Apologies for not telling you that I had created a case for this (and all your CSS / W3C / HTTPS thrads) a couple days ago. Below is the response:
These CSS warnings are not a problem. We include some CSS properties that are only compatible with some browsers (such as border-radius). Browsers that support them will render the CSS, and others will ignore the properties. This is very common practice in building complex web pages and is not a threat to the performance or stability of the site.
I will follow up with our hosting team on this issue as they are more well versed in these types of issues.
One of the helpful checks posted was regarding broken links. The Link Checker found that the first announcement in the Security Awareness, Home, and Help spaces include a link to the space Community Interface Feedback which does not actually exist. This announcement should be updated with a correct link.
Let me know if you have questions about the details above.
I am goingto forward this case on to our hosting team to address the concerns with using RFC 5746 for SSL.
Let me know if this is enough info for you or sufficiently explains things.
April, thanks for taking this seriously and investigating it. The messages that came back from the W3C validation process clearly included a large number of trivial or non-critical issues, but in amongst them were some that looked as if they might be important. The point about the HTML messages I thought was that any invalid page coding instructions might be having an impact on Internet Explorer's ability to parse the code and render the page. Firefox seems much more tolerant of the code, possibly because it just skips over anything it doesn't understand or can't process. I'm not so sure about IE (and I don't know at all about Chrome).
As for the links and redirects, they probably would have an impact on page load times. Not huge, but probably measurable.
The major problems with the pages still seem to be
1) Calls to other servers to get items needed for the pages. I'll have to see what's coming from where.
2) Scripts downloaded with the pages. *Something* is running on my PC and chewing up the CPU all the time a Community webpage gets displayed in IE. Not so, I think, in Firefox. But Firefox on my PC has NoScript and QuickJava as extensions, meaning i can fine-tune what's allowed to run, and NoScript is currently blocking something from addthis.com (and was blocking sacore until I let it through).
btw, the Error Console in FF (which I am using right now) isn't showing the RFC messages today, so perhaps that's been fixed.
April maybe you need to remote into Peter's PC and see this happening?
From April's reply it's clear that someone at Jive is looking at this and I think that the RFC issue for a start has been dealt with. Possibly also something has changed on the webpages, because although loading a page is still slow (and it seems to load in segments, with pauses in between) once loaded the CPU activity in IE has noticeably fallen. At present the CPU is in the range 6 - 15 per cent when a page is displayed but is otherwise inactive, instead of the 15 - 40 per cent that it was a day or two ago. It still soars to around 100% whenever I try scrolling a page.