Strong words Shai, I don't know if you think anything has changed during
2018 regarding Flutter.
I've been evaluating it recently and there are a lot of things I like about
it and while its certainly beta (with a humble version 0.35) it already has
produced apps that have been featured on the
Hi,
Rebuilt this same test case in the past couple of hours and run on 2
different devices and the Simulator all showing the problem. But I think I
see where the confusion is. You can edit the last text area on the screen
(the 12th on an iPhone 6) and it will scroll into view fine but there
Hi Steve,
This build was made 2 days ago. Are you running it from the build server?
How do you test on the xcode simulator?
Cheers,
Nick
On Wednesday, March 21, 2018 at 5:21:56 AM UTC+13, Steve Hannah wrote:
>
> I'm unable to reproduce this issue on iPhone 6S running 11.1.2 or with
> Xcode
Yes, anything after the 6th text box will be covered by the keyboard when
it opens, you can scroll to see some but it springs back when take your
finger off to start typing.
iPhone 6s
iOS 11.2.6 but had users with older version report this issue too.
What were your findings from running the
public class Main {
private Form current;
private Resources theme;
public void init(Object context) {
theme = UIManager.initFirstTheme("/theme");
// Enable Toolbar on all Forms by default
Toolbar.setGlobalToolbar(true);
// Pro only feature, uncomment if
If you are experiencing an issue please mention the full platform your
issue applies to:
IDE: NetBeans
Desktop OS: Linux
Sending builds requests a log in for intercom.io which I don't have and
don't need in my app. It didn't used to do this.
--
You received this message because you are
I have a BoxLayout Y with a variable number of Containers in it. It is
scrollableY and can have 10 or so Containers each one have some buttons and
a TextArea. On Android when you edit the bottom TextArea the screen resizes
and you can see what you are typing, on iOS it does not and the TextArea
I can log into the dashboard and see my builds from yesterday and check
that my subscription is still active etc., but in Netbeans sending builds
fails saying 'Login failed' it prompts for the account and password after
that, this has been triple checked and is correct.
--
You received this
Copying the code from Capture.java and ensuring it runs on the EDT seems to
make a significant difference in stability on iOS.
On Monday, October 9, 2017 at 5:34:38 PM UTC+13, Shai Almog wrote:
>
> OK, I see the problem. I'll fix the log so it doesn't make noise.
>
> This isn't a problem in
Looks like getRGB is called for each pixel in the image and checkEDT is
called for each call to getRGB = a lot of EDT violations.
I think whatever is going on under the hood for Capture.capturePhoto should
make sure its on the EDT and this issue would go away and hopefully the
side effect
Adding a breakpoint before saveImage is called shows that the log attached
is generated only from this line:
String imagePath = Capture.capturePhoto(960, -1);
And saveImage is only called once per image.
I can provide a minimal example that shows this too, but the steps are
fairly easy:
1.
I don't have a reference, just from experience on a number of devices. The
3-4 that are updated to the latest versions of Android and Chrome will not
lauch a download directly from an intent, the ones that still have the old
version of chrome still do.
Entering the build server URL on the
It looks like a recent update to Chrome on Android prevents intents from
download files (or at least APKs) directly. This means that we can no
longer scan the QR code on the build server to download the apk to our
phones. You can download APK files if its the result of clicking a link in
the
I'm implementing Estimote Beacons as a cn1lib with background region
monitoring. It requires that the beaconmanager is initialised in
applicationDidFinishLoading and it I can only get it working this way:
ios.beforeFinishLaunching = self.beaconManager = [ESTBeaconManager new];
And then access
I can't send any iOS builds, the error log is different depending on the
project:
BluekeyInspect-dummy.o)) was built for newer iOS version (7.0) than being
linked (6.0)
Undefined symbols for architecture armv7:
"_com_codename1_components_FloatingHint___INIT_com_codename1_ui_TextField",
I use a third party analytics and crash reporting solution that catches
unhandled exceptions (full app crashes) and reports them its working fine,
integrating their SDKs via native code.
It has functionality to work with handled exceptions as well but it needs
the exception passed.
I can
I agree that using exceptions for normal flow isn't ideal but none of the
regex I tried worked with the Codename One regex class properly. I've tried
again and found a simple Regex did the job for my simplified use case. The
number is generated by a count of user actions so is always an integer
I need to tokenise a string (split by space) to see if it ends in a number
or a word. I'm using this hack to do it:
public static boolean isNumeric(String str) {
try {
Double.parseDouble(str);
} catch (NumberFormatException nfe) {
return false;
I have some older and newer devices. The newer devices are running iOS 11
beta. I can install apps on either type via TestFlight but its much quicker
to test using debug builds. However if I include ios.debug.archs=armv7 then
I can ONLY install it on 32 bit devices and if I don't include that
No harm in having an alternative that does work if the current system
doesn't. However this particular app has no server so I'd really like to
keep it that way if at all possible which means the only way I can get logs
are through the Codename One systems.
On Friday, July 28, 2017 at 4:56:35
If you change skins the database ends up open twice or something and you
get the following error. This also happens if you run tests from the
TestRecorder.
[EDT] 0:0:0,236 - Exception: java.lang.UnsatisfiedLinkError -
org.sqlite.NativeDB._open(Ljava/lang/String;I)V
The workaround below sends emails but they are blank on new iOS devices -
works fine on the simulator and old iOS devices... I think the log isn't
being written at all.
On Thursday, July 27, 2017 at 3:58:59 PM UTC+12, nickk...@gmail.com wrote:
>
> This works as a Log.sendLog() workaround but
This works as a Log.sendLog() workaround but I'd still prefer logs off the
iOS device to be working again:
Message m = new
Message(Util.readToString(Storage.getInstance().createInputStream("CN1Log__$")));
m.sendMessageViaCloud(.)
On Tuesday, July 25,
I've checked my spam settings, anything marked as spam is delivered to a
folder and not deleted, and I've just white listed @codenameone.com (is
this the right domain name?)
Note that I still get emails but only from older apps - so if its the spam
filter then something would have changed in
Devices I could get logs from a week or so ago I can no longer see logs on
builds sent this week.
I haven't had any where near 200 emails in the past 48 hours.
The only email logs I get are from apps released last year that haven't
been updated since - they come in at a rate of a couple per
Actually in making a minimal example to send to you I found out that if
permission hasn't been granted then Display.execute MUST be on the EDT,
once permission has been granted it doesn't matter which thread its called
on. This is probably related to those lines that you highlighted but
I have a project that is crashing hard since last Friday's library update
but only on iOS.
I have put Log.p thoughout to try and get to the bottom of it, but no Log.p
is output on the running console - this used to work for the devices
running iOS 9 (it hasn't ever worked on iOS 10 devices as
Methods like Row.getString throw this error.
java.io.IOException: ResultSet closed
at com.codename1.impl.javase.SECursor.getString(SECursor.java:269)
at
com.codename1.db.ThreadSafeDatabase$RowWrapper$7.run(ThreadSafeDatabase.java:216)
at
I ended up using BrowserComponent.executeAndReturnString and used a
synchronous javascript call to make the same call that the
ConnectionRequest was doing. Its a bit of a hack, requires a synchronous
call and injects logic in javascript in the middle of the app but it does
get the response
I haven't tried that, how can I get the cookie from the BrowserComponent in
order to set it on the ConnectionRequest header?
On Friday, June 16, 2017 at 6:22:40 PM UTC+12, Shai Almog wrote:
>
> I didn't deal much with the native store cookies feature maybe Steve has
> an idea.
> Have you tried
I've spent days working with a helpful tester and there seems to be a
problem in persisting Cookies between the BrowserComponent and
ConnectIonRequest. The use case is pretty simple. Sorry about all the
details, I'm desperate for a solution as soon as possible and don't want to
go back and
On Android M devices the first time you run the app with native Google Maps
it is black (with a google logo, with no map content) the next time the app
is run then it is Ok.
This happens even if the location permission box is loaded a number of
forms before any form that uses the map - it
Ok might be finally making progress on this.
The documentation says that the deadline is the time limit by which time
the background fetch must be completed, it also mentions that its a 30
second limit, however in the implementation its hardcoded to send
System.getTimeMillis() + 25 * 60 *
I don't actually want the menu to hide / show, I want it always visible,
but if I have commandBehaviour=Bar and a Toolbar I always get a Side Menu.
On Monday, August 29, 2016 at 5:24:06 PM UTC+12, Shai Almog wrote:
>
> Not with general customization.
>
> The menu is wired to the overflow. Some
I've been made aware of some issues on my QRScanning library and looking
into this now its the same issue here as well. onActivityResult is never
called.
Going back though build and testing logs it looks like this was working
fine on builds that were created two weeks ago, some time between
I want to use the Toolbar API so I can have good control over buttons and
elements in the title bar, but want to have the main menu of the app in a
bar at the bottom of the screen. This combination doesn't seem to be
possible with theme constants and adding commands to the form.
--
You
I used SplunkMINT for this purpose and have written a cn1lib wraps their
Android and iOS SDKs and adds breadcrumbs, events, transactions etc., to
the reporting and can be managed within Codename One code.
I haven't released it because my request for clarification on whether I
could put it on
Is it going to be possible to edit left aligned but display centered
without the placholder text remaining while editing - this would be an
acceptable solution. Or does choosing center aligned mean that all editing
is broken and unsupported?
Trying it on other android phones some have this
I'm revisiting some older projects and a new build with the latest library
has added an ugly scroll bar to my scrolling containers and lists. I put in
theme entries for 'Scroll' and "ScrollThumb" with transparency 0 and border
empty. It removed some but not all, ticking or unticking scroll
The design I'm required to match has input fields that are center aligned.
TextFields can't be center aligned so I'm using single line TextArea fields
instead.
When tapping the field the placeholder text flickers on and off a few times
and sometimes remains while typing and sometimes clears
I'm drawing clocks for a list of events that have different times that need
to be displayed.
I am caching them so they aren't generated every time but they do need to
be generated at some point. Currently there are about 20 and it takes under
a second so it just 'sticks' the UI briefly.
I
In the UI designer I've added a multi button to a form, if I show that form
I get a null pointer. Stepping through shows this error:
Component not found use UIBuilder.registerCustomComponent(MultiButton, class
);
Being thrown from UIBuilder line 413.
As far as I know libraries and plugins are
I need to use 'iphone_new' for the bluetooth cn1lib support but it fails
codesigning. Using the original 'iphone' has build errors.
To confirm its not to do with the certificates / private key / provisioning
profile setup I took an app that doesn't use bluetooth and builds fine on
the 'iphone'
If I have two separate cn1libs that add the same build hints
in codenameone_library_appended.properties they get added twice and then
won't build due to duplicates (i.e., android.xpermissions,
android.xapplication)
Unfortunately the only workaround I've found not practical for a sensible
I'm having multiple issues with the web browser component on Android, it
these issues don't appear on iOS and in the simulator it is generally OK as
well.
The set up is a form with a border layout and the only component is a
webbrowser in the center and a number of other forms that don't have
Does the new Bluetooth LE library support iBeacons and scanning in the
background (with the app closed)? If not, would it be possible to support
this via other cordova libraries such as these:
https://github.com/petermetz/cordova-plugin-ibeacon
OR
46 matches
Mail list logo