Re: [Evolution] Cannot see folders, evolution-mapi

[Patrik Magnusson <Patrik Magnusson comhem se>]

On Wed, Jan 04, 2012 at 10:56:27AM +0100, Milan Crha wrote:
> On Tue, 2012-01-03 at 22:11 +0100, Patrik Magnusson wrote:
> > But I could also see the folders being listed at one point - not with
> > names, but there was the correct number of them there.
> >
> > I would be happy to provide further details, but I don't think I could
> > post the whole log without quite a bit of redacting because I don't
> > know how much information is in the hexdumps.
>
> How do you know it's those folders if there are not names? 
As a somewhat more definite answer, I can compare the hexes with those
I get from 'openchangeclient -m' - which gives my a list of folders
with names and hexes.

> Try to search
> the debug output for GetHierarchyTable call, which is for the directory
> listing. It's followed by SetColumns call and then couple of
> QueryPosition and QueryRows calls. Is any of them failing?
The only one of these I find in the log is QueryRows (as part of
NspiQueryRows), the other ones are missing completely. In the reply to
this there are some rows that look slightly suspect but the result field
is MAPI_E_SUCCESS.

 lpProps: struct SPropValue
   ulPropTag                : PidTagAttachNumber_Error (0xE21000A)
   dwAlignPad : 0x00000000 (0)
   value : union SPropValue_CTR(case 10)
   err : MAPI_E_NOT_FOUND (0x8004010F)
 lpProps: struct SPropValue
   ulPropTag : UNKNOWN_ENUM_VALUE (0x6613000A)
   dwAlignPad : 0x00000000 (0)
   value : union SPropValue_CTR(case 10)
   err : MAPI_E_NOT_FOUND (0x8004010F)
 lpProps: struct SPropValue
   ulPropTag : UNKNOWN_ENUM_VALUE (0x800F101E)
   dwAlignPad : 0x00000000 (0)
   value : union SPropValue_CTR(case 4126)
   MVszA: struct StringArray_r
      cValues : 0x00000009 (9)
      lppszA : *
        lppszA: ARRAY(9)
         lppszA : *
           lppszA : 'x500:/o=CorpMail/ou=First Administrative Group/cn=Recipients/cn=Magnusson04'
         lppszA : *
           lppszA : 'x500:/o=<snip>/ou=First Administrative Group/cn=Recipients/cn=Magnusson04'
         lppszA : *
           lppszA : 'smtp:<snip>'
         lppszA : *
           lppszA : 'smtp:<snip>'
         lppszA : *
           lppszA : 'smtp:<snip>'
         lppszA : *
           lppszA : 'x400:C=SE;A= ;P=CorpMail;O=Exchange;S=Magnusson;G=Patrik;'
         lppszA : *
           lppszA : 'X400:C=SE;A= ;P=<snip>;O=Exchange;S=Magnusson;G=Patrik;'
         lppszA : *
           lppszA : 'smtp:<snip>'
         lppszA : *
           lppszA : 'SMTP:<snip>'
 result : MAPI_E_SUCCESS (0x0)

> Also, check ~/.cache/evolution/mail/<uid-of-your-mapi-account>/.summary
> file, which contains list of known folders. If that's mostly empty, then
> it didn't find any folder. Note the file is a binary file, not a text
> file.
The folder .cache/evolution/mail is empty.

> By the way, do you know the server version you connect to?
Specifically, no. But it definitely is the kind that insists on
using MAPI. I have checked that with a wireshark trace during an
outlook session.

In the log I find this:
rgwServerVersion: ARRAY(3)
  rgwServerVersion         : 0x0803 (2051)
  rgwServerVersion         : 0x80c0 (32960)
  rgwServerVersion : 0x0001 (1)

I have also tried the samba make-target in openchange (though I still
had to do a log manually because our firewall blocks git), and
this made no difference. 

(I also found that compiling openchange with
libmapi+ causes the compilation of evolution-data-server to fail
with 
make[2]: *** No rule to make target `e-gdbus-marshallers.c', needed by
`all'.  Stop.
I didn't run into that earlier because then I didn't have boost
installed.)

When I think about the steps I do to compile everything, I think the 
largest room for error is during configure of the evolution-packages. 
Could you possibly tell me a safe/conservative combination of options to
use in those steps?

Best Regards,

/Patrik Magnusson