I’ve added a CSS fix to the site to work around the Internet Explorer expanding box bug. In a nutshell, IE will incorrectly expand a fixed-width block element if the block is populated with oversized content. On this site, that means that the sidebar can drop below the enter content block.
I employed the overflow:hidden workaround on my left-hand content block as described in the linked article on Position Is Everything. Thanks go to Titus for reporting the problem.
As of approximately 9:30 PM this evening, the entire site is officially running on a combination of the Smarty template engine and the PEAR DB database abstraction layer. I expect productivity gains in the long term via cleaner, more portable code. I quickly found that Smarty is particularly effective at changing the way you think about scripting your pages. Thanks go to Titus for bringing Smarty and Pear to my attention.
In addition to the aforementioned changes, roughly half the site now feeds from a PostgreSQL backend rather than MySQL. The rest of the site will soon follow. I’ve planned the migration as a direct precursor to a comments system. Rejoice! (Or don’t.)
I picked up a couple tickets for the Cowboy Mouth concert at the Roxy Theatre on October 30th. Ticketmaster wanted an obscene 30% service charge online so I made a special trip to the Roxy’s box office this afternoon to save some money.
After trying to wade through complex examples in the Apache guide, I was able to grasp the basics of Apache’s mod_rewrite capabilities with the help of a straightforward guide at HTMLSource. I wrote and tested some simple rules on the development site before discovering an unforseen interaction with my web host’s preset rewriting rules.
I think it’s time to look for new hosting. My hosting expires at the end of this month, and I’m finding myself wanting more control over the internals of my site. BlueHost and 1&1 look interesting, but I need to do more research. I’m open for suggestions.
Progress on the site’s basic image browser functionality went well. (I’m about 95% of the way there.) I got past most of the stumbling blocks today. Images must be scaled to fit inside the DIV elements (otherwise, images extend beyond the boundaries of the DIV). Scaling the images in itself produced inconsistencies between browsers. The Mozilla and Safari browsers exhibited no apparent problems, but Internet Explorer failed to scale the image properly until the browser window was resized.
A severe thunderstorm rolled in abruptly in the late afternoon. A couple brief power failures put web development on hold until the evening. Mom was out running errands during the entire thing.
The time has come! I’ve finally transitioned the site to XHTML. After toiling over the blog entries for the better part of the last two days, all 252 entries validated for XHTML 1.0 Strict; it was a rather arduous process.
I started with a bunch of regular expressions to supplant HTML markup with Wiki markup. This took care of the bulk of the work. I finished the rest by hand (fixing inconsistencies and grammar along the way).
Perhaps the biggest hurdle involved special HTML entities like the em dash and the ellipsis. TextWiki inexplicably spat out these entities as plain text and I feared some unforeseen incompatibility. After some research and a little trial and error, I discovered that the problem, in fact, lie not with TextWiki but with PHP’s htmlentities() function, which apparently fails to include a lot of the higher ASCII characters. I was able to resolve the issue by writing my own simple function to restore these special entities after TextWiki and htmlentities() have run their course.
I also had to disable TextWiki’s anchor targets (e.g. target="_blank") to validate as XHTML 1.0 Strict. I don’t typically use the target attribute anyway and I was slightly irked that TextWiki took it upon itself to insert one. I couldn’t figure out how to disable the rule with TextWiki’s disableRule() so I just disabled it manually using:
unset($wiki->rules['url']['conf']['target']);
I’m still having issues with character set encoding that need to be resolved. The problem lies with child directories not inheriting the proper php.ini file. (Every since my hosting installed suEXEC it’s been all screwed up.) I’ll continuing working on this and any unresolved validation errors over the next few days.
Wikipedia resolved my confusion regarding the correct usage of the em dash and the ellipsis (including HTML entities!).
Mom also showed me a good book by Patricia T. O’Conner entitled Woe Is I. It’s a thorough grammar handbook covering everything from word usage (oral vs. verbal) to punctuation (what to include inside quotation marks). I was interested because I’ve started preparing my database for the move to XHTML and TextWiki and this will be a good opportunity to fix problems.
Titus introduced me to a couple cool PHP libraries: Smarty and Text_Wiki. Smarty is a templating engine designed to completely separate the presentation layer (i.e. HTML) and the application layer (i.e. PHP code). Text_Wiki is a PHP library capable of transforming text in Wiki format to valid XHTML. I messed around a little with both but had neither working at the end of the day. I don’t have full access to my hosting so installation could require special considerations.
I also tried to do a little 3042 homework but didn’t really finish that either. Where does the day go?
I added an RSS feed to the site at http://rss.acpstudios.net/. RSS is a spin-off of XML designed for web content syndication, which makes RSS ideal for news and weblogs. This site is primarily a vehicle for my blogging so in a sense the addition was inevitable. A little friendly competition from Titus also helped to motivate the process. To take full advantage of RSS you must have an RSS aggregator, otherwise you will just see an unformatted XML file. I use the freeware NetNewsWire Lite for Mac OS X.
Edit: RSS link removed
I feel old all of a sudden. I did some investigation into the issue of whether two spaces or one space should be inserted after a period at the end of the sentence. I have always been taught that the former approach was correct and this site has reflected that; but, it seems the times have changed. I read a thorough report on the subject, which would indicate overwhelming support for the usage of only one space. The report establishes that the two space approach was a relic of the typewriter days; a hack, if you will, to improve readability. You can still see the effects of using two spaces when dealing with monospaced fonts. They make some valid points for the usage of a single space and I have updated this page accordingly.
In the morning and early afternoon I made final preparations for statistics before heading out to take the exam. The problems were daunting at first, but I struggled through it and came out feeling cautiously optimistic. After the test I went to the third floor ECE labs in the CoC to work on my VLSI project. I only worked for about an hour but I took care of some minor design changes and general housekeeping so it was time well spent. The next deadline for the VLSI project is now less than two weeks away, which has me somewhat worried.
Having all but ran out of food stores, I begrudgingly made a trip to Publix in the evening. I grabbed a tuna melt from Blimpie on the way home for dinner. Titus and I finished the D.N.Angel series. The ending was good and typical of the series with only a few small surprises. Looking back, it’s a very enjoyable series with a little action, light humor, and a cute romance at its heart. Highly recommended.
I went through the CSS yet again and changed all the font-size attributes to use em‘s for relative sizing as described in the Sane CSS Typography tutorial. This technique should yield consistent sizing across most browsers. It’s also easier to manage as a designer than using the smaller/larger inheritance model.
The front page post info text has also been fixed in IE browsers. The absolute text positioning properties were removed. This also fixes a case where the post info could overlap with the subject text if the browser window was too small.
In addition, all pages should now validate without error for HTML 4.01 Strict. Visitors can use the links to the appropriate validators at right to validate the current page at any time.
