I can't remember the exact date that the seminar was held, but the Math department at the University of Waterloo held a tutorial on creating your own web page. I thought that it was a pretty cool idea even though there wasn't actually much out there, but I had heard about the internet and was interested to learn.

I believe CN (who later became CN) was the one who gave the tutorial. The tutorial itself was fairly short and simple, but the tools at the time and the language was also very simple. We were working with HTML (version 1.0) and there was a limited number of browsers available to view our work (like Lynx and Mosaic).

The website itself was very simple: There was only one page and it was basically a "Who am I" type of website (like most beginner websites). I chose to name the site "The PIT" as I was spodding at the time and had a room on a talker called "Cat's Cradle" which I had called "The PIT" and figured that it was as good as a name as any. The background was black, and the font was a dark green, and I really didn't have any good ideas as to what I was going to do with this site..

During those first couple of months, I decided to turn the website into a quotes resource, starting with my list of Murphy's Laws that I had gotten from Dee. This actually took quite some time as I was never satisfied as to the layout or colour scheme that I was using. After several months of intensive work, I stopped working on the site and it began to gather dust..

I started to have some extra free time again in my life and I decided to start working on my website again. I had met a lot of people while spodding and got a lot of e-mail addresses and websites from those people and wanted a place to list these links. I realized that the website would be an ideal location to host this information and began working on setting up a few pages to allow me to show these links in style.

Around this time, version 3.0 of the HTML spec had come out and FRAMES had been introduced. I had a "great" idea involving this new technology where I could set up pages for all the people I knew and include their personal information so that I would never have to remember anything (like birthdates and the like). It would use frames to reduce the amount of work I would have to do (if I ever needed to change anything) and I'd make it look fancy and modular.

Unfortunately, I found out later that frames were a bear to use, and even worse to maintain as not all browsers were equipped to deal with them. In addition, I never had enough information to make these pages really useful. A few lines really was all I needed for each person even though I wanted to include a picture. After a few of months of banging my head against the wall, I took down what I had on there and kept the old list of addresses up on the site.

I did, however, update the site enough so that it at least wasn't annoying. In my first incarnation I went crazy with the "bling bling". Blinking tags, graphics (I used a rainbow divider instead of the horizontal rule), animated gifs, and a graphical counter. I toned it down a bit, just enough so that my eyes wouldn't get sore looking at my site, but it still hadn't changed much from when I first started a couple of years back, so it was a little depressing.

I also updated my quotes (I added quite a few) and I also converted my bookmarks into a webpage. I figured that if I had all of these bookmarks and wasn't at the computer I usually used, then I'd be screwed. So I made them world readable so that I can access them whenever I got myself on the internet.

I had an epiphany. I was exceptionally depressed that day from a combination of factors that I cannot recall anymore. However, I mentioned that very few people visited my site anymore (most were fellow spodders who I hadn't talked to in years by this point) and I realized that I hadn't written a journal in a very long time.

I figured that I should start writing one again. This time, using this website as the medium for my message. I figured that if nobody was visiting this site, nobody would be reading my thoughts now right? It was basically a "pay attention to me" sort of thing. Writing up stuff in such a public forum but not expecting anyone to listen.

Oh how wrong I was.

I think it was FlyingS's suggestion. Soon after I had started up my constant stream of semi-coherent tripe, I decided to find out what CSS was. I found out how easy it was to set up and realized how much cleaner (and easier to maintain) my site would be if I didn't have to set up the colours, fonts, sizes, etc. on each page.

After an initial attempt that failed, I finally got the thing up and running about a month after I started looking into this change.

Laz also began to use style sheets and was able to set up a system by which the user could select from a list of styles to view his page with. With the ability to change style sheets at will, one can tailor certain styles for certain situations or environments. Not to be outdone, I set up a method to select from multiple styles on my own site soon after.

In addition to steering me towards using CSS, FlyingS also suggested that I rewrite my code so that it was more standards compliant and to push the text formatting from the HTML code to CSS. I began working on that more than a year after I started using CSS. But I spent quite some time to rewrite my entire site for compliance, but the work paid off when I later converted my code to XHTML.

When I was studying the abilities of CSS, I realized how powerful style sheets could be. One can literally rearrange a page by setting the location of objects to various positions on the page. Unfortunately, I was still using tables to format my site, so when I attempted to change the format of my site that year, I had to rewrite my scripts and most of my pages. The results were unsatisfactory and the page loaded slowly due to changes in the navigational menu. This, combined with the poor choice of colours, forced me to revert back to the old formatting, although I did keep some small improvements. It would take a few years before I was able to reformat my page in such a way to make it standards compliant and easily modifiable by style sheets.

CSS has made my life a lot easier with upgrading and modifying my site, and I must thank FlyingS for pointing me in the right direction.

My journal setup originally had my entries on the front page. I would slowly move the entries down on the page as I entered in the next entry. This made the front page awfully long, so I started cropping them every month. The files were set up so that each file had the entries of an entire month stored. This made it easier for me (and later on for my scripts) to search for a specific entry. A little later down the pipe, on MJO's suggestion, I stopped showing the entire month on the front page and only displayed the most recent entry. This made load times for the most up to date entry much shorter although it made my workload a little bit bigger.

Eventually, I started working on a script that would allow users to go to a specific journal entry by selecting a date on a form. Anchors were placed on each day's entries so the script basically pulled up that month's file and went directly to that date (if it existed). Otherwise it would display an error of the file didn't exist or went to the top of the page if the entry didn't exist. It was a very simple script and didn't do any sort of file or text manipulation but it worked, and with the fairly small size of my entries at the time, it wasn't too much of a hassle to download the entire month's entries all at once.

However, as my file sizes grew and my skill with perl improved, I eventually updated my journal script months later to only pull up the entry for the date that was being investigated. This required a bit of data massaging on my part as well as learning some regular expressions. This made entry selection much faster and much more specific and I was very happy with the result.

I eventually moved away from displaying the most recent entry on the front page as it required too much effort. I moved the entries into a folder of it's own and had links to the journal script to pull up the most recent entry (as well as integrated the counter and date routines into it). This meant that I only had to update the archive files rather than both the archive and the main page. This made my life a lot easier, although it didn't push the journal stuff as much as it did before.

The setup hasn't changed much from then. There have been minor improvements with search capabilities, the addition of comments, some reformatting and the like, and I even changed the data storage from a monthly format to a yearly format, but the basic structure is still the same. However, unless the site goes down, the journal will keep evolving..

I can't really say if I remember when I first started using javascript on the website. The site has been up long before javascript came about, and I had started using it in the beginning because it was cool to have the most recent gadgets and additions and what not displayed on your page at that time (considering that nobody was really reading at that time either didn't make a difference). I did start to notice issues with one of my first scripts in the beginning, but that didn't stop me from using it later. I was eventually happy with my javascript clock after a few more weeks of tweaking. The original design showed the date and time in a LED billboard format but I eventually found it not as useful as I had hoped and eventually made it text only.

I have since done a number of small projects with javascript which I usually find originally on someone else's site. A list would include: opening up pop up windows, dropdown or expanding menu (which I made list and CSS based later), hiding objects, manipulating style sheets, form manipulation, cookie manipulation, and a chess game script.

I actually wanted a Guestbook for my website for quite some time. There was no way (other than e-mail) for me to know how people thought of my site, and obviously no way for them to share their thoughts with others. I began thinking of installing one in March but I didn't really get the project off the ground until I ran the idea by weirDo a month later. He was a Linux guru and got a sample Guestbook running on his own site and was able to help me get it up and running on mine.

With the success of my journal script I became very confident in my perl writing ability and began writing many perl scripts. There was the picture script I put together several days later (and eventually replaced with javascript a few months later), the finger script which I later scrapped due to security issues, a script that randomizes my e-mail sig file every time I load the mail program, a graphical date script, a calendar script, a search engine with text highlighting (which I later dropped as search engines were indexing my journal entries, cookies, a `Mech comment script (which I quickly scrapped), a link submission form (which I also dumped), a chess script which uses javascript to animate the game, but I dropped the script in favor of PHP.

One of my most useful scripts is the journal update script which I wrote soon after coming back from the 2000 World Broomball Championships.

As most of my scripts had some common functions, I put together a package which held these functions. I wanted to convert it into a real module, but because it was so complex, and so many scripts relied on it, I had a lot of trouble trying to convert it (not to mention trying to make it strict) that I didn't get to update it until I started moving away from perl. Once I was reduced to the bare minimum number of scripts, I finally made the conversion.

As my site evolved, newer browsers appeared in the market with newer features, and different standards, and this required constant work on my site to keep the code up to date and workable with the various browsers that were floating around at the time. When I had started the blog, many browsers had become too forgiving with errors on web pages, and authors had gotten lazy and sloppy when it came to designing web pages, which meant more work for the browser developer and web author. Thus, a brilliant idea was put forward by the W3C to introduce HTML compliance. By making sure that a web page was compliant with a certain version of HTML, you can be guaranteed that the page will show (more or less) how you want it to be shown on all browsers. This movement required both the authors to alter their code to meet the language's standards, and the browser developers to make sure that they support the standards. It has come a long way, and has made upgrading (like to XHTML) much easier.

I started looking at HTML compliance after having a discussion with GBC and weirDo about the transition to XML that people were talking about. It took me a couple of months before I made site HTML 4.0 (transitionally) compliant. I then starting looking at closing paragraph tags which made things easier years later when I upgraded to HTML 4.01 Transitional and then moving to HTML 4.01 Strict. From here, having the site validate as XHTML 1.1 Strict was very simple to do, so I began changing my pages over to XHTML 1.1 Strict with most of the site updated a couple of weeks later. The remaining pages weren't updated until the next year.

There were still some issues though, as changes I would make would break the validation. (Mostly typos on my end.) For example, I ended up breaking the XHTML validation because of a little bug with PHP.

Soon after I had put together my Guestbook, I decided that I wanted to take anything off the site that I couldn't control directly. The first thing that stuck out was the graphical counter which I got from the web counter site (the original site is down).

I was using one of their graphical counters, but I wanted one that looked more like one of those LED billboards. I did a bit of searching and found a very good graphical counter (from Matt's Script Archive). In addition to providing a counter, he provided scripts to take logs and also display reports from those logs. I decided to implement this a few weeks after I finished my Guestbook and was finished after a few days of work.

I continued adding log routines for a number of my scripts (like the search script, comment script, etc.). The logs didn't change much until I converted the formatting to use time stamps instead of a date format. It makes it easier to deal with in the code, but the values are not as intuitive when reading the raw information.

In addition to implementing this new counter, I updated all my pages so that they were HTML 4.0 compliant and made sure that my style sheet was nearly CSS 2.0 compliant. I've tried to maintain compliance of my pages while still trying to make them work (more or less) with as many browsers I could get my hands on. It hasn't always been an easy task, but maintaining code compliance has made later migrations to more recent versions of HTML (and XHTML) much easier.

My usual routine was to go to school late, hang around with the boys, and then do some serious work at night when people were leaving. Late at night, when my brain was fried, I would work on that day's entry. It gave me plenty of time to write up about the day's events, and it was in an environment that allowed me to write freely.

This was a good setup for me for quite some time, but I quickly determined the limitations of this. Some days there was so much going on that I would have no way of recalling all of the day's events. Furthermore, what scraps of information I was able to dig up from my memory banks never did the original event any justice. Quotes were only approximated, and I would always be thinking "you just had to be there".

For days like those, I tended to write up an entry and then slowly modify it as the day progressed. I found that a little restrictive as sometimes something I wrote earlier in the day contrasted with what happened afterward and I wanted to keep a record of that, so I used the <h4> tag (for the first time in March 1999) to denote a separate entry in the same day.

Eventually I realized that I didn't even know when those later entries were being written (so I began to time stamp my entries a year later. This proved to be a very useful change to the site as I could now keep track of my thoughts a bit more concisely and give some sort of indication as to when I was updating my site.

However, the times I used were somewhat subjective as I would use the time from my computer when I first started writing the entry. As my entries could take an hour to compose, these time stamps weren't exactly accurate, but it was a start, and later changes would address this fact..

Although the Guestbook was very useful at allowing visitors to make comments on the site, it wasn't the best forum by which one could make comments on the actual entries that I was posting in the journal. I didn't have any way of linking the Guestbook entries with the journal entries, so I began thinking of setting up some sort of script and data structure that would allow people to make date specific comments.

It didn't take me very long to come up with a storage mechanism (I was using the same method I used for my journal entries) and I had something up and running pretty quickly. To make it a bit more functional (and allow for mistakes), I added a preview option which took a bit of time to code.

FlyingS had always had a good idea as to the ins and outs of web design. So he played around with adding comments to my page and broke it. So I spent some time to figure out how to disable tags, and eventually allowed some formatting and links to be entered in with the comment as well.

As I slowly got sick of seeing pop ups (which included those on my own web site) I eventually moved away from having the comment form show up in a pop up window and moved it directly on to the journal page. (Hiding it with the help of javascript.)

My guestbook was the first script that I had running on my web site. I was pretty proud of that achievement and very happy with the way it turned out. It was the spark that fired up my desire to write scripts, to learn perl, and to continue improving on my site. It also allowed viewers to comment on my page publicly and/or just to say hello, without the need to write a more personal e-mail.

I didn't get very many people writing into the guestbook, and after I had completed my comment script it sort of fell by the wayside. Because the blog was the main draw for the site, it was much more convenient for people to post topics or comments that were more specific to an entry that I had written, rather than to go to the guestbook to write a more generic statement. The advantage of the guestbook was that the entries listed there would not disappear the next time I threw up the entry, but as with all guestbooks, more recent entries will take precedence over older ones, so a busy guestbook will have the same effect.

I began thinking of dropping my guestbook because of those reasons. I didn't really need to keep it around and I am a neat freak when it comes to my site. I hate to have extra scripts and pages lying around to maintain when I could easily clear them out and make space for other projects.

A few weeks later, I cut out the guestbook. I moved the existing entries into the journal comments (where the date of entry was the journal entry that they were linked to) and cleared out that directory. It shrank my navigational bar and cleared out some files, so in a sense, the site was a bit cleaner.

I talked about my actions a few days later and the reason behind those changes. I don't regret taking it down. As the site evolves, items will come and go, and things are expected to change.

Perhaps sometime in the future I may revive the guestbook idea, although along the thread similar to that of FlyingS's page where he uses it as a miscellaneous comment entry rather than as a true guestbook. Something like that does seem appealing, but it will be a project for some future time.

With the many forums that had been popping up around the net, I was interested in writing one for myself. I didn't have much of a reason to, until LXC and ButtChunk flooded one of my entries with comments. I realized that if more people started making comments, especially if it nothing to do with the entries themselves, it would make sense to have a separate forum to post them. So I decided to try my hand at programming one myself. Unfortunately, I was limited to the tools and knowledge that I had available to me at the time: perl and javascript. My knowledge of databases were limited at best, so I stored the information the same way I stored my journal entries: in text files.

I spent the better part of the year working on this project but I didn't really spend all that much effort on it. In addition, after I had a working version up, very few people actually used it. I realized that the forum idea had gone the same way as my guestbook and eventually stopped working on it.

It was an interesting little project though. I may come back to it someday, but there are so many free programs out there that can do so much more that writing one for myself will only be useful if I want to learn a new language or methodology.

Since it's inception, my site was always being hosted on one of the university computers. At first it was on one of the undergrad math servers, and then I moved it to the undergrad science server, before I brought it to our theoretical chemistry server. (This was all before the blog was started.) However, I realized that I wouldn't be at the university forever, and that I would have to find a more permanent venue for my site. I wasn't sure how well my website would port, but I had a taste of what would happen when we upgraded our theoretical chemistry server. I had to move my site back to the science undergrad server for about a week and a half. Unfortunately, the upgrade broke all of my scripts and they didn't all work until a month later.

With these possible issues over my head, I was a little worried about what might happen during the transfer, but then I remembered that the server it was currently residing on wasn't going down anytime soon, so I had plenty of time to prepare, and I could have both sites up running concurrently so I could test it out before making the final switch over.

I was given some impetus when I was told that I may need to move my site off the server because the computer was being shared with other researchers and it's use should really be limited to academic functions (for those that have paid for that time). I had to find a new home for my site.

It was soon after I moved back home that I blew the dust off of my old Pentium 133 system which I had gotten after my third year (which girl pilfered when I upgraded to my Celeron 300a) and decided that it was time to learn how to use Linux. The flavour of choice was Red Hat because weirDo was using it, and he was the best resource at my disposal for learning how to use this OS. His help was instrumental in getting my system up and running. Without him I probably would've kicked my machine into small bits and given up on the whole ordeal.

As I had plenty of spare time (being out of work and not really looking for a job), I was able to get my machine up and running and to figure out how to host my own website. With the help of some free domain name registration (which I heard of from weirDo) I was able to get my webserver and website up and running very quickly. Porting the site to my new box took a little longer as I had to make all of my pages more robust for transport, but a few simple perl scripts was all I needed to fix some potential linking issues.

Things continued smoothly until we started moving our office and I got my hands on a P200 MMX processor and compatible motherboards so I upgraded the server the next day. I also upgraded the hard drive when I switched to Debian and I'm pretty happy with this setup for now.

After a bit of research and some thoughts of using sessions, I decided to consider using PHP on my website. I wanted a different type of server side scripting other than perl. After I found out that my server was already set up for PHP so I started using PHP and spent the better part of a year upgrading my site. I finished most of my static pages in a day. I did have a hiccup when the pages weren't valid XHTML but I was able to figure it out and started flipping the switch to PHP a couple of days later. I started working on my perl scripts a week later with some of the navigational stuff done the next week and some of my admin functions done a month later (with some date and time improvements). I started working on the last of the scripts after a long break.

I had some issues when I upgraded to Debian, and Apache. I had to hold off of upgrading PHP to PHP5 until it came out for Apache 2, so that actually made my life easier.. I eventually started improving on the site with some custom PHP error messages and automatically generated thumbnails (with the help of GD). I also needed to have functionality to read files and write files so I started changing groups and permissions and converted most of the rest of the perl scripts to use more PHP. I was down to a few remaining perl scripts a few days later. Due to the nature of those scripts though, I doubt I'll be able to convert them easily. I have no worries though, the site is basically completely running on PHP now.

When PHP5 became available as a module for Debian, I jumped at the chance to install it. I only found one small bug with the code that I'm using, and it was pretty easy to fix, so I was pretty happy with the upgrade.

I eventually decided to start putting together an ongoing history of this website. There have been so many changes made here over the years that I didn't want to forget where I had started from. Fortunately, most of the information is stored in the journal, but because of the nature of my blog, it can be quite difficult to extract the web specific information that I was interested in. Thus I decided to write a separate section just for changes to the website, and keep like changes together under the same header.

PRQ suggested one day that I should consider using RSS or Atom for my blog. I had no idea what it was or what it was for, but knowing PRQ, it probably was something useful. I started doing some research on it the next day and had quite a bit of help from PRQ.

I started looking for some sort of feed aggregator a couple months later and began work on my own RSS and Atom feeds the day after. However, I wasn't particularly happy with what I had, so I shelved the idea and left it on my list of jobs to do for the next year.

I eventually came back to setting up a feed after Laz asked me about it. I did a bit of looking around and found out that Atom 1.0 had come out so I looked up the specifications for Atom 1.0 and after a comparison of RSS 2.0 and Atom 1.0, figured that I should just set up a Atom 1.0 feed and ignore RSS.

I also added a link to the Atom feed a week later.

The feed is still under development, but once I've reorganized my data, I should be able to utilize the feed to it's full extent.. *cue maniacal laughter*

Once I realized that the size and number of my Shallow Thoughts and Revisions was going to be so large as to be unpractical to display on a single web page, I started separating the entries out into their own pages. I noticed that maintaining this type of structure would become difficult over time, so I modified the entries and my journal script so that my data could be stored in a single file, but displayed individually. The chances that one would want to need to display multiple entries at once was small, so I figured that this was the best way to go.

I never liked keeping a separate directory to hold thumbnails of my photographs as well as directory to keep the original photos. I considered it a waste of space and resources. My ideal situation was to have the thumbnails generated automatically but there were two big problems: I didn't know how to do it, and I didn't know if my server could handle it.

I began thinking of generating my thumbnails on the fly with PHP (with the help of GD) and started testing it out a few days later. After some thought, I decided to keep the changes. The increased download time due to the time required to generate the thumbnails was noticeable, but I didn't think that there would be a lot of people looking at those photographs anyway, so I figured that I'd save myself some space and clean out those thumbnail images from the site.

Ever since I set up my Linux server, I've been using the Red Hat distribution since weirDo was using it himself and I wanted to have someone around who knew how the OS worked. I started with version 6.2 and was quite happy with it. Eventually, I decided to upgrade to version 7.3 a year later, just before version 8.0 came out.

Over a year later, I heard the news that Red Hat was not supporting their Red Hat Linux line. I realized that I had to find a new distribution of Linux to use on my server. A few months earlier I had already done some research into finding Linux distributions that I could use for some of my older computers and decided to go and check them out. After a bit of research, I decided to try out Debian GNU/Linux. I played around with it a bit with installing Debian twice: once on a P166, and then again on a P133 Eventually, I found some time and spent most of the long weekend to install Debian. While I was installing Debian I noticed that Apache 2 was out so I installed it as well. One thing lead to another and I was ended up replacing all of the software that I was using on my Red Hat server with open source versions from Debian. The biggest change was switching from Pine to Mutt, but in the end, I think it was all for the better.

I had originally kept all of my data files, scripts, and front end pages together in the same public directory. Although files that probably should've been inaccessible by the public were exposed to the world, it made backing up my site very easy as I only had one directory to worry about. I did put together an elaborate permissions setup which, in conjunction with .htaccess, I was able to prevent access to some critical files that, if changed, could severely damage the site.

Eventually I realized that I would have to move the data out of the public directory anyway, as I plan to move most of the content onto a database. As such, I began to splice out the files that could be moved out of the public directory into a more secure folder, away from prying eyes.

There are still some administration pages that I've left on the site, but these pages are required if I plan to continue to maintain the site in an effective manner from insecure locations. There is protection of course, But a good and dedicated cracker can always find a way in..

I have, for quite some time, been using post processing to clean up my data before displaying it to the end user. This allowed me to include mouse over tricks and then titles to links to add more functionality to the site. It's also allowed me to alter some of the information as it gets processed so that I can eliminate anything I didn't want displayed.

I eventually realized that I could also use it to automatically update the entry with any commonly used links instead of storing the information in the entry itself. By doing this, I could significantly reduce the amount of space required for storage, standardize the link for a specific topic/contact, and also highlight those words even when I forget to assign the link (which was fairly often).

I've spent some time clearing up the archives and also started using it to label some frequently recurring nicknames so that I don't forget that person's real name. However this is a never ending task as I meet more people and forget about some older acquaintances.

I've wanted to use XML and XSLT to improve my site and make it a bit more modular, but I hadn't had the time, nor did I have any samples that I had easy access to to play around with. This changed when Derelict put together a statistics page for the Quake III games we had a couple months earlier. His page was very simple, but it was also very powerful, so I spent some time putting together an XSLT file for the 'Mech XML record sheets that I had made from The Drawing Board. It seemed like an ideal choice, and after I was done, I was able to replace the old static pages I had for those 'Mechs I designed (oh so long ago).

I've been meaning to set up some sort of spell checking method for my site for years. I looked into getting some sort of spider to do the work for me, and eventually found out that vim could do it for me but that only worked for pages that were already written. I knew that I needed something that could spell check my entries as I was entering them but it took me a long time before I got around to installing the necessary libraries and extensions to get it to work. Once that was done, the actual function itself was pretty simple, although it doesn't have that much in the way of fancy bells and whistles..

I've been thinking of writing up a script to generate a dynamic graph using the GD library ever since I got rid of gnuplot and the graph that displayed the number of hits per day on my statitics page. It didn't take long for me to figure out the proper calculations and formatting for the image so that I could get something close to what I had before, but the amount of work I need to do at this point to make it any better is a little prohibative, so I'll be holding off on doing any more changes to this script until after I've done work on some of the other projects I've got on hold.

Over the years, I have been collected quite a number of links that people have forwarded to me. These links are usually sent to me frequently and often, so I had trouble compiling them into a list. In addition, I only had passing interest in the vast majority of these links so keeping track of them was not a very high priority. Therefore, I created a new subject header in the blog titled "Chain Links" which was used to differentiate my normal entries from entries which were just a list of links.

I eventually got annoyed with the fact that these "Chain Link" entries were starting to dominate my blog so I got around to writing up a script to allow me to enter these links into a file, and then use Atom to display the resulting information as a syndicated feed.

Ever since I came across a website that had a dropdown or expandable menu, I knew that I wanted to implement one on my own site. The first menu I created used some javascript and a lot of HTML structures which made the entire menu large and cumbersome. This was difficult to implement at the time due to bandwidth and space constraints.

I shelved the whole idea until I came across a second menu which used only CSS (and a bit of javascript). In addition, the menu itself was structured around the standard HTML list, which allowed older browsers to render it fairly nicely. I decided to implement the new menu and see how it would fare. I liked the menu as it scrolled with the screen and allowed me to access all the pages on my site without needing to switch between subsections. Unfortunately, I found out that it had a tendency to screw up on some types of browsers.

After several months, I removed horizontal dropdown menu, and while I was in the process of reorganizing the page layout, I reimplemented a simple expandable menu (like my first one) instead. With my bandwidth issues less of a concern, this method seemed acceptable to me. In addition to replacing the dropdown menu, I was able to remove the submenu, which freed up some space on the page.

I'm happy with it so far so we'll see how long this menu lasts.

I have been meaning to try out Ajax for quite some time now. The idea has merit, but I did not know if it would help me or improve my web page.

Well, Laz beat me to the punch when he put up a list of friends which links to their syndication feeds using Ajax (or at least a method very similar to it). I decided to take some time during my vacation to understand how it works.

I eventually found a way to make it work using javascript and PHP only, but I think this may lead to an overloaded server on my end. I may consider going back to the usual XmlHttpRequest method if this becomes a problem.

At least I know it can be done, and relatively quickly as well. Now to sit back and watch the bugs roll in..

I have spent the past couple of weeks going through my old data and assigning time stamps to every entry and comment so that I can eventually do two things: move towards entry based comments (rather than date based comments), and move the data into a relational database. Towards that end, I have also removed a lot of the line feeds that were in the files, so everything should be fairly consistent now.

The next step is to set up the entry based retrieval, and then allow for entry based comments. This would also require me to go through the old comments and assign links from the comments to the specific entries, but at least I am getting there. I thought it would have been a lot harder than this. I just hope I did not corrupt any data.

I had been debating with myself whether or not I should continue hosting my own photos, and I eventually decided not to. I decided to do two things. One was to move the photos off site and use an image hosting site to store and display those pictures. This would reduce my bandwidth and storage requirements. Secondly, I wanted to drop the pictures section and incorporate the images with my journal. This would provide context for both the pictures as well as add visual cues for the entry itself.

I eventually settled on using Picasa to host the images (instead of flickr which I was originally going to use). It did not take as long as I had expected it to, since I decided not to upload a large portion of the photos that I originally had displayed. Also, I put together a list of entries which have images attached to them so that it would be easy for people to see them. We'll see how this experiment goes.

I was thinking about switching my comments from using my home grown markup method to something a bit more standard. Eventually I decided to use BBCode and made a quick switch. My implementation of BBCode is mostly standard, except for a few changes I made here and there (like the quotes).

Why did I make the change? I am not too sure myself. I just like to follow standards.

While working on my BBCode changes, I inadvertantly triggered another upgrade for the website. I downloaded a Firefox plugin to validate HTML which led to a number of validation errors and warnings. Specifically, the W3C validator warned that my XHTML should be served as a Mime type of application/xhtml+xml. I was able to get a PHP fix but it broke some of my javascript.

I discovered that the upgrade broke the document.write call. With some outside help, it took me a couple of weeks to fix it, only to find out that the innerHTML function was not supported, so I found another way of dealing with it.

With all of those changes done, I flipped the switch, only to discover that the javascript still did not work. It did not take me long to discover that it was due to the comments used to hide the javascript form browsers. Once I discovered this, I was able to get the javascript to work. I am still not out of the woods yet. I am pretty sure that there is some hidden code that will break the site (like using -- within a comment).

I have been taking some time to move some of the functions that I have been using in some of my code into classes so that I can more easily access them from other pages. I also added some more validation and processing to my entry forms including Safe HTML checker, which I will need to modify to catch a few glaring XHTML errors.

I never really had a problem with spam, so when I started getting some spam comments at the end of last year, I paid little attention; It was just a one comment.

However, the next year gave me no reprieve as I got another comment next month but again, nothing happened for quite some time so I did not bother worrying about it.

Then a couple of days ago I started getting a torrent of spam on a particular entry. This time the spam did not stop. Every hour a new comment would be entered and I would have to remove it from the system.

I started looking at ways to reduce the spam so I took a look at CAPTCHA and it's successor reCAPTCHA as a possible method. But I never liked using these images as it is just plain annoying! When FlyingS suggested that I alter my comment form to fool the simple bots, I decided to try it out, and so far it seems to have worked..

I figured that it was time for me to start using Google Calendar in earnest because I was not updating my own scheduler, and the convenience of Google Calendar far outweighed the time and effort it would take for me to do something even remotely similar on my own box.

As such, I moved all of my appointments over to it. Since I did not have too many things booked, it did not take long, and now anyone can see what I am doing over the next year!

Well, maybe not anyone?

My server died around this time. It was getting old and I had not touched it for years, which is normal for a system that is rock stable. Unfortunately, it also means that I do not generally tend to the machine so failures like this will catch me off guard.

I need to replace the machine, but since the blog has not been updated very often due to the wedding preparations, I don't think that it is that high a priority.

Besides, who is still reading this?

Well, with the server down, my cron job that occasionally updated my DNS settings stopped, and I allowed my control of the BEGghILOS2.ath.cx domain lapse. Normally, they would send me warning e-mails to remind me that the domain name will expire if untended, but those e-mails had been rerouted to my spam folder, and it was not until I had set up my new server before I discovered the problem.

I tried to renew my domain name, and was even prepared to pay to have it reinstated, but apparently the .ath.cx domains are no longer available. Christmas Island has decided not to allow other parties to use their top level domain and it is impossible to get it back.

So, what to do now?

I have purchased a couple of virtual servers from Cloud At Cost to play around with and to rehost The PIT. The cheapest servers they have allow for a single lifetime payment of $35, which is ridiculously cheap, considering that the specs on the machine are way more than enough to host my website. In fact, it's good enough to host The PIT as well as our wedding website without a hitch.

Now, when we downloaded our wedding website, we were given a file which basically had a static version which one could effectively view without requiring a web server. This made setup incredibly easy and allowed for a much simpler web server to be installed to host the site. Now that I realize that I will not be working on The PIT anymore, I am deciding on how much effort I would like to put into setting it up again.

Our wedding website has been up and running for a year now and I figured that it was high time to get The PIT back up on the internet. The first thing I needed to do was to set up the web server again, and thanks to some notes I took and Google, I was able to quickly get the website up, but had to spend some time to change some of the settings so I could get the final 10% of the website back into working order.

I have come across a number of methods of effectively downloading a static version of a website. I have decided to try doing this to The PIT using some open source software called HTTrack. Initial testing looks pretty good, although I have quite a bit of dynamic content which does not lend itself well to generating static pages in this way. I'll be spending the rest of the week converting some of my data to XML or manually creating static pages so that I can finally put this project to rest and rehost my website as a monument to my efforts from the early part of this millenium.