loverso@Xylogics.COM (John Robert LoVerso) (10/13/89)
In article <69987@uunet.UU.NET> version@uunet.UU.NET writes: > I'm compiling a list of what versions of news are in use by > what people. A summary will appear in news.admin in a month or so I did the same a few months ago (before B news patch 18). The results weren't surprising: most sites are running old code. Two surprising (annoying?) things is that neither B2.11 nor C news put the current "patch" version in the return message. I.e., for B news, a system tracked by patch numbers, you only get the patch date. For C news, a system tracked by patch dates, you get no other information than the single character "C". While its a valiant effort for Geoff & Henry to want only "one C news", there are (now) different `versions' in use in the field, and it would be nice to know who's got an out of date version! John [comments I added (B news patch levels) are in brackets] Count Version 2 B 2.10.1 6/24/83 2 B 2.10.3 4.3bds beta 6/6/85 3 B 2.10.3 4.3bsd-beta 6/6/85 1 B 2.10.3 BETA 3/19/86 3 B 2.11 beta 9/22/86 13 B 2.11 10/30/86 [patchlevel 0] 2 B 2.11 11/15/86 [patchlevel 1] 1 B 2.11 3/10/87 [patchlevel 5] 1 B 2.11.8-Pyramid OSx 4.1 87/07/14 11 B 2.11 4/10/87 [patchlevel 8] 4 B 2.11 10/15/87 [patchlevel 11] 5 B 2.11 11/04/87 [patchlevel 12] 3 B 2.11.13 11/19/87 3 B 2.11 11/19/87 [patchlevel 13] 61 B 2.11 12/1/87 [patchlevel 14] 34 B 2.11 1/24/89 [patchlevel 17] 1 B3.0 (beta level 5) 1 B3.0 (beta level 7.7) 14 C 1 P 1.2, Release 06/13/89, Patch 98 [IBM VM] ---- 166 total for "ne" distribution -- John Robert LoVerso Xylogics, Inc. 617/272-8140 loverso@Xylogics.COM Annex Terminal Server Development Group
coolidge@brutus.cs.uiuc.edu (John Coolidge) (10/13/89)
loverso@Xylogics.COM (John Robert LoVerso) writes: >For C news, >a system tracked by patch dates, you get no other information than >the single character "C". While its a valiant effort for Geoff & >Henry to want only "one C news", there are (now) different `versions' >in use in the field, and it would be nice to know who's got an >out of date version! I just changed our C news to report the patch date we're at (22-Aug-89, for the curious. 24-Aug didn't seem to add much, and I really don't want to rebuild history right now). Anyway, it would be nice for people to patch their version responder to say the right thing. A better question is: what sorts of information should version give? Just the news version? News and nntp version? How about a 'system' control that returns the system type, news and nntp version, system name, and contact info. This may become more important as the number of pure-nntp systems (therefore not listed in the maps) grows. Yet another gripe about control messages: the granularity of control seems a little too low. If I send out a sendsys, I (almost without exception) mean to get just one particular machine. Posting to 'to.<machine>' works if it's a neighbor, and Distribution: helps a lot. But it seems like it should be possible to say 'sendsys <machine>' and have only <machine> respond, while others propagate the message ignoring the sendsys otherwise. --John -------------------------------------------------------------------------- John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge Of course I don't speak for the U of I (or anyone else except myself) Copyright 1989 John L. Coolidge. Copying allowed if (and only if) attributed. You may redistribute this article if and only if your recipients may as well.
chip@ateng.com (Chip Salzenberg) (10/13/89)
According to coolidge@brutus.cs.uiuc.edu (John Coolidge): >Yet another gripe about control messages: the granularity of control seems >a little too low. If I send out a sendsys, I (almost without exception) >mean to get just one particular machine. There already is such a mechanism: $ mail usenet@machinename Subject: A request Please send my your news "sys" file. adTHANKSvance. . -- You may redistribute this article only to those who may freely do likewise. Chip Salzenberg at A T Engineering; <chip@ateng.com> or <uunet!ateng!chip> "'Why do we post to Usenet?' Naturally, the answer is, 'To get a response.'" -- Brad "Flame Me" Templeton
coolidge@brutus.cs.uiuc.edu (John Coolidge) (10/14/89)
chip@ateng.com (Chip Salzenberg) writes: >According to coolidge@brutus.cs.uiuc.edu (John Coolidge): >>Yet another gripe about control messages: the granularity of control seems >>a little too low. If I send out a sendsys, I (almost without exception) >>mean to get just one particular machine. >There already is such a mechanism: > $ mail usenet@machinename > Subject: A request > Please send my your news "sys" file. > adTHANKSvance. > . Yes, that _does_ accomplish about the same thing. The difference is that the sendsys method is automatic and requires no sysadmin action, while the mail to usenet method requires active cooperation of the sysadmin. This might be good or bad, but in general I think the sendsys extention method is better because: since sendsys already exists, it's better to allow a finer scope of control, and because some sysadmins simply ignore mail to usenet for long stretches of time. --John -------------------------------------------------------------------------- John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge Of course I don't speak for the U of I (or anyone else except myself) Copyright 1989 John L. Coolidge. Copying allowed if (and only if) attributed. You may redistribute this article if and only if your recipients may as well.
henry@utzoo.uucp (Henry Spencer) (10/14/89)
loverso@Xylogics.COM (John Robert LoVerso) writes: >>For C news, >>a system tracked by patch dates, you get no other information than >>the single character "C". While its a valiant effort for Geoff & >>Henry to want only "one C news", there are (now) different `versions' >>in use in the field, and it would be nice to know who's got an >>out of date version! I am (reluctantly) coming to agree that patch date in the version output would be a useful thing. We'd really be much happier with code that just came out with either "current" or "out of date", but there isn't any way to do that without other complications. In article <1989Oct12.233424.28224@brutus.cs.uiuc.edu> coolidge@cs.uiuc.edu writes: >... A better question is: what sorts of information >should version give? Just the news version? News and nntp version? How >about a 'system' control that returns the system type, news and nntp >version, system name, and contact info... Interesting point. Anybody else have any thoughts on this? C News can include anything that is readily found by a Unix command, since the code that actually builds the version message is a shell file. -- A bit of tolerance is worth a | Henry Spencer at U of Toronto Zoology megabyte of flaming. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
campbell@redsox.bsw.com (Larry Campbell) (10/14/89)
In article <1989Oct13.212209.7537@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) writes: -In article <1989Oct12.233424.28224@brutus.cs.uiuc.edu> coolidge@cs.uiuc.edu writes: ->... A better question is: what sorts of information ->should version give? Just the news version? News and nntp version? How ->about a 'system' control that returns the system type, news and nntp ->version, system name, and contact info... - -Interesting point. Anybody else have any thoughts on this? C News can -include anything that is readily found by a Unix command, since the code -that actually builds the version message is a shell file. How about including the output of a "what inews" command? It's a bit verbose, but has the benefit that you *know* exactly what version of *everything* went into it. -- Larry Campbell The Boston Software Works, Inc. campbell@bsw.com 120 Fulton Street wjh12!redsox!campbell Boston, MA 02146
coolidge@brutus.cs.uiuc.edu (John Coolidge) (10/16/89)
campbell@redsox.bsw.com (Larry Campbell) writes: >[In the results of Control: version] >How about including the output of a "what inews" command? It's a bit >verbose, but has the benefit that you *know* exactly what version of >*everything* went into it. This doesn't work well if inews is a shell script, which (in C news) it is... --John -------------------------------------------------------------------------- John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge Of course I don't speak for the U of I (or anyone else except myself) Copyright 1989 John L. Coolidge. Copying allowed if (and only if) attributed. You may redistribute this article if and only if your recipients may as well.
epsilon@wet.UUCP (Eric P. Scott) (10/16/89)
In article <1463@redsox.bsw.com> campbell@redsox.UUCP (Larry Campbell) writes: >How about including the output of a "what inews" command? It's a bit >verbose, but has the benefit that you *know* exactly what version of >*everything* went into it. Why do I feel like Johhny Carson about to rebut Ed McMahon? "You are wrong..." I have about 6 patches to netnews B 2.11.17 and have never diddled the SCCS ids. I suspect I'm not alone... -=EPS=-
tim@brspyr1.BRS.Com (Tim Northrup) (10/17/89)
On the idea of how to report versions: How about a "versions" file in the news/lib directory. This could be sent out the same way as a sendsys request sends the sys file. Have the makefile for news update a set line in the file (or add it if it doesn't find its entry in the file). The makefile could update on each compile based on the contents of patchlevel.h. In this way, versions of virtually anything could be contained there, and any site specific comments could be added. My initial thought is to use something simple. For example, use a fixed name for the package as the first element per line, followed by the version, patch level, compilation date, etc. Here is what it might look like (with some arbitrary comments): # Versions file for site brspyr1.BRS.Com # Contact tim@brspyr1.BRS.Com for more info Bnews 2.11 PL17 1/22/89 nn 3.6 PL5 10/1/89 This would allow news readers, map unpackers, etc. to all be recorded and reported. It ain't bullet-proof, but I thought it might be worth discussing. -- Tim -- Tim Northrup +------------------------------------------+ +---------------------------------+ GEnie: T.Northrup | UUCP: uunet!crdgw1!brspyr1!tim | Air Warrior: "Duke" | ARPA: tim@brspyr1.BRS.Com +------------------------------------------+
clewis@eci386.uucp (Chris Lewis) (10/19/89)
In article <1989Oct15.190004.29046@brutus.cs.uiuc.edu> coolidge@cs.uiuc.edu writes: >campbell@redsox.bsw.com (Larry Campbell) writes: >>[In the results of Control: version] >>How about including the output of a "what inews" command? It's a bit >>verbose, but has the benefit that you *know* exactly what version of >>*everything* went into it. >This doesn't work well if inews is a shell script, which (in C news) it >is... Shore, no problem, just have Henry insert something along the lines of: #@(#)cnews <patchlevel> into inews. -- He's a consultant: | Chris Lewis, Elegant Communications Inc. Lend him your watch | UUCP {uunet!attcan,utzoo}!lsuc!eci386!clewis and he'll tell you the time. | Moderator of the Ferret mailing list. Don Munroe, Cosmic Commander|