thad@cup.portal.com (Thad P Floryan) (01/12/91)
forrie@morwyn.uucp (/dev/w1) in <1991Jan7.035239.877@morwyn.uucp> writes:
I am wondering about the KSH. Someone tells me that the one
distributed with the 2.0 fixes is 'braindamaged' ... will someone
please tell me which verion WORKS right? Why is it brain damaged,
what can be done to change it, etc etc...
I like some of the features of the KSH, but since everyone always says
that there are a lot of differences in compatibility with the SH, I
have avoided it... but realistically, there's not much sense in
avoiding it if you can fix it, or find out what's wrong...
I'm the one who claimed it's braindamaged, and it is (even the version that
accompanied the FIXDISK2). I posted the original details years ago and
cannot quickly find them, but among the problems were incorrect evaluation
of arithmetic expressions and some others ...
The version you WANT (and which, interestingly, has the fixes that were made
to the FIXDISK2 ksh) is the one from osu-cis in the att7300/STORE directory
which is version "06/03/86". You could use "strings" to ascertain the version
but the easiest is to be in emacs mode and type a ^V (control-V).
As far as any alleged incompatibilty with "sh", I really haven't found any,
and "ksh" is even root's default shell on ALL my systems. The only quirk using
ksh for root is that some custom crontab entries may cause the message
"Nonstandard shell" to appear in the cronlog and abort; the "fix" to this is
to have entries in crontab like (which, by the way, IS my /usr/lib/crontab):
#sccs "@(#)fndcmd:crontab 1.7"
#
#Mn Hr Da Mo Da (0=SUN, 1=MON, 2=TUE, 3=WED, 4=THU, 5=FRI, 6=SAT)
#of of of of of
#Hr Da Mo Yr Wk Command
#
16,46 * * * * /bin/su uucpadm -c "/usr/lib/uucp/uudemon.poll >/dev/null"
18,48 * * * * /bin/su uucpadm -c "/usr/lib/uucp/uudemon.hour >/dev/null"
00 4 * * * /bin/su uucpadm -c "/usr/lib/uucp/uudemon.admin >/dev/null"
59 23 * * * /bin/sh -c "/usr/local/bin/hdwarn.day >/dev/null"
59 23 * * * /bin/sh -c "/usr/local/bin/du-logs.day >/dev/null"
03 3 * * 0 /bin/sh -c "/etc/clockupd.wk >/dev/null"
30 5 * * 1 /bin/su uucpadm -c "/usr/lib/uucp/uudemon.cleanu >/dev/null"
30 5 * * 1 /bin/sh -c "/etc/cleanup.wk >/dev/null"
Thad Floryan [ thad@cup.portal.com ]wwm@wa8tzg.mi.org (Bill Meahan) (01/13/91)
In article <37920@cup.portal.com> thad@cup.portal.com (Thad P Floryan) writes: >forrie@morwyn.uucp (/dev/w1) in <1991Jan7.035239.877@morwyn.uucp> writes: > > I am wondering about the KSH. Someone tells me that the one > distributed with the 2.0 fixes is 'braindamaged' ... will someone > please tell me which verion WORKS right? Why is it brain damaged, > what can be done to change it, etc etc... > >I'm the one who claimed it's braindamaged, and it is (even the version that >accompanied the FIXDISK2). I posted the original details years ago and >cannot quickly find them, but among the problems were incorrect evaluation >of arithmetic expressions and some others ... > >The version you WANT (and which, interestingly, has the fixes that were made >to the FIXDISK2 ksh) is the one from osu-cis in the att7300/STORE directory >which is version "06/03/86". You could use "strings" to ascertain the version >but the easiest is to be in emacs mode and type a ^V (control-V). > > > >Thad Floryan [ thad@cup.portal.com ] Does this mean that I should get the version from the STORE even if I have FIXDISK2 ?? [Thad, you're not quite clear] Has anyone ported ksh88e to the 3b1? [and is it legal to snarf the ksh88e sources on OSU ??] -- Bill Meahan (WA8TZG) | Programming is simple: wwm@wa8tzg.mi.org OR | uunet!mailrus!sharkey!wa8tzg!wwm | All you have to do is put the right "Home for Cybernetic Orphans" | numbers in the right memory locations!
thad@cup.portal.com (Thad P Floryan) (01/13/91)
wwm@wa8tzg.mi.org (Bill Meahan) in <1991Jan12.220430.1078@wa8tzg.mi.org>
writes:
+ >I'm the one who claimed it's braindamaged, and it is (even the version that
+ >accompanied the FIXDISK2). I posted the original details years ago and
+ >cannot quickly find them, but among the problems were incorrect evaluation
+ >of arithmetic expressions and some others ...
+ >
+ >The version you WANT (and which, interestingly, has the fixes that were made
+ >to the FIXDISK2 ksh) is the one from osu-cis in the att7300/STORE directory
+ >which is version "06/03/86". You could use "strings" to ascertain the versi
on
+ >but the easiest is to be in emacs mode and type a ^V (control-V).
+ >
+ >Thad Floryan [ thad@cup.portal.com ]
+
+ Does this mean that I should get the version from the STORE even if I have
+ FIXDISK2 ?? [Thad, you're not quite clear]
Yes, you want the "STORE" version of ksh. That version works with the stuff
documented in "THE KORN SHELL" book; the version on FIXDISK2 does NOT work.
+ Has anyone ported ksh88e to the 3b1? [and is it legal to snarf the ksh88e
+ sources on OSU ??]
WHAT!?!?!?!?!?!?! If that's on osu-cis, someone's gonna be in deep doo-doo.
ksh88e is a COMMERCIAL PRODUCT of AT&T and costs $3,000 for source and some
$10,000 for binary (you figure WHY binary costs more than source! :-)
Are you SURE you're not referring to the "PD ksh" clone that recently was
posted to alt.sources?
Thad Floryan [ thad@cup.portal.com ]thad@cup.portal.com (Thad P Floryan) (01/13/91)
wwm@wa8tzg.mi.org (Bill Meahan) in <1991Jan12.220430.1078@wa8tzg.mi.org>
writes:
Has anyone ported ksh88e to the 3b1? [and is it legal to snarf the
ksh88e sources on OSU ??]
If you DID see the ksh88e sources at osu, they're not there now. Just 10
minutes ago I checked all three:
tut.cis.ohio-state.edu [IP 128.146.8.60]
cheops.cis.ohio-state.edu [IP 128.146.8.62]
saqqara.cis.ohio-state.edu [IP 128.146.8.98]
by doing an "ls -lR" from the root directory, then "grep -in" the dir lists for
all of "korn","88e","ksh","shell","sh" and didn't see anything suspicious.
What I found were (on cheops):
csh-src.tar.Z presumably "csh" source
KSH+IN.Z 3B1 installable binary
And I also noticed the pub/att7300 directory on cheops hasn't been updated
in awhile, but did notice that lavg-daemon.Z was clobbered on Dec 31 12:51;
it's length is now zero.
What makes(made) you believe the sources to ksh88e are at OSU?
Thad Floryan [ thad@cup.portal.com ]rstevens@noao.edu (Rich Stevens) (01/13/91)
>ksh88e is a COMMERCIAL PRODUCT of AT&T and costs $3,000 for source and some >$10,000 for binary (you figure WHY binary costs more than source! :-) The $10k is a license from AT&T to sell the binary, not the cost of the binary.
john@banzai.PCC.COM (John Canning) (01/14/91)
Someone asked, and then Thad answered: > >+ Has anyone ported ksh88e to the 3b1? [and is it legal to snarf the ksh88e >+ sources on OSU ??] > >WHAT!?!?!?!?!?!?! If that's on osu-cis, someone's gonna be in deep doo-doo. > >ksh88e is a COMMERCIAL PRODUCT of AT&T and costs $3,000 for source and some >$10,000 for binary (you figure WHY binary costs more than source! :-) > The source code for ksh88e costs $3,000. The right to sell/distribute the binaries costs $10,000; the source comes with this right. That is why the binary seems to cost more than the source.
wwm@wa8tzg.mi.org (Bill Meahan) (01/14/91)
In article <37974@cup.portal.com> thad@cup.portal.com (Thad P Floryan) writes: >wwm@wa8tzg.mi.org (Bill Meahan) in <1991Jan12.220430.1078@wa8tzg.mi.org> >writes: > > Has anyone ported ksh88e to the 3b1? [and is it legal to snarf the > ksh88e sources on OSU ??] > >If you DID see the ksh88e sources at osu, they're not there now. Just 10 >minutes ago I checked all three: > > tut.cis.ohio-state.edu [IP 128.146.8.60] > cheops.cis.ohio-state.edu [IP 128.146.8.62] > saqqara.cis.ohio-state.edu [IP 128.146.8.98] > >by doing an "ls -lR" from the root directory, then "grep -in" the dir lists for >all of "korn","88e","ksh","shell","sh" and didn't see anything suspicious. > >What I found were (on cheops): > > csh-src.tar.Z presumably "csh" source > KSH+IN.Z 3B1 installable binary > >And I also noticed the pub/att7300 directory on cheops hasn't been updated >in awhile, but did notice that lavg-daemon.Z was clobbered on Dec 31 12:51; >it's length is now zero. > >What makes(made) you believe the sources to ksh88e are at OSU? > >Thad Floryan [ thad@cup.portal.com ] A ls-lR file uucp'ed from osu-cis on Nov 11, 1990 contained the following: >>> receive: >>> total 2 >>> drwxrwxrwx 3 uucp 2048 Nov 8 02:00 perlman >>> >>> receive/perlman: >>> total 2 >>> drwxrwxrwx 3 uucp 2048 Nov 8 05:45 toolchest >>> >>> receive/perlman/toolchest: >>> total 10 >>> -rwxrwxrwx 1 uucp 7374 Nov 8 05:45 INSTALL >>> drwxrwxrwx 4 uucp 2048 Nov 8 05:32 ksh-i >>> >>> receive/perlman/toolchest/ksh-i: >>> total 4 >>> drwxrwxrwx 2 uucp 2048 Nov 8 05:31 src >>> drwxrwxrwx 2 uucp 2048 Nov 8 05:44 unfmt >>> >>> receive/perlman/toolchest/ksh-i/src: >>> total 810 >>> -rwxrwxrwx 1 uucp 136 Nov 8 02:00 1thank_you >>> -rwxrwxrwx 1 uucp 95662 Nov 8 02:06 ksh-88e.0b >>> -rwxrwxrwx 1 uucp 105512 Nov 8 03:01 ksh-88e.1b >>> -rwxrwxrwx 1 uucp 108133 Nov 8 03:06 ksh-88e.2b >>> -rwxrwxrwx 1 uucp 129195 Nov 8 05:07 ksh-88e.3b >>> -rwxrwxrwx 1 uucp 109959 Nov 8 05:16 ksh-88e.4b >>> -rwxrwxrwx 1 uucp 121299 Nov 8 05:24 ksh-88e.5b >>> -rwxrwxrwx 1 uucp 114092 Nov 8 05:27 ksh-88e.6b >>> -rwxrwxrwx 1 uucp 35540 Nov 8 05:31 ksh-88e.7b >>> >>> receive/perlman/toolchest/ksh-i/unfmt: >>> total 166 >>> -rwxrwxrwx 1 uucp 136 Nov 8 05:32 1thank_you >>> -rwxrwxrwx 1 uucp 4915 Nov 8 05:33 PROMO >>> -rwxrwxrwx 1 uucp 97966 Nov 8 05:39 sh.1 >>> -rwxrwxrwx 1 uucp 63348 Nov 8 05:44 sh.memo I **THOUGHT** something seemed amiss here since (the last I'd heard) ksh88e was a commercial product. Someone at OSU should be more careful! the 'receive' directory was world-accessible (777) so that anyone who logged in that day (or until it was discovered and removed) could have slurped up a copy! [For the file police - I DID NOT THEN OR EVER DO THAT!!!!] I'm quite glad that this has been straightened up. Hate to get anyone in trouble but I'd also hate to see the OSU archives become unavailable! -- Bill Meahan (WA8TZG) | Programming is simple: wwm@wa8tzg.mi.org OR | uunet!mailrus!sharkey!wa8tzg!wwm | All you have to do is put the right "Home for Cybernetic Orphans" | numbers in the right memory locations!
thad@cup.portal.com (Thad P Floryan) (01/14/91)
wwm@wa8tzg.mi.org (Bill Meahan) in <1991Jan14.011752.2384@wa8tzg.mi.org>
writes:
[...]
I **THOUGHT** something seemed amiss here since (the last I'd heard)
ksh88e was a commercial product. Someone at OSU should be more
careful! the 'receive' directory was world-accessible (777) so that
anyone who logged in that day (or until it was discovered and removed)
could have slurped up a copy! [For the file police - I DID NOT THEN
OR EVER DO THAT!!!!]
I'm quite glad that this has been straightened up. Hate to get anyone
in trouble but I'd also hate to see the OSU archives become
unavailable!
Whew! You're not kidding. And I should have remembered from my signing onto
the Toolchest last year that when one orders something the order is shipped
via uucp to a site of one's choosing.
If someone has Karl Kleinpaste's email address handy, he should be advised of
this to PREVENT any problems in the future; seems like "maybe" a simple umask
change is all that's required to prevent us all from getting more gray hairs
over the thought of losing a resource like osu-cis.
Thad Floryan [ thad@cup.portal.com ]jfischer@cbnewsm.att.com (james.fischer) (01/15/91)
Thad mentioned that he did not know why BINARY for Korn Shell costs $10K, while source rights are $3K.... The $3K gives you the right to use the source as you see fit on a single machine. This is "Right To Use". The $10K puts you in business to SELL the Korn Shell to others in Binary form. Of course, the Toolchest prices are ALL YOU EVER PAY! This means that for a lousy $13K, you can sell an UNLIMITED number of copies, and never pay UNIX System Labs another cent (once you pay the $13K). You have no royalties to pay on copies EVER. Not such a bad deal... james fischer UNIX System Labs
bruce@balilly.UUCP (Bruce Lilly) (01/16/91)
In article <37920@cup.portal.com> thad@cup.portal.com (Thad P Floryan) writes: > >which is version "06/03/86". You could use "strings" to ascertain the version >but the easiest is to be in emacs mode and type a ^V (control-V). what(1) also works, if you have it. >As far as any alleged incompatibilty with "sh", I really haven't found any, >and "ksh" is even root's default shell on ALL my systems. The only quirk using >ksh for root is that some custom crontab entries may cause the message >"Nonstandard shell" to appear in the cronlog and abort; the "fix" to this is >to have entries in crontab like (which, by the way, IS my /usr/lib/crontab): Linking /bin/ksh to /bin/sh also works (the documentation said it was OK with this version, so I tried it on one machine about 5 months ago -- no problems (yet)). However, ksh does use a startup file ($ENV), which can slow things down, and may cause unexpected things to happens if the startup file writes anything to stdout. -- Bruce Lilly blilly!balilly!bruce@sonyd1.Broadcast.Sony.COM
lefko@vaxwaller.UUCP (Marty Lefkowitz) (01/18/91)
I've been using the GNU Born Again SHell (BASH) for a few months now. While I admit I'm no developer, it does seem to work for me. It's got functions, or whatever they call it, just like KSH, forground and background processes and command line history and editing. I'd love to get a manual for it so I could delve deeper into it, but I don't have the time or space to deal with tex, and as far as I know the manual is not really formalized. However it is enough like KSH, CSH and SH for me to use it. The only bug that bothers me, is one concering the abort function "^g" in emacs. It seems to cause some sort of error bad enough to log me out. It was not that hard to get it running either.