Fri, 29 May 2020 12:16:15 +0200 Roker log what happens on connection close. JSON-143 tip
Fri, 29 May 2020 12:12:08 +0200 Roker set callback for connection close, to remove pending eventListener entries. JSON-143
Fri, 29 May 2020 11:15:37 +0200 Roker add more debug log :-/ JSON-143
Thu, 28 May 2020 23:38:34 +0200 Roker add request counter and thread ID to the JSON-RPC results JSON-143
Thu, 28 May 2020 11:03:55 +0200 Roker make Logger's thread_id() function accessable by other modules JSON-143
Mon, 25 May 2020 09:56:35 +0200 Roker re-structured In<> and InOut<> class templates. I hope, MSVC is happy now. JSON-143
Wed, 20 May 2020 18:36:02 +0200 Roker make compiler happy. And remove no longer needed, but dangerous code. JSON-143
Wed, 20 May 2020 18:33:16 +0200 Roker the same for notifyHandshake() JSON-143
Wed, 20 May 2020 18:29:55 +0200 Roker Hah! Here was the culprit: 2 In<> or InOut<> wrapper on the same struct message! But it is unecessary here; direct to_json<> is what we need here. JSON-143
Wed, 20 May 2020 18:29:05 +0200 Roker set value to nullptr. Should be unecessary, but avoids double-free problems, if 2 In<>, InOut<>... wrappers are assigned to the same raw C object. JSON-143
Tue, 19 May 2020 12:52:01 +0200 Roker oops, I forgot this change to commit. My fault. Sorryyyyy... JSON-143
Tue, 19 May 2020 12:24:44 +0200 Roker okay, joerg sais, MSVC does not need them anymore. so I throw them away. JSON-143
Mon, 18 May 2020 19:06:27 +0200 Roker merge double head. JSON-143
Mon, 18 May 2020 19:05:44 +0200 Roker kaum macht man es richtig, schon geht's! :-D JSON-143
Mon, 18 May 2020 12:07:34 +0200 Jorg Knobloch Add session_registry.cc/hh to Windows build. JSON-143
Fri, 15 May 2020 15:40:23 +0200 Roker add this function to test JSON-143
Fri, 15 May 2020 11:33:30 +0200 Roker implement pollForEvents(). :-D JSON-143
Thu, 14 May 2020 16:26:24 +0200 Roker Yeah, it compiles again. Big refactoring completed. :-DDDD JSON-143
Thu, 14 May 2020 16:21:38 +0200 Roker add some more methods... :-D JSON-143
Thu, 14 May 2020 10:22:20 +0200 Roker why 3 methods when 2 are better? combine add() and get(). JSON-143
Thu, 14 May 2020 01:31:23 +0200 Roker ongoing work. will quit for today. JSON-143
Thu, 14 May 2020 01:30:22 +0200 Roker JsonAdapter becomes abstract class. pEp::mini::Adapter will implement it. JSON-134 will also profit from it. :-) JSON-143
Thu, 14 May 2020 01:27:36 +0200 Roker (un)registerEventListerner() no longer exist. JSON-143
Thu, 14 May 2020 01:16:28 +0200 Roker event delivery to clients is easier now! JSON-143
Thu, 14 May 2020 00:18:58 +0200 Roker re-define SessionRegistry completely. JSON-143
Wed, 13 May 2020 22:28:04 +0200 Roker merge 'postrelease' into JSON-143 again. JSON-143
Wed, 13 May 2020 20:08:19 +0200 Roker accept is not necessary, because evhtp_bind_socket() does it already. AHA! JSON-147
Wed, 13 May 2020 15:04:42 +0200 Roker merge "sync" into "JSON-147" JSON-147
Tue, 12 May 2020 17:17:44 +0200 Roker make stupid MSVC happy again. postrelease
Tue, 12 May 2020 16:08:50 +0200 Roker redefine In_Pep_Session, so it uses the session_registry of the JsonAdapter. sync
Tue, 12 May 2020 16:08:50 +0200 Roker redefine In_Pep_Session, so it uses the session_registry of the JsonAdapter. postrelease
Fri, 08 May 2020 18:24:51 +0200 Roker switch code from ev_http to libevhtp in JsonAdapter and ev_server JSON-147
Thu, 07 May 2020 17:23:25 +0200 Roker add libevhtp to Makefile.conf and requirements in README.md JSON-147
Thu, 07 May 2020 13:08:27 +0200 Roker start new branch JSON-147 "Switch from ev_http to libevhtp" JSON-147
Wed, 06 May 2020 16:48:19 +0200 Roker stupid compilers require templates that are never used. "if constexpr" in C++17 might help here, because it forces compilers to be less dumb. postrelease
Wed, 06 May 2020 16:48:19 +0200 Roker stupid compilers require templates that are never used. "if constexpr" in C++17 might help here, because it forces compilers to be less dumb. sync
Wed, 06 May 2020 13:56:33 +0200 Roker merge parts of JSON-142 (per_user_directory()) into sync branch. the next revision would create a merge conflict due to Volker's -D command line switch is used in different ways. sync
Mon, 04 May 2020 18:25:23 +0200 Roker declare JsonAdapter::pollForEvents() and add necessary template specializations for js::Array JSON-143
Mon, 04 May 2020 17:03:35 +0200 Roker start branch JSON-143 "Implement long polling for event delivery" JSON-143
Mon, 04 May 2020 16:00:19 +0200 Roker InLength<NoInput> needs to set the value of the base class postrelease
Sun, 03 May 2020 17:40:32 +0200 Volker Birk merging JSON-142
Sun, 03 May 2020 17:38:18 +0200 Volker Birk adding -D JSON-142
Sat, 02 May 2020 21:54:27 +0200 Roker according to Krista, neither the bool parameter of mime_encode_message() nor the bool* of mime_decode_message are part of the API for the cloents. So I don't expose them to the JSON-RPC clients. postrelease
Sat, 02 May 2020 21:31:38 +0200 Roker merge changes of JSON-142 into "postrelease" branch. JSON-142 is not finished, yet! postrelease
Sat, 02 May 2020 21:27:06 +0200 Roker menno... JSON-142
Thu, 30 Apr 2020 19:02:06 +0200 Roker use per_user_directory() on WIN32 and UNIX JSON-142
Tue, 21 Apr 2020 11:38:29 +0200 Roker ticket rejected. close branch. JSON-62
Tue, 28 Apr 2020 17:40:24 +0200 Volker Birk merging in PER_USER_DIRECTORY postrelease
Tue, 28 Apr 2020 17:39:48 +0200 Volker Birk ignore more postrelease
Tue, 28 Apr 2020 17:38:49 +0200 Volker Birk removing this postrelease
Tue, 28 Apr 2020 17:29:36 +0200 Volker Birk postrelease API postrelease
Sat, 25 Apr 2020 08:13:05 +0200 Volker Birk make daemonize optional like with sshd sync
Tue, 21 Apr 2020 15:51:19 +0200 Jorg Knobloch JSON-142: Use engine's per_user_directory() function to determine storage location for json-token (Windows part). JSON-142
Mon, 20 Apr 2020 18:46:46 +0200 Roker document event delivery to clients sync
Mon, 13 Apr 2020 01:02:11 +0200 Jorg Knobloch Adjusted comment after ENGINE-735. sync
Sun, 12 Apr 2020 18:48:03 +0200 Jorg Knobloch Merge of rev. f61ec0225fcd (which was forgotten). sync
Sun, 12 Apr 2020 18:41:51 +0200 Jorg Knobloch ENGINE-735 usage of timestamp* is now mandatory, take 2. sync
Wed, 04 Mar 2020 15:15:34 +0100 Jorg Knobloch Added mini-adapter-impl.hh back to the library project since it's used in ev_server.cc. sync
Sun, 12 Apr 2020 15:36:41 +0200 Volker Birk merging sync
Sun, 12 Apr 2020 15:36:15 +0200 Volker Birk ENGINE-735 usage of timestamp * is now mandatory sync