Solaris 9/Ultra 1/gcc 2.95.3
==== cmdMZ-1.4 Tcl_PwdObjCmd: failure FAILED
==== Contents of test case:
# We don't want this test to run on Linux because
they do a
# permissions caching trick which causes this to
fail. The
# caching is incorrect, but we have no control over
that.
file delete -force foo
file mkdir foo
set cwd [pwd]
cd foo
file attr . -permissions 000
set result [list [catch {pwd} msg] $msg]
cd $cwd
file delete -force foo
set result
---- Result was:
0 /usr/local/src/tcl/unix/foo
---- Result should have been (exact matching):
1 {error getting working directory name: permission denied}
==== cmdMZ-1.4 FAILED
==== http-4.14 http::Event FAILED
==== Contents of test case:
set code [catch {
set token [http::geturl $badurl/?timeout=10
-timeout 10000 -command {#}]
if {[string length $token] == 0} {
error "bogus return from http::geturl"
}
http::wait $token
http::status $token
} err]
# error code varies among platforms.
list $code [regexp {(connect failed|couldn't open
socket)} $err]
---- Result was:
0 0
---- Result should have been (exact matching):
1 1
==== http-4.14 FAILED
==== unixInit-3.1 TclpSetInitialEncodings FAILED
==== Contents of test case:
set env(LANG) C
set f [open "|[list $::tcltest::tcltest]" w+]
fconfigure $f -buffering none
puts $f {puts [encoding system]; exit}
set enc [gets $f]
close $f
unset env(LANG)
set enc
---- Result was:
iso8859-15
---- Result should have been (exact matching):
iso8859-1
==== unixInit-3.1 FAILED
Logged In: YES
user_id=80530
are these new failures?
What was the latest revision of Tcl
you tested under the same
conditions where these failures
did not occur?
Logged In: YES
user_id=399655
cmdMZ did not fail on 8.4a4
Logged In: YES
user_id=399655
cmdMZ did not fail on 8.3.4.
http and unixInit failed on 8.3.4. I believe you said http was connection related and unixInit was a holdover from
pre-ttf on unix days? Should I assume the only real issue is the cmdMZ?
Logged In: YES
user_id=399655
cvs_nightly 30Jun02: cmdMZ did not fail.
Logged In: YES
user_id=399655
cvs_nightly 30Jun02: cmdMZ did not fail.
Logged In: YES
user_id=80530
As noted in the test, cmdMZ-1.4 is
not portable to systems that do
weird caching things on permissions
in their filesystems.
The http test fails frequently, though
I've never analyzed it. Again, though,
it reveals something about the user's
system, not a bug in http.
the unixInit failure looks new to me.
Assigning to that category.
Logged In: YES
user_id=80530
TclpSetInitialEncodings is not a routine I know.
Logged In: YES
user_id=72656
This looks like an unimportant aspect of Solaris9 updating
their default C encoding from iso8859-1 to iso8859-15, which
is a good thing. 15 is a superset that includes the euro char,
and maybe a couple others.
Logged In: YES
user_id=80530
relaxed the test to accept either encoding.
fix committed to 8.4b2.