Trn is the best method I’ve seen for making long discussions usable on-line.
The .newsrc (a record of your preferences and what you’ve read) is the most crucial part.
Usenet was set up (at least by the time I was reading it) to be read on a telnet screen. Your computer probably has telnet, and you’ve probably never heard of it. Telnet screens were ASCII-only, and they could load pretty fast, even on a 56K dial-up connection.
The result was that it made sense to display one article (post) at a time, and that meant that a record could be kept of which posts you’d read, and that meant you’d never see the posts you’d already read unless you’d asked for them. [1]
This meant that you’d never have to hunt around for new posts in a big discussion.
LW could have something like that feature just by adding “recent comments” at the original post level, instead of just having it for all the original posts simultaneously.
However, trn had more than that. It had commands for navigating up and down the tree structure of a discussion, and the ‘t’ command for displaying the tree structure. That was only useful for medium-sized discussions—a browser-based system could do much better.
When you opened a newsgroup (discussion community), it displayed a list with the authors and subjects of each post you hadn’t seen before, and you could mark which ones you wanted to see.
I didn’t use killfiles (marking posts to read satisfied me), but they added a lot of flexibility. You could tell the .newsrc to never show you anything by a given poster ever again. You could “thunderplonk”—never see anything by that poster, replying to that poster, or mentioning that poster. You could never see posts with specific words in the subject lines, posts which had been sent to more than some number of newsgroups.
Killfiles could also be used to make sure you did get shown articles with specific features, but there was less drama about that, hence the name.
Slrn added scoring to trn.
[1]I don’t think there was any way to keep such a record efficiently for browsers until JavaScript came into common use.
Trn is the best method I’ve seen for making long discussions usable on-line.
The .newsrc (a record of your preferences and what you’ve read) is the most crucial part.
Usenet was set up (at least by the time I was reading it) to be read on a telnet screen. Your computer probably has telnet, and you’ve probably never heard of it. Telnet screens were ASCII-only, and they could load pretty fast, even on a 56K dial-up connection.
The result was that it made sense to display one article (post) at a time, and that meant that a record could be kept of which posts you’d read, and that meant you’d never see the posts you’d already read unless you’d asked for them. [1]
This meant that you’d never have to hunt around for new posts in a big discussion.
LW could have something like that feature just by adding “recent comments” at the original post level, instead of just having it for all the original posts simultaneously.
However, trn had more than that. It had commands for navigating up and down the tree structure of a discussion, and the ‘t’ command for displaying the tree structure. That was only useful for medium-sized discussions—a browser-based system could do much better.
When you opened a newsgroup (discussion community), it displayed a list with the authors and subjects of each post you hadn’t seen before, and you could mark which ones you wanted to see.
I didn’t use killfiles (marking posts to read satisfied me), but they added a lot of flexibility. You could tell the .newsrc to never show you anything by a given poster ever again. You could “thunderplonk”—never see anything by that poster, replying to that poster, or mentioning that poster. You could never see posts with specific words in the subject lines, posts which had been sent to more than some number of newsgroups.
Killfiles could also be used to make sure you did get shown articles with specific features, but there was less drama about that, hence the name.
Slrn added scoring to trn.
[1]I don’t think there was any way to keep such a record efficiently for browsers until JavaScript came into common use.
The .newsrc also let you switch clients, while keeping the same records of what had been read.