This weekend marks the ten year anniversary of the mrsharkey.com domain name. It was December 12, 1998 that I registered the name, and Dec 14th when the first log entries of visits to the site began. Here's an overview of the background, and a preview of the start of the next ten or so...
Way back when I composed my first web page, coding the entire thing by hand, I had no purpose other than to have a subject while teaching myself the basics of HTML. When I needed an image to practice inserting graphics, I had to teach myself to use a scanner, *then* figure out how to put it into the page. The first version would fit on a floppy disk, and was hosted in my front shirt pocket so that I could plug it into friend's computers to view it.
In late 1996, the SYSOP at the local community free network provided me with a dial-up account that included 2Mb of disk space. I put up the pages I had created and added a few. The graphics were small because I was using .gif files, which didn't have compression. That and the fact that the dial-up connection was slow and took forever to load images onto the 640x480 screen that was the default resolution for 14" monitors of the time. 2Mb didn't go all that far.
Shortly thereafter, I moved much of the site (this was all pre-mrsharkey.com domain) to a sub directory on a server hosted for my employer's radio station. Although disk space wasn't a major issue any longer, the dial-up connection when away from work was. While in my office I had what we considered to be ripping fast service, 56k Frame Relay.
Originally, my idea was to scan the entire book Rolling Homes in order to get images of house trucks and buses, and to present the book, which was long out of print. Copyright fears put this to an end, but my friend Greg saved the concept by offering to allow me to scan and post many of the photos from his albums of trucks and buses. Files were saved as .jpg's to save disk space and bandwidth (I couldn't go hog wild with the disk space, as I was kind of piggy-backing my site on the radio station's account. At some point the host's server would begin to strain and I might get booted off.)
Out of the blue, I began to get emails from people who had viewed the pages, some of them from New Zealand, where housetrucks are still a popular means of alternative housing. When I asked how they had found me, I was told that my pages had some up during a search for "housetrucks". Obviously, the busy little search robots had found and cataloged my pages, and they were now being spread over the planet.
Back then, searching for "housetruck" would yield about four pages on Alta Vista, one from NZ called "New Year's Day Buzzwords" (I just searched, it's not out there anymore), a couple of pages about house moving, and my page. Obviously, I had the corner on something unique.
Early 1998, I designed a web site for a local computer store. They weren't very impressed with it, and decided to keep doing their own web pages (which I thought were exceedingly awful, but that was their problem.) Trying to decide how to salvage the many hours that I had into the design and concept resulted in using the template for my site, completely redesigned to resemble Windows NT4.0.
December 1998, I registered the mrsharkey.com domain, and secured domain hosting on a server that was now owned by the former SYSOP for the old free community network. Gone were disk quotas, and at the time, bandwidth on the server was nil, and so was the traffic on the new domain. I began to add pages, scanning yet more from Greg's albums, and putting together book reviews that I thought were relevant.
In September of 2004, I became fed up with sporadic failures associated with the existing server, and the lackadaisical attitude of it's SYSOP, and moved the entire domain over to yet a different server, paradoxically the same one that hosts the page of the radio station where it all began.
From there, the site traffic grew, more pages were added, and you see the result on the site. As it is now, the site comprises over 93Mb in file size. Weekly bandwidth is about 2 gigabytes, serving 13,000+ pages of information.
Last winter, I ran the page code from the pages on this site though the markup validator at the World Wide Web Consortium (W3C). The point of the exercise was to detect and correct syntax and grammar errors in the HTML code from which the pages of this site are composed.
Many of the pages here go back to my original design. The basic framework of the page design was composed in Netscape Navigator Composer. As new pages were added, I would open a page, edit its contents, and save it under a new file name. Errors in the code were carried forward in this manner, and editing the code introduced new errors. Netscape Composer was rather experimental, and the developers invented markup tags that were proprietary to Netscape browsers. This code remained in the pages long after Netscape ceased to be one of the leaders in browser software.
Once I got access to Microsoft Front Page (starting with FP 98, and working up through FP 2000 and then FP 2003), some, but not all of the Netscape errors got edited out of some of the pages, but not others. Fragments of orphan code floated around in the pages, sometimes ignored by browsers, sometimes not. Occasionally, the page code would produce such a whacked-out mess that I'd have to go in and hand edit raw code to get the look I wanted.
All of this resulted in a real "dog's breakfast" (to quote our Aussie friends) of page code.
The World Wide Web Consortium validator service examines code and returns error messages, advising what parts of the pages are broken. As I ran pages though it the results were anywhere from 2 to 150 errors per page. I then hand-edited the pages, testing the validation along the way, until the validator reported that there were no problems with the code on the page. I then saved the file and ran it through a browser to see if the layout has been affected, and perhaps edited the code further and re-validate if necessary. That process resulted in most of the pages on the site being HTML 4.01 Transitional compliant, with key pages, (Home page, Site Map, Busbarn) being edited to a higher standard of XHTML 1.0 Transitional.
Even while I was validating all the pages on the site, I was sensing that the look and feel of the site was dated, looking a lot like Windows 95 in an era of Vista. Little, if any new content has been added to the site, partly because Housetrucks is a rather static subject for the most part, but also because composing, editing and uploading web pages isn't all that much fun. What I wanted was to be able to type in the info, click "submit", and get on with something else.
About five weeks ago, I did some research on CMS (Content Management Systems), and after consulting a couple of online reviews, picked one to try out in a specially-created directory on the server. After playing around with it for a week or so, I saw immense possibilities for migrating the entire mrsharkey.com domain over to the new database-driven system. All of the page titles and site directory structure could be preserved, and I was able to find a bare-bones theme that I could customize to resemble the existing site's color scheme and presentation, so that the change wouldn't be too much of a shock to the viewers. That theme is named ZEN. I've typed that name into so many file names while composing, uploading and adjusting that I've decided to name this THE ZEN PROJECT!
So far, I've only managed to port over about 10% of this site's pages to the new CMS. Each page has to be opened, its formatting code stripped away, then the content is pasted into a form on the ZEN site. Most pages then have to have their structure adjusted in a significant manner to display properly. Page titles, menu headings, filename aliases, shortcut icons, robots tags, and meta tag information have to be entered on each page.
When that's done, the page gets run though the W3C validator and the code adjusted to the very stringent XHTML 1.0 Strict standard. Once it passes validation, I move on to the next page.
I've had to learn a lot about Cascading Style Sheets, and the latest techniques of formatting page layout in order to properly validate the pages. It's no longer acceptable to use old-school tags such as font, center, and align="*" to control the viewable aspects of pages. Everything is now done with "style" and "class" statements that pull the formatting out of a master style sheet(s) for the page. Since the CMS is rather generic, and because I am attempting to get some specific behavior out of it, I've also had to resort to modifying the PHP code that runs ZEN as well as write custom scripts to provide content I think is important.
The end result is, the site will (eventually) be brought into this century, if not in content, at least in presentation. The use of the CMS will make addition of pages easier and navigation of the site much more reliable through active menus that list nearly ~every~ page on the site. The new ZEN also has installable modules that make new formats simple to implement. There is a blog module, and I have been making use of the book module to create new content. Additional publishing possibilities will come into play as I get more familiar with the CMS system and it’s capabilities.
OK, so you get the idea, the site is evolving and (I hope) heading towards a better future. It's A LOT of work, but otherwise, I could be vegging out in frot of the tube, watching DVD's to while away the winter.
