web/2493: GNATSweb should discourage huge attachments

Joseph Myers jsm28@cam.ac.uk
Thu Apr 5 09:56:00 GMT 2001

>Number:         2493
>Category:       web
>Synopsis:       GNATSweb should discourage huge attachments
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    gerald
>State:          open
>Class:          change-request
>Submitter-Id:   net
>Arrival-Date:   Thu Apr 05 09:56:01 PDT 2001
>Originator:     Joseph S. Myers
>Release:        3.0 20010405 (prerelease)
System: Linux digraph 2.2.19 #1 Wed Mar 28 16:01:38 UTC 2001 i686 unknown
Architecture: i686

host: i686-pc-linux-gnu
build: i686-pc-linux-gnu
target: i686-pc-linux-gnu
configured with: ../gcc-3cvs/configure --prefix=/opt/gcc/snapshot --disable-shared --enable-threads=posix --with-system-zlib

Sometimes people submit huge attachments via GNATSweb, which then get
sent to gcc-prs.  It would be nice if some heuristics could be applied
to discourage such attachments when they are larger than necessary.
Examples of inappropriate usages:

* Attaching a multimegabyte source tarball rather than the individual
  preprocessed file that is relevant.
* Attaching a huge uncompressed .i or .ii file instead of a compressed


Observe that unnecessarily multimegabyte messages arrive at gcc-prs
from time to time (e.g. a 6 Mbyte one today).  They don't arrive at
gcc-bugs because of the message size limit on that list, so not many
people will see them.


If someone tries to submit a large tarball, ask them "are you sure"
before accepting the PR.  If someone submits a large file which bzip2
or gzip reduces in size substantially, silently compress it with bzip2
or gzip (adjusting the name accordingly) before attaching it to the
new PR.  (Since GNATSweb attachments don't appear as MIME attachments
in the mailing list messages and so need some manual decoding, this
doesn't significantly affect the usability of the message to gcc-prs.)

More information about the Gcc-bugs mailing list