Back to the Web Roots

One step forward, two Steps back.

creator: x14km2d | build: 2022-01-01 | update: 2022-01-06

Intro

There are projects that never end and are always being rebuilt. My website is such a project. In 2010 I created my first real website/blog on another domain and kept changing it. I like to try new things, techniques and methods to evolve. But what always bothered me was how lumpy and slow the world wide web became. I also didn’t like other developments at all, so I turned away from some corporations and prefer to take care of my little digital garden. This is more important to me personally and I want readers to feel comfortable here. I don’t compete with other websites here, I just want to make it comfortable and cozy for interested people. This also includes that I always question the software, frameworks or ideas I use. I will write that down here in this project diary.

Let’s just start

Sometimes you want to try things out and after a while you realize that it didn’t work as well as you thought it would. When I made my first attempts with Markdeep, I was really excited (I still am). There is nothing negative that I can say about the project, except that it doesn’t fit me and my ideas any more. My idea of a good website is that it (almost) always works. To implement a website or blog with CommonMark is a good idea, but only as long as JavaScript is turned on. I don’t want to force people to turn on JS in their web browsers, if they want to look at my site. If other websites don’t work, that’s not my problem. Without JS even text browsers like Lynx can display my website. In addition the articles from the low tech magazine to design everything back to the roots. All in all, my project had already become too big again, so I left the videos out until I found a better idea for it. Also, the site is self-scaling again and therefore (to some extent) displayable on different devices. There is no real mobile version yet and my target group isnt the typical smartphone user. What I know from some conversations, my user tends to over sized monitors or typical heavy landscape laptops. But since I only have a small amount of data at my disposal that people send to me freely, I can only give a rough guess. Personally I also like it better and that is somehow an important reason for a change. The loading times have improved a lot, because not so many external files have to be loaded any more. But I still have to optimize small problems in some places. For the framework I chose Skeleton instead of Bootstrap. I wanted to keep the design as simple as possible and if I need more features, I can also add them later on.

There were a lot of changes in the links. I took out almost all Wikipedia articles and preferred to look for original sources. The market dominance of the encyclopaedia has become a bit too big for me in the last years and I also have no control over the links within an article. There are too many professionals who rewrite articles for corporations, governments and so on according to a demanded propaganda. My page can only be rewritten by me and with that I have solved the problem. I also removed another maker webpages links completely, because the cookie messages really annoyed me every time I opened the websites. This also applies to any other website that doesn’t want to comply with fair website use. You have been excluded from digital inclusion. Your bad luck. Frameworks, CommonMark and JS are really nice developments, but if we look at it soberly a simple website with good content and interesting projects is enough. My projects will be published again, but just like the articles I have to revise them first. There are many more details on these pages that I have to control and improve.

I have to rework them now, otherwise there will be too many of them and I don’t want that either. Especially because I want to publish some more content here. To improve the overview of my articles, I have reworked the HTML a little. Instead of the links I now show the sources as footnotes. This makes my source code more readable and also in text browsers you can read the text more fluently now. I personally like this better because I can now concentrate more on the content. When links are displayed normally, I tend to quickly open a new tab and get distracted. It also looks a little more professional now. I’m just clicking through the project of a CERN team. They have rebuilt the first web browser and there are some points I really like. The design is not so bad and the different fonts are easy to read. I like it when the windows adapt to the content. I also like it when a window always opens for a new page, so you don’t get tempted to have too many pages open at the same time. Although the CERN browser has been developed for a desktop interface, it should also work in dwm. What struck me was that the web pages are loaded really fast. Of course there are no pictures available, which reduces the loading time, but still it works really nifty. Why I write all this down here is another reason. I would like it again if websites were more modular. In Minetest, you have a common basis as a framework and the mods are then the individual websites. If you want to include a page in your project, you can simply enter a command in the terminal and download the part from the other project. Since all users have a common base, this can be automatically added by the system. Unfortunately this will probably not be possible, because people always want to have individual websites. Nevertheless I keep the idea in the back of my mind.

In the last ten years I had tried a lot with CSS, fonts and design on my website. That was fun, too. But I noticed with the website of some websites how much I love brutal/minimal web design. The loading times are great and a website can be displayed in almost any browser. Therefore I removed the CSS from my project. Also all the art and design pictures are gone again, because I can’t represent it any more and just want to put an end to this chapter. I dumped the pictures into a online account and there they were saved under the public domain license. In the next days I will completely rework the folder structure and re-sort everything. Especially with the tool photos I have so many double pictures with me that I have to re-sort everything again. Altogether I want to reduce the size of my website as much as possible. I also adjusted the folder structure afterwards. It is now created like it says in a computer scientist textbook. All pictures were named new and fit to the respective article now and I delete the bookmarks file. I had toyed with an idea how to reduce the file size of my project. I was thinking about CommonMark and my hoster offers something there as well. I created a hello.md file for test purposes and tried that. My hoster converts the hello.md in the browser to the hello.html. After that I took a look at the source code of the converted html file. The generator is Jekyll and the SEO tag fills everything in. I don’t need all this stuff and I don’t want it in my source code. But what really bothers me is the embedded JS at the end of the file. Why should I use CommonMark when the converted file is so much bigger?

Wouldn’t it be better if CommonMark is rendered in the browser instead of taking the detour via html? Do we really need html anymore? I was just thinking about who actually benefits if we follow rules. I mean now from network technologies, standardization and he computer science and programming etc. Search engines could only become so successful because millions of people have been involved in search engine optimization. Who has the better keywords, the better links. We have not worked for our own success, but always only for others. We have always strictly adhered to the rules as they were prescribed for us. But isn’t it sometimes better to consciously break the rules. Must a website always have a menu on the left side. No, it must not. It must be easy for visitors to read. No, it does not. This is my project and I have to be able to find my way around. If you don’t like my website, there are countless alternatives in the data jungle. Reclaim the Web. We should try more, test, experiment, fail, successfully implement something and dare something again. We are responsible for making the web look so consistent, so boring, because we have followed these rules. I have just read the article Web of Documents by Daniel Janus and agree with many of his statements. But if we want to create a Web of Documents, we have to ask ourselves a question of principle. Which file format should our WOD have? In my personal opinion HTML is outdated and not up to date anymore. CommonMark seems to be going in the right direction, but there are still some points in the syntax that bother me. These are only little things, but I am a programmer and at this point a little bitchy.

This sentence In my personal opinion HTML is outdated was my last standpoint. A lot has happened in the meantime. For example, I found a website that suggested that I should work with HTML. Why? Unlike what you learn in the HTML courses you can reduce this language to a minimum and the pages will still be displayed. This makes the websites incredibly small. Nothing has to be rendered in the browser, like it is the case with CommonMark. In addition even the simplest browser can display HTML. So it is downward compatible and I want to reach as many people as possible with my website. For my single documents I only use the tt and pre element. Otherwise href and img. In most cases there is no more in the source code. No search engine optimization, no JS or other controlling elements. Even CSS I left out and leave it to my web browser. I also switched to a new hoster and will see how it develops. Everything is in the data flow.

As I described a paragraph before, I have completely rewritten my website to pure HTML. This all took a little bit of time and in some places I try out even more. I have noticed that this is also the case with other (by me strongly frequented) websites. I got used to it when the font is displayed until the end of the window or the background is always white. Also the blue links do not bother me anymore. I don’t need design on websites anymore and I really like that because I want to concentrate more on the content. I have also noticed with myself how I can invest much more time in my texts and not always focus on the layout. Maybe that’s why Wikipedia is still so ugly. So people can concentrate on the articles. The use of plain HTML has changed a lot. Not only in the way I work with my articles, but also how I look at other websites. Plain-HTML is in moment a pretty good indicator for clearly good content.

Two Years later

Two years have passed and a lot has changed. I started writing my articles and project pages entirely in Markdown. I am a trained programmer, but with my website I want to create content. HTML is really bad to read and for example the square brackets get in the way of my content. In Markdown everything is clearer and I can focus on the important aspect of my work. Content. Through Pandoc I can then convert the raw data into HTML files, which are then uploaded to my web hoster. The original files stay on my desktop and I prefer this new freedom since a few weeks. I don’t want to be dependent on any provider and store my data in data silos. So that everything doesn’t look too simple I have integrated a new template into my work process, which takes care of the CSS and the structure of the HTML file. I find it quite good to have at least a minimalistic design. Besides there is only the darkmode to save elictricity.

Something I also want people to understand is the context between HTML and Markdown. For this it is necessary to provide the source files. These .md files can be downloaded and then converted to other formats, such as e.g. .pdf, using Pandoc. I think it’s good to have the possibility to learn more about a topic without having to go to the trouble of finding these files. Since after two years I still think that HTML is obsolete, I could even imagine that a new web browser will render out only the markdown files and thus the .html files are no longer necessary. I’m not talking about a crappy browser extension that is published by a developer and not openly documented as a open source project. This should already happen on the level of the web engine.

Since I was able to save a lot of work with Pandoc, I finally managed to realize a small personal heart’s desire that I’ve been working on for a long time. RSS. I still love RSS, even if people want to kill it, because you can’t implement ads and it’s open. By chance I found a RSS template for Pandoc, which I implemented in my processes. But in the process I also made completely new decisions. I don’t want to deliver content via RSS, because that would be too complex for me. I see RSS as a ping with which I send changelogs. A changelog is something that tells an interested reader what and when something was changed in a project. So you subscribe to the RSS of a project or article of mine and when there is a new update, I ping you in your feedreader software. So you don’t have to visit my website every day, but you can only come when something changes that you are really interested in.

On most websites you can subscribe to one main feed at most, but then you get spammed with articles that don’t interest you. If you subscribe to the tabletop or fashion category, you will only get news about these topics and not when I update e.g. pixel art or electrical engineering. The last thing I want is to waste the very important lifetime of my readers with nonsense. You can also subscribe to single articles to reduce the number of messages. This is something I always wished for on other websites, but never got. Now I do it myself on my website. Maybe this idea is so interesting that other developers will adopt it and spread it. RSS is anything but dead and my media data got a view boost par excellent by the built-in RSS every day.

Then there is another idea. Very few people comment on an article and that’s ok, I welcome every lurker here. if someone still wants to say something about an article, it can easily be done by email in the future. Email is a classic and not a data silo. No Discord, no Slack pi, pa, po super fancy startup which then collects data again. Mail. Very simple. For a short time, I also had this communication channel encrypted, but I need to read more into this topic. All in all, it didn’t all work out as I had imagined. So far you can only end up with comments as unencrypted mails. Please keep that in mind when you send comments. That’s where I’m at with the website so far. It’s the first of January 2022 and I have a lot to do. Shall we just get started?


Download the web_roots.md source file or follow this page with RSS. Some readers will want to know what Equipment I use. Who wants can Support my work or leave a Comment. Please use my PGP-Key for your submission.

x14km2d.com | 2021-2022 | Made for the Web1.1 with a Terminal and Pandoc.