Website Revival

In the Spring of 2014, I had finally built the website I had always wanted. Using the Tapestry framework and my web development experience, I wrote a site that could host all of my blog content I had written over the years and be extensible enough to easily change and build whatever I wanted into it.

I wasn’t manually building pages anymore, like I had been when I first made my website in 2003. My blog content wasn’t orphaned on another server, like when I moved to Movable Type in 2005. And, I was no longer stuck within the constraints of someone else’s design, like when I moved to Wordpress in 2006.

With my new website, I had built everything from scratch, which gave me complete full control over everything, which gave me the freedom to do anything. It was perfect.

Almost, anyways.

Tapestry is a very powerful framework, but for my particular setup at the time, it often crashed. Tapestry ran in Java, so to deploy the site on the Internet, it needed to be run in Tomcat. Once running Tomcat would serve the pages to whomever requested them. But, every two or three days, something within Tapestry or Tomcat would cause the site to be unresponsive. To fix this, I would turn it off and turn it back on again. Not a lot of people actually visited my site, so I don’t think anyone noticed.

For almost three years, things were pretty good. Though, as usual, my posting became less frequent and I stopped visiting my site as often as before. One day, when I power cycled my website, it was gone. There was nothing on my server. Apparently, a month or two prior, my web server had failed its migration. Well, whatever. All of my code was a mirror of what I had locally. I’ll just copy everything over again and start it up. Not so. My web server had also dropped Java and Tomcat support after this migration. No word about this at all. They just deleted all of my stuff, dropped support, but still billed me every month.

I looked around a little bit for a new web server. I tried one service, but I couldn’t get the site running. I continued to look around, but Tomcat support had become expensive for some reason. Maybe demand had dropped. I had thought about using cloud servers with Google which would homogenize my Internet eco-system further, but that was also more expensive that what I wanted. Also, it took me a while to figure out how to do it, and I was getting no where.

So, the site was dead.

In the last couple of months, I have been doing a lot of programming and using GitHub thanks to the Analogue Pocket. One day I saw that someone had “hosted” their blog on Github. Looking further, you could also have your web domain point to these pages and mask itself as a real website.

Oh shit. Could I do this?

GitHub is not a webserver, so it can’t run anything on the backend. Any webpage it hosts, would have to be static. Despite my website running Tapestry, almost all of my content is static. Yes, it uses a database, and Yes, the server is generating pages. But, the result is always the same. Anything that was dynamically generated from the server, could be changed to a separate page or rewritten in Javascript.

I had often thought that if I could simply make my website a huge mess of HTML pages, then I could easily find hosting and it would absolutely never crash or be inaccessible. And GitHub was answer, because not only could it host these html pages, but it could do it for free.

It didn’t take me long to find the command line program, WGET. When run, it will crawl through a website and save everything locally. It didn’t take me long to write a terminal script to save my entire website in HTML into a GitHub directory that could push back to the Github to be hosted.

There were still some broken links, dynamic links that needed to be changed, and the image gallery that needed to be re-written. A few weeks later, those were all solved. I cleaned up some old posts, and removed some problematic ones from 15+ years ago. I'm also hosting images directly from Google Drive. Now, I’ve written a few new posts to cover some topics from the last year. There are still some posts that I haven't written yet. You can see that progress on GitHub. I even made a custom error page!

I think everything is good enough to consider my website reopened. :)

I first opened theschlock.com in August of 2003, so in just a few more months, it'll be the 20-year anniversary. My life has changed considerably since then, from college, to studying abroad, to working as a software developer, to moving to Japan and working in public schools, to now working in a private school. It's interesting to be able to look back and recall my thoughts during these times.