c88646d...
by
Tor Lillqvist <email address hidden>
Work in progress on a re-working of the iOS app plumbing
No more use of FakeSocket, yay.
Functionality still quite limited compared to the main branch. Many of
the messages sent by JS are ignored. Most callbacks from core are
ignored yet. But at least tiles show up.
The app is temporarily renamed in this branch while work is ongoing,
so that one can have both versions on the same device.
e633dd7...
by
Tor Lillqvist <email address hidden>
e64bbeb...
by
Tor Lillqvist <email address hidden>
Pass rendered tiles as uncompressed BMP files in the iOS app
Avoids the need for PNG encoding (takes significant amount of CPU
time) and Base64 encoding in the app process, transfer to JavaScript
(running in a WebKit process of its own), and corresponding decoding
(in the WebKit process). Instead simply pass the URL of each tile file
to the JavaScript. Remove each BMP file once it has been loaded.
07bf598...
by
Tor Lillqvist <email address hidden>
Make objects and threads go away more reliably in the iOS app
We probably used to have circular references that made KitSocketPoll
and KitWebSocketHandler objects hang around forever, or something.
(Not a problem in web-based Online where kit processes have a
restricted lifetime.)
ef7e79a...
by
Tor Lillqvist <email address hidden>
Use LOG_INF_NOFILE() for the FakeSocket logging callbacks
(Why not call LOG_INF directly in FakeSocket.cpp instead? Good
question. I guess my idea was originally to keep FakeSocket separately
testable without all the Online logging stuff.)
137981f...
by
Tor Lillqvist <email address hidden>
Add LOG_INF_NOFILE that doesn't display the source file location
Useful (like LOG_TRC_NOFILE) in cases where the source file location
of the LOG_INF call is uninteresting. (Like if the message to be
logged has originated somewhere else.)