All around blogspace, blogrolls are down as it seems rpc.blogrolling.com isn’t responding to requests, even though the main site is still working. All I had to do was increase the $cacheminutes variable in my script to a high number and now it’s returning the cache from the last time it was up. I need to make it so it does that automatically, but that’s a task for when I have more juice left in my laptop.

Well before I was delivering application/xhtml+xml to all Gecko-based browsers, but there seems to be a bizarre problem with Mozilla on this page. I had been doing my Gecko testing in Phoenix, which has no problems at all, and is now the only browser receiving the proper content-type for XHTML 1.1. Mozilla 1.0.1, 1.2, and 1.3a all seriously mess things up. First it shows the hostile XML error when it loads the page, which I brought on myself by putting it into super-strict mode, but the problem is there are no errors. I’ve triple-checked the page, and everything validates. It reports the error at a different place every reload, and it points to a random line from inside the blogroll, which is just basic XHTML. Then completely randomly it will show the page, but then stop displaying the content div after the “M” in the third paragraph from the last post, but then it shows the menu, which is after that in the code. When I change the content-type to be sent as text/html it displays every time, but still stops at the same place. Wait, as I’ve been writing this it looks like it’s working again in Mozilla, but only with the old content-type. Sigh. Hat tip: Mike Little.

In tweaking the structure of things here in my quest for perhaps the perfect(ly structured) weblog I came to the problem of the blogroll, which is obviously a list but there is no obvious way to format it in list format (I know that’s the XHTML2 module documentation but it’s well-written and applicable). What follows is an explanation of how to create a semantically correct blogroll. I am a paid subscriber to the service, so there may be options available to me that aren’t available to everyone, or you may just need to do things a little differently to get the same result.

1. First you can’t have anything in the first prepended or appended to updated links, because as we will see in a moment these are actually put outside the link "goodies" we’re going to play with later. So leave that blank. You can order your links however you’d like, I used to have them alphabetical but now that I’ve lost the visual indication of recently updated ones I’ve ordered them using the "Recent" option. (Should I be using an ordered list because of this? I haven’t decided yet)
2. Set there to be no container, we’re going to have to make our own container in a minute.
3. Originally I had thought that the opening and closing list item tags (<li>) could be put in the link "goodies" section, but it doesn’t work like expected, so for right now use the contributer only "Optional Linebreaks" option and put </li> in the field.
4. For global targets I set it to "NOTARGET" but you may or may not need/want to depending on your personal philosophy on this issue.
5. In the "Link Goodies" section prepend <li> to the link.
6. You’re done! Almost. Check out what we have so far.

We’re not quite done yet, because we don’t actually have a list yet, all we have is a set of list items with no enclosure. This will work with lazy UAs, but it’s not correct. To enclose it I’m going use a modified version of the PHP method given on the site for including the blogroll, but if you don’t feel like you need the extra feature I’m going to lay out, you can just wrap the output from the stock script above in a ordered or unordered list tag.

Mike Little just dropped his blogroll in favor of his own link manager. He may be biased, but the issue he brought up of the blogrolling server provoked some thought. In terms of features, there’s no reason why I shouldn’t be using Mike’s script, but what really keeps me coming back to blogrolling is the community features, which aren’t easily duplicated without (you guessed it) a large community. So with that in mind, I began to think of ways I could cache things on my end, thereby making things easier on the blogrolling server and faster on my end. Two benefits that come to mind immediately is that it will eliminate an unnecessary request on the client-side (I’m using the javascript version currently.) and that the blogroll can be gzipped with the rest of my page for clients that support it. Here is the code I’m currently using. The code can either be embedded directly or included from another file. It caches using the unique blogroll id, so multiple blogrolls should not be a problem.

Also please don’t think that formatting your blogroll as a proper list entails having it look boring, the power of CSS allows you to customize the look of your list like never before, and though my blogroll isn’t styled (yet) the button-looking menus on your right are actually unordered lists. In fact, you will have more control over the appearance now that your blogroll is formatted as a list than if it was just a paragraph seperated by line breaks. If you have any questions about this leave a comment and I’ll be happy to help out. Of course Jason could add a list option tomorrow and make everything but the cache redundant. C’est la vie!

If you are browsing with Mozilla or its progeny, you are now receiving the entire site sans the photolog as application/xhtml+xml, rather than the harmful text/html. Those of us on less-capable browsers are still doing things the old-fashioned way.

I think I may have just found a way to get the <abbr> tag to work right under Internet Explorer for Windows. Unfortunately if this works I’ll have to deal with abbreviations now in my acronym code.

Finally! As of about five minutes ago the AMA Crystal Awards site is now live and open to the public. Check it out and let me know what you think. This is actually the second iteration of the design, the first was entirely done in CSS and the barest XHTML and it worked quite well cross-browser but it was decided that we needed more uniformity across the print and web pieces and so this is what I came up with.

If you’re interested in seeing more of Miss Richfield 1981, he has a very informative website where you can find out all about him. I’ve heard rumors that he might be flown down for this year’s awards as well. We’ll see . . .

I’ve been having some really weird problems with my mime types lately, which has put a hamper on some of the RSS and standards compliance stuff I want to do here. Every document generated by PHP was being sent with a Content-Type: text/html; encoding: iso-8859-1; regardless of whatever was being specified with header in the code. I’m still not entirely clear on what fixed it, but I believe it had something to do with no default content type being specified in the php.ini file, but a default encoding specified. This confused things somewhere and it’s all resolved now. There is the default content type of text/html which works for 99% of stuff, and the overrides now work for the rest. Now I can rest easy.