Too many bugs in 9500 Java

[garzotto]

Developing in Java for the 9500 is really frustrating. There are just too many bugs. I was so glad to finally have a portable machine that can run Java programs. Enthusiastically, I began to write code for the 9500 - and now get more and more frustrated as I am mainly trying to find work arounds (if possible at all). The bugs that annoy me most so far:

* File f.lastModified() returns garbage
* BufferedInpuStream b.readLine() hangs forever
* No way to handle key presses from the cursor pad
* Runtime.exec() does not work
* No way to find out the IP address of the machine (always returns 127.0.0.1)
* Clipboard does not work for text bigger than 255(?) characters and appends odd line breaks at the end.
* Selection in a textField does not work and neither placement of the caret (so the caret is always at the beginning of the text in the field when opened)
* TextAreas are painfully slow (so slow that they cannot be used if more than a few words are in them)
* Menu accelerators only work with letters (Ctrl-A, Ctrl-B, etc.) but not with anything else (Ctrl-1, Ctrl-Up, etc.)

I wonder if there is any chance Nokia is following this forum (or different ones) and will react by improving the Java implementation on the 9500...

Andreas

[FreddyOne]

I have many problems in Java on the Nokia 9500 too. The programs works fine in the emulator. But on the device its failed.
Mainly i thinks there is ab big bug by the java String-functions.
if you habe strings lower then 265 byte, its okay, but by strings
higher then 256 bytes, you get by the length-function always 256 or by using then indexOf-function for searching the string you get always 256 if the string you are looking for is on a position higher than 256.

[garzotto]

Maybe this is the reason for why the clipboard does not work, i.e. is limited in size.

I really hope that somebody from Nokia will eventually care about Java working reasonably on the 9500...

Andreas

[jprosti]

Atleast the lack of "keypresses from cursor pad" is just a "design decision" - 9500 has separate arrow keys (which produce valid key events) in addition to the "joystick type" pad (which generates mouse events)!

As for the bug fixing - the Java on 9500 is IBM sw... as in 9200 it was Symbian... and the responsibility?

[zgabor43]

Joystick pad doesn't generate anything. It can only navigate the virtual cursor (if it's active). It's not a design decision, it's a bug (or to be more precise: lameness)

[garzotto]

There was no reaction from Nokia so far and the device is available for half a year now. Unfortunately, I tend to come to the conclusion:

Nokia does have zero interest in making Java available on the 9500/9300, but only had an interest in writing on the box that these devices support Java :-(

[jprosti]

Originally posted by zgabor43
Joystick pad doesn't generate anything. It can only navigate the virtual cursor (if it's active). It's not a design decision, it's a bug (or to be more precise: lameness)

Well, just tested that and actually it generates mouse events (of course only when the v-cursor is enabled).

I'm not saying to be very happy with the thing, but atleast it works some way ...even if not the one intuitively supposed.

The biggest thing that bugs me here is the ambiguity between native and Java apps - the 'joystick pad' works as 'arrow keys' in f.eg. File Manager, but in Java apps it doesn't, so it's somewhat confusing for the user... :(

And what comes to Nokia supporting Java... starting from 92xx I've been told "do it with C++..." ;)

[zgabor43]

Originally posted by jprosti


The biggest thing that bugs me here is the ambiguity between native and Java apps - the 'joystick pad' works as 'arrow keys' in f.eg. File Manager, but in Java apps it doesn't, so it's somewhat confusing for the user... :(

Me and Garzotto are talking about this all the time. The virtual cursor thing woth nothing at all. Or am I mistaken? There is no desing decision here, it's just a stupid, unfinished personal profile implementation that can't be used for implementing good java application. I can't tell you how disappointed I am.

[jprosti]

Originally posted by zgabor43
... The virtual cursor thing woth nothing at all...

Eh, did not quite get that? "The ... is worth nothing" perhaps...

As I wrote - the 'joystick-kind-a-pad' works as a mouse-replacement when the virtual cursor is enabled. Yes, the v-cursor is not very handy and in style guides it is advised not to be used.

There are these 'arrow keys' in addition though - they seem to work as expected ...once you notice them on the keyboard - like me just pounding on the 'non-working' joystick ;)

[zgabor43]

Originally posted by jprosti
Eh, did not quite get that? "The ... is worth nothing" perhaps...

As I wrote - the 'joystick-kind-a-pad' works as a mouse-replacement when the virtual cursor is enabled. Yes, the v-cursor is not very handy and in style guides it is advised not to be used.

Yes, so that it woth nothing and behaves quite strangely compared to the native apps. I get that. You call that as a 'desing decision...'

There are these 'arrow keys' in addition though - they seem to work as expected ...once you notice them on the keyboard - like me just pounding on the 'non-working' joystick ;)

Yeah, I noticed too. But will your users notice too? Our users think we are stupid because they can navigate in all applications with the joypad except our application. And our users think we are lamers because we can't implement this 'feature'.

[jprosti]

Yes I call it a 'design decision' as they do in hundreds of miscellaneous docs (mostly release notes of course...) by MS, Sun, IBM, Nokia etc. ...so don't think it as anything like positive :D

[malbry]

* File f.lastModified() returns garbage
* BufferedInpuStream b.readLine() hangs forever
* No way to handle key presses from the cursor pad
* Runtime.exec() does not work
* No way to find out the IP address of the machine (always returns 127.0.0.1)
* Clipboard does not work for text bigger than 255(?) characters and appends odd line breaks at the end.
* Selection in a textField does not work and neither placement of the caret (so the caret is always at the beginning of the text in the field when opened)
* TextAreas are painfully slow (so slow that they cannot be used if more than a few words are in them)
* Menu accelerators only work with letters (Ctrl-A, Ctrl-B, etc.) but not with anything else (Ctrl-1, Ctrl-Up, etc.)

I wonder if there is any chance Nokia is following this forum (or different ones) and will react by improving the Java implementation on the 9500...

It's now 13 months since these bugs were first reported. Has anyone from Symbian and/or Nokia investigated?

Thanks,
Malcolm
www.freepoc.org

[Nokia Ron]

It's now 13 months since these bugs were first reported. Has anyone from Symbian and/or Nokia investigated?

Thanks,
Malcolm
www.freepoc.org

Hi Malcom,

I'm new here, but I am here just for cases like this. I'll see if I can find out something. It may take me a bit of time to find the right contact.

Ron

[garzotto]

None of the bugs have been addressed by Nokia so far (and they were also reported to Nokia development using a "more direct" channel). As Nokia has fired all Series 80 developers - at least that is what I heard - it seems to be save to assume none of the bugs will ever be fixed.

[karstenliebert]

Hi,

there are some more problems in IBMs new J9VM on 9300 (and of course on 9500 too).
One example is the not handled function, if the control button is pressed only.
Otherwise if released, there will be a handling. So what does this use for?

Another one is following:
Not all countries in the world does use an american keyboard for input.
(e.g. here in germany y and z are turn around)
But there is no way the VM is coresponding with the country properties of the device. So if you want to check the button z is pressed you have to check for button y and other way arround.
(the old VM on 9210i does function with standard european keybaord input)

Are there some workarounds from Nokia or IBM to fix these problems?
Especialy for the first one i need a fix or same hints to solve it other way very urgent.

Thanks

Karsten Liebert