Since the doShowReader will run in nextTick, it will get picked up by
checkTask, which is in the beginning of next UI loop and blocks
UIManager from redrawing the screen for the info popup.
bring back the old behavior: when width and height are set and all the
zooming options are false, zoom the image to that size. The last commit
to Imagewidget.lua caused it ignores the set width and height and use
the image's original size instead when no zoom option is set. This
caused #1979. Although #1979 can be fixed by setting autostretch=true,
but I think it's good to fix here.
to make koreader on Android more stable
and with these opt params:
```
require("jit.opt").start("sizemcode=64","maxmcode=64", "hotloop=10000")
```
The strategy here is that we only use precious mcode memory (jitting)
on deep loops like the several blitting methods in blitbuffer.lua and
the pixel-copying methods in mupdf.lua. So that a small amount of mcode
memory (64KB) allocated when koreader is launched in the android.lua
is enough for the program and it won't need to jit other parts of lua
code and thus won't allocate mcode memory any more which by our
observation will be harder and harder as we run koreader.
Matches the reader setting we're reading/writing at that point,
and avoids confusion related to scoping because it doesn't necessarily
match powerd.is_fl_on at that point.
Well, at least I had to wrap my mind around it ^^
* Drops support for mocking the frontlight setting internally which may
cause incorrect in-memory values.
* Adds new supported value for `KOBO_LIGHT_ON_START` (-2), which sets
'Kobo eReader.conf' as the source to update `settings.reader.lua`'s
brightness setting on startup, thus using the value from it
indirectly.
* Adds the `KOBO_SYNC_BRIGHTNESS_WITH_NICKEL` configuration variable
which updates 'Kobo eReader.conf' every time the brightness setting is
changed within koreader.
* Fixes missing call to save brightness when modifying via two-finger
swipe.
Closes#1523.
_checkTasks first get number of tasks in the stack and does a numeric
for loop to go through each task. The problem is a task might call
schedule or unschedule, which will reorder tasks in the stack. This will
invalidate many of the table indexes used in the for loop.
This patch turns the task stack into an ordered queue, so _checkTasks
only pops one item out of the queue each time instead of setting up a
for loop at the beginning. This should avoid the race condition
mentioned above.
Making sure to always show up to date info.
Also fix the battery polling cache invalidation logic to actually
work?
Should make it behave properly on devices where we handle resuming
ourselves.
Currently only tested on Ubuntu-touch emulator with framework
ubuntu-sdk-14.10 for armhf.
The ubuntu-touch port is binary compatible with the Kobo port
major changes in this PR are:
1. rename the emulator device to sdl device since both the emulator
and the ubuntu-touch target use libsdl to handle input/output.
2. ubuntu-touch app has no write access to the installation dir so
all write-outs should be in a seperate dir definded in `datastorage`.
And also probably prevent a crash when trying to enable WiFi there.
We really shouldn't even show the WiFi stuff on some of these old
devices, but at least now it doesn't crash ;).
use coroutine to support ad-hoc callback in login button so that when
the username/password is incorrect it will retry automatically
currently tested with COPS on apache
This should fix#1001.
and get rid of the ugly baseurl workaround for "Internet Archive",
actually the url building of all these catalogs is perfectly
handled with `url.absolute(base, relative)`.
Without this fix, self.dimen is shared among all inputcontainers
intances, which breaks some of the UI rendering. All widget should
set/initialize their own self.dimen in self:init() method.
Allow one to go further in the FM's setting menu on legacy devices...
... and potentially shoot oneself in the foot because of the unusable
confirmdialogs (among other things ;p).
To avoid potential issues with USBMS.
Namely, make sure CWD doesn't point to the userstore for persitent
processes.
Make sure wpa_supplicant doesn't try to use our own OpenSSL lib, which,
again, is in the userstore area.
Properly kill our dhcp client invocation on when killing WiFi
Also, some related fixes to the startup script:
Properly kill wpa_supplicant & the dhcp client when we kill
wifi before rebooting to nickel.
More closely matches native behavior on REAGL devices.
Closing those widgets should still trigger a partial refresh though,
because we usually get back to the reader, and text, so we want REAGL
;).
The VFAT file system of `/mnt/us` is case insensitive so that
require("json") from Spore which intends to load luajson module in
rock directory actually will load `frontend/JSON.lua`, the two JSON
parser libraries have slightly different APIs.
This patch remove the pure Lua JSON library in favor of `luajson` which
uses LPeg pattern marching backend for better performance.
it uses non-blocking turbo I/O looper to process http request
so that multiple http request can be handled simultaneously and http request
won't block user input, and most importantly, in Lua's way.
Change from the commands from `wifi.sh` to those from `wifi_enable_dhcp.sh` from Tshering's Start Menu. See Issue #939. Should resolve hanging of Kobo Aura H2O.
