None other than that I tried again reproducing the problem and failed..
Regards
Henrik
On mån, 2008-11-03 at 12:41 +0100, Malte Schröder wrote:
Hi,
any news on this topic?
On Thu, 23 Oct 2008 14:53:49 +0200
Henrik Nordstrom [EMAIL PROTECTED] wrote:
On fre, 2008-10-24 at 00:48 +1300
Nice work you did with the grabbing the change history from
RELEASENOTES.html. Very nice! Even if it gets slightly confusing
regarding 3.0 and how it compares to earlier/lower numbered releases.
The subdirs have now been changed to symlink the template, which means
that however template.dyn gets
On tor, 2008-10-30 at 10:49 -0600, Alex Rousskov wrote:
I do not have gssapi.h and it should not be required to build Squid with
my configure options. Sorry for not being explicit: I need the trunk
code to build cleanly. I do not care if the, hopefully, optional
squid_kerb_auth is built.
On ons, 2008-10-29 at 15:57 +1100, Mark Nottingham wrote:
I'm looking at a dump from the wire, and it doesn't make any sense;
e.g,. something logged as taking 79ms takes less than .5ms from GET to
last ack.
The interesting thing is that this seems related to client persistent
On tor, 2008-10-30 at 01:05 +1300, Amos Jeffries wrote:
The info comes from the latest current global cfgman so the default
values of many settings are already wrong referring to values only
present in 3.1 and 3.HEAD. Think http_port 80 sslbump in 2.6, 2.7,
3.0 and listed default
for opt_send_signal == 0.
On 29/10/2008, at 12:13 PM, Henrik Nordstrom wrote:
On ons, 2008-10-29 at 11:32 +1100, Mark Nottingham wrote:
I'm not seeing that; no crash, and -k check still correctly finds
syntax errors in squid.conf. Behaviour appears the same as without
the
patch.
squid
Thanks. Fixed.
A full build matrix is planned before merging to 2.7.
On tor, 2008-10-30 at 10:09 +1100, Mark Nottingham wrote:
Small typo in comm_kqueue.c:157;
if (ignoreErrno(saved_errnoerrno))
On 29/10/2008, at 7:51 PM, Henrik Nordstrom wrote:
On ons, 2008-10-29 at 09:27
On ons, 2008-10-29 at 11:48 +0100, Kinkie wrote:
I'm mulling over the possibility of importing IRC logs from
selected channels (FreeNode #squid and #squiddev) into the wiki, as
done in http://www.moinmo.in/MoinMoinChat/Logs/moin-dev
-1
There is too much noise in this traffic to provide
On tor, 2008-10-30 at 12:24 +1300, Amos Jeffries wrote:
I think we should do it when we hold meetings. But not as a matter of
routine.
That's better accomplished by copy-paste and clean up a log from one of
the participants.
collaborations to get something fixed fast. Not exactly important
On mån, 2008-10-27 at 10:32 -0400, [EMAIL PROTECTED] wrote:
access.log seems working but connection is very slow and getting error
messages from cache.log like i said before..
1225128451.805 1395 127.0.0.1 TCP_MISS/200 536 GET
https://mail.google.com/mail/rc? - DIRECT/72.14.247.83
The first part next to checkRunningPid makes sense.
But the second part crashes squid -k check, by always returning true
if the pid file doesn't exists..
On tis, 2008-10-28 at 20:16 +1100, Mark Nottingham wrote:
I realised I never sent the patch without the option;
On 12/08/2008, at
On ons, 2008-10-29 at 11:32 +1100, Mark Nottingham wrote:
I'm not seeing that; no crash, and -k check still correctly finds
syntax errors in squid.conf. Behaviour appears the same as without the
patch.
squid -k check is for checking if Squid is running.
squid -k parse is the one parsing
On sön, 2008-10-26 at 11:11 -0400, Steve Snyder wrote:
Example #1:
acl snapcom dstdomain .snap.com
http_access deny snapcom
Example #2:
acl pogotime time SMTFA 00:00-23:59
acl pogosite dstdomain .pogo.com
http_access deny pogosite pogotime
Where are these in relation to your other
On sön, 2008-10-26 at 14:23 -0400, Steve Snyder wrote:
On Sunday 26 October 2008 02:11:28 pm Henrik Nordstrom wrote:
On sön, 2008-10-26 at 11:11 -0400, Steve Snyder wrote:
Example #1:
acl snapcom dstdomain .snap.com
http_access deny snapcom
Example #2:
acl pogotime time
On lör, 2008-10-25 at 10:05 -0400, [EMAIL PROTECTED] wrote:
and another error i get is fallows;
2008/10/25 19:55:32| fwdNegotiateSSL: Error negotiating SSL
connection on FD 15: error:140770FC:SSL
routines:SSL23_GET_SERVER_HELLO:unknown protocol (1/-1/0)
What does access.log say?
bzr on squid-cache.org has been updated to version 1.8
Regards
Henrik
signature.asc
Description: This is a digitally signed message part
On tor, 2008-10-23 at 15:45 +1100, Mark Nottingham wrote:
I've been thinking a while that it would be useful to make the
collapsed_forwarding timeout configurable; right now it's hard-coded
at 30 seconds. Would reducing this time help manage the depth (and
perhaps load implications)?
On tor, 2008-10-23 at 15:58 +1100, Mark Nottingham wrote:
Bit more on this;
2008/10/22 21:48:31| fd_open FD 21 HTTP Request
2008/10/22 21:48:31| httpAccept: FD 21: accepted port 3128 client
127.0.0.1:44244
[...]
2008/10/22 21:48:31| aclCheck: checking 'always_direct allow
+1300
Amos Jeffries [EMAIL PROTECTED] wrote:
Malte Schröder wrote:
On Thu, 23 Oct 2008 01:10:58 +0200
Henrik Nordstrom [EMAIL PROTECTED] wrote:
On ons, 2008-10-22 at 11:31 +0200, Malte Schröder wrote:
Not normal.
Squid version?
3.0.STABLE10
And how did you
On fre, 2008-10-24 at 00:48 +1300, Amos Jeffries wrote:
Thanks.
...um... I was expecting to see the peer:port/domain listed uniquely
with each pconn FD.
Domain is only used on originserver class peers.
Regards
Henrik
signature.asc
Description: This is a digitally signed message part
On sön, 2008-10-19 at 14:09 -0600, Alex Rousskov wrote:
I know this is an old thread, but if somebody is going to work on this,
please add some logic/options that would allow the peer to stay up
after N sequential failures. The current Squid2 monitoring code is not
usable in busy environments
On ons, 2008-10-15 at 12:04 +1100, Mark Nottingham wrote:
FWIW, my original use case was to allow passing arguments to cachemgr
functions, so we can do things like:
This is completely unrelated to changing the scheme. cache_object://
also theoretically supports parameters just as http://
tis 2008-10-14 klockan 16:28 +0200 skrev Kinkie:
I agree.
What about restructuring to
- core
Instead of core I would use general, for everything not having a
explicit component.
- authentication and authorization
- docs
- content adaptation
- testsuite
Looks good.
Regards
Henrik
On tis, 2008-10-14 at 12:17 +0200, Kinkie wrote:
Hi all,
IMVHO Bugzilla is missing a couple of component specifiers:
- for the squid project, it misses a possibility of reporting bugs to
the unit tests suite
- for the website, the missing component is bugzilla itself.
If there's
On tis, 2008-09-30 at 22:57 -0700, sandiphw wrote:
Hi,
Is it possible to arrage groupbased-timebased acl in squid? Eg, I want to
create to groups of users by their initials, say G-1 G-2. G-1 should have
all time access to internet but G-2 should have access for a fixed time
perod say 10 Am
On fre, 2008-10-10 at 21:02 +1300, Amos Jeffries wrote:
* I've omitted the files touched by Alexes patches from this run.
Reformat for them will be done and committed when the patches are
verified or updated.
Ok,
* I may run reformats irregularly on trunk manually afterwards, like we
On tor, 2008-10-09 at 16:01 +1300, Amos Jeffries wrote:
revno: 9270
committer: Henrik Nordstrom [EMAIL PROTECTED]
branch nick: trunk
timestamp: Thu 2008-10-09 00:39:31 +0200
message:
Properly fix squid_kerb_auth configure
On mån, 2008-09-29 at 08:53 +0200, kinkie wrote:
src/Makefile.am is missing one include location, and this prevents
out-of-tree builds due to a missing include in cf_gen.cc.
Please add
-I$(top_builddir)/src
to INCLUDES
or just apply the attached patch.
Odd. I always do out-of-tree builds
On mån, 2008-10-06 at 23:37 +1300, Amos Jeffries wrote:
For now, if people agree to the above, I can make _just_ the current
'ntlm_auth' helper name match that scheme.
Or, given the current +3/-1 votes change it to smb_lm_auth if no
agreement on a naming scheme alteration.
Don't count my
On mån, 2008-10-06 at 11:55 +0200, Guido Serassio wrote:
What about something like this ?
fake_ntlm_auth
mswin_ntlm_auth
no_check_ntlm_auth.pl
lm_ntlm_auth
Perfect ;-)
Regards
Henrik
signature.asc
Description: This is a digitally signed message part
Done.
On mån, 2008-10-06 at 21:55 +0200, Guido Serassio wrote:
Hi Henrik,
At 13.24 25/09/2008, Henrik Nordstrom wrote:
2.7.STABLE5 is being planned for release, and if you know any relevant
patches or important bugs in bugzilla or otherwise please speak up now.
This Windows fix is also
On sön, 2008-10-05 at 14:29 +1300, Amos Jeffries wrote:
The users who have needed to use it recently pretty much overrule
complete removal in 3.1.
Does squid_smb_lm_auth cover everyones understanding of it?
Like to keep the ntlm part if possible as it's a good indication of what
auth_param
On lör, 2008-10-04 at 08:22 +0200, Kinkie wrote:
We already are, see /squid-internal-static, cache digests and friends.
Rather than configuring a different port/interface I'd prefer making
the access path configurable (/squid-internal-*)
Well.. not really. See global_internal_static
On fre, 2008-10-03 at 15:19 +1300, Amos Jeffries wrote:
Actually, what reasoning has gone into using ntlm_smb_auth as the new
name?
Because it uses SMB for the NTLM verification, not Windows domain
services.. Works to any SMB server who is a member or trust member.
Regards
Henrik
New attempt to batch Bug #2447. This time it actually builds in trunk
and the new segfault error spotted by Amos corrected.
# Bazaar merge directive format 2 (Bazaar 0.90)
# revision_id: [EMAIL PROTECTED]
# kt0uvz3ppaaha0ec
# target_branch: http://www.squid-cache.org/bzr/squid3/trunk/
#
On tor, 2008-10-02 at 11:14 +0200, Kinkie wrote:
../../src/src/dns_internal.cc: In function 'void
idnsInitVCConnected(int, comm_err_t, int, void*)':
../../src/src/dns_internal.cc:721: error: 'class IPAddress' has no
member named 'sin_addr'
../../src/src/dns_internal.cc: In function 'void
On tor, 2008-10-02 at 16:04 +1300, Amos Jeffries wrote:
Henrik,
did you check my comment about the fatal bug in that patch before
committing?
No, didn't see that. Sorry.
if(!vc) ... nameservers[vc-ns] ...
Will crash with a different more useless trace in the middle of a debug
context
On fre, 2008-09-26 at 13:25 +1200, Amos Jeffries wrote:
logdaemon, internal redirectors, and netdb are big component changes. But
not squid-wide.
Yes. Isolated feature changes. Number of lines added/deleted does not
matter much, the complexity is in the amount of changes to existing
code.
The
On tor, 2008-09-25 at 15:15 +1200, Amos Jeffries wrote:
So we have
1. Branch when trunk is considered a suitable startingpoint for getting
to stable, and tag a x.y.0 release at the branchpoint (or at least set
this version in configure.in).
2a. Release X.Y.0.1 when ready for testing
On tor, 2008-09-25 at 22:41 +1200, Amos Jeffries wrote:
So I should simply say 'X is how we are doing it'? I think not.
Kinkie has somehow convinced Alex of a different way of numbering then
simply omitting the word 'STABLE', I'm pondering it but need a good
reason to agree.
No, it's
2.7.STABLE5 is being planned for release, and if you know any relevant
patches or important bugs in bugzilla or otherwise please speak up now.
What's collected so far can be seen in the changesets. Not much, but one
rather critical one.. (2464)
On tor, 2008-09-25 at 23:36 +1200, Amos Jeffries wrote:
Hmm, so rolling DEVEL + PRE + RC into one version called a.b.0.N
Well, I'm a little doubtful but can't argue against that yet.
No, not quite.
There is no point of DEVEL releases in the current form of the
development cycle. So we skip
On tor, 2008-09-25 at 08:11 -0600, Alex Rousskov wrote:
This is a different issue, but I do not think that limiting access is a
good idea until the branch becomes stable. When 3.1 is branched (which
you want to happen soon), there will be a lot of cleanup work still
going on. I see no reason
On tor, 2008-09-25 at 08:40 -0600, Alex Rousskov wrote:
I do not want the meaningful words like RC or STABLE to be in the
version string. And I want a way to release many times with a specific
version (not an automated snapshot) before the branch is marked as
stable.
Same here, but I am not
On tor, 2008-09-25 at 12:40 -0600, Alex Rousskov wrote:
Understood. Well, if we branch now, I think you will not have cycles to
work on the above for at least a month, because of the back/forward
porting overhead, _especially_ if the trunk becomes open again.
trunk being open is not that big
On tor, 2008-09-25 at 12:57 -0600, Alex Rousskov wrote:
Understood. What do you think is better:
A) Branch and commit waiting stuff to trunk now. Deal with a large
volume of complicated commit traffic between the 3.1 branch and trunk.
B) Branch and commit waiting stuff later. Focus on 3.1
On tor, 2008-09-25 at 13:32 -0600, Alex Rousskov wrote:
If we use RC label for selected/last 4-digit releases, then we will not
have the above problems. If a 4-digit release proves stable, we post
3.1.1 (an exact copy of the last successful 4-digit release, except for
the version number) and
On ons, 2008-09-24 at 14:14 +1200, Martin Langhoff wrote:
Good hint, thanks! If we did have such a control, what is the wired
memory that squid will use for each entry? In an email earlier I
wrote...
For on-disk objects about 100 bytes.
In-memory objects obviously uses a lot more. Probably
suqid-cache.org bzr has been upgraded to 1.6.1 (was 1.5)
Regards
Henrik
signature.asc
Description: This is a digitally signed message part
On ons, 2008-09-24 at 13:56 -0600, Alex Rousskov wrote:
Cool. So, if there are no objections, we have:
0) Trunk is usually open for new stuff. Make daily snapshots.
1) Branch X.Y when major features are committed. Snapshots.
2a) Release X.Y.0.0 when all major bugs are fixed.
2b) Release
On ons, 2008-09-24 at 21:26 +0200, Kinkie wrote:
The version X.Y.Z for all Z W approach means that it's not easy to
tell what the state of a release is without knowing what W is for each
X,Y.
Even then you don't know as the notion of something being stable is very
temporal in nature and
On tis, 2008-09-23 at 14:57 +0800, Adrian Chadd wrote:
You can limit the amount of cache_mem which limits the memory cache
size; you could probably modify the squid codebase to start purging
objects at a certain object count rather than based on the disk+memory
storage size. That wouldn't
Talked with Aaron about this, and most likely the merge request did not
have a proper public branch location and as result Bundlebuggy failed to
select the proper project for the merge request.
voting and approval is per merge request, and is governed by the project
assigned to the merge...
the *_header_access directives can only change the external view of the
headers sent by Squid, not how Squid parses received headers.
To override bad things sent by client or server you need to use
refresh_pattern overrides. In your case both the ignore-reload and
ignore-private overrides.
Be
On tor, 2008-09-11 at 23:07 -0600, Alex Rousskov wrote:
It looks like I mistook the Reply-To: email header for a To: header. I
was not really sending those commit messages to squid-dev, just to
squid-cvs. So I was not really spamming. Until now...
No problem.
Got me a little confused there
On tor, 2008-09-11 at 09:54 -0600, Alex Rousskov wrote:
You mean as a large blob of comments at the end of some .h file? The
cons are:
I think Adrian is thinking more of blobs next to the implementations
(.cc)
As time has passed a little I have this slightly refined view:
* Overview type
On tor, 2008-09-11 at 10:03 -0600, Alex Rousskov wrote:
You made the right choice. The modifying methods modify their object
approach is more efficient and should be expected by most C++
developers.
Yes, the other approach (return a new object) is used mainly for
operators.
C = A + B;
On fre, 2008-09-12 at 20:39 +1200, Amos Jeffries wrote:
+#define WCCP2_FORWARDING_METHOD_GRE WCCP2_METHOD_GRE
+#define WCCP2_FORWARDING_METHOD_L2 WCCP2_METHOD_L2
+#define WCCP2_PACKET_RETURN_METHOD_GRE WCCP2_METHOD_GRE
+#define WCCP2_PACKET_RETURN_METHOD_L2WCCP2_METHOD_L2
entities and using them in
this fashion makes it clearer for people working on the code.
Adrian
2008/9/13 Henrik Nordstrom [EMAIL PROTECTED]:
On fre, 2008-09-12 at 20:39 +1200, Amos Jeffries wrote:
+#define WCCP2_FORWARDING_METHOD_GRE WCCP2_METHOD_GRE
+#define
On ons, 2008-09-10 at 11:38 -0600, Alex Rousskov wrote:
* It is not a bug to cancel something late, but the canceling code
should be aware of that possibility. Canceling something late is a
no-op.
Can you define late?
Is this after the call has been queued, or after it has been dialed?
On ons, 2008-09-10 at 11:33 -0600, Alex Rousskov wrote:
Since all notifications are asynchronous, it is possible for a read
or write notification that was scheduled before comm_close was
called to arrive at its destination after comm_close was called.
Such notification will arrive
On ons, 2008-09-10 at 15:45 -0600, Alex Rousskov wrote:
Take an HttpStateData, for example. When the socket descriptor is
closing, HttpStateData may still have a lot of work to do. The old code
(some of which remains) assumed that comm_close is pretty much the final
word. After that, there is
On ons, 2008-09-10 at 16:30 -0600, Alex Rousskov wrote:
The fd_t conversion work should not go into v3.1, right?
I don't have an opinion on that. It depends on when it gets done and
what the 3.1 release maintainer thinks. But I don't see it likely to get
included in 3.1 no.
Regards
Henrik
On tis, 2008-09-09 at 14:59 +1200, Amos Jeffries wrote:
revno: 9176
committer: Alex Rousskov [EMAIL PROTECTED]
branch nick: trunk
timestamp: Mon 2008-09-08 17:52:06 -0600
message:
Fixed typo: Config.Addrs.udp_outgoing was
On ons, 2008-09-03 at 16:53 +0200, Kinkie wrote:
I didn't really think of different buffer types. Do you have in mind
any scenario where it would be useful?
One example is if KBuf gets implemented using a mallocator that may
reallocate the memory area to reduce fragmentation.
On the other
ons 2008-09-03 klockan 16:30 +1200 skrev Amos Jeffries:
Declaring variables as needed rather than in advance makes code
maintenance easier and might help the compiler to optimize. Not a big
deal though.
But it does needlessly break the oldest-possible-compiler policy.
In Squid-2 the base
That patches the MSNT/SMB helper (msnt_auth), not the session helper..
ons 2008-09-03 klockan 22:25 +1200 skrev Amos Jeffries:
Advanced testing reveals that the session helper does not use the
string-safe functions Squid should be using.
This patch corrects that and a bit of formatting
On sön, 2008-08-31 at 17:43 +1200, Amos Jeffries wrote:
Henrik Nordstrom wrote:
The attached patch is unverified, but attemtpts to add proper checks on
TCP DNS connections to detect log when there is some problem
connecting to the nameserver.
Additionally it corrects a bug
On sön, 2008-08-31 at 14:07 +0200, Kinkie wrote:
Now I'm a bit at a loss as to how to best integrate with iostream.
For what?
We need to be able to dump the contents in debugs which is easily
implemented as an operator using the already defined interface. What
other uses of iostream do we
On sön, 2008-08-31 at 22:40 +0200, Kinkie wrote:
src=src.substr(pos,-1); //to end-of-string
This is not valid in your proposed interface. The substr() arguments is
defined as unsigned.
Regards
Herik
signature.asc
Description: This is a digitally signed message part
The attached patch is unverified, but attemtpts to add proper checks on
TCP DNS connections to detect log when there is some problem
connecting to the nameserver.
Additionally it corrects a bug in connection management if there is more
than one nameserver configured. The code forgot to assign
Just realized that bundlebuggy does not track patches submitted for the
stable branch branches/SQUID_3_0. I think it should..
signature.asc
Description: This is a digitally signed message part
On fre, 2008-08-29 at 10:44 +1000, Mark Nottingham wrote:
I'm not convinced it's a great solution, but something like URISpace
may be appropriate;
http://www.w3.org/TR/urispace.html
Not very different in function, besides being XML..
However, the URISpace xml tag space is a bit too
On fre, 2008-08-29 at 08:38 +0800, Adrian Chadd wrote:
Make sure you can collapse those ACLs down to something sensible for
software processing before you go down that path!
It's relatively easy to make a unified lookup tree of such structure,
and even if you don't it's still as fast or faster
On tis, 2008-08-26 at 19:34 -0600, Benno Rice wrote:
benno 2008/08/26 19:34:42 MDT
Modified files:
src url.c
Log:
Fix breakage caused by recent GCC fixes.
The fixes in question prevented the returning of pre-defined method_t
structures as well as
On tor, 2008-08-28 at 23:26 +1200, Amos Jeffries wrote:
+const char *
+urlAbsolute(const HttpRequest * req, const char *relUrl)
+{
+LOCAL_ARRAY(char, portbuf, 32);
+LOCAL_ARRAY(char, urlbuf, MAX_URL);
LOCAL_ARRAY is unfortunately not thread safe. Someone else may argue,
but
On tor, 2008-08-28 at 22:39 +1000, Benno Rice wrote:
Just as a side question, once I've reworked my code to address Amos'
concerns, what's the workflow for resubmitting? Do I just bzr send
again? Is there anything I need to do to tell bundlebuggy (or
whatever) that it's a resubmit?
On tor, 2008-08-28 at 22:09 +0200, Henrik Nordstrom wrote:
I have always wanted to see them dead for many reasons.
* On reasonable systems (i.e. most non-dos) the stack is a very suitable
place for temporary allocations.
* Using a static local array as return value will bite you from time
ons 2008-08-27 klockan 11:10 +0200 skrev Kinkie:
I'm unconvinced. What's the purpose of a MemoryRegion if you can't
manipulate it?
You can manipulate it, but at a rather low level. You can also append to
it by adding a MemoryRegion (or String), you can access raw octets. But
you can't do any
On ons, 2008-08-27 at 11:43 -0600, Alex Rousskov wrote:
My vote is not to vote until it is clear what you are implementing.
The original discussion was about meaningful strings and opaque buffers
they will use. Then you said that you are not implementing a string,
just a buffer. Right now,
On tor, 2008-08-28 at 01:25 +1200, Amos Jeffries wrote:
I read your definitions of memory buffer (simple case) and string
(complex case) as being definition of two separate non-interchangeable
things:
In my world they are fully interchangeable when you (as the programmer)
know the data
On tis, 2008-08-26 at 23:11 +0800, Adrian Chadd wrote:
The first merge will probably be after I've completed some more HTTP
parser reworking and complete the IPv6 core. Oh, and after I bootstrap
subversion or setup a bzr-CVS repository mirror for guido to continue
windows-related development.
On tis, 2008-08-26 at 09:27 +0200, Kinkie wrote:
Does your internal buffer point back to each String using it? (QQQ)
Shouldn't be needed, so I plan not to.
It's very useful for debugging and analysis however... but a complete
waste for production.
Regards
Henrik
signature.asc
On tis, 2008-08-26 at 09:35 +0200, Kinkie wrote:
In my opinion there is not that much of a difference between Strings
and Buffers, and the latter could use the services of the former to
delegate the issues of memory management, while concentrating on
different aspects - joining, chaining,
On tis, 2008-08-26 at 16:24 -0600, Alex Rousskov wrote:
... but it should be passed by reference whenever possible to avoid the
hidden overheads of refcounting. Every time we pass or return a String
by value, we create and later destroy at least one copy of a String,
which implies locking and
On ons, 2008-08-27 at 00:24 +0200, Kinkie wrote:
This is quite different from my current approach, by which Strings get
created and drive the instantiations of Bufs (MemoryRegions).
I feel that you'd be trying to reimplement parts of the memory
manager. Maximum efficiency, at the expense of
On tis, 2008-08-26 at 16:24 -0600, Alex Rousskov wrote:
The class names feel familiar. Have you posted some related code long
time ago? Was it during Squid3 birth?
No idea if I used the same class names before. I just picked what felt
natural to me based on the conceptual thoughts on what they
Sorry, MemoryChunk should read MemoryBlob to be consistent with my first
post...
On ons, 2008-08-27 at 01:33 +0200, Henrik Nordstrom wrote:
On ons, 2008-08-27 at 00:24 +0200, Kinkie wrote:
This is quite different from my current approach, by which Strings get
created and drive
On fre, 2008-08-22 at 23:01 +0800, Adrian Chadd wrote:
Generally, Squid's event driven methodology involves registering
callbacks and using cbdata pointers, then relying on pointer
validation/invalidation to determine whether to run a callback or not.
This means that an owner or anyone else
On fre, 2008-08-22 at 23:16 +0800, Adrian Chadd wrote:
I've been staring at the comm code in Squid-2 to get an idea of how to
implement properly async copy-free network IO. Its tricky to say the
least.
I have said it many times over the years in different words: Any data
should be stored in
On mån, 2008-08-25 at 18:51 +0800, Adrian Chadd wrote:
Henrik, what could possibly cause a httpState leak in the store update code?
No idea. There should be no httpState reference or dependency from store
update.
But it may have effects on the quick abort logics, as the store update
adds a new
On ons, 2008-08-20 at 19:03 +1200, Amos Jeffries wrote:
Here is a quick patch I think builds a clean squid.conf file for use
The gist of it is to build to shift the existing squid.conf.default to
the name squid.conf.documented (anyone got a better one?)
Or integrade the cfgman script and
On ons, 2008-08-20 at 20:18 +1200, Amos Jeffries wrote:
It also force-installs squid.conf.documented and squid.conf.default.
Assuming that squid.conf is the one user edits to be left untouched.
.default is supposed to be force installed already. If it's not thats a
bug.
Regards
Henrik
On ons, 2008-08-20 at 11:43 -0600, Alex Rousskov wrote:
Please note that this will cause headaches for distributions that
include squid.conf patches (e.g., CentOS Squid RPMs).
Do they really patch squid.conf and not cf.data.pre? The RPMs I have
seen patches cf.data.pre as they should.
Regards
mån 2008-08-18 klockan 09:05 -0600 skrev Russell Suter:
Hello,
Let me try this again. Due to some kind of foul-up on my outbound
email, it appears I've missed the sign-up window for this.
Not your fault. Somehow none of the moderators confirmed your
subscription after the introduction on
On lör, 2008-08-16 at 02:16 +1200, Amos Jeffries wrote:
Henrik, Guido;
The parser in 3.0 is just different enough from 3.1 for me to be
unsure whether the behavior would be the same. Does this need to be
back-ported?
The 3.0 parser looks fine to me. It was just HEAD that was twisted..
On lör, 2008-08-16 at 02:12 +1200, Amos Jeffries wrote:
There are now two approaches possible for this:
a) unzipping the shar into gif and storing those in bzr directly.
+1
Something we should do for Squid-2 as well.. (as previously discussed)
- This is the preferred method. Are there
On fre, 2008-08-15 at 18:11 +0800, Adrian Chadd wrote:
SVN allows me to port over the restructuring I've been doing in
Cacheboy and work with Guido to start sorting out more sensible Win32
support without having to mess about.
I'm fed up waiting. I'd like to do this now.
So start preparing
On fre, 2008-08-15 at 15:44 +0100, Micheal Wilkinson wrote:
I have a very limited programming knowledge so would appreciate some
help here.
I am wanting to alter squid so that instead of querying local databases
it queries a script which draws the information from a centralised server.
On ons, 2008-08-13 at 16:25 +1200, Amos Jeffries wrote:
Or maybe = should be equivalent to SetEmpty() in which case the
original code works, and can be simplified a lot... not sure how
IPAddress is supposed to work here, if it at all is supposed to be used
in = ..
It should be capable
301 - 400 of 2137 matches
Mail list logo
