To make 0.99.2 easier to reach, telepathy-glib generates tp_cli_channel_type_call_call_hangup() instead of tp_cli_channel_type_call1_call_hangup(), and so on. We should include the "1" in non-core interfaces so that we can actually version them. Motivation for doing this now: I want to move a bunch of our non-spec interfaces into the main spec. Since many of them have names like .DRAFT, they need renaming *anyway*, so we might as well version them properly, even in 0.x; but if I do that and merge it into 1.0 before fixing this, I'll have to delete their ABI versions and then reinstate them later.
Created attachment 87462 [details] [review] [tp-glib next] Don't strip ABI versions from generated code --- 90 files changed, 2881 insertions(+), 2894 deletions(-) This was the least interesting patch I've done for a long time. The interesting part is in tools/.
Created attachment 87463 [details] [review] [idle next] Adapt for use of ABI versions in generated code
Created attachment 87464 [details] [review] [haze next] Adapt for use of ABI versions in generated code
Also needs patches for the rest of the usual suspects: Empathy (unless it only uses high-level API) Farstream (unless it only uses high-level API) Folks (unless it only uses high-level API) Logger (unless it only uses high-level API) Gabble (please do Bug #69431 first) MC Rakia Salut
Comment on attachment 87462 [details] [review] [tp-glib next] Don't strip ABI versions from generated code Review of attachment 87462 [details] [review]: ----------------------------------------------------------------- ++
Comment on attachment 87463 [details] [review] [idle next] Adapt for use of ABI versions in generated code Review of attachment 87463 [details] [review]: ----------------------------------------------------------------- ++ assuming tests pass
Comment on attachment 87464 [details] [review] [haze next] Adapt for use of ABI versions in generated code Review of attachment 87464 [details] [review]: ----------------------------------------------------------------- ++ assuming tests pass
Created attachment 87743 [details] [review] [MC next] Adapt for ABI versions in telepathy-glib codegen --- Same idea.
Created attachment 87744 [details] [review] [Salut next] Update for telepathy-glib codegen putting ABI versions in names
Comment on attachment 87743 [details] [review] [MC next] Adapt for ABI versions in telepathy-glib codegen Review of attachment 87743 [details] [review]: ----------------------------------------------------------------- ++
Comment on attachment 87744 [details] [review] [Salut next] Update for telepathy-glib codegen putting ABI versions in names Review of attachment 87744 [details] [review]: ----------------------------------------------------------------- ++
I merged all these patches to 'next'.
Created attachment 88113 [details] [review] [gabble] Adapt for use of ABI versions in generated code
Created attachment 88114 [details] [review] [rakia] Adapt for use of ABI versions in generated code
Created attachment 88215 [details] [review] [logger] Adapt for use of ABI versions in generated code
Created attachment 88216 [details] [review] [logger] tests/lib: sync with tp-glib next
Created attachment 88221 [details] [review] [tp-fs] Adapt for use of ABI versions in generated code
Empathy: https://bugzilla.gnome.org/show_bug.cgi?id=710992 Folks: https://bugzilla.gnome.org/show_bug.cgi?id=710996
Thanks, checking...
Comment on attachment 88113 [details] [review] [gabble] Adapt for use of ABI versions in generated code Review of attachment 88113 [details] [review]: ----------------------------------------------------------------- Gabble in "rather large" shock... Looks good. ::: src/conn-aliasing.c @@ +1053,4 @@ > g_value_take_string (val, alias); > > tp_contacts_mixin_set_contact_attribute (attributes_hash, > + handle, TP_IFACE_CONNECTION_INTERFACE_ALIASING1"/alias", Definitely not a blocker (quite the opposite: please merge first!) but it would be good to grep these projects for "/alias" and use TP_TOKEN_CONNECTION_INTERFACE_ALIASING1_ALIAS. ::: src/conn-avatars.c @@ +612,4 @@ > g_value_set_string (val, ""); > > tp_contacts_mixin_set_contact_attribute (attributes_hash, handle, > + TP_IFACE_CONNECTION_INTERFACE_AVATARS1"/token", val); ... and grep for "/token" too ::: src/conn-client-types.c @@ +148,4 @@ > val = tp_g_value_slice_new_take_boxed (G_TYPE_STRV, types); > > tp_contacts_mixin_set_contact_attribute (attributes_hash, handle, > + TP_IFACE_CONNECTION_INTERFACE_CLIENT_TYPES1 "/client-types", val); ... and "/client-types" ::: src/conn-contact-info.c @@ +1057,4 @@ > TP_ARRAY_TYPE_CONTACT_INFO_FIELD_LIST, contact_info); > > tp_contacts_mixin_set_contact_attribute (attributes_hash, > + contact, TP_IFACE_CONNECTION_INTERFACE_CONTACT_INFO1"/info", ... and "/info" ::: src/conn-location.c @@ +622,4 @@ > TP_HASH_TYPE_STRING_VARIANT_MAP, location); > > tp_contacts_mixin_set_contact_attribute (attributes_hash, > + handle, TP_IFACE_CONNECTION_INTERFACE_LOCATION1"/location", val); and "/location" ::: src/connection.c @@ +3429,4 @@ > > tp_contacts_mixin_set_contact_attribute (attributes_hash, > handle, > + TP_IFACE_CONNECTION_INTERFACE_CONTACT_CAPABILITIES1"/capabilities", this too ::: src/media-factory.c @@ +840,4 @@ > if (G_UNLIKELY (qc_gtalk_p2p == 0)) > { > qc_gtalk_p2p = g_quark_from_static_string ( > + TP_IFACE_CHANNEL_TYPE_CALL1 "/gtalk-p2p"); I forget, do we have tokens for these too?
Comment on attachment 88114 [details] [review] [rakia] Adapt for use of ABI versions in generated code Review of attachment 88114 [details] [review]: ----------------------------------------------------------------- Looks good. What I said about "/alias" for Gabble applies here too.
Comment on attachment 88215 [details] [review] [logger] Adapt for use of ABI versions in generated code Review of attachment 88215 [details] [review]: ----------------------------------------------------------------- Sure.
Comment on attachment 88216 [details] [review] [logger] tests/lib: sync with tp-glib next Review of attachment 88216 [details] [review]: ----------------------------------------------------------------- Sure.
Comment on attachment 88221 [details] [review] [tp-fs] Adapt for use of ABI versions in generated code Review of attachment 88221 [details] [review]: ----------------------------------------------------------------- Sure ::: examples/call-handler.c @@ +578,4 @@ > > tp_handle_channels_context_accept (handler_context); > > + tp_cli_channel_type_call1_call_accept (proxy, -1, Certainly not a blocker, but shouldn't this be using tp_call_channel_accept_async() since we have it? @@ +632,5 @@ > tp_base_client_add_handler_capabilities_varargs (client, > + TP_IFACE_CHANNEL_TYPE_CALL1 "/video/h264", > + TP_IFACE_CHANNEL_TYPE_CALL1 "/shm", > + TP_IFACE_CHANNEL_TYPE_CALL1 "/ice", > + TP_IFACE_CHANNEL_TYPE_CALL1 "/gtalk-p2p", Not a blocker, but, tokens would be good.
All merged, I'll look at the extra comments now.
I insta merged the patches to use TOKEN macros, except some in Gabble because of bug #70962
Created attachment 88236 [details] [review] [tp-fs] call-handler: use tp_call_channel_accept_async()
Comment on attachment 88236 [details] [review] [tp-fs] call-handler: use tp_call_channel_accept_async() Review of attachment 88236 [details] [review]: ----------------------------------------------------------------- ::: examples/call-handler.c @@ +578,4 @@ > > tp_handle_channels_context_accept (handler_context); > > + tp_call_channel_accept_async (TP_CALL_CHANNEL (proxy), NULL, NULL); Assuming it works, ++
All merged; closing.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.