************************     THIS FILE IS OBSOLETE!     ***********************


Bugs for CSSC are recorded on the CSSC bug tracking system at
http://sourceforge.net/projects/CSSC.





















The Time-stamp: line indicates when this file was last modified.
Time-stamp: <1999-04-18 18:16:55 james>

List of known bugs (bugs are moved to FIXED as they are fixed so it is
valid but unlikely for there to be nothing below this line)


Sun Apr 18 18:14:07 1999: James Youngman <jay@gnu.org>
In version 0.10beta.pl0.  Unfixed.
 sccs unedit /tmp/SCCS/s.foo will cause "./s.foo" to be unlinked
 (though this file probably does not exist).  The actual gotten file
 will not be deleted.

Mon Feb  9 23:41:03 1998: Lars Hecking <lhecking@nmrc.ucc.ie>
In version 0.05-pl0.  Unfixed.
 Then, Irix 6.3. I have installed gcc-2.8.0, which supports 6.3/32-bit
 API proper. I get tons of warnings like these at linking stage:

 ld: WARNING 131: Multiply defined weak symbol:(__as__Q210sccs_pfile9edit_lockRCQ
 210sccs_pfile9edit_lock) in get.o and pf-add.o (2nd definition ignored).
 ld: WARNING 131: Multiply defined weak symbol:(__as__Q210sccs_pfile9edit_lockRCQ
 210sccs_pfile9edit_lock) in get.o and pfile.o (2nd definition ignored).
 ld: WARNING 131: Multiply defined weak symbol:(destroy__t4list1Z7release) in rl-
 merge.o and sccsfile.o (2nd definition ignored).
 ld: WARNING 131: Multiply defined weak symbol:(copy__t4list1Z7releaseRCt4list1Z7
 release) in rl-merge.o and sccsfile.o (2nd definition ignored).
 ld: WARNING 131: Multiply defined weak symbol:(__as__t4list1Z7releaseRCt4list1Z7
 release) in rl-merge.o and sccsfile.o (2nd definition ignored).

[JAY: looks like a template instantiation problem.  Perhaps this is
 unavoidable given GCC-2.8's instantiation system and the IRIX linker.
 Check the GCC release notes and FAQ]


Wed Nov  5 18:56:06 1997: James Youngman <jay@gnu.org>
In version 0.02-pl6.  Unfixed.
 GCC version 2.7.2.2 compiled for an Alpha running NetBSD fails to
 compile delta.cc.  It falls over with an "Unrecognised insn" error.


Tue Jul 22 08:07:29 1997: Lars Hecking <lhecking@nmrc.ucc.ie>
In version 0.02-pl3.  Unfixed.
 Irix 6.3, gcc-2.7.2.2 configured for Irix 5.3 (32bit-64bit issue):
  All tests passed except for test-admin. test-admin fails in
 init-mr.sh:
laredo (53) sh ./init-mr.sh 
I1...passed 
I2...passed 
I3...passed 
I4...passed 
I5...passed 
I6...1c1
< 2.1
---
> 1.1
FAIL init-mr.sh I6: stderr format error with ../../get -p s.new.txt

This indicates that the -r option of "admin -i" has failed to produce
an initial SID of 2.1, but has instead produced the default, 1.1.
Examination of the SCCS file bears this out:-

begin 644 s.new.txt
M`6@P.#0U.0H!<R`P,#`P,2\P,#`P,"\P,#`P,`H!9"!$(#$N,2`Y-R\P-R\R
M,2`Q,#HU-3HR-"!L:&5C:VEN9R`Q(#`*`6,@9&%T92!A;F0@=&EM92!C<F5A
M=&5D(#DW+S`W+S(Q(#$P.C4U.C(T(&)Y(&QH96-K:6YG"@%E"@%U"@%5"@%T
2"@%4"@%)(#$*)4TE"@%%(#$*
`
end


Mon Jul 7 22:17:59 1997: James Youngman <jay@gnu.org>
In version [all]: Unfixed.
 On systems with no working fork() or spawn(), we have to use system()
 to execute commands.  The return value of system() is
 implementation-defined unless its argument is NULL.  Hence we cannot
 place any interpretation on the return value.  This means that we will
 be unaware of any return value from the command and (for example) MR
 validation will not happen reliably.

Local Variables:
mode: Text
End: