diff options
| -rw-r--r-- | TODO | 2 | ||||
| -rw-r--r-- | doc/FAQ | 7 | ||||
| -rw-r--r-- | g10/ChangeLog | 5 | ||||
| -rw-r--r-- | g10/keyedit.c | 5 | ||||
| -rw-r--r-- | g10/trustdb.c | 10 |
5 files changed, 27 insertions, 2 deletions
@@ -1,4 +1,6 @@ + * g10/trustdb.c (make_sig_records): fix the fixme. + * at least an option to prefer DSA keys over RSA when selecting the key to use. Depending on creation time would be nice too. I think this is already done for the subkeys - check it. @@ -398,4 +398,11 @@ revoke all the subkeys installed on that machine and install new subkeys. + Q: In the edit meu the trust values is not displayed correctly after + signing uids - why? + A: This happens because the some informations are stored immediately + in the trustdb, but the actual trust calculation can be done after + the save command. This is a not easy to fix design bug which will be + addressed in GnuPG 1.1 + diff --git a/g10/ChangeLog b/g10/ChangeLog index 01c6fb181..0d783d072 100644 --- a/g10/ChangeLog +++ b/g10/ChangeLog @@ -1,3 +1,8 @@ +Mon May 1 17:08:14 CEST 2000 Werner Koch <[email protected]> + + * trustdb.c (propagate_validity): Fixed the bug that only one uid + gets fully trusted even when all are signed by an ultimate key. + Mon May 1 15:38:04 CEST 2000 Werner Koch <[email protected]> * getkey.c (key_byname): Always returned a defined context. Fixed diff --git a/g10/keyedit.c b/g10/keyedit.c index 17764b543..424b412a6 100644 --- a/g10/keyedit.c +++ b/g10/keyedit.c @@ -783,6 +783,11 @@ keyedit_menu( const char *username, STRLIST locusr, STRLIST commands, if( !sign_uids( keyblock, locusr, &modified, cmd == cmdLSIGN ) && sign_mode ) goto do_cmd_save; + /* Actually we should do a update_trust_record() here so that + * the trust gets displayed correctly. however this is not possible + * because we would have to save the keyblock first - something + * we don't want to do without an explicit save command. + */ break; case cmdDEBUG: diff --git a/g10/trustdb.c b/g10/trustdb.c index 1520d9d6c..cc769d1f8 100644 --- a/g10/trustdb.c +++ b/g10/trustdb.c @@ -1243,6 +1243,12 @@ make_sig_records( KBNODE keyblock, KBNODE uidnode, siglid = find_or_create_lid( sig ); /* smash dups */ + /* FIXME: Here we have a problem: + * We can't distinguish between a certification and a certification + * revocation without looking at class of the signature - we have + * to see how we can store the sigclass in the sigrecord.. + * Argg- I hope I can get rid of this ugly trustdb ASAP. + */ for( s2 = s; s2 ; s2 = s2->next ) { for(i=0; i < sigidx; i++ ) { if( s2->r.sig.sig[i].lid == siglid ) @@ -2014,10 +2020,10 @@ propagate_validity( TN root, TN node, int (*add_fnc)(ulong), unsigned *retflgs ) } /* loop over all user ids */ - for( ur=node->list; ur && max_validity < TRUST_FULLY; ur = ur->next ) { + for( ur=node->list; ur && max_validity <= TRUST_FULLY; ur = ur->next ) { assert( ur->is_uid ); /* loop over all signators */ - for(kr=ur->list; kr && max_validity < TRUST_FULLY; kr = kr->next ) { + for(kr=ur->list; kr && max_validity <= TRUST_FULLY; kr = kr->next ) { if( propagate_validity( root, kr, add_fnc, retflgs ) ) return -1; /* quit */ if( kr->n.k.validity == TRUST_ULTIMATE ) { |