paul@vixie.UUCP (Paul Vixie Esq) (05/11/87)
I just installed the recently posted 'Sys V.[23]' patches, and it worked
pretty well, considering... The patches were in 'diff' format, not 'diff -c'
format (okay, so SysV doesn't have 'diff -c'...). Any lines that were
changed will have the old and new versions of the text; Patch can figure
out where to stick the change by looking at *my* source, in case the line
numbers vary -- which they *do*. On additions, however, 'diff' outputs
an 'a' command with some line numbers, and Patch knows only to stick the
new text after the specified line number. If the line numbers vary (as
mine did), the text will be added in the wrong place.
In my specific case, defs.h was patched to include an '#ifdef BSD' around
the '#include <string.h>'. I had already made this change from a previous
note in this group; this patch gave something like
#ifdef BSD
#include <strings.h>
#else
#ifdef BSD
#include <strings.h>
#else
#endif
#include <string.h>
#endif
(or something utterly bizarre and wrong that would make my cc(1) try to
include <string.h> even though BSD was set...)
Patch cannot know that the line numbers differ unless 'diff -c' was used
to make the patch. I could have installed the patch by hand, and in fact
I have had to do that a lot, since Patch just can't figure out some of the
combinations of versions I've sent it from this group.
I can't suggest that people without 'diff -c' refrain from posting patches;
that would be rude, unrealistic, and counterproductive. What *would* work
would be to serialize the patches in some way -- ideally, get someone to
keep a definitive version, accept & approve & broadcast patches. It would
have to be someone who liked ELM a lot -- which lets me out, sorry, I can
suggest this, but I can't implement it.
I know that the entire collection of ELM users would benefit from serialized
patches, and most everyone would fully support anyone who wanted the job.
The person who does this needn't make any improvements to the code themselves;
the minimum need in this job is just to coordinate the efforts of other folks.
Any volunteers? Reply to the net...
--
Paul A. Vixie {ptsfa, crash, winfree}!vixie!paul
329 Noe Street dual!ptsfa!vixie!paul@ucbvax.Berkeley.EDU
San Francisco
CA 94116 paul@vixie.UUCP (415) 864-7013daemon@hplabsc.UUCP (05/12/87)
In article <1790@hplabsc.HP.COM>, paul@vixie.UUCP (Paul Vixie Esq) writes: > . . . . . . . . . . . -- ideally, get someone to > keep a definitive version, accept & approve & broadcast patches. It would > have to be someone who liked ELM a lot -- This is a good idea, but I doubt if anyone has a clean copy of ELM anymore after the recent deluge of patches. ELM definitely does need someone at the helm to keep the program consistent, so that all will benefit from the patches. Whoever this person is they should have access to AT LEAST a System V and a Berkeley UNIX system. When using the original ELM, I got the distinct impression that the program had never even been run on a BSD system; there were bugs that couldn't have possibly escaped notice otherwise. Any volunteers? kyle jones <kyle@xanth.cs.odu.edu> old dominion university, norfolk, va
geoff@desint.UUCP (Geoff Kuenning) (05/13/87)
In article <1790@hplabsc.HP.COM> ptsfa!vixie!paul@ames (Paul Vixie Esq) writes: > I just installed the recently posted 'Sys V.[23]' patches, and it worked > pretty well, considering... The patches were in 'diff' format, not 'diff -c' > format (okay, so SysV doesn't have 'diff -c'...). Volume 1 of mod.sources contained "diffc", which provides context-diff capabilities to System V sites. See the mod.sources index for info on how to get it. -- Geoff Kuenning geoff@ITcorp.com {hplabs,ihnp4}!trwrb!desint!geoff
baur@spp3.UUCP (Steven L. Baur) (05/13/87)
in article <1790@hplabsc.HP.COM>, paul@vixie.UUCP (Paul Vixie Esq) says: > ...What *would* work > would be to serialize the patches in some way -- ideally, get someone to > keep a definitive version, accept & approve & broadcast patches. It would > have to be someone who liked ELM a lot ... > Any volunteers? Reply to the net... I would volunteer to keep the official version of ELM. Not only do I use ELM as a preferred mailer, I have gone into the source code and fixed bugs. -- -Steve Baur (Altair) {ihnp4,decvax,ucbvax,hplabs}!trwrb!trwspp!spp3!baur The opions expressed do not represent the opions of my employers, or my employers employers. -- -Steve Baur (Altair) {ihnp4,decvax,ucbvax,hplabs}!trwrb!trwspp!spp3!baur People without college degrees are people too.
wedgingt@udenva.UUCP (Will Edgington) (05/14/87)
In article <1797@hplabsc.HP.COM> kyle@xanth (kyle jones) writes: >In article <1790@hplabsc.HP.COM>, paul@vixie.UUCP (Paul Vixie Esq) writes: >> . . . . . . . . . . . -- ideally, get someone to >> keep a definitive version, accept & approve & broadcast patches. It would >> have to be someone who liked ELM a lot -- > >This is a good idea, but I doubt if anyone has a clean copy of ELM anymore >after the recent deluge of patches. I do. Someone else here unpacked it and started modifying it before I knew anyone else had it. I've kept the original postings, still shar'ed. >Whoever this person is they should >have access to AT LEAST a System V and a Berkeley UNIX system. Unfortunately, all I have access to are BSD systems and one Pyramid (which doesn't have enough disk space for ELM right now) ... >Any volunteers? ... and I can't volunteer, since I have too much to do as it is (see my .sig). However, I *can* volunteer to send someone the original, unchanged articles ... ----- Will Edgington, Computing and Information Resources, University of Denver System Administrator for udenva (== dueos), dutyche, duorion, dunike, ... {{hplabs,seismo}!hao,ucbvax!nbires,boulder,cires,ssds}!udenva!wedgingt, WEDGINGT@DUCAIR (BITNET), wedgingt@ccndu (CSN/CCN), ... COMING SOON: wedgingt@du.edu (all nets)|"No two addresses are the same ..."
barad@othello.usc.edu (Herb Barad) (05/14/87)
In article <1811@hplabsc.HP.COM> trwrb!trwspp!spp3!baur%sdcrdcf.UUCP@sdcsvax.ucsd.edu (Steven L. Baur) writes: >I would volunteer to keep the official version of ELM. Not only do I >use ELM as a preferred mailer, I have gone into the source code and >fixed bugs. >-- > -Steve Baur (Altair) >{ihnp4,decvax,ucbvax,hplabs}!trwrb!trwspp!spp3!baur Yeah - I definitely think that someone should take on control of versions, patches, etc. I don't think that he has to moderate the group, but I think that official patches should come from just 1 source (take some of Larry Wall's stuff for example). Unless there are any other volunteers, I think we can make Steve the controller! Huh, how about it... -- Herb Barad [USC - Signal and Image Processing Institute] email: barad@brand.usc.edu or ...!sdcrdcf!oberon!brand!barad
daemon@hplabsc.UUCP (05/14/87)
In art <1811@hplabsc.HP.COM> baur%sdcrdcf@sdcsvax.ucsd.edu (Steven L. Baur): >in article <1790@hplabsc.HP.COM>, paul@vixie.UUCP (Paul Vixie Esq) says: >> ...What *would* work would be to serialize the patches in some way... >> Any volunteers? Reply to the net... > >I would volunteer to keep the official version of ELM. Not only do I >use ELM as a preferred mailer, I have gone into the source code and >fixed bugs. >-- > -Steve Baur (Altair) >{ihnp4,decvax,ucbvax,hplabs}!trwrb!trwspp!spp3!baur Okay, people, we have a volunteer. Steve, I hope you know what you're in for; remember, this is the job that exhausted the enthusiasm of Dave Taylor, a man KNOWN for his extraordinary enthusiasm (among other qualities). I'd suggest that you become the moderator of this group -- get with Dave to see what he wants to do. I'm pretty sure he won't mind. If this all works out, I suggest that anyone who plans any mods to the ELM source gets with Steve first, to make sure he (or someone else) isn't already working on it. Steve, you may want to keep a list of "work in progress", and a list of "work needing to be done" -- if someone wants to pitch in, send 'em the list and let them pick their area. I know that someone is working on 4.3 BSD locking; I would like to see this done, as well as numerous changes to the Configure script, and some changes to the prompts (for a friendly mailer, there are certainly a few GOTCHAs for neophyte users). The option editor needs to be able to edit all the options, too. I'm sure there are lots more ... Steve, I think the next thing to do is get with Dave and try to move the moderation address -- there's a mailing list as well, that will need to be forwarded from hplabs to your machine -- I think some people are still mailing their submissions to elm-info@hplabs instead of comp-mail-elm@[backbone]. Anyone object to this move? I obviously think it's a good idea, but someone might not. Discussion? -- Paul A. Vixie {ptsfa, crash, winfree}!vixie!paul 329 Noe Street dual!ptsfa!vixie!paul@ucbvax.Berkeley.EDU San Francisco CA 94116 paul@vixie.UUCP (415) 864-7013
klr@hadron.UUCP (Kurt L. Reisler) (05/15/87)
The next time that a batches of patches (8-}) are posted, it would be nice to include explanations of what these patchs are supposed to fix. I know, read the code. But it would be a little less time consuming to know that patchx was something that had already been taken care of, even if in another manner.