On Thursday, 03 August 2006 at 11:16, Rocco Rutte wrote: > Hi, > > * Rocco Rutte [06-07-13 12:59:12 +0000] wrote: > >Now I had time to track it down: the server sends different responses: > >only IMAP4rev1 and AUTH=... before authentication but much more items > >after authentication. > > Here is a debug snippet: > > imap_cmd_step: grew buffer to 512 bytes > < * OK [CAPABILITY IMAP4REV1 LITERAL+ SASL-IR LOGIN-REFERRALS AUTH=PLAIN > AUTH=LOGIN] mail IMAP4rev1 2004.357 at Thu, 3 Aug 2006 13:05:17 +0200 > (MEST) > Handling CAPABILITY > imap_authenticate: Using any available method. > local ip: 192.168.0.113;53468, remote ip:XXX;993 > External SSF: 256 > External authentication name: pdmef > mutt_sasl_cb_authname: getting authname for XXX:993 > mutt_sasl_cb_authname: getting user for XXX:993 > mutt_sasl_cb_pass: getting password for pdmef@XXX:993 > > a0000 AUTHENTICATE PLAIN XXX > < a0000 OK [CAPABILITY IMAP4REV1 LITERAL+ IDLE NAMESPACE > MAILBOX-REFERRALS BINARY UNSELECT SCAN SORT THREAD=REFERENCES > THREAD=ORDEREDSUBJECT MULTIAPPEND] User pdmef authenticated > SASL protection strength: 0 > SASL protection buffer size: 65536 > Communication encrypted at 256 bits > > a0001 LIST "" "" > < * LIST (\NoSelect) "/" "" > Root delimiter: / > < a0001 OK LIST completed > > I'm using SASL and as there's no second 'Handling CAPABILITY' in there, > I guess the full CAPABILITY line is eaten by SASL and not seen by mutt. Mutt just wasn't checking for capabilities on tagged responses. Does the following patch work for you? > Any objections against the previous patch? If the one I posted works, I'd be happier with it. Asking for CAPABILITY just after receiving a gratuitous CAPABILITY is kind of silly. > >The second patch adds more verbose debugging to the socket layer: print > >which connection gets what file descriptor and use that when logging > >socket traffic to distinct between different hosts with identical > >protocols (it's really hard to debug otherwise). > > Any objections against this one? For me it's highly usefull. Lately > someone posted info for IMAP browser bug with traffic to different > servers within a session and this patch improves readability. The second patch is certainly worthwhile.
Index: imap/command.c
===================================================================
RCS file: /home/roessler/cvs/mutt/imap/command.c,v
retrieving revision 3.40
diff -u -p -r3.40 command.c
--- imap/command.c 6 Jul 2006 18:18:31 -0000 3.40
+++ imap/command.c 5 Aug 2006 00:38:13 -0000
@@ -119,6 +119,7 @@ int imap_cmd_step (IMAP_DATA* idata)
size_t len = 0;
int c;
int rc;
+ char* s;
if (idata->status == IMAP_FATAL)
{
@@ -177,6 +178,9 @@ int imap_cmd_step (IMAP_DATA* idata)
/* tagged completion code. TODO: I believe commands will always be completed
* in order, but will have to double-check when I have net again */
+ s = imap_next_word (idata->buf);
+ if (!ascii_strncmp (s, "OK [CAPABILITY", 14))
+ cmd_parse_capability (idata, imap_next_word (s));
rc = IMAP_CMD_CONTINUE;
if (!ascii_strncmp (idata->buf, idata->cmds[idata->lastcmd].seq, SEQLEN))
{
Attachment:
pgpCB7PV2WYDO.pgp
Description: PGP signature