jkh@meepmeep.pcs.com (Jordan K. Hubbard) (10/28/89)
I don't mean to make a public example of Bruce, but I hope that posting
this once will deter future boneheads.. I'll be brief:
------
To: bruce@ic.cmc.ca
Subject: ARGHHHHHHHHHHHHH!!
First it was Cindy, then (perhaps encouraged by her example) you!
PLEASE PLEASE PLEASE PLEASE do not send 85K CPP dumps to the entire
gnu.gcc.bug list! We're just not interested in reading it, much
less paying twenty bucks to ship it across the Atlantic! (by which
time it's too late, of course). Cull the relevant portion of the
code you think is broken, then send it to the list. Better yet,
if you feel a great need to send the entire raw (not even
blank-compressed, geez) cpp output to someone, send it to RMS. He
can then flame you back much more cheaply..
Only great restraint prevents me from using stronger language here.
Please do not do this again.
Jordan Hubbard
West Germany.
PCS Computer Systeme GmbH, Munich, West Germany
UUCP: pyramid!pcsbst!jkh jkh@meepmeep.pcs.com
EUNET: unido!pcsbst!jkh
ARPA: jkh@violet.berkeley.edu or hubbard@decwrl.dec.comRaeburn@MIT.EDU (Ken Raeburn) (10/31/89)
To: bruce@ic.cmc.ca
Subject: ARGHHHHHHHHHHHHH!!
First it was Cindy, then (perhaps encouraged by her example) you!
PLEASE PLEASE PLEASE PLEASE do not send 85K CPP dumps to the entire
gnu.gcc.bug list! We're just not interested in reading it, much
less paying twenty bucks to ship it across the Atlantic! (by which
time it's too late, of course). Cull the relevant portion of the
code you think is broken, then send it to the list. Better yet,
if you feel a great need to send the entire raw (not even
blank-compressed, geez) cpp output to someone, send it to RMS. He
can then flame you back much more cheaply..
Only great restraint prevents me from using stronger language here.
Please do not do this again.
Jordan:
I beg to differ. Read section 6.2 of "Using and Porting GNU CC"
again. (Did you read it a first time?) It asks that reports be sent
to bug-gcc. It also asks that, if the problem is not with the
preprocessor, the preprocessor output be sent. That inevitably means
that some of the reports are going to be very large.
If you're tired of getting these huge messages, you've got several
choices:
* Move to info-gcc from bug-gcc. Or stop reading the corresponding
newsgroup. As RMS says in the above-mentioned section, "Most users
of GNU CC do not want to receive bug reports. Those that do have
asked to be on `bug-gcc'."
* Filter your mail, and discard large messages to bug-gcc that have
lots of whitespace.
* Convince the people running the list/newsgroup gateway to do the
filtering.
* Give RMS changes to cccp.c to automagically suppress all the extra
whitespace (keeping only what's needed), and convince him that it's
a good thing. (If you want to suppress the `#' lines that are
followed by just whitespace and another `#' line, that will save
more space, but might lead to some difficulty reading the
preprocessor output; if only superfluous whitespace is omitted, it
would probably be acceptable.) This would make the reports much
smaller, I'm fairly sure.
* Suffer.
(In case you couldn't guess, I've thought of doing the fourth. I've
been doing the last without much difficulty, though.)
The Right Thing to do is not to flame people for sending in cpp
output, when they've been asked to do so; it is not to make
`rms@prep.ai.mit.edu' become another publicized bug list for those not
interested or knowledgable enough to reduce their problem files to a
minimal case.
Perhaps you could convince RMS that it would be good for later
versions of his documentation should instruct all users reporting bugs
to trim down the cc1 input to a minimal case that still demonstrates
the problem, but I doubt it. But if you do, then you can start
complaining at people....
-- Ken Raeburn
MIT Project Athena