tor 2010-02-25 klockan 11:20 +1300 skrev Amos Jeffries:
What would be nice is that if the Kerberos libraries received NTLM input
they should handle it as NTLM instead of immediately rejecting it. When
that happens the Squid kerberos helper (or an extended one) should be able
to handle both
tor 2010-02-25 klockan 11:52 -0700 skrev Alex Rousskov:
What would this mean for Squid code? Can we still use simple O_APPEND
logging in SMP mode?
Yes, plus remove the built-in log rotation complexity, at least on
UNIX.. Not entirely sure about Windows.. (have some memory about
problems
tor 2010-02-25 klockan 11:47 -0700 skrev Alex Rousskov:
Main.cc installs a sig_child handler that catches SIGCHILD and calls
waitpid(). Watch_child() in main.cc also calls waitpid(). Why are we
waiting for the same thing in two places? Does not that create a race
condition, possibly not
tor 2010-02-25 klockan 13:45 -0700 skrev Alex Rousskov:
Will the simplified code just close and open the log file? Anything else
needed to integrate with external log rotation programs?
No, that's it. And done by setting logfile_rotate 0 in squid.conf so
no code changes is needed for it.
ons 2010-02-24 klockan 22:45 +0800 skrev Adrian Chadd:
On 24 February 2010 18:06, Adrian Chadd adr...@squid-cache.org wrote:
Uhm, is O_APPEND defined as an atomic write? I didn't think so. It may
be under Linux and it may be under certain FreeBSD versions, but it's
likely a side-effect of
ons 2010-02-24 klockan 17:03 +0100 skrev Livio B:
For example, even when kerberos auth succeeds (AF message), still
squid acls can deny access to that authenticated user. And the helper
has no way to know that. Depending on the scenario, it may make sense
to prompt the user again or not.
mån 2010-02-22 klockan 10:20 +0100 skrev Kinkie:
Unfortunately it's implementation-dependent. For all we know
posix_spawn may be a library convenience wrapper around fork/exec..
(vfork if we're lucky).
Well, it's been around for quite some time now.
If we want to argue about oldish systems
tis 2010-02-23 klockan 14:44 +0100 skrev Livio B:
The way this works is that the server first returns a 407 page with
e.g. 2 headers: Authenticate: Negotiate and Authenticate: Basic; then
the browser tries with its kerberos ticket or current user's ntlm
credentials (without user interaction);
tis 2010-02-23 klockan 15:52 +0100 skrev Kinkie:
This bug http://sources.redhat.com/bugzilla/show_bug.cgi?id=10354
shows that in glibc many cases posix_spawn() is exactly a wrapper for
(v)fork/exec .
So it is no worse than fork/exec, may be as good as vfork/exec
depending on the conditions,
mån 2010-02-22 klockan 20:38 -0700 skrev Alex Rousskov:
For logs, can we just use O_APPEND and have each process write to the
logs as we do now?
I would funnel the logs all via the same log_daemon..
but I guess O_APPEND would work out reasonably well, but it's a high
likelyhood that one
tis 2010-02-23 klockan 17:00 +1300 skrev Amos Jeffries:
Maybe officially offload to external log rotation software? They seem to
be widely available by default. The ones I've seen all provide some modular
config where we can drop-in a rotate config file and it's done. Some distro
packages do
tis 2010-02-23 klockan 08:47 -0700 skrev Alex Rousskov:
Let's compare the pros of a dedicated logging daemon with those of O_APPEND.
Daemon: Simplifies log rotation implementation.
And simplifies a whole lot when trying to send logs anywhere else than a
plain file, or if wanting to perform
tis 2010-02-23 klockan 12:50 -0700 skrev Alex Rousskov:
Do you think we need --disable-smp or even --enable-smp?
Can't say before I see a little more of how it unfolds, but it's most
likely easy to add if/when needed.
Regards
Henrik
tis 2010-02-23 klockan 12:45 -0700 skrev Alex Rousskov:
Actually, in my experience, many folks that want SMP disable transaction
logging completely. For them, we just need to support cache.log :-).
Yes, was more thinking of the selected few who actually want log as
well..
For others, there
mån 2010-02-22 klockan 15:20 +1300 skrev Amos Jeffries:
Just the case you keep kicking me about whenever I propose this:
That we also have to cater for broken clients when sending 417 back for
Expect: 100-continue requests.
That's towards clients, not towards servers.
Regards
Henrik
mån 2010-02-22 klockan 09:43 +0100 skrev Kinkie:
Hm.. do you think we need better documentation so we can just answer to RTFM?
Or is it something which can't be covered by documentaiton, or too
many different cases to be able to synthetize?
If you can come up with a better explanation than
mån 2010-02-22 klockan 11:34 +0100 skrev Kinkie:
[...]
If you can come up with a better explanation than the bucket and refill
rate..
A pair of delay parameters is written restore/maximum, where restore
is
the number of bytes (not bits - modem and network speeds are
sön 2010-02-21 klockan 20:37 +0200 skrev Tsantilas Christos:
Please take care to the #if WHEN_SQUID_IS_HTTP1_1 block in
HttpMsg.cc file which was disabled. This block allow squid to consider
a connection to the HTTP server as persistent in the case the
Connection: close header does not
sön 2010-02-21 klockan 20:37 +0200 skrev Tsantilas Christos:
There are broken servers out there that insist on receiving HTTP/1.1
requests.
Yes..
and even more which are broken when receiving HTTP/1.0 requests.
This patch adds a new force_http_1p1_request acl-driven option to
squid.conf.
lör 2010-02-20 klockan 17:12 +1300 skrev Amos Jeffries:
It has been found necessary in certain corner cases with PHP helpers
(which have system imposed limited lifetimes) where a proxy has
previously been under some load and helpers started then are timed out
later under low load as a
lör 2010-02-20 klockan 18:25 -0700 skrev Alex Rousskov:
The reasons you mention seem like a good justification for this option
official existence. I do not quite get the fork bomb analogy because
we are not creating more than a configured number of concurrent forks,
are we? We may create
sön 2010-02-21 klockan 00:52 -0700 skrev Alex Rousskov:
I agree that number and name do not match well. I struggled with this. I
did not want to use process_id to avoid the clash with system PID. Just
process sounds too generic and difficult to extend though. Would
process_number be better
lör 2010-02-20 klockan 19:14 -0700 skrev Alex Rousskov:
Hello,
If you recall, I am working on Squid that starts multiple processes,
each doing similar things. Even with this simple design, folks want to
(a) have differently configured processes (e.g., a process that is
dedicated to a
sön 2010-02-21 klockan 15:14 -0700 skrev Alex Rousskov:
Please note that the CPU affinity hack where every process gets its own
core would not work if ${process_name} is not a number, but we can come
up with another hack to support that without enumerating all
processes/cores.
From what I
mån 2010-02-22 klockan 11:44 +1100 skrev Robert Collins:
command protocol for it would be pretty similar to the SHM disk IO
helper, but for processes. Something like:
squid-helper: spawn stderrfd argv(escaped/encoded to be line NULLZ
string safe)
helper-squid: pid, stdinfd,
sön 2010-02-21 klockan 17:10 -0700 skrev Alex Rousskov:
The only inter-process cooperation I plan to support initially is N
processes monitoring the same http_port (and doing everything else).
I guess there will be no shared cache then?
I am working on option (B). While more complex, I think
Fedora now gets the following odd errors trying to build 3.1.0.16
g++ -I/usr/include/libxml2 -Wall -Wpointer-arith -Wwrite-strings
-Wcomments -Werror -D_REENTRANT -m32 -D_LARGEFILE_SOURCE
-D_FILE_OFFSET_BITS=64 -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2
-fexceptions -fstack-protector
Partner
Via Lucia Savarino, 110098 - Rivoli (TO) - ITALY
Tel. : +39.011.9530135 Fax. : +39.011.9781115
Email: guido.seras...@acmeconsulting.it
WWW: http://www.acmeconsulting.it
-Messaggio originale-
Da: Henrik Nordström [mailto:hen...@henriknordstrom.net
sön 2010-02-14 klockan 11:01 +1100 skrev Mark Nottingham:
* Make storeurl_rewriter work with Vary - bug 2678
Ok.
* Make miss_access a slow lookup - bug 2688
Applied to 2.HEAD, but not going into 2.7 at this time. Both new feature
and needs testing to make sure there is no flames on aborted
Release slightly delayed due to additional issues needing fixes..
fre 2010-02-12 klockan 22:36 +0100 skrev Henrik Nordström:
New 2.7 2.6 releases with patches for the resent security issues is
currently being prepared.
If I have overlooked any other patches you think should have been
fre 2010-02-12 klockan 09:06 +0100 skrev Kinkie:
On Fri, Feb 12, 2010 at 8:38 AM, Robert Collins
robe...@robertcollins.net wrote:
JFDI :)
+1
+1 here as well.
It's well in line with my own thinking.
Regards
Henrik
New 2.7 2.6 releases with patches for the resent security issues is
currently being prepared.
If I have overlooked any other patches you think should have been
included then speak up now. You have approximately 16 hours before the
releases is frozen.
Regards
Henrik
tor 2010-02-11 klockan 08:33 -0500 skrev Matthew Morgan:
I may be misunderstanding you here, but not specifying
range_offset_limit makes it default to 0, which only downloads what the
client asks for no matter what. -1 makes it download the whole object
regardless of the range asked for
Something seems to have gone wrong with error page generation in the
3.1.0.16 tarball. errors/ca/ERR_CACHE_ACCESS_DENIED is empty.
Regards
Henrik
sön 2010-02-07 klockan 15:43 +1300 skrev Amos Jeffries:
Yes, there was for a few hours a bundle that did not build.
The signature got as far as HTTP east/west before I erased everything
and started again.
And I picked up the first when building 3.1.0.16 for Fedora, which
caused some major
Hmm..
was there two signed 3.1.0.16 releases?
please do not respin a signed release. Once it is signed the path
forward is bumping the version..
Regards
Henrik
This is one of those odd bug reports which users encounter in the field
when running a packaged Squid (Fedora).
If Squid is build with IPv6 support enabled (which Fedora 12 Squid is),
but the host for some reason do not have IPv6 support enabled then the
binary will fail with comm_open: socket
lör 2008-03-08 klockan 16:39 +1300 skrev Amos Jeffries:
Got this far. But one thing is still making me think...
The official sigs Duane has up for STABLE1 look like this:
File: squid-3.0.STABLE1.tar.gz
Date: Fri Dec 14 00:26:31 GMT 2007
Size: 2406601
MD5 :
fre 2008-03-07 klockan 22:40 -0700 skrev Alex Rousskov:
I think we should link with group/libgroup.la libraries and compile with
group/name.h headers rather than adding Squid prefixes everywhere. If we
want Squid prefix as an extra protection, we could use
squid/group/name.*
fre 2008-03-07 klockan 21:49 +0100 skrev Henrik Nordström:
The SF devel CVS tree will continue, just haven't switched over the
update source from CVS to bzr yet.
It's now switched to pull the updates from bzr so future bzr commits
will show up in the SourceForge CVS tree as if nothing has
ons 2008-03-05 klockan 01:48 +1300 skrev Amos Jeffries:
Partway through.
Now the its I have never done before hit me ...
anything in particular to look for in the diff?
Only that it seems to match the changes which is expected to be there,
and that nothing else sticks out which isn't
ons 2008-03-05 klockan 03:20 +1100 skrev Robert Collins:
Where should I send the commit emails to for bzr? squid-commits ?
The list we are all on is [EMAIL PROTECTED]
Should probably get renamed to squid-commits.
Regards
Henrik
tor 2008-03-06 klockan 00:58 +1300 skrev Amos Jeffries:
A few questions now that I have thought out the consequences of the move:
1) Sourceforge Future. are SF CVS addition commits being sealed as
well? what about merging up to the migrated state of HEAD?
The SF devel CVS tree will
lör 2008-03-08 klockan 10:52 +1300 skrev Amos Jeffries:
In line with the 3.1 cleanup and re-arranging I'd like to propose a
change to the naming scheme of the *.la files squid produces.
Namely that they get called libsquid-X.la instead of just libX.la.
This would mean:
sön 2008-02-24 klockan 22:23 -0700 skrev Alex Rousskov:
On Sun, 2008-02-24 at 21:48 +0100, Henrik Nordström wrote:
Patches with a bugzilla reference: close the bug report but have it
targeted for the stable release.
Only if the person closing the bug report thinks the fix should
mån 2008-02-25 klockan 10:59 -0700 skrev Alex Rousskov:
The following patches are listed as to be merged (I think), but should
not be blindly merged until they are reviewed, polished, and committed.
11462 Bug #2230 possible fix: segmentation fault with a pure
virtual method
tis 2008-02-26 klockan 13:01 +1300 skrev Amos Jeffries:
Henrik:
is there something special to do with individual patches not to be
merged from a group of patches which as a whole are?
The include directive group which I have just done had 3 in the middle
which you may recall
tis 2008-02-26 klockan 01:08 +0100 skrev Henrik Nordström:
tis 2008-02-26 klockan 13:01 +1300 skrev Amos Jeffries:
Henrik:
is there something special to do with individual patches not to be
merged from a group of patches which as a whole are?
The include directive group which I
mån 2008-02-25 klockan 01:31 +1300 skrev Amos Jeffries:
Just done a short span of updates.
One thats standing out now is the myportname ACL mini-feature.
I'm doubtful, but agnostic. Are there any actual needs for this in 3.0?
It's trivial and a feature found in 2.7. But it's also a
Guido,
what's actually in the NT branch today? Is it only the makefiles, or
is there any actual source changes which should not be merged to the
main branch?
If it's only the makefiles then I propose those are stored in the main
branch (HEAD and SQUID_3_0) directly
Looking at a diff...
port
fre 2008-01-25 klockan 08:42 +0200 skrev Razard:
Question about basic user authentication on proxy.
If user get http web page first times, the proxy respond to
authenticate them unsecure as default, so what described on login
windows on browser. But if user get https page, browser creates
This message is primarily to the developers who commit stuff in the main
tree.
To get a little structure in the release process it helps if only the
release manager commits stuff on STABLE branches. The exception being if
you have been specifically asked to backport something. It's normally no
mån 2008-02-18 klockan 15:58 -0700 skrev Alex Rousskov:
Since this problem did not go away since January and Henrik may not be
available, perhaps you can log a ticket with SourceForge?
That's the proper way.
SourceForge site services support requests is filed here:
sön 2008-02-24 klockan 22:54 +0100 skrev marc impini:
as it was suggested on #squiddev, i'm writting this little
introduction to myself:
Welcome.
Actually i'm particullary interested in https proxying (ssl_bump squid
3 feature) as it's related to my employment (need to filter ciphered
(12.28.36) amosjeffries: I've added a script to the 3-HEAD changesets
you might like - .group - to save typing ln -s all the time. I have
yet to add 'merge' to the .group script, but it will only do all the
inks for a nomerge group. but it saves half the typing already anyway.
Good idea.
http://www.squid-cache.org/Versions/v3/HEAD/changesets/11438.patch
This is plain wrong use of cbdataReference(). cbdataReference returns a
reference to the cbdata object, and you are meant to use
cbdataReferenceDone on that reference when done.
The current implementation happens to return the
The first pass of grouping and classifying the 3 HEAD changesets since
3.0 branches has now been completed by me and Amos.
Please go to
http://www.squid-cache.org/Versions/v3/HEAD/changesets/merge.html and
inspect the list of patches to merge and not merge and give your opinion
if you think some
lör 2008-02-23 klockan 13:34 +0100 skrev Henrik Nordström:
The idea with cbdata is that when you are given a cbdata enabled object
use cbdataReference on it to get a reference safe to be saved between
call events. Then use cbdataReferenceValid to verify it's validity if
needed
fre 2008-02-22 klockan 11:29 -0700 skrev Alex Rousskov:
On Fri, 2008-02-22 at 19:23 +0100, Guido Serassio wrote:
Changing the case of files/dir will not be a problem if we will avoid
upper/lower case collisions.
This only applies to files in the same directory, right? AFAICT,
tor 2008-02-14 klockan 09:09 -0700 skrev Alex Rousskov:
1) Expose Squid internals: Publish/install Squid headers and
libraries to give direct access to Squid resources. This
approach will most likely require installing pretty much all
headers because the module
fre 2008-02-15 klockan 09:07 +1100 skrev Robert Collins:
Its more work both at code and at runtime. The only thing it really
allows that 1) doesn't is non-GPL eCAP modules.
I don't see how 2) can allow non-GPL eCAP modules. We can't add a
linking excemption to the license even if there is a
lör 2008-02-16 klockan 18:14 -0700 skrev Alex Rousskov:
On Fri, 2008-02-15 at 14:05 +, Arthur Tumanyan wrote:
Update of cvs.devel.squid-cache.org:/cvsroot/squid/squid3/errors/Armenian
Modified Files:
ERR_ESI ERR_ICAP_FAILURE
Log Message:
Another SourceForge HEAD commit?
sön 2008-02-17 klockan 11:46 +0100 skrev Guido Serassio:
Hi,
Im resending again this message, because the problem is still here,
since 21 January ..
18'th actually..
# Set a temporary tag at the new version we are merging towards...
cvs -q rtag -F -r HEAD Z-nt_merge-new_HEAD squid
lör 2008-02-23 klockan 18:35 +0100 skrev Guido Serassio:
Hi Henrik,
At 14:35 23/02/2008, Henrik Nordström wrote:
The first pass of grouping and classifying the 3 HEAD changesets since
3.0 branches has now been completed by me and Amos.
Please go to
http://www.squid-cache.org/Versions/v3
ons 2008-02-20 klockan 12:04 +1300 skrev Amos Jeffries:
Henrik:
It was suggested I ask you about how-to for editing the changesets.
Seeing as I have stood up for 3.1 maintainer, I think I might get have
some practice on 3.0.
There is no editing, just classification grouping. It's all
sön 2008-02-10 klockan 12:28 +0100 skrev Guido Serassio:
Hi Henrik,
At 03:01 08/02/2008, Henrik Nordstrom wrote:
hno 2008/02/07 19:01:16 MST
Modified files:
src cache_cf.c
Log:
Make include support wildcard patterns
Revision ChangesPath
sön 2008-02-03 klockan 19:05 +1300 skrev Amos Jeffries:
The problem with MD5 is what the side-effect of altering the MD5 in
store will do. Would it make older caches after upgrade 'loose' all
their content as never-matching-again objects?
I don't know enough at this point to answer that
This has been discussed a couple of times before, but there was no final
conclusion, so lets try again.
I propose storing cf.data.pre split in one file per directive in the CVS
repository, as generated by the scripts/split-cf.data.pre.pl
The purpose is to ease maintenance when merging squid.conf
tis 2008-01-15 klockan 18:13 +0900 skrev Adrian Chadd:
I'm trying to get tproxy working, and I'm stuck trying to build a kernel
with both capabilities -and- the latest tproxy-2 stuff that Squid supports.
So far I've got a 2.6.20 vanilla kernel which I've patched tproxy-2 into.
That works
sön 2008-01-13 klockan 04:44 +0100 skrev Bernhard Schmidt:
Although, why does it say NONE/- instead of printing the server it
fetched the data from? This is only happening occasionally, the whole
update looks like this.
1200195061.914328 2001:1b10:100:3::1:2 TCP_MISS/200 514 GET
ons 2008-01-09 klockan 17:56 +0900 skrev Adrian Chadd:
My security fix which went into 2.6.stable17 had an off-by-one in it
and thus occasionally would assert() incorrectly.
Henrik, could we please roll a 2.6.stable18 release specifically
for that bug fix?
2.6.STABLE18 rolled yesterday.
ons 2008-01-09 klockan 21:13 +1300 skrev Amos Jeffries:
That this code is currently depending on that implicit guarantee when it
should not be. This change is the perfect time to drop that implicit
dependency and make it clean (bug free!). Having one precursor call
schedule its successor
tor 2008-01-10 klockan 20:40 +1300 skrev Amos Jeffries:
Doh! thanks. I think I see the problem, its should be checking tmpbuf is
NULL when no records are present, non-NULL when they are.
The assert was there to stop the safe free dying nastily. If it's
working for you better without
tor 2008-01-10 klockan 22:45 +1300 skrev Amos Jeffries:
query 1:
response: CNAME + NX additional
result: was returns no-results shortcut in DNS.
(identical if A returns CNAME-only)
You can only shortcut if you get a NXDOMAIN, not when you get a positive
response but with no answer of
tis 2008-01-08 klockan 23:39 -0700 skrev Alex Rousskov:
The ICAPXaction::readBuf data member stores incoming headers. It is
initialized as follows:
readBuf.init(SQUID_TCP_SO_RCVBUF, SQUID_TCP_SO_RCVBUF);
Ok. So it should then be limited to TCP_SO_RCVBUF before seeing the same
problem.
ons 2008-01-09 klockan 08:15 -0700 skrev Alex Rousskov:
My understanding is that you can replace the second SQUID_TCP_SO_RCVBUF
constant with a larger one to allow the buffer to grow. No other changes
should be needed except taking care of this assertion:
Yes, provided you actually try to
tor 2008-01-10 klockan 01:18 +0100 skrev Bernhard Schmidt:
Amos Jeffries schrieb:
Bernhard:
Can you get a cache.log trace of section 14 please?
debug_options 14,9 gives:
2008/01/10 01:12:59.532| ipcacheParse: edge.jobthread.com CNAME
This is a IPv4-only CNAME host. An query
tis 2008-01-08 klockan 02:17 +0900 skrev Adrian Chadd:
I've been thinking about how to fix that particular issue whilst hacking on
the
s27_adri branch. The main requirement is modifying the store to allow
arbitrary
sized pages rather than just 4k pages, and then find a way to read the
tis 2008-01-08 klockan 13:53 +0100 skrev Bernhard Schmidt:
Hi,
I've been running Squid 3.0-ipv6 branch and now 3.HEAD for quite some
time and found that after an upgrade my Squid cache was wildly crashing
with
assertion failed: ipcache.cc:995: tmpbuf
This is related to recursive
tis 2008-01-08 klockan 00:47 +0900 skrev Adrian Chadd:
My next trick will be breaking up cf.data.pre along the documentation
boundary lines and have seperate configuration file chunks for various
tasks.
cf.data.pre should be split in one file per directive, using the
existing scripts.
the
ons 2008-01-09 klockan 11:41 +1300 skrev Amos Jeffries:
It has worked for A records because the trend there has been for people
to use A records commonly. These days of records and long-IP have
seen a major increase in CNAME records pointing at shared instead
of multiple
tis 2008-01-08 klockan 16:36 -0700 skrev Alex Rousskov:
This may be a little too aggressive. I expect AsyncCalls and eCAP to
become more-or-less stable by the end of January. On the other hand,
perhaps PRE1 does not mean a merge freeze.
PRE means feature complete but there may still be known
ons 2008-01-09 klockan 11:30 +0900 skrev Adrian Chadd:
You don't need to release the buffer. Whats the difference in having
a 4k buffer of reply status, headers and body with 300 odd bytes
of metadata, versus ~4k worth of allocated strings representing the reply
status and headers? probably
ons 2008-01-09 klockan 15:36 +1300 skrev Amos Jeffries:
Henrik Nordström wrote:
tis 2008-01-08 klockan 16:36 -0700 skrev Alex Rousskov:
This may be a little too aggressive. I expect AsyncCalls and eCAP to
become more-or-less stable by the end of January. On the other hand,
perhaps PRE1
ons 2008-01-09 klockan 16:28 +1300 skrev Amos Jeffries:
In those cases its a broken server as you point out, and will respond
with naked CNAME whether asked for A or .
If your DNS resolver server responds with a naked CNAME in response to a
A query then the requested node has a CNAME, but
Hmm.. I guess we can delay the A lookup until connection timeout, which
would both improve performance and compliance.
ons 2007-10-10 klockan 00:10 -0600 skrev Adrian Chadd:
Interesting! I'd suggest leaving it on by default though and logging
statistics
showing the number of requests which
Squid-3 has lots of problems with response headers 4KB. (Bug #2001)
3.0.STABLE hangs and stops reading the response. Cleared by abort or
timeout only.
3.HEAD at least grows it's header receive buffer and continues to read
the data and feed the response down the chain, but this then triggers
mån 2008-01-07 klockan 02:30 -0800 skrev Arthur Tumanyan:
Arthur Tumanyan wrote:
Hi,I'm currently working above mysql logging feature for
squid3(cleanly,not a plugin like previous version for squid.2.6).
And the trouble is the message in subject when i do ./bootstrap.sh.
I don't
Plese file a bug report if you haven't already.
Regards
Henrik
sön 2008-01-06 klockan 18:37 +0100 skrev Pawel Worach:
There seems to be another tiny leak related to vary accept-encoding
processing:
==69932== 56 bytes in 4 blocks are definitely lost in loss record 2 of 23
==69932==at
sön 2008-01-06 klockan 12:27 +0200 skrev Tsantilas Christos:
Hi Henrik,
It is not a cbdata problem.
Are you sure?
cbdata exists to help invalidate callbacks when they are no longer
meaningful. A callback being made to something no longer meaningful is a
clear sign of a missing cbdata
sön 2007-12-30 klockan 11:24 +1100 skrev Robert Collins:
This is most of a changeover of scripts for squid 3 trunk to use bzr;
the missing bit appears to need a bzr 1.1 (to do 'rdiff' basically) or
thereabouts, I'll look into that in a bit. I'm not sure that the
unconverted cvs calls will
fre 2007-12-28 klockan 16:16 +0200 skrev Tsantilas Christos:
Hi Robert,
Which are the required steps from developers?
Will we open our computers a morning and find all our development
branches in the new system :-) ?
What about the sourceforge developers repository? Will be hosted in the
fre 2007-12-28 klockan 10:42 +1100 skrev Robert Collins:
On Fri, 2007-12-28 at 08:25 +0900, Adrian Chadd wrote:
I've been following the VCS debate a little.
Guys, I'm not an enormous fan of CVS, but what we have works
*badly*. 150 emails when autoconf changes are made for instance.
lör 2007-12-29 klockan 05:24 +1100 skrev Robert Collins:
Other projects using distributed VCS tools often do not have a dedicated
development area, preferring to let individuals publish their own
branches; I had been assuming something like that - but its a good
question to raise and discuss.
tor 2007-12-27 klockan 07:52 +1100 skrev Robert Collins:
Do I have sufficient access to do this? I'm not a BSD afficiondo anyhow,
so perhaps its best if I say 'install the bzr port' as long as its at
version 1.0 now. (If its not we can just install by hand). Same for
loggerhead which a
fre 2007-12-28 klockan 08:25 +0900 skrev Adrian Chadd:
I've been following the VCS debate a little.
Guys, I'm not an enormous fan of CVS, but what we have works, and
I think we have bigger things to work on right now than a VCS migration.
Unless someone can show how we'll improve
fre 2007-11-30 klockan 14:18 +0900 skrev Adrian Chadd:
The netdb code stuffs the netdb_state file in the first swap dir; which
obviously fails if you're running a cache with a null swapdir.
Hmm.. why does it do that? It should stuff it into a StoreEntry like
everything else..
Regards
Henrik
501 - 597 of 597 matches
Mail list logo
