ronald@robobar.co.uk (Ronald S H Khoo) (10/03/90)
A question for the C News authors: Could you say what you consider to be the future of the included newsreader ? You don't seem to have abandoned it completely seeing that it actually has been patched since the release :-) Presumably it has already diverged sufficiently from the UNSW news package that backwards compatibility with UNSW is no longer of major concern? As it stands, it seems to be useable, but incomplete. I think I'd like to see a few features REMOVED from it (like -u -- this will shrink it a LOT :-) but made useable as a complete newsreading system. The stated main problem (lack of pure USG support) doesn't *seem* to be a problem -- (though I haven't tested it on a pure USG yet but I will) it works _almost_ perfectly if you stick the sgttyb declaration into the #ifdef and compile WITHOUT any terminal ioctls at all -- I actually prefer it this way -- leaves no chance of screwing up the tty settings :-) (it's slightly better if you remove the manual echoing code in this case) The unresolved issues which bother me currently are: 1) readnews expects postnews to handle some weird arguments for followups (which it doesn't :-). If these arguments are going to stay put, there's a case for teaching the shell script postnews (and also rn's Pnews maybe) to understand at least what readnews spews out. 2) The mail reply addresses are total rubbish. I'd like to see its logic conformant with newsreply's. Really, a proper mail reply system related to postnews is needed, although this requires the mail equivalent of inews -h which some systems may not have :-( (I use mush -h) 3) cancellation (if postnews worked :-) is a bit abrupt and permissive :-) 4) The 'p'ost command shouldn't really issue an "include this article" request to postnews should it ? 5) reading the .newsrc isn't done with fgetmfs :-( At some stage I'd like to hack 1-4 of the above into my own copy, (5) is really necessary, but is also a VERY major hack. I'd like to know if a) anyone else has done it b) whether anyone else is GOING to do it c) what the C News authors would like to happen d) what others would like to see happen. I bet most of you never ever compiled it though :-) Before Henry chips in, (:-) I'd like to remind every one else that my purpose is purely to make it work, add NOTHING that is not ABSOLUTELY required for it to completely work. (I'd say that bits that can't be made to work in a simple manner should be REMOVED!) If you want more features, you know where to find the source to "nn" :-) Now, who's first to point out where I've gone wrong and completely missed where all this stuff is ? :-) -- ronald@robobar.co.uk | +44 81 991 1142 (O) | +44 71 229 7741 (H) | YELL! "Nothing sucks like a VAX" -- confirmed after recent radiator burst! Hit 'R' <RETURN> to continue .....
henry@zoo.toronto.edu (Henry Spencer) (10/04/90)
In article <1990Oct3.004953.9387@robobar.co.uk> ronald@robobar.co.uk (Ronald S H Khoo) writes: >Could you say what you consider to be the future of the included >newsreader ? ... It needs to be there, partly for the sake of completeness and partly for the benefit of people who want to use news as a way of supplementing motd for seriously naive users. (Readers like rn are awfully intimidating to such people.) In its future -- someday -- are better ties to our setup and configuration stuff, and probably an effort to strip out unused features and improve portability. Be warned that this is not a high priority with us at present. >Presumably it has already diverged sufficiently from the UNSW news package >that backwards compatibility with UNSW is no longer of major concern? Correct. >The stated main problem (lack of pure USG support) doesn't *seem* to be a >problem -- (though I haven't tested it on a pure USG yet but I will) >it works _almost_ perfectly if you stick the sgttyb declaration into the >#ifdef and compile WITHOUT any terminal ioctls at all -- I actually >prefer it this way -- leaves no chance of screwing up the tty settings :-) I'm inclined to agree that it has no real need to mess with the tty stuff and should probably be fixed not to. That would improve portability considerably too. > 1) readnews expects postnews to handle some weird arguments for > followups (which it doesn't :-). > 2) The mail reply addresses are total rubbish. > 3) cancellation (if postnews worked :-) is a bit abrupt ... The posting and replying interface definitely needs work. We haven't run into this much because our users tend to graduate to rn once they stop being passive observers, but it's a valid point. > 5) reading the .newsrc isn't done with fgetmfs :-( Reading of the .newsrc is another area that definitely needs work! >I'd like to know if a) anyone else has done it b) whether anyone else is >GOING to do it c) what the C News authors would like to happen d) what >others would like to see happen... I don't know of anyone who's pursued it. Personally, we'd be happy to see someone competent clean the thing up; we don't really have time to deal with its rough edges just now. -- Imagine life with OS/360 the standard | Henry Spencer at U of Toronto Zoology operating system. Now think about X. | henry@zoo.toronto.edu utzoo!henry