|Projects That Weren't
||[Sep. 30th, 2004|03:56 am]
This community is really quiet and dead, so I hope no-one will object to a bit of nostalgia. A little while back I was poking through the archives for this community looking for a specific entry that didn't come up on Google from what I remembered about it. It was quite entertaining to look through the old entries and see the projects that never went anywhere despite the excitement about them at the time, so I thought I'd share! :) In no particular order:
- ESN System
The Event Subscription/Notification system was a project to generalise the mechanism for letting users know when specific things happen. Right now there is special code for sending comment notification messages and support messages and no support for subscribing to a specific entry or comment that you otherwise have no relationship with. I briefly wandered back into this a few months ago but didn't get especially far with it so I didn't post anything about it. My repeated insistance that this should be implemented instead of more special cases for different events is one of the main things that caused the phrase “Mart bug death” to be coined, I believe.
- NNTP Interface
Brad was once enthusiastic about providing a newsreader interface to LiveJournal, but no-one else seemed to be interested in writing it. Since then, the need has mostly been supplanted by RSS/Atom aggregators.
- Entry File Attachments
I don't have a good link for this, but I was reminded of it by reading about some of the S2 stuff. Just before S2 came about, it was planned to add support for attaching files to LiveJournal entries. In retrospect, it's probably a good thing that this didn't happen then because it could be done so much more nicely now with the general blobserver. It might be nice to revisit this at some point as it would certainly make posting entries with incidental images easier and nicer, and would generalise what's currently done with phoneposts.
- Separating Friend edge into Friend, Watch and Trust
As far as I'm aware this is still a low-priority goal. The “friend” relationship is overloaded, meaning “This user is my friend”, “I want to read this user's journal” and “I trust this user” all at once. Currently we work around this pretty badly using hacks with friend groups. Back when Brad originally proposed it there was no general user relationship graph, but we have that now so this is really just an issue of how to present this to users and how to migrate the existing data.
- Trust System
The trust system was intended to rate how “trustworthy” users are using a trust-graph system similar to the trust metric on Advogato. This idea was junked as it was too complicated and some users got scared of it for various reasons.
- The Great English Removal Project
This was never quite as great as it was supposed to be. After an initial burst of excitement, the English-removal patches started conflicting with other changes and generally going stale and getting ignored, so to this day there are still parts left untranslatable as well as languages whose maintainers got bored and left the text mostly English anyway.
- LiveJournal 1.0
A few times there were an attempts to make a 1.0 release of the LiveJournal Server software. However, these days it's generally accepted that LiveJournal's development is continuous and so it'll never really be “finished”. The offshoot components, such as memcached, S2, BML, MogileFS and Perlbal could, on the other hand, be packaged into releases, but it involves someone taking the time to write docs and package it up. memcached is the only component with a stable release at this time.
It's interesting to see how things have changed. Most of the idle projects are still viable, but unfortunately LiveJournal doesn't have the resources to do “cool” things so much these days as it did in the early days when it was small and there was time for Brad (and later, the rest of us) to mess around. Maybe one day…