Charles Sullivan posted on Sat, 19 Dec 2009 05:27:38 -0500 as excerpted: > Older versions of PAN allowed you to create internal folders and save > articles in them. Clicking on a folder then displayed the articles > saved in that folder just as if the folder were a newsgroup. > > This feature was omitted in PAN version 2. My opinion that this was one > of the most useful features of PAN is apparently not shared by many > others as requests for its restoration have fallen on deaf ears.
I wouldn't say "fallen on deaf ears" so much, as after the original just- over one-year sprint with pan-0.90 thru 0.131, there's been little development at all. 0.132 and 0.133 were very minimal updates, mostly fixes to keep them compiling against the latest glib and other libraries, with the latest gcc, using patches already developed tested and deployed by the various distributions. There were a couple other small patches, including one to properly add pan to the freedesktop.org standard menus, but no major new code at all. Charles hasn't had a lot of time for development, so what has happened recently is that K Haley's community repository has been where the new patches are going, with a number of us running sources straight off his git repo. After they're tested there some, Charles can pull them into the official git repo, and will ultimately create a 0.134 and there's talk of finally getting a 1.0. But even that doesn't have, at this time, a lot of new features, mainly because nobody with the skills seems to be interested enough in the pan code to take the time to develop the patches and post them so khaley can incorporate them. khaley doesn't seem to have the time to go to that level either, tho he does run a community repo and do patch integration, for which we're all grateful, as at least pan's being maintained at some level now, not almost abandoned. But there's a big difference between maintaining current code, doing small bug fixes and incorporating patches as others create them, and actually having the skill and being willing to invest the time in coding major new features, which is what something like local storage folders would be. Then there's the fact that the old implementation had a few notable bugs, one of which was that as pan tracks messages by message-id, in ordered to store copies of sent posts, it had to add a message-id before sending them, and then, it would know it already had that message so wouldn't download it again. Thus it was impossible to see your message as it actually appeared on the server (the multiple pan instances feature would make it slightly possible now, but still too difficult to be practical), only as you sent it. If it was mangled in transit, you, the poster, had absolutely no way of knowing that. Personally, I'd rather have my posts downloaded along with all the others from the server, and see how they actually appear to others, than have the local folders feature but get that bug that came with the previous implementation as a result. Also note that in the previous implementation, expiration was locked to that of the server. New-pan allows you to set your own expiration. Here, I set no-expire and have posts going back several years, tho that does mean setting the cache size manually, directly in the config file, since there's no GUI for that and the default 10 MB would fill up quickly even with text-only. Thus, it's possible to keep posts as they appear, directly in the groups in question, if desired, rather than having to put them in local folders, in ordered to save them. That was the major reason for local folders before -- so you could keep stuff without having it expire when it did on the server. That's no longer needed since expiry is entirely under user control now, and you can keep stuff indefinitely, directly in the newsgroup itself, if desired (and if you have disk space). So (1) the biggest reason for local folders has disappeared, expiry is entirely under user control, and (2), we avoid various bugs by leaving the posts directly in the newsgroups they download to, instead of trying to do the local folders thing. Together, that means there's both little need for the feature, and a discouragement for reimplimentation as it was, so it's little surprise that it has remained somewhere down below actual implementation priority, when people DO have the time to code up and test new features. It's not deaf ears, it's a question of prioritizing all too rare real new-feature development time, on stuff that will really make a difference, without reintroducing the bugs of the previous version. You are right, tho, in that not so many people really need local folders now, so the request level is lower, because the real necessity of it in terms of having a place to keep messages where they wouldn't expire is now gone. -- Duncan - List replies preferred. No HTML msgs. "Every nonfree program has a lord, a master -- and if you use the program, he is your master." Richard Stallman _______________________________________________ Pan-users mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/pan-users
