This calls for an upgrade. And since I won't change my interests and bookmarking habits, something else needs to be done. That's where I need your advice. Crowdsourcing the concept and stuff.
This is probably one of the greatest fails of the project. At first, I thought of it as an extension of my blog. Hence the account @stritar_net, the name Stritar's chronolog, together with the description it has. Should I rename it and try to make it a standalone "brand"? Should I openly say it's a bot (Stritar's bot or something)?
The bot currently posts ALL my bookmarks to Twitter (as mentioned, around 15 per day or 500 per month) without any selection. But it could be done. Since my bookmarks are originally tagged, it could leave out those with too few tags (since I use the same method of counting tags to determine the initial weight of content for the magazine). Or specialize in specific segments according to tags. There could even be more of them bots. What do you think?
Currently, the bookmarks I make go into a queue. I hate those Twitter accounts that post 10 tweets in 5 minutes and go silent for a day. I wanted to make it more smooth. So the queue always knows how many items it holds and adapts the frequency of posting according to it (less bookmarks in queue mean less frequent tweets). But that produces the situation where most of them are already a few hours or days out of date when they are published. A higher publishing frequency would solve some of it, but it opens a great dilemma: what's the lesser evil, over-spamming or being out of date?
The order of bookmarks posted on Twitter is determined by two factors. The number of tags and the date they were published. More tags equals more importance. Older bookmarks get published sooner, otherwise they would get even more out of date. Should I do it the other way around and post more recent links sooner? This would make some of them more interesting and up-to-date, but other worse. Breaking content or consolidated content?
So many decisions… In this case, the best way probably doesn't exists, but trade-offs can always be decided for the better. Please let me know what you think, your help would be more than appreciated. I could help back if I can.