I haven't looked at my own web site much recently -- it's also overdue for a round of edits that I keep putting off -- but tonight I typed in one of my URLs to make sure that what I'd written down for someone I met at the nail salon was right (I still have to remind myself to use the new location). And I was horrified to discover that my musical instruments page appeared to have gotten corrupted somehow! The last two lines of the bodhran section appearing between baryton and bowed psaltery but without the bohran <h2> heading or the rest of the text, a fragment of the krumhorn entry similarly standing by itself, and, I eventually realized, some instuments missing entirely! Aiiieee -- how long had it been like this?
At the moment, due entirely to round tuit acquisition failure, there are two additional entire copies of my web site currently accessible (which I need to replace with redirect pages Real Soon Now, but ...) so I looked at the same page in each of those, hoping to find a clean copy to re-upload. Alas and alack, they were corrupted in the same way. I fired up 'vi' on the Panix copy, and ... uh ... yeah, those quotes are closed; uh huh, there's the closing tag for that href; yup, the <img> tag has its closing angle-bracket; is that comment terminated? Yes, yes it is ... uh ... and I couldn't see what was wrong.
realinterrobang had phoned me, so she heard my
exclamation of dismay and asked for the URL. She saw the same
thing, so it wasn't just my computer. What browser was she using?
Firefox 1.0.5 (under Windows, I think?). I was using Firefox
1.5.0.7 under Linux. Wait, wait, lemme look at this in Konqueror.
And Lynx. And Opera (gotta wait for it to reload the pages I'd
been in the middle of the last time I shut Opera down). How's it
look in those?
It looks fine in each of those. Meanwhile on the other end of the phone, a similar experiment with similar results, using MSIE (I've forgotten the absurdly long version number), except that the drawings didn't show up until she hit 'reload'. Can I try it in IE? Let's fire up VNC and see what browsers (other than the obvious copy of Opera ("obvious" because that's my main browser under all the operating systems I use except for Mac Classic, where it takes second place to iCab)) the Win2K machine has on it. Netscape Communicator (4.7) and IE 5.50.4807.2300. The page looked fine in both of those as well. Oh, some browsers insist on drawing a blue rectangle around images that are links and others don't (and for obvious reasons Lynx did not display the images), but all the expected elements of the page are present and in the right order in every browser except, so far, two versions of Firefox.
I think this warrants a simultaneous "*whew*" and "WTF? WTFingF?!". Right? Firefox? Not the browser I'd pick as most likely to glitch on a page everybody else can handle!
So if I might be permitted to impose upon a few of you for a favour, could y'all take a look at http://www.dglenn.org/defs/inst.html if nobody else with the same browser has already commented, and let me know what browser you're using and: whether you see headings and text for ashiko and cornamuse, and whether there are complete entries with headings for bodhran and krummhorn or just fragments seemingly attached to other instruments? I'm interested in browsers I don't have handy, such as Safari, and I'm interested in whether any versions of Firefox do display the page correctly. I'd appreciate the data.
Possibly a clue: when I select "view page source" in Firefox, it looks like the entire file is showing up, but the syntax highlighting in the source window seems to be saying that the "-->" end-of-comment strings are sometimes being correctly interpreted as terminating a comment, and sometimes overlooked with the comment being "closed" by the end of the next comment, or the one after that, or by some random <a> or </p> tag.
I'm scared to look at the rest of my web site in Firefox now. Tho' I hadn't noticed it obviously misbehaving on any other sites on the web (that is, if it left out entire chunks of content on other sites, it picked stuff I didn't already know would be there and it didn't leave odd fragments lying around the page. Maybe my HTML comments are magic, anti-flaming-vulpine comments that the HTML tag salesman gave me by mistake.
Maybe I should go 'od -ah' or 'hexdump -C' the file and see whether there's anything different about the lines that confuse Firefox. But maybe I should do that in the morning, as I'd planned to be asleep three hours ago. That's when I'll go hunting for my HTML-validator bookmarks as well.
I'd planned to write about my day. Maybe I'll get to that later. Or maybe I'll just count on a puzzle being more interesting anyway.