[imp] Replying to pgp-encrypted messages: no body

Michael M Slusarz slusarz at bigworm.colorado.edu
Sun Apr 25 04:24:16 PDT 2004


Quoting Oliver Kuhl <okuhl at netcologne.de>:

> Michael M Slusarz wrote:
>
>> The only location where we have the decrypted MIME data from the PGP message
>> that has been parsed into its correct pieces is in the 
>> MIME_Contents:: object
>> located in _outputPGPEncrypted() in imp/lib/MIME/Viewer/pgp.php.  It takes a
>> whole heck of a lot of effort to get to that point.
>>
>> That function call is so far removed from any kind of composition 
>> code (if the
>> text is displayed on the message page, not only is it so far removed with
>> respect to how many function calls it takes just to get to that 
>> point, but it
>> is  located in an entirely different page load from where we determine the
>> reply text (!)) that it would take some major hard-coding of this 
>> single case
>> in the code to be able to resurrect it when determining the reply text.
>>
>> One potential future solution is to move the code in the IMP pgp MIME_Viewer
>> that creates the MIME_Message structure out of the encrypted PGP 
>> data to a new
>> location (i.e. Horde_Crypt_pgp::) so that the reply text determination code
>> could recreate the unencrypted data without seriously distorting the 
>> space/time
>> continuum.
> Maybe another way would be (I really don't know if it is possible, just
> a simple idea) to get the decrypted data from the cache. So maybe
> implementing the possibility to reuse already dectryted mime parts like
> diskussed in thread "pgp-encrypted emails blowing up session" would help
> here, too.

True, the decrypted email parts could easily live as cached objects.  The
problem still remains how to identify these objects when trying to reuse them
later.  Once again, the issue is that there is no way of passing this
information back through the MIME rendering system to the code that is 
creating
the reply links.  Without identifying the cached object at this point, 
there is
no way the reply page, in a completely separate session access, can access the
cached information since there is no way it will be able to find the cached
information in the cache store (since, as a MIME_Contents object, the pgp
cached object can *not* be accessed via a mailbox/index lookup since
MIME_Contents:: objects have no knowledge of this information.)

so this is unfortunately a completely separate issue from reusing 
cached objects
as talked about in the above mentioned thread.

michael

______________________________________________
Michael Slusarz [slusarz at bigworm.colorado.edu]
The University of Colorado at Boulder


More information about the imp mailing list