citadel.git
16 years ago* migrated to new hash create signature
Wilfried Göesgens [Sat, 26 Apr 2008 09:01:37 +0000 (09:01 +0000)]
* migrated to new hash create signature
* bumped min libcitadel version to 1.10
* first starts for hashing url-handlers

16 years ago* bugfix in the hash implementation; expanding the hash wasn't done properly
Wilfried Göesgens [Sat, 26 Apr 2008 09:00:15 +0000 (09:00 +0000)]
* bugfix in the hash implementation; expanding the hash wasn't done properly
* added possibility to have list with duplicate primary keys
* added ability to specify custom "hashing" algorithm
* bumped version number to 1.10 since the newhash function has a different signature now.

16 years ago* fix URL termination bug
Wilfried Göesgens [Fri, 25 Apr 2008 20:59:26 +0000 (20:59 +0000)]
* fix URL termination bug

16 years agoBegan work on vNote based display for sticky notes
Art Cancro [Thu, 24 Apr 2008 21:01:28 +0000 (21:01 +0000)]
Began work on vNote based display for sticky notes

16 years agoFix '@AUTOCONF@ command not found' build problem
Art Cancro [Thu, 24 Apr 2008 19:43:31 +0000 (19:43 +0000)]
Fix '@AUTOCONF@ command not found' build problem

16 years ago* busting the prototype package isn't that easy as its swapping places.
Wilfried Göesgens [Wed, 23 Apr 2008 06:25:21 +0000 (06:25 +0000)]
* busting the prototype package isn't that easy as its swapping places.

16 years ago* added brazillian portuguese update
Wilfried Göesgens [Wed, 23 Apr 2008 06:24:10 +0000 (06:24 +0000)]
* added brazillian portuguese update

16 years ago* THIS IS 7.35
Wilfried Göesgens [Tue, 22 Apr 2008 17:18:18 +0000 (17:18 +0000)]
* THIS IS 7.35

16 years ago* THIS IS 7.35
Wilfried Göesgens [Tue, 22 Apr 2008 17:16:01 +0000 (17:16 +0000)]
* THIS IS 7.35

16 years ago* THIS IS 7.35 / 1.09
Wilfried Göesgens [Tue, 22 Apr 2008 17:13:16 +0000 (17:13 +0000)]
* THIS IS 7.35 / 1.09

16 years ago* move utility binaries to /usr/lib/citadel-server/ so we don't get name clashes...
Wilfried Göesgens [Tue, 22 Apr 2008 11:20:57 +0000 (11:20 +0000)]
* move utility binaries to /usr/lib/citadel-server/ so we don't get name clashes with other software.

16 years ago* added Portoguese Brazillian translation done by Marco Goncalves
Wilfried Göesgens [Mon, 21 Apr 2008 21:42:12 +0000 (21:42 +0000)]
* added Portoguese Brazillian translation done by Marco Goncalves

16 years ago* more carefully render urls
Wilfried Göesgens [Mon, 21 Apr 2008 19:23:27 +0000 (19:23 +0000)]
* more carefully render urls

16 years agoMinor formatting cleanup
Art Cancro [Mon, 21 Apr 2008 15:39:38 +0000 (15:39 +0000)]
Minor formatting cleanup

16 years ago* regulary shut down libcitadel, so we don't leak the icondir hash on shutdown.
Wilfried Göesgens [Sun, 20 Apr 2008 20:55:17 +0000 (20:55 +0000)]
* regulary shut down libcitadel, so we don't leak the icondir hash on shutdown.

16 years ago* init vars before using them.
Wilfried Göesgens [Sun, 20 Apr 2008 20:29:15 +0000 (20:29 +0000)]
* init vars before using them.

16 years ago* fixed several compiler warnings.
Wilfried Göesgens [Sun, 20 Apr 2008 19:23:25 +0000 (19:23 +0000)]
* fixed several compiler warnings.

16 years ago* removed possibly dangerous code
Wilfried Göesgens [Sun, 20 Apr 2008 11:44:04 +0000 (11:44 +0000)]
* removed possibly dangerous code

16 years ago* implemented realloc
Wilfried Göesgens [Sun, 20 Apr 2008 11:38:05 +0000 (11:38 +0000)]
* implemented realloc
* just encode emailaddresses that actually contain umlauts

16 years agovnote.c is now (for the most part) complete.
Art Cancro [Sun, 20 Apr 2008 03:41:45 +0000 (03:41 +0000)]
vnote.c is now (for the most part) complete.
However, I will not be touching WebCit until after the next release.

16 years agomore vnote work
Art Cancro [Sun, 20 Apr 2008 03:09:19 +0000 (03:09 +0000)]
more vnote work

16 years agovnote_serialize() is now (probably) complete.
Art Cancro [Sat, 19 Apr 2008 04:28:45 +0000 (04:28 +0000)]
vnote_serialize() is now (probably) complete.

16 years agopropset on vnote.c
Art Cancro [Sat, 19 Apr 2008 04:04:10 +0000 (04:04 +0000)]
propset on vnote.c

16 years agovnote serialization
Art Cancro [Fri, 18 Apr 2008 15:28:43 +0000 (15:28 +0000)]
vnote serialization

16 years agoIn the text client, <Q>uote is now reply<Q>uoted
Art Cancro [Fri, 18 Apr 2008 03:02:12 +0000 (03:02 +0000)]
In the text client, <Q>uote is now reply<Q>uoted

16 years agomore vnote hacking
Art Cancro [Fri, 18 Apr 2008 02:36:34 +0000 (02:36 +0000)]
more vnote hacking

16 years ago* QP encode Realnames if they are non-ascii
Wilfried Göesgens [Thu, 17 Apr 2008 21:52:46 +0000 (21:52 +0000)]
* QP encode Realnames if they are non-ascii

16 years ago(no commit message)
Wilfried Göesgens [Thu, 17 Apr 2008 21:50:49 +0000 (21:50 +0000)]

16 years agomore vnote skeleton code
Art Cancro [Thu, 17 Apr 2008 21:00:21 +0000 (21:00 +0000)]
more vnote skeleton code

16 years agoAdded a skeleton 'vnote.c' module to libcitadel
Art Cancro [Thu, 17 Apr 2008 17:04:42 +0000 (17:04 +0000)]
Added a skeleton 'vnote.c' module to libcitadel

16 years agoWhen assembling lists of thread references, if the list
Art Cancro [Thu, 17 Apr 2008 14:32:29 +0000 (14:32 +0000)]
When assembling lists of thread references, if the list
exceeds 10 messages or 900 bytes, eliminate the second reference
in the list.  See http://cr.yp.to/immhf/thread.html for more info.

16 years agoDoh! Thread references read left-to-right, not right-to-left.
Art Cancro [Wed, 16 Apr 2008 15:34:52 +0000 (15:34 +0000)]
Doh!  Thread references read left-to-right, not right-to-left.

16 years ago* just modify /etc/services in batch mode if we are told so by the environment variab...
Wilfried Göesgens [Mon, 14 Apr 2008 21:06:07 +0000 (21:06 +0000)]
* just modify /etc/services in batch mode if we are told so by the environment variable ALTER_ETC_SERVICES
* make our debs do this.

16 years ago* don't deliver stress in the debs.
Wilfried Göesgens [Mon, 14 Apr 2008 21:01:18 +0000 (21:01 +0000)]
* don't deliver stress in the debs.

16 years ago* commented out debug url parsing
Wilfried Göesgens [Mon, 14 Apr 2008 20:44:39 +0000 (20:44 +0000)]
* commented out debug url parsing

16 years ago* use the debian prototypejs package
Wilfried Göesgens [Mon, 14 Apr 2008 14:57:49 +0000 (14:57 +0000)]
* use the debian prototypejs package

16 years ago* remove stress binary from the install target, since its just a developer tool.
Wilfried Göesgens [Mon, 14 Apr 2008 09:41:38 +0000 (09:41 +0000)]
* remove stress binary from the install target, since its just a developer tool.

16 years agoVersion number to 7.35 in preparation for an upcoming release.
Art Cancro [Fri, 11 Apr 2008 14:26:13 +0000 (14:26 +0000)]
Version number to 7.35 in preparation for an upcoming release.

16 years agoAdded the words 'SMTP client:' to some of the log messages
Art Cancro [Thu, 10 Apr 2008 18:48:21 +0000 (18:48 +0000)]
Added the words 'SMTP client:' to some of the log messages
in serv_smtp.c so I can grep for them.

16 years agoRemoved the 'Gateway Domain' function from the entire Citadel system. This function...
Art Cancro [Mon, 7 Apr 2008 04:01:04 +0000 (04:01 +0000)]
Removed the 'Gateway Domain' function from the entire Citadel system.  This function has long since been superseded by the Global Address Book.  Absolutely no one is using it.  In fact, as far as I can tell, its implementation does not even exist anymore -- I couldn't find the back end code to remove.  So instead of confusing site administrators with a superfluous configuration option, we are removing it.

16 years ago A minor fix to some cosmetics in ARTV.
Dave West [Sun, 6 Apr 2008 21:15:40 +0000 (21:15 +0000)]
 A minor fix to some cosmetics in ARTV.
This does NOT, repeat NOT affect the operation of ARTV at all.

16 years agoGive the user contained within masterCC a name.
Dave West [Sun, 6 Apr 2008 21:14:34 +0000 (21:14 +0000)]
Give the user contained within masterCC a name.
Its not a real user but it is stored in the DB and does contain info
like how many messages the system has created.
We call it Citadel.
This make "citadel" an illegal user name.
Now this user has a name we need to make sure the auto purger doesn't
purge it and make sure no one can login by that name.

16 years agoAdded some CtdlThreadCheckStop() calls to exit rbl checks and stop the
Dave West [Sat, 5 Apr 2008 21:40:46 +0000 (21:40 +0000)]
Added some CtdlThreadCheckStop() calls to exit rbl checks and stop the
SMTP negotiation if we are shutting down.
Also added a bit of code to log the time it takes to do an RBL check. If
it takes a long time maybe we can cache the results in a hash with
something like a 30 second life span.

16 years agoQuieten the user 0 message to the Aide room.
Dave West [Sat, 5 Apr 2008 19:51:53 +0000 (19:51 +0000)]
Quieten the user 0 message to the Aide room.
User 0 is created as a side effect of the system posting messages from
the masterCC. This side effect has the benefit of recording the number
of messages the system has generated.
Should we make use of this somewhere?
The actual code that creates the user 0 is in CtdlSubmitMsg() and reads
as
        /* Bump this user's messages posted counter. */
        CtdlLogPrintf(CTDL_DEBUG, "Updating user\n");
        lgetuser(&CCC->user, CCC->curr_user);
        CCC->user.posted = CCC->user.posted + 1;
        lputuser(&CCC->user);

Its around msgbase.c:2624

16 years ago* fixed some hickups in the urlencode function.
Wilfried Göesgens [Sat, 5 Apr 2008 18:23:11 +0000 (18:23 +0000)]
* fixed some hickups in the urlencode function.

16 years agoReduced the output of the SMTP HELP command.
Art Cancro [Fri, 4 Apr 2008 16:51:23 +0000 (16:51 +0000)]
Reduced the output of the SMTP HELP command.
Now it just points to http://www.ietf.org/rfc/rfc2821.txt

16 years ago* we should send the buffer we printed the command into to the server.
Wilfried Göesgens [Fri, 4 Apr 2008 14:02:28 +0000 (14:02 +0000)]
* we should send the buffer we printed the command into to the server.

16 years ago* we have now several bstrs:
Wilfried Göesgens [Fri, 4 Apr 2008 10:44:32 +0000 (10:44 +0000)]
* we have now several bstrs:
  * bstr: simply gives the string
  * xbstr: gives string and its length
  * lbstr: gives the long value of the item
  * ibstr: gives the int value of the item
  * havebstr: whether this bstr and if it wasn't set empty
  * yesbstr: returns true if the value equals "yes"
there also is a CAPITAL version of each, for dynamicaly calculated keys.
* replaced all atol(bstr()), strcmp(bstr(), "yes") IsEmptyStr(bstr()) by the above.

16 years ago* as its rather noisy, the dezonify logging is #ifdef DBG_ICAL now.
Wilfried Göesgens [Fri, 4 Apr 2008 10:39:25 +0000 (10:39 +0000)]
* as its rather noisy, the dezonify logging is #ifdef DBG_ICAL now.

16 years agoFixed a memory allocation problem
Art Cancro [Fri, 4 Apr 2008 03:52:55 +0000 (03:52 +0000)]
Fixed a memory allocation problem

16 years ago* Restore the ability of addurls() to delimit on '?' as well as '&'
Art Cancro [Fri, 4 Apr 2008 02:37:42 +0000 (02:37 +0000)]
* Restore the ability of addurls() to delimit on '?' as well as '&'
* Handle thread references when posting message replies.

16 years agowebcit - generate Reply/ReplyAll/ReplyQuoted URL's that contain thread references
Art Cancro [Thu, 3 Apr 2008 21:54:42 +0000 (21:54 +0000)]
webcit - generate Reply/ReplyAll/ReplyQuoted URL's that contain thread references

16 years agoKeep track of thread references, in the server and in the text client. This required...
Art Cancro [Thu, 3 Apr 2008 20:37:33 +0000 (20:37 +0000)]
Keep track of thread references, in the server and in the text client.  This required amending the ENT0 protocol command.

16 years ago* due to demand of debian, we rename the webcit binary from webserver to webcit.
Wilfried Göesgens [Thu, 3 Apr 2008 19:43:00 +0000 (19:43 +0000)]
* due to demand of debian, we rename the webcit binary from webserver to webcit.
* for migration make install wil create a symlink.

16 years ago* xbstr now will give the length of the value to its key, so working with it is more...
Wilfried Göesgens [Thu, 3 Apr 2008 18:51:32 +0000 (18:51 +0000)]
* xbstr now will give the length of the value to its key, so working with it is more easy.

16 years ago* moved url argument list to hash.
Wilfried Göesgens [Wed, 2 Apr 2008 23:45:08 +0000 (23:45 +0000)]
* moved url argument list to hash.
* carry length of parameter value in the hash for later use
* added Bstr, BSTR and bstr. bstr mustn't be used for dynamic keys, use BSTR instead.
* use BSTR where apropriate.
* fixed crash condition with message posting Mrs. Foobar found out

16 years ago* add convenience macro for constant hashkeys, as the compiler can do sizeof instead...
Wilfried Göesgens [Wed, 2 Apr 2008 23:41:47 +0000 (23:41 +0000)]
* add convenience macro for constant hashkeys, as the compiler can do sizeof instead of the program doing strlen.

16 years ago* rather trunkate data then letting escape chars through.
Wilfried Göesgens [Tue, 1 Apr 2008 22:35:35 +0000 (22:35 +0000)]
* rather trunkate data then letting escape chars through.

16 years ago* rewrote urlstring tokenizer
Wilfried Göesgens [Tue, 1 Apr 2008 22:02:58 +0000 (22:02 +0000)]
* rewrote urlstring tokenizer
* fixed crash condition in de-escaper

16 years ago* removed nonfree component.
Wilfried Göesgens [Mon, 31 Mar 2008 21:48:52 +0000 (21:48 +0000)]
* removed nonfree component.

16 years agoICAL 'putics' should NOT do server-generated invitations.
Art Cancro [Mon, 31 Mar 2008 21:43:41 +0000 (21:43 +0000)]
ICAL 'putics' should NOT do server-generated invitations.

16 years ago* we need to find ?Q? or ?B? before searching for the end. else we might catch the...
Wilfried Göesgens [Mon, 31 Mar 2008 21:28:54 +0000 (21:28 +0000)]
* we need to find ?Q? or ?B? before searching for the end. else we might catch the wrong location and struggle.

16 years agoRefrain from transmitting untagged IMAP responses when the
Art Cancro [Mon, 31 Mar 2008 20:59:09 +0000 (20:59 +0000)]
Refrain from transmitting untagged IMAP responses when the
command in progress is FETCH, STORE, or SEARCH.  RFC3501 requires this.
Fixes bug #355.

16 years ago* lengthen the subject buffer so it can take utf8 glyphs
Wilfried Göesgens [Sun, 30 Mar 2008 23:33:28 +0000 (23:33 +0000)]
* lengthen the subject buffer so it can take utf8 glyphs

16 years agoENABLE_NLS must wrap HAVE_USELOCALE, else configure will make us produce compile...
Wilfried Göesgens [Sun, 30 Mar 2008 23:18:01 +0000 (23:18 +0000)]
ENABLE_NLS must wrap HAVE_USELOCALE, else configure will make us produce compile errors in certain situations.

16 years agoFix the artv export bug that crept in while I was fixing artv import.
Dave West [Sun, 30 Mar 2008 02:13:56 +0000 (02:13 +0000)]
Fix the artv export bug that crept in while I was fixing artv import.

16 years agoNo code changes, just cleaned some warnings and some comments.
Dave West [Sat, 29 Mar 2008 14:19:05 +0000 (14:19 +0000)]
No code changes, just cleaned some warnings and some comments.

16 years agoFix up ARTV import so that imports actually work.
Dave West [Sat, 29 Mar 2008 01:15:32 +0000 (01:15 +0000)]
Fix up ARTV import so that imports actually work.
Fixed various client_getln loops so that they don't go endless if the
client disappears.
ARTV import now makes use of the internal base64 decoder so it should be
quicker.
ARTV import now sends lots of dots to the client. This is to kick the
watchdog timer in sendcommand so that it doesn't expire while we process
a big message.
Changes in threads.* are in preperation for cancelling threads and using
signals to interrupt threads. They are #ifdef excluded.

16 years agoImporting REALLY BIG messages can also cause the watchdog in sendcommand
Dave West [Fri, 28 Mar 2008 15:09:19 +0000 (15:09 +0000)]
Importing REALLY BIG messages can also cause the watchdog in sendcommand
to expire. So now we walk the dots for each line of a message as well.

16 years agoFix the artv import.
Dave West [Fri, 28 Mar 2008 13:03:37 +0000 (13:03 +0000)]
Fix the artv import.
It couldn't handle the blank lines between sections and now it does.
Also added code to send dots during the room import to keep the user
informed that something is happening if we are importing a BIG room.
This has the added advantage of kicking the watchdog in sendcommand so
it doesn't time out unexpectedly.
Also added code to exit the artv early if we are shutting down.

16 years agoOn my slow system a watch dog time out of 5 seconds is not always
Dave West [Fri, 28 Mar 2008 12:59:05 +0000 (12:59 +0000)]
On my slow system a watch dog time out of 5 seconds is not always
enough.
I added a command line arg to increase it if needed.
Watchdog defaults to 5 seconds but you can use -w to extend it if you
need to.

16 years agoFixed a potential infinite-loop condition in previous commit
Art Cancro [Thu, 27 Mar 2008 02:55:57 +0000 (02:55 +0000)]
Fixed a potential infinite-loop condition in previous commit

16 years agoHandle RFC822 'References:' field and Citadel W (Wefewences) field
Art Cancro [Wed, 26 Mar 2008 21:15:32 +0000 (21:15 +0000)]
Handle RFC822 'References:' field and Citadel W (Wefewences) field

16 years agoThread log was too verbose when writing to disk
Art Cancro [Wed, 26 Mar 2008 04:30:19 +0000 (04:30 +0000)]
Thread log was too verbose when writing to disk

16 years ago* use the utf8 functions to cut subjects at the border of glyphs
Wilfried Göesgens [Wed, 26 Mar 2008 00:04:22 +0000 (00:04 +0000)]
* use the utf8 functions to cut subjects at the border of glyphs

16 years ago* add functions to evaluate utf8 strings as described at the instructables
Wilfried Göesgens [Wed, 26 Mar 2008 00:03:12 +0000 (00:03 +0000)]
* add functions to evaluate utf8 strings as described at the instructables

16 years agosmall update to hack.txt
Art Cancro [Tue, 25 Mar 2008 21:53:48 +0000 (21:53 +0000)]
small update to hack.txt

16 years agoBump the citadel rev number to 733. It got missed in the last release.
Dave West [Tue, 25 Mar 2008 19:35:14 +0000 (19:35 +0000)]
Bump the citadel rev number to 733. It got missed in the last release.
Bump the EXPORT_REV_MIN to 733 since I added the extra POP3 config
options.

16 years agoSending an instant message via Jabber now resets the idle timer.
Art Cancro [Mon, 24 Mar 2008 02:22:14 +0000 (02:22 +0000)]
Sending an instant message via Jabber now resets the idle timer.

16 years ago* added aditional licenses of the non-citadel created parts of libcitadel.
Wilfried Göesgens [Sun, 23 Mar 2008 20:42:23 +0000 (20:42 +0000)]
* added aditional licenses of the non-citadel created parts of libcitadel.

16 years agoDid away with lprintf all together now its called CtdlLogPrintf()
Dave West [Fri, 21 Mar 2008 21:42:54 +0000 (21:42 +0000)]
Did away with lprintf all together now its called CtdlLogPrintf()
This is a step towards the coding convention listed on the web site.

16 years agosvn_revision() now declared in svn_revision.h
Art Cancro [Fri, 21 Mar 2008 20:09:41 +0000 (20:09 +0000)]
svn_revision() now declared in svn_revision.h

16 years agoBootstrap now calls mk_svn_revision.sh for the package builders.
Dave West [Thu, 20 Mar 2008 20:48:13 +0000 (20:48 +0000)]
Bootstrap now calls mk_svn_revision.sh for the package builders.

16 years agoNow the text client and webcit show the servers svn revision too.
Dave West [Thu, 20 Mar 2008 19:22:26 +0000 (19:22 +0000)]
Now the text client and webcit show the servers svn revision too.

16 years agoNow the server records the SVN revision in the log at startup.
Dave West [Thu, 20 Mar 2008 18:32:07 +0000 (18:32 +0000)]
Now the server records the SVN revision in the log at startup.

16 years agoA test fix for the thread cancellation routine.
Dave West [Thu, 20 Mar 2008 15:39:58 +0000 (15:39 +0000)]
A test fix for the thread cancellation routine.
We only log situations where we would have cancelled a thread, we don't
actually cancel them.
With this code we should not see cancellation requests during normal
running.
Cancelation requests during normal running were due to threads being
asked to stop when load decreased but the idle thread detection aloud a
thread to be asked to stop just as select gave it a task and made it not
idle.
Now if select gives a thread a job to do and the thread was asked to
stop during the select the thread changes its state back to running and
refuses the stop request. This only happens if the system is NOT
shutting down.

16 years agoPrevent cancellation of threads but log that we would have cancelled it.
Dave West [Thu, 20 Mar 2008 13:26:47 +0000 (13:26 +0000)]
Prevent cancellation of threads but log that we would have cancelled it.
Its causing problems on Uncensored.

16 years ago* use snprintf to avoid stack corruption.
Wilfried Göesgens [Thu, 20 Mar 2008 09:57:42 +0000 (09:57 +0000)]
* use snprintf to avoid stack corruption.

16 years ago* init command buf empty to avoid fancy side effects
Wilfried Göesgens [Thu, 20 Mar 2008 09:08:38 +0000 (09:08 +0000)]
* init command buf empty to avoid fancy side effects

16 years ago* this way arround some compilers like the casts better.
Wilfried Göesgens [Wed, 19 Mar 2008 23:19:38 +0000 (23:19 +0000)]
* this way arround some compilers like the casts better.

16 years agooops ... minor fix to above
Art Cancro [Wed, 19 Mar 2008 19:41:58 +0000 (19:41 +0000)]
oops ... minor fix to above

16 years agoEVT_LOGOUT hooks are now called with the thread's current
Art Cancro [Wed, 19 Mar 2008 19:28:48 +0000 (19:28 +0000)]
EVT_LOGOUT hooks are now called with the thread's current
CitContext set to that of the session being logged out, even when
killing a session administratively.  This allows the identity of the
user to be passed along to module hooks.  This fixes the problem in
the Jabber server where buddy 'ghosts' whould remain visible to
clients if the user disconnected without logging off.

16 years agobmstrcasestr() now returns NULL if either of its supplied
Art Cancro [Wed, 19 Mar 2008 17:46:00 +0000 (17:46 +0000)]
bmstrcasestr() now returns NULL if either of its supplied
arguments is NULL ... instead of passing NULL to strlen() and making
the program crash.

16 years agoAuto purger thread was declaring itself as the indexer_thread() in the
Dave West [Wed, 19 Mar 2008 17:31:31 +0000 (17:31 +0000)]
Auto purger thread was declaring itself as the indexer_thread() in the
log

16 years agoWhen doing shutdown we allow each thread up to 5 seconds to exit on its
Dave West [Tue, 18 Mar 2008 23:28:55 +0000 (23:28 +0000)]
When doing shutdown we allow each thread up to 5 seconds to exit on its
own after asking it nicely to do so. After 5 seconds we will cancel()
the thread. Its possible that we may loose messages in this manner
since its similar to a SIGTERM. Because of this we write messages to the
log about this event. If we are seeing these entries in the log we need
to track down why the thread is strugling to exit and fix it.

16 years agoMoved remaining POP3 options onto the POP3 configuration tab
Dave West [Tue, 18 Mar 2008 21:58:56 +0000 (21:58 +0000)]
Moved remaining POP3 options onto the POP3 configuration tab

16 years agoAdded new POP3 client config options
Dave West [Tue, 18 Mar 2008 21:50:25 +0000 (21:50 +0000)]
Added new POP3 client config options

16 years agoAdded code to handle new POP3 client config options.
Dave West [Tue, 18 Mar 2008 21:46:56 +0000 (21:46 +0000)]
Added code to handle new POP3 client config options.

16 years agoOops, managed to commit a partial change regarding the POP3 client
Dave West [Tue, 18 Mar 2008 21:12:03 +0000 (21:12 +0000)]
Oops, managed to commit a partial change regarding the POP3 client
interval. Here is the rest of it.
Still to do are text client and documentation.

16 years agoAdded some more CtdlThreadCheckStop() calls to speed shutdown.
Dave West [Tue, 18 Mar 2008 21:05:27 +0000 (21:05 +0000)]
Added some more CtdlThreadCheckStop() calls to speed shutdown.