W3I has been offline for over a month now. Which, even though there are only about 2.5 people who browse this god-forsaken link directory, is quite a while. This news post addresses the server's downtime, why there are only so few links, and the overcomplicated structure of the directory.
The downtime part is pretty simple: I just didn't pay for my server, lol. It cost me around 6 euros per month, and I used Gandi's (my old registrar) hosting services. And, well, I didn't have root access to the system. However, I recently bought a VPS from Vultr for $5 per month (the cheapest option for my London server), and I have root access to my server. What are the benefits? Well, it's cheaper, I can break the system more easily, I can host multiple websites on it, and I just have more control over my server. During this downtime, I also switched registrars from Gandi to Epik, because I had heard that it was better. Plus, Gandi's interface was way too confusing and didn't seem to work properly half of the time. It might've just been my own lack of understanding, but Epik's interface has worked perfectly thus far.
Now to address why links are added so rarely. You see, my process for updating W3I has always been an ineffective one. I have always focused too much on minute details, like the placement of commas, perfect sentences, and the like. That might sound stupid to you, and if it does, then I completely agree with you. I have never gotten out of this habit before, but I aim to do so from this point onward. This part of creating W3I took a lot of time which impeded my speed and discouraged me from working on W3I.
Another big problem of W3I is its overcomplicated structure. There are far too many directories. I have tried to create "the perfect directory", containing virtually every little subdirectory. But, in reality, that's just not necessary. Instead of creating an extremely detailed directory, I just make the browsing part a burden on any visitor. Let's look at one example. To get to machine learning, I have to first go to Science/, then to Formal_Sciences/, then Computer_Science/, then Artificial_Intelligence/, then, finally, I get to Machine_Learning/, but wait, there's a catch! Guess what? There are two subdirectories to Machine_Learning/... Those are Artificial_Neural_Networks/ and General_Adversarial_Networks/. (I've probably already made a typo just writing them out.) Do ya see the problem? It's way too overcomplicated! I don't even know what either of those networks are! Here's my solution. Firstly, we delete the machine learning subdirectory altogether and replace it with your old trusty artificial intelligence, then we remove the Formal_Sciences/ subdirectory. So, the new way to get to AI would be to go to Science/, then to Computer_Science/, then to Artificial_Intelligence/. This seems waaaay more intuitive to me than the previous method. Don't you agree? Anyhow, this problem plagues most of the directory and is on par with my grammatical autism. Just for fun, let's look at another example. This one's my method for going to digital maps. I'll just write out the full directory location: Science/ Social_Sciences/ Geography/ Human_Geography/ Cartography/ Maps. Absolutely insane, isn't it? I can't really think of a solution for it off the top of my head since I'm focusing on writing this news post, but I guess I'll figure it out eventually.
All in all, the website is back up and running. The only thing I don't really have yet is an e-mail server, but I will probably create that in a couple days time. As I mentioned before, I will be focusing on fixing these core issues with the directory and will be adding more links in the near future. And, as with everything I do, this post was written at 3 AM.