[4992] in SIPB bug reports
Re: mosaic outputs cruft to stdout
daemon@ATHENA.MIT.EDU (jhawk@MIT.EDU)
Thu Oct 27 21:43:15 1994
From: jhawk@MIT.EDU
To: yandros@MIT.EDU, bug-outland@MIT.EDU
Cc: bug-sipb@MIT.EDU, eichin@MIT.EDU, yoav@MIT.EDU
In-Reply-To: Your message of "Thu, 27 Oct 1994 13:42:36 +0500."
<9410271742.AA07418@infocalypse.MIT.EDU>
Date: Thu, 27 Oct 1994 21:41:54 EDT
In message <9410271742.AA07418@infocalypse.MIT.EDU>, yandros@mit.edu writes:
>Perhaps you all should read bug-outland. :-)
>
>Look at /mit/outland/decmipsbin/lynx for an example of a script
>that seems to work reasonably well (after I was reminded to not
>echo the `ultrix lossage' message to stdout. :-).
Fie!
Aside from actually using fmt, which is questionable (:-0), and
using a here document where an echo would suffice (one-line comment),
and resorting to an ``if;then fi'' construct which seems to be right
inefficient under Ultrix,
[well, perhaps, I'm overdoing it?]
but the most importantly:
1) Yoav should consider including a ``sleep 2'' in his mosaic
script, and perhaps using ``nice'' on ``Mosaic''. I must admit
to being rather perplexed as to why there's a sleep 2 in the lynx
script...
2) The lynx script shouldn't use:
exec /afs/sipb.mit.edu/project/outland/decmipsbin/lynx.real $@
because this does the wrong thing when you do a
lynx a b "c d" e
(it loses track of the fact that word 3 is c\ d and not just c).
Instead it should be:
exec /afs/sipb.mit.edu/project/outland/decmipsbin/lynx.real "$@"
which solves this problem quite well (and is the standard solution).
>Re: what Mark said, this sounds suspiciously like something I
>wrote for DCNS this summer. I've been thinking abuot integrating
>it into the sipb locker for a little while now (even since someone
>started installing untested binaries into the sipb locker. Grumble.),
>I'll look at it this weekend.
Grumble, grumble.
--jhawk
