3denpa

RE: A blogboard...

$>000014

Okay, I am sober now. My previous blog was about a certain idea that I had and while it was interesting I already notice that there is an issue. Basically the system I am presenting is dependent on keeping the same amount of entries. If there is a change on the RSS feeds that reduces the number of entries then the PostID would not match, in other words, if you want to use a system like this the PostID must be properly assigned and contained within a database that links it to a particular URL and Name and when a feed is fetched and parsed there must be a check against the database, so entries are not added again and the ID matches.

Of course this would not be a problem if there are no changes in the RSS in terms of deleting entries. But some RSS generators have a limited number of items, for example... Bearblog. I managed to make a poor version without the database thanks to vibecoding the fuck out of it.

image

It updates automatically by fetching the feeds every 15 minutes, and it generates the HTML table, then there is a main page with a CSS file and an iframe which contains the location for the table only. I also did a test and clicking a reply marks the post it is replying to using :target in CSS. After all it is done and the unnecessary files are deleted, it is pushed to GitHub pages, which allows me to automate it with a bash script which indefinitely loops with a waiting time of 15 minutes.

As you can see though, due to the fact that this is a Bearblog site there is the problem that the last entry that can be fetched is VRMMORPG and Unrealism. Well these are my limits1, if I knew what I was doing it'd be better... well it'd work at least. Right now it is basically useless other than indexing recent feeds from a list. Which... I guess it's kind of cool?

If there was a web-ring for blogs they should have something like this...

  1. I lied, I implemented a change and now there shouldn't be an issue with changing PostID. Basically each PostID is saved as a text-file, and thus they always link to the same thing. There is a check for duplicates based on name, as you can imagine... that's another issue. If someone changes the name of their post the fetcher will think it is a new entry... but well... there's little to be done there. At least the replies are consistent now and changing the title of the item is a rare occurrence.