4 ** Make parsing more robust
5 Currently we don't cope with overlong lines in the best way.
6 ** Check that we really release the ksba reader/writer objects.
9 ** Some code should go into import.c
10 ** When we allow concurrent service request in gpgsm, we
11 might want to have an agent context for each service request
12 (i.e. Assuan context).
15 ** Try to keep certificate references somewhere
16 This will help with some of our caching code. We also need to test
17 that caching; in particular "regtp_ca_chainlen".
20 ** replace leading zero in integer hack by a cleaner solution
23 ** mark all unimplemented commands and options.
24 ** Implement --default-key
25 ** support the anyPolicy semantic
29 ** Check file permissions
30 ** Check that all error code mapping is done.
31 ** Remove the inter-module dependencies between gpgsm and keybox
32 ** Add an source_of_key field
35 ** If we detect that a private key has been deleted
36 Bump the key event counter.
39 ** Make sure that secure memory is used where appropriate
41 * agent/pkdecrypt.c, agent/pksign.c
44 * Move pkcs-1 encoding into libgcrypt.
46 * Use a MAC to protect sensitive files.
47 The problem here is that we need yet another key and it is unlikely
48 that users are willing to remember that key too. It is possible to
49 do this with a smartcard, though.
52 ** Return an error code or a status info per user ID.
55 The parse_sexp function should not go into this file. Check whether
56 we can change all S-expression handling code to make use of this
60 ** Application context vs. reader slot
61 We have 2 concurrent method of tracking whether a reader is in use:
62 Using the session_list in command.c and the lock_table in app.c. IT
63 would be better to do this just at one place. First we need to see
64 how we can support cards with multiple applications.
65 ** Detecting a removed card works only after the ticker detected it.
66 We should check the card status in open-card to make this smoother.
67 Needs to be integrated with the status file update, though. It is
68 not a real problem because application will get a card removed
69 status and should the send a reset to try solving the problem.
71 ** Add a regression test to check the extkeyusage.
74 ** No card status notifications.
75 ** Regex support is disabled
76 We need to adjust the test to find the regex we have anyway in
77 gpg4win. Is that regex compatible to the OpenPGP requirement?
81 ** check that we issue NO_SECKEY xxx if a -u key was not found
82 We don't. The messages returned are also wrong (recipient vs. signer).
84 Replace by an estream based implementation.
87 ** provide jnlib_malloc and try to remove all jnlib_xmalloc.
90 ** issue a NO_SECKEY xxxx if a -u key was not found.
93 * When switching to libgcrypt 1.4
94 ** scd#encode_md_for_card, g10#encode_md_value, sm@do_encode_md
95 Remove the extra test for a valid algorithm as libgcrypt will do it
96 then in gcry_md_algo_info.
98 Remove the hard coded constant in random_is_faked.
99 ** g10/ Replace DIGEST_ALGO_SHA224
100 We can't do that right now because it is only defined by newer
101 versions of libgcrypt.
103 Remove these definitions.
107 * Extend selinux support to other modules
109 * Remove -sat PGP2 compatibility hack
111 * UTF-8 specific TODOs
115 We do not yet support P15 applications. The trivial thing using
116 ASCII characters will be easy to implement but the other cases need
120 ** After disabling scdaemon and sending a HUP
121 scdaemon stays as a zombie and gpg-agent does not perform any more
125 ** Migrate OpenPGP keys to another system
128 Although we pass LC_MESSAGE from gpgsm et al. to Pinentry, this has
129 only an effect on the stock GTK strings (e.g. "OK") and not on any
130 strings gpg-agent generates and passes to Pinentry. This defeats
131 our design goal to allow changing the locale without changing
132 gpg-agent's default locale (e.g. by the command updatestartuptty).
134 * RFC 4387: Operational Protocols: Certificate Store Access via HTTP