|
Replies:
15
-
Last Post:
Aug 4, 2008 6:46 AM
by: Patrick Julien
|
|
|
|
|
|
|
TextField on the BlackBerry
Posted:
Aug 1, 2008 7:15 AM
|
|
|
|
|
Any hints on how to make this work properly on the blackberry would be
appreciated. I have several problems.
1. First, back space never works, in fact, it includes a forward space
2. I can assume a Qwerty device every time, yet, even then, I can't get
proper keys into the text field.
I was using TextArea at first and this was working properly. However, upon
seeing the final product, the customer wanted inline editing. I've also
tried setting the input mode to all the different modes ("abc", etc.)
--
http://www.spectrumdt.com
http://codepimps.org
[att1.html]
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 1, 2008 7:22 AM
in response to: Patrick Julien
|
|
|
|
|
Hi,
> Any hints on how to make this work properly on the blackberry would
> be appreciated. I have several problems.
>
> 1. First, back space never works, in fact, it includes a forward space
Please read my previous response to the deleteChar post.
>
> 2. I can assume a Qwerty device every time, yet, even then, I can't
> get proper keys into the text field.
While qwerty input is unsupported officially try setting
setQwertyInput(true).
> I was using TextArea at first and this was working properly.
> However, upon seeing the final product, the customer wanted inline
> editing. I've also tried setting the input mode to all the
> different modes ("abc", etc.)
>
> --
> http://www.spectrumdt.com
> http://codepimps.org
>
Shai Almog
http://lwuit.blogspot.com/
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 1, 2008 7:28 AM
in response to: Shai Almog
|
|
|
|
|
On Fri, Aug 1, 2008 at 10:22 AM, Shai Almog <Shai.Almog@sun.com> wrote:
> Hi,
>
> Any hints on how to make this work properly on the blackberry would be
> appreciated. I have several problems.
>
> 1. First, back space never works, in fact, it includes a forward space
>
>
> Please read my previous response to the deleteChar post.
>
OK, set SoftKey-Clear, I get it.
>
>
> 2. I can assume a Qwerty device every time, yet, even then, I can't get
> proper keys into the text field.
>
>
> While qwerty input is unsupported officially try
> setting setQwertyInput(true).
>
Already done. This does not work very well at all.
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 1, 2008 8:03 AM
in response to: Patrick Julien
|
|
|
|
|
On Fri, Aug 1, 2008 at 10:28 AM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>
>
> On Fri, Aug 1, 2008 at 10:22 AM, Shai Almog <Shai.Almog@sun.com> wrote:
>
>> Hi,
>>
>> Any hints on how to make this work properly on the blackberry would be
>> appreciated. I have several problems.
>>
>> 1. First, back space never works, in fact, it includes a forward space
>>
>>
>> Please read my previous response to the deleteChar post.
>>
>
> OK, set SoftKey-Clear, I get it.
>
Hmm, can get this to work with other platforms, but not the BB.
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 6:40 AM
in response to: Patrick Julien
|
|
|
|
|
When you jump out to the native text field, what class is that? Is it the
class from lcdui? How do actually mingle the two widget toolkits?
I'm asking because I really need to find a solution to the inline editing
problem. On a gamecanvas on the BB, the key events don't actually
correspond to what is found on the keys. This is because the BB on a game
canvas is simulating a regular MIDP device. So even if the #2 key is for
the symbols 2, T and Y, it will product 2, a, b and c characters for the
game canvas.
So if I had a better idea on how you interoperate between the native widget
and the LWUIT, I think I could find a solution to this.
On Fri, Aug 1, 2008 at 11:03 AM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>
>
> On Fri, Aug 1, 2008 at 10:28 AM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>
>>
>>
>> On Fri, Aug 1, 2008 at 10:22 AM, Shai Almog <Shai.Almog@sun.com> wrote:
>>
>>> Hi,
>>>
>>> Any hints on how to make this work properly on the blackberry would be
>>> appreciated. I have several problems.
>>>
>>> 1. First, back space never works, in fact, it includes a forward space
>>>
>>>
>>> Please read my previous response to the deleteChar post.
>>>
>>
>> OK, set SoftKey-Clear, I get it.
>>
>
>
> Hmm, can get this to work with other platforms, but not the BB.
>
>
>
--
http://www.spectrumdt.com
http://codepimps.org
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 7:38 AM
in response to: Patrick Julien
|
|
|
|
|
> When you jump out to the native text field, what class is that? Is
> it the class from lcdui? How do actually mingle the two widget
> toolkits?
Yes, the native text editing is javax.microedition.lcdui.TextBox.
> I'm asking because I really need to find a solution to the inline
> editing problem. On a gamecanvas on the BB, the key events don't
> actually correspond to what is found on the keys. This is because
> the BB on a game canvas is simulating a regular MIDP device. So
> even if the #2 key is for the symbols 2, T and Y, it will product
> 2, a, b and c characters for the game canvas.
>
> So if I had a better idea on how you interoperate between the
> native widget and the LWUIT, I think I could find a solution to this.
There is no technical way to interpolate between the native widget
and LWUIT that I am aware of.
>
>
> On Fri, Aug 1, 2008 at 11:03 AM, Patrick Julien
> <pjulien@spectrumdt.com> wrote:
>
>
> On Fri, Aug 1, 2008 at 10:28 AM, Patrick Julien
> <pjulien@spectrumdt.com> wrote:
>
>
> On Fri, Aug 1, 2008 at 10:22 AM, Shai Almog <Shai.Almog@sun.com>
> wrote:
> Hi,
>
>> Any hints on how to make this work properly on the blackberry
>> would be appreciated. I have several problems.
>>
>> 1. First, back space never works, in fact, it includes a forward
>> space
>
> Please read my previous response to the deleteChar post.
>
> OK, set SoftKey-Clear, I get it.
>
>
> Hmm, can get this to work with other platforms, but not the BB.
>
>
>
>
>
> --
> http://www.spectrumdt.com
> http://codepimps.org
>
Shai Almog
http://lwuit.blogspot.com/
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 8:08 AM
in response to: Shai Almog
|
|
|
|
|
But you do it with TextBox no? You're killing me here, at this point, I'm
going to have to ditch LWUIT and re-write using the native BB ui toolkit
On Sat, Aug 2, 2008 at 10:38 AM, Shai Almog <Shai.Almog@sun.com> wrote:
>
> When you jump out to the native text field, what class is that? Is it the
> class from lcdui? How do actually mingle the two widget toolkits?
>
>
> Yes, the native text editing is javax.microedition.lcdui.TextBox.
>
> I'm asking because I really need to find a solution to the inline editing
> problem. On a gamecanvas on the BB, the key events don't actually
> correspond to what is found on the keys. This is because the BB on a game
> canvas is simulating a regular MIDP device. So even if the #2 key is for
> the symbols 2, T and Y, it will product 2, a, b and c characters for the
> game canvas.
>
> So if I had a better idea on how you interoperate between the native widget
> and the LWUIT, I think I could find a solution to this.
>
>
> There is no technical way to interpolate between the native widget and
> LWUIT that I am aware of.
>
>
>
> On Fri, Aug 1, 2008 at 11:03 AM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>
>>
>>
>> On Fri, Aug 1, 2008 at 10:28 AM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>>
>>>
>>>
>>> On Fri, Aug 1, 2008 at 10:22 AM, Shai Almog <Shai.Almog@sun.com> wrote:
>>>
>>>> Hi,
>>>>
>>>> Any hints on how to make this work properly on the blackberry would be
>>>> appreciated. I have several problems.
>>>>
>>>> 1. First, back space never works, in fact, it includes a forward space
>>>>
>>>>
>>>> Please read my previous response to the deleteChar post.
>>>>
>>>
>>> OK, set SoftKey-Clear, I get it.
>>>
>>
>>
>> Hmm, can get this to work with other platforms, but not the BB.
>>
>>
>>
>
>
> --
> http://www.spectrumdt.com
> http://codepimps.org
>
>
> Shai Almog
> http://lwuit.blogspot.com/
>
>
--
http://www.spectrumdt.com
http://codepimps.org
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 9:17 AM
in response to: Patrick Julien
|
|
|
|
|
The text field is lightweight and does all the key interception caret
positioning itself. It is written in such a way to allow adaptability
to other devices (such as qwerty devices or devices of other
theoretical input modes). However, if said device doesn't send any
key event for a key press/release it is impossible to implement
something like that correctly.
I don't have an immediate solution for you.
However, since LWUIT itself is very portable it is feasible to port
LWUIT to the RIMlet API once the LWUIT source code becomes available
which might workaround some of the issues in the device. This is not
a port we currently plan to perform.
> But you do it with TextBox no? You're killing me here, at this
> point, I'm going to have to ditch LWUIT and re-write using the
> native BB ui toolkit
>
> On Sat, Aug 2, 2008 at 10:38 AM, Shai Almog <Shai.Almog@sun.com>
> wrote:
>
>> When you jump out to the native text field, what class is that?
>> Is it the class from lcdui? How do actually mingle the two widget
>> toolkits?
>
> Yes, the native text editing is javax.microedition.lcdui.TextBox.
>
>> I'm asking because I really need to find a solution to the inline
>> editing problem. On a gamecanvas on the BB, the key events don't
>> actually correspond to what is found on the keys. This is because
>> the BB on a game canvas is simulating a regular MIDP device. So
>> even if the #2 key is for the symbols 2, T and Y, it will product
>> 2, a, b and c characters for the game canvas.
>>
>> So if I had a better idea on how you interoperate between the
>> native widget and the LWUIT, I think I could find a solution to this.
>
> There is no technical way to interpolate between the native widget
> and LWUIT that I am aware of.
>
>>
>>
>> On Fri, Aug 1, 2008 at 11:03 AM, Patrick Julien
>> <pjulien@spectrumdt.com> wrote:
>>
>>
>> On Fri, Aug 1, 2008 at 10:28 AM, Patrick Julien
>> <pjulien@spectrumdt.com> wrote:
>>
>>
>> On Fri, Aug 1, 2008 at 10:22 AM, Shai Almog <Shai.Almog@sun.com>
>> wrote:
>> Hi,
>>
>>> Any hints on how to make this work properly on the blackberry
>>> would be appreciated. I have several problems.
>>>
>>> 1. First, back space never works, in fact, it includes a forward
>>> space
>>
>> Please read my previous response to the deleteChar post.
>>
>> OK, set SoftKey-Clear, I get it.
>>
>>
>> Hmm, can get this to work with other platforms, but not the BB.
>>
>>
>>
>>
>>
>> --
>> http://www.spectrumdt.com
>> http://codepimps.org
>>
>
>
> Shai Almog
> http://lwuit.blogspot.com/
>
>
>
>
> --
> http://www.spectrumdt.com
> http://codepimps.org
>
Shai Almog
http://lwuit.blogspot.com/
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 9:41 AM
in response to: Shai Almog
|
|
|
|
|
On Sat, Aug 2, 2008 at 12:17 PM, Shai Almog <Shai.Almog@sun.com> wrote:
> The text field is lightweight and does all the key interception caret
> positioning itself. It is written in such a way to allow adaptability to
> other devices (such as qwerty devices or devices of other theoretical input
> modes). However, if said device doesn't send any key event for a key
> press/release it is impossible to implement something like that correctly.
>
I know that and that is what I am attempting right now. The problem however
is that the keys that are sent to the game canvas don't follow the same key
mapping that lcdui or something similar.
If I install a key listener using RIM's API, I can see the real codes, but
those don't match the ones that are sent to game canvas
>
> I don't have an immediate solution for you.
> However, since LWUIT itself is very portable it is feasible to port LWUIT
> to the RIMlet API once the LWUIT source code becomes available which might
> workaround some of the issues in the device. This is not a port we currently
> plan to perform.
>
Trying to do that now just by extending TextField. However, considering the
market share BB has and the interest that LWUIT + BB has generated on this
list, it just boggles my mind that BB is not tier 1 platform for LWUIT.
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 10:01 AM
in response to: Patrick Julien
|
|
|
|
|
> The text field is lightweight and does all the key interception
> caret positioning itself. It is written in such a way to allow
> adaptability to other devices (such as qwerty devices or devices of
> other theoretical input modes). However, if said device doesn't
> send any key event for a key press/release it is impossible to
> implement something like that correctly.
>
> I know that and that is what I am attempting right now. The
> problem however is that the keys that are sent to the game canvas
> don't follow the same key mapping that lcdui or something similar.
>
> If I install a key listener using RIM's API, I can see the real
> codes, but those don't match the ones that are sent to game canvas
If there are unique and consistent keys you can create a key map that
translates these keys to proper abc codes for LWUIT.
> I don't have an immediate solution for you.
> However, since LWUIT itself is very portable it is feasible to port
> LWUIT to the RIMlet API once the LWUIT source code becomes
> available which might workaround some of the issues in the device.
> This is not a port we currently plan to perform.
>
>
> Trying to do that now just by extending TextField. However,
> considering the market share BB has and the interest that LWUIT +
> BB has generated on this list, it just boggles my mind that BB is
> not tier 1 platform for LWUIT.
While we want to support the BB as much as feasible we are hindered
by one of the worst MIDP implementations and an uncooperative company
(RIM).
Outside of north america it is physically impossible to buy a RIM
device in most territories e.g. here... We have a device which is
just not functional on our network, this means we can't even testany
of the problems. So it is just not technically feasible for us to
make this a priority platform.
Since we integrated workarounds for several BB specific issues into
LWUIT itself I think that indicates that we take the platform
seriously but we are limited by technical reasons. Porting LWUIT into
a RIMlet would be a duplication and would require more effort than
the one we keep for every single other supported platform, since we
have no way to test on devices we couldn't even test the results of
this effort.
Shai Almog
http://lwuit.blogspot.com/
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 10:36 AM
in response to: Shai Almog
|
|
|
|
|
I apologize, I did not understand the level of desire on the part of LWUIT
to support BB and the obstacles it's facing.
On the other hand, I have something almost working now with my TextField
based class. Now just need to figure out the sensitivity of the wheel.
On Sat, Aug 2, 2008 at 1:01 PM, Shai Almog <Shai.Almog@sun.com> wrote:
> The text field is lightweight and does all the key interception caret
>> positioning itself. It is written in such a way to allow adaptability to
>> other devices (such as qwerty devices or devices of other theoretical input
>> modes). However, if said device doesn't send any key event for a key
>> press/release it is impossible to implement something like that correctly.
>>
>
> I know that and that is what I am attempting right now. The problem
> however is that the keys that are sent to the game canvas don't follow the
> same key mapping that lcdui or something similar.
>
> If I install a key listener using RIM's API, I can see the real codes, but
> those don't match the ones that are sent to game canvas
>
>
> If there are unique and consistent keys you can create a key map that
> translates these keys to proper abc codes for LWUIT.
>
> I don't have an immediate solution for you.
>
>> However, since LWUIT itself is very portable it is feasible to port LWUIT
>> to the RIMlet API once the LWUIT source code becomes available which might
>> workaround some of the issues in the device. This is not a port we currently
>> plan to perform.
>>
>
>
> Trying to do that now just by extending TextField. However, considering
> the market share BB has and the interest that LWUIT + BB has generated on
> this list, it just boggles my mind that BB is not tier 1 platform for LWUIT.
>
>
> While we want to support the BB as much as feasible we are hindered by one
> of the worst MIDP implementations and an uncooperative company (RIM).
> Outside of north america it is physically impossible to buy a RIM device in
> most territories e.g. here... We have a device which is just not functional
> on our network, this means we can't even testany of the problems. So it is
> just not technically feasible for us to make this a priority platform.
> Since we integrated workarounds for several BB specific issues into LWUIT
> itself I think that indicates that we take the platform seriously but we are
> limited by technical reasons. Porting LWUIT into a RIMlet would be a
> duplication and would require more effort than the one we keep for every
> single other supported platform, since we have no way to test on devices we
> couldn't even test the results of this effort.
>
>
> Shai Almog
> http://lwuit.blogspot.com/
>
>
--
http://www.spectrumdt.com
http://codepimps.org
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 1:49 PM
in response to: Patrick Julien
|
|
|
|
|
Shai,
I've gotten this to work properly, the only real downside is that it doesn't
know anything about locales. Might be an interesting contribution once
LWUIT opens up if you're interested.
On Sat, Aug 2, 2008 at 1:36 PM, Patrick Julien <pjulien@spectrumdt.com>wrote:
> I apologize, I did not understand the level of desire on the part of LWUIT
> to support BB and the obstacles it's facing.
>
> On the other hand, I have something almost working now with my TextField
> based class. Now just need to figure out the sensitivity of the wheel.
>
>
> On Sat, Aug 2, 2008 at 1:01 PM, Shai Almog <Shai.Almog@sun.com> wrote:
>
>> The text field is lightweight and does all the key interception caret
>>> positioning itself. It is written in such a way to allow adaptability to
>>> other devices (such as qwerty devices or devices of other theoretical input
>>> modes). However, if said device doesn't send any key event for a key
>>> press/release it is impossible to implement something like that correctly.
>>>
>>
>> I know that and that is what I am attempting right now. The problem
>> however is that the keys that are sent to the game canvas don't follow the
>> same key mapping that lcdui or something similar.
>>
>> If I install a key listener using RIM's API, I can see the real codes, but
>> those don't match the ones that are sent to game canvas
>>
>>
>> If there are unique and consistent keys you can create a key map that
>> translates these keys to proper abc codes for LWUIT.
>>
>> I don't have an immediate solution for you.
>>
>>> However, since LWUIT itself is very portable it is feasible to port LWUIT
>>> to the RIMlet API once the LWUIT source code becomes available which might
>>> workaround some of the issues in the device. This is not a port we currently
>>> plan to perform.
>>>
>>
>>
>> Trying to do that now just by extending TextField. However, considering
>> the market share BB has and the interest that LWUIT + BB has generated on
>> this list, it just boggles my mind that BB is not tier 1 platform for LWUIT.
>>
>>
>> While we want to support the BB as much as feasible we are hindered by one
>> of the worst MIDP implementations and an uncooperative company (RIM).
>> Outside of north america it is physically impossible to buy a RIM device
>> in most territories e.g. here... We have a device which is just not
>> functional on our network, this means we can't even testany of the problems.
>> So it is just not technically feasible for us to make this a priority
>> platform.
>> Since we integrated workarounds for several BB specific issues into LWUIT
>> itself I think that indicates that we take the platform seriously but we are
>> limited by technical reasons. Porting LWUIT into a RIMlet would be a
>> duplication and would require more effort than the one we keep for every
>> single other supported platform, since we have no way to test on devices we
>> couldn't even test the results of this effort.
>>
>>
>> Shai Almog
>> http://lwuit.blogspot.com/
>>
>>
>
>
> --
> http://www.spectrumdt.com
> http://codepimps.org
>
>
--
http://www.spectrumdt.com
http://codepimps.org
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 4:09 PM
in response to: Patrick Julien
|
|
|
|
|
Patrick,
Would you be willing to share what you've done to make this work? I would
like to get this working correctly in my BB app as well.
Thanks!
--Bill
On Sat, Aug 2, 2008 at 4:49 PM, Patrick Julien <pjulien@spectrumdt.com>wrote:
> Shai,
>
> I've gotten this to work properly, the only real downside is that it
> doesn't know anything about locales. Might be an interesting contribution
> once LWUIT opens up if you're interested.
>
>
> On Sat, Aug 2, 2008 at 1:36 PM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>
>> I apologize, I did not understand the level of desire on the part of LWUIT
>> to support BB and the obstacles it's facing.
>>
>> On the other hand, I have something almost working now with my TextField
>> based class. Now just need to figure out the sensitivity of the wheel.
>>
>>
>> On Sat, Aug 2, 2008 at 1:01 PM, Shai Almog <Shai.Almog@sun.com> wrote:
>>
>>> The text field is lightweight and does all the key interception caret
>>>> positioning itself. It is written in such a way to allow adaptability to
>>>> other devices (such as qwerty devices or devices of other theoretical input
>>>> modes). However, if said device doesn't send any key event for a key
>>>> press/release it is impossible to implement something like that correctly.
>>>>
>>>
>>> I know that and that is what I am attempting right now. The problem
>>> however is that the keys that are sent to the game canvas don't follow the
>>> same key mapping that lcdui or something similar.
>>>
>>> If I install a key listener using RIM's API, I can see the real codes,
>>> but those don't match the ones that are sent to game canvas
>>>
>>>
>>> If there are unique and consistent keys you can create a key map that
>>> translates these keys to proper abc codes for LWUIT.
>>>
>>> I don't have an immediate solution for you.
>>>
>>>> However, since LWUIT itself is very portable it is feasible to port
>>>> LWUIT to the RIMlet API once the LWUIT source code becomes available which
>>>> might workaround some of the issues in the device. This is not a port we
>>>> currently plan to perform.
>>>>
>>>
>>>
>>> Trying to do that now just by extending TextField. However, considering
>>> the market share BB has and the interest that LWUIT + BB has generated on
>>> this list, it just boggles my mind that BB is not tier 1 platform for LWUIT.
>>>
>>>
>>> While we want to support the BB as much as feasible we are hindered by
>>> one of the worst MIDP implementations and an uncooperative company (RIM).
>>> Outside of north america it is physically impossible to buy a RIM device
>>> in most territories e.g. here... We have a device which is just not
>>> functional on our network, this means we can't even testany of the problems.
>>> So it is just not technically feasible for us to make this a priority
>>> platform.
>>> Since we integrated workarounds for several BB specific issues into LWUIT
>>> itself I think that indicates that we take the platform seriously but we are
>>> limited by technical reasons. Porting LWUIT into a RIMlet would be a
>>> duplication and would require more effort than the one we keep for every
>>> single other supported platform, since we have no way to test on devices we
>>> couldn't even test the results of this effort.
>>>
>>>
>>> Shai Almog
>>> http://lwuit.blogspot.com/
>>>
>>>
>>
>>
>> --
>> http://www.spectrumdt.com
>> http://codepimps.org
>>
>>
>
>
> --
> http://www.spectrumdt.com
> http://codepimps.org
>
>
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 9:14 PM
in response to: Bill Snyder
|
|
|
|
|
I would, unfortunately, it seems the code doesn't actually work on a real
device, it only works in the emulator(4.2 and 4.3). I still have work to
do it seems.
On Sat, Aug 2, 2008 at 7:09 PM, Bill Snyder <wsnyder6@gmail.com> wrote:
> Patrick,
>
> Would you be willing to share what you've done to make this work? I would
> like to get this working correctly in my BB app as well.
>
> Thanks!
>
> --Bill
>
>
> On Sat, Aug 2, 2008 at 4:49 PM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>
>> Shai,
>>
>> I've gotten this to work properly, the only real downside is that it
>> doesn't know anything about locales. Might be an interesting contribution
>> once LWUIT opens up if you're interested.
>>
>>
>> On Sat, Aug 2, 2008 at 1:36 PM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>>
>>> I apologize, I did not understand the level of desire on the part of
>>> LWUIT to support BB and the obstacles it's facing.
>>>
>>> On the other hand, I have something almost working now with my TextField
>>> based class. Now just need to figure out the sensitivity of the wheel.
>>>
>>>
>>> On Sat, Aug 2, 2008 at 1:01 PM, Shai Almog <Shai.Almog@sun.com> wrote:
>>>
>>>> The text field is lightweight and does all the key interception caret
>>>>> positioning itself. It is written in such a way to allow adaptability to
>>>>> other devices (such as qwerty devices or devices of other theoretical input
>>>>> modes). However, if said device doesn't send any key event for a key
>>>>> press/release it is impossible to implement something like that correctly.
>>>>>
>>>>
>>>> I know that and that is what I am attempting right now. The problem
>>>> however is that the keys that are sent to the game canvas don't follow the
>>>> same key mapping that lcdui or something similar.
>>>>
>>>> If I install a key listener using RIM's API, I can see the real codes,
>>>> but those don't match the ones that are sent to game canvas
>>>>
>>>>
>>>> If there are unique and consistent keys you can create a key map that
>>>> translates these keys to proper abc codes for LWUIT.
>>>>
>>>> I don't have an immediate solution for you.
>>>>
>>>>> However, since LWUIT itself is very portable it is feasible to port
>>>>> LWUIT to the RIMlet API once the LWUIT source code becomes available which
>>>>> might workaround some of the issues in the device. This is not a port we
>>>>> currently plan to perform.
>>>>>
>>>>
>>>>
>>>> Trying to do that now just by extending TextField. However, considering
>>>> the market share BB has and the interest that LWUIT + BB has generated on
>>>> this list, it just boggles my mind that BB is not tier 1 platform for LWUIT.
>>>>
>>>>
>>>> While we want to support the BB as much as feasible we are hindered by
>>>> one of the worst MIDP implementations and an uncooperative company (RIM).
>>>> Outside of north america it is physically impossible to buy a RIM device
>>>> in most territories e.g. here... We have a device which is just not
>>>> functional on our network, this means we can't even testany of the problems.
>>>> So it is just not technically feasible for us to make this a priority
>>>> platform.
>>>> Since we integrated workarounds for several BB specific issues into
>>>> LWUIT itself I think that indicates that we take the platform seriously but
>>>> we are limited by technical reasons. Porting LWUIT into a RIMlet would be a
>>>> duplication and would require more effort than the one we keep for every
>>>> single other supported platform, since we have no way to test on devices we
>>>> couldn't even test the results of this effort.
>>>>
>>>>
>>>> Shai Almog
>>>> http://lwuit.blogspot.com/
>>>>
>>>>
>>>
>>>
>>> --
>>> http://www.spectrumdt.com
>>> http://codepimps.org
>>>
>>>
>>
>>
>> --
>> http://www.spectrumdt.com
>> http://codepimps.org
>>
>>
>
--
http://www.spectrumdt.com
http://codepimps.org
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 4, 2008 6:46 AM
in response to: Patrick Julien
|
|
|
|
|
OK, now I got this to work, but just like Sun, I can't release code just
like that.
Here's the algorithm nonetheless
1. extend text field and implement focus listener, add this to the focus
listener list in the ctor, e.g.,
addFocusListener(this);
2. have an internal class implement KeyListener from RIM
3. On focus in and out, register and un-register the KeyListener with the
singleton UiApplication object
4. Only keyDown and keyUp will ever get called, hence why no locales will
work except en.
5. On keyDown, use Keypad to extract both the char and int from the
key-code.
6. Implement keyDown to have key repeating, cursor positioning and either
call deleteChar, keyPressed or nothing on the text field itself
On Sun, Aug 3, 2008 at 12:14 AM, Patrick Julien <pjulien@spectrumdt.com>wrote:
> I would, unfortunately, it seems the code doesn't actually work on a real
> device, it only works in the emulator(4.2 and 4.3). I still have work to
> do it seems.
>
>
> On Sat, Aug 2, 2008 at 7:09 PM, Bill Snyder <wsnyder6@gmail.com> wrote:
>
>> Patrick,
>>
>> Would you be willing to share what you've done to make this work? I would
>> like to get this working correctly in my BB app as well.
>>
>> Thanks!
>>
>> --Bill
>>
>>
>> On Sat, Aug 2, 2008 at 4:49 PM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>>
>>> Shai,
>>>
>>> I've gotten this to work properly, the only real downside is that it
>>> doesn't know anything about locales. Might be an interesting contribution
>>> once LWUIT opens up if you're interested.
>>>
>>>
>>> On Sat, Aug 2, 2008 at 1:36 PM, Patrick Julien <pjulien@spectrumdt.com>wrote:
>>>
>>>> I apologize, I did not understand the level of desire on the part of
>>>> LWUIT to support BB and the obstacles it's facing.
>>>>
>>>> On the other hand, I have something almost working now with my TextField
>>>> based class. Now just need to figure out the sensitivity of the wheel.
>>>>
>>>>
>>>> On Sat, Aug 2, 2008 at 1:01 PM, Shai Almog <Shai.Almog@sun.com> wrote:
>>>>
>>>>> The text field is lightweight and does all the key interception caret
>>>>>> positioning itself. It is written in such a way to allow adaptability to
>>>>>> other devices (such as qwerty devices or devices of other theoretical input
>>>>>> modes). However, if said device doesn't send any key event for a key
>>>>>> press/release it is impossible to implement something like that correctly.
>>>>>>
>>>>>
>>>>> I know that and that is what I am attempting right now. The problem
>>>>> however is that the keys that are sent to the game canvas don't follow the
>>>>> same key mapping that lcdui or something similar.
>>>>>
>>>>> If I install a key listener using RIM's API, I can see the real codes,
>>>>> but those don't match the ones that are sent to game canvas
>>>>>
>>>>>
>>>>> If there are unique and consistent keys you can create a key map that
>>>>> translates these keys to proper abc codes for LWUIT.
>>>>>
>>>>> I don't have an immediate solution for you.
>>>>>
>>>>>> However, since LWUIT itself is very portable it is feasible to port
>>>>>> LWUIT to the RIMlet API once the LWUIT source code becomes available which
>>>>>> might workaround some of the issues in the device. This is not a port we
>>>>>> currently plan to perform.
>>>>>>
>>>>>
>>>>>
>>>>> Trying to do that now just by extending TextField. However,
>>>>> considering the market share BB has and the interest that LWUIT + BB has
>>>>> generated on this list, it just boggles my mind that BB is not tier 1
>>>>> platform for LWUIT.
>>>>>
>>>>>
>>>>> While we want to support the BB as much as feasible we are hindered by
>>>>> one of the worst MIDP implementations and an uncooperative company (RIM).
>>>>> Outside of north america it is physically impossible to buy a RIM
>>>>> device in most territories e.g. here... We have a device which is just not
>>>>> functional on our network, this means we can't even testany of the problems.
>>>>> So it is just not technically feasible for us to make this a priority
>>>>> platform.
>>>>> Since we integrated workarounds for several BB specific issues into
>>>>> LWUIT itself I think that indicates that we take the platform seriously but
>>>>> we are limited by technical reasons. Porting LWUIT into a RIMlet would be a
>>>>> duplication and would require more effort than the one we keep for every
>>>>> single other supported platform, since we have no way to test on devices we
>>>>> couldn't even test the results of this effort.
>>>>>
>>>>>
>>>>> Shai Almog
>>>>> http://lwuit.blogspot.com/
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> http://www.spectrumdt.com
>>>> http://codepimps.org
>>>>
>>>>
>>>
>>>
>>> --
>>> http://www.spectrumdt.com
>>> http://codepimps.org
>>>
>>>
>>
>
>
> --
> http://www.spectrumdt.com
> http://codepimps.org
>
>
--
http://www.spectrumdt.com
http://codepimps.org
[att1.html]
|
|
|
|
|
|
|
|
|
Re: TextField on the BlackBerry
Posted:
Aug 2, 2008 8:10 PM
in response to: Patrick Julien
|
|
|
|
|
Locales aren't handled by text field and much as input methods which
allow you to add an input method for additional locales.
While LWUIT itself hasn't been published yet, you can start
contributing and the contribution process through LWUIT-Incubator.
The project has instructions on how to contribute and collaborate
with us (we need a signed document the SCA) without putting Sun's IP
at risk.
Thanks.
> Shai,
>
> I've gotten this to work properly, the only real downside is that
> it doesn't know anything about locales. Might be an interesting
> contribution once LWUIT opens up if you're interested.
>
> On Sat, Aug 2, 2008 at 1:36 PM, Patrick Julien
> <pjulien@spectrumdt.com> wrote:
> I apologize, I did not understand the level of desire on the part
> of LWUIT to support BB and the obstacles it's facing.
>
> On the other hand, I have something almost working now with my
> TextField based class. Now just need to figure out the sensitivity
> of the wheel.
>
>
> On Sat, Aug 2, 2008 at 1:01 PM, Shai Almog <Shai.Almog@sun.com> wrote:
>> The text field is lightweight and does all the key interception
>> caret positioning itself. It is written in such a way to allow
>> adaptability to other devices (such as qwerty devices or devices
>> of other theoretical input modes). However, if said device doesn't
>> send any key event for a key press/release it is impossible to
>> implement something like that correctly.
>>
>> I know that and that is what I am attempting right now. The
>> problem however is that the keys that are sent to the game canvas
>> don't follow the same key mapping that lcdui or something similar.
>>
>> If I install a key listener using RIM's API, I can see the real
>> codes, but those don't match the ones that are sent to game canvas
>
> If there are unique and consistent keys you can create a key map
> that translates these keys to proper abc codes for LWUIT.
>
>> I don't have an immediate solution for you.
>> However, since LWUIT itself is very portable it is feasible to
>> port LWUIT to the RIMlet API once the LWUIT source code becomes
>> available which might workaround some of the issues in the device.
>> This is not a port we currently plan to perform.
>>
>>
>> Trying to do that now just by extending TextField. However,
>> considering the market share BB has and the interest that LWUIT +
>> BB has generated on this list, it just boggles my mind that BB is
>> not tier 1 platform for LWUIT.
>
>
> While we want to support the BB as much as feasible we are hindered
> by one of the worst MIDP implementations and an uncooperative
> company (RIM).
> Outside of north america it is physically impossible to buy a RIM
> device in most territories e.g. here... We have a device which is
> just not functional on our network, this means we can't even
> testany of the problems. So it is just not technically feasible for
> us to make this a priority platform.
> Since we integrated workarounds for several BB specific issues into
> LWUIT itself I think that indicates that we take the platform
> seriously but we are limited by technical reasons. Porting LWUIT
> into a RIMlet would be a duplication and would require more effort
> than the one we keep for every single other supported platform,
> since we have no way to test on devices we couldn't even test the
> results of this effort.
>
>
> Shai Almog
> http://lwuit.blogspot.com/
>
>
>
>
> --
> http://www.spectrumdt.com
> http://codepimps.org
>
>
>
>
> --
> http://www.spectrumdt.com
> http://codepimps.org
>
Shai Almog
http://lwuit.blogspot.com/
[att1.html]
|
|
|
|
|
java.net RSS
