Re: Remote Debugger Anomalies

[Trevor DeVore via use-livecode]

Yes, it is part of the repo.

https://github.com/trevordevore/levure/tree/master/utils/external_editor_server

-- 
Trevor DeVore
ScreenSteps

On Mon, Sep 17, 2018 at 9:58 AM Sannyasin Brahmanathaswami via use-livecode
 wrote:

> Do you have that "script acting as a server" available in Levure?
>
>
> On 9/17/18 2:31 AM, Trevor DeVore via use-livecode wrote:
> > When working on a project I always work in ST with the LC IDE open. LC
> has
> > a script acting as a server running and ST is the client. ST sends a
> > notification to LC whenever a file is saved and if a stack with a
> matching
> > name/filename exists then the stack script is updated in memory. This has
> > been a very reliable workflow for me.
>
>
>
> ___
> use-livecode mailing list
> use-livecode@lists.runrev.com
> Please visit this url to subscribe, unsubscribe and manage your
> subscription preferences:
> http://lists.runrev.com/mailman/listinfo/use-livecode
>
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Sannyasin Brahmanathaswami via use-livecode]

Do you have that "script acting as a server" available in Levure?


On 9/17/18 2:31 AM, Trevor DeVore via use-livecode wrote:
> When working on a project I always work in ST with the LC IDE open. LC has
> a script acting as a server running and ST is the client. ST sends a
> notification to LC whenever a file is saved and if a stack with a matching
> name/filename exists then the stack script is updated in memory. This has
> been a very reliable workflow for me.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Trevor DeVore via use-livecode]

On Sun, Sep 16, 2018 at 11:54 PM Sannyasin Brahmanathaswami via
use-livecode  wrote:

>
> I suppose Trevor would chime here. If you to switch Sublime text and
> stay away for SE all together, all the problems go away? Because when
> you save the script you know it is in the state of your recent
> modification, confirmed, on disk.
>
> Maybe he could describe how he works.
> He must have IDE open and Sublime text at the same times.
>

I haven't been following the specific details of this thread so I don't
know if using Sublime Text (ST), or other external script editor, would
make the problems go away. It is true that when you save the scripts you
know that your changes are being saved directly to disk.

When working on a project I always work in ST with the LC IDE open. LC has
a script acting as a server running and ST is the client. ST sends a
notification to LC whenever a file is saved and if a stack with a matching
name/filename exists then the stack script is updated in memory. This has
been a very reliable workflow for me.

Now, if I switch branches in my git repo then I close the LC IDE and reopen
it. I don't have anything set up that notifies LC to reload all stacks in a
project when switching branches and there are potentially a number of files
that have been changed on disk.

-- 
Trevor DeVore
ScreenSteps
www.screensteps.com
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Sannyasin Brahmanathaswami via use-livecode]

On 9/16/18 1:38 PM, J. Landman Gay via use-livecode wrote:
> Stacks named that way were the ones I tried deleting when the whole IDE 
> wen

I suppose Trevor would chime here. If you to switch Sublime text and 
stay away for SE all together, all the problems go away? Because when 
you save the script you know it is in the state of your recent 
modification, confirmed, on disk.

Maybe he could describe how he works.
He must have IDE open and Sublime text at the same times.

BR


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[J. Landman Gay via use-livecode]

Stacks named that way were the ones I tried deleting when the whole IDE 
went down on my next test build. I decided not to do that any more. :)

--
Jacqueline Landman Gay | jac...@hyperactivesw.com
HyperActive Software | http://www.hyperactivesw.com
On September 16, 2018 6:20:13 PM Monte Goulding via use-livecode 
 wrote:


On 16 Sep 2018, at 1:47 am, Sannyasin Brahmanathaswami via use-livecode 
 wrote:


OK, given that we understand the problem, to be solved one day.

Until then, how do we work around it?

Give an error, and the Script Editor "becomes"

stack "d2dc63f5-efeb-412c-a6ae-741d60b569de"
card id 1002

What is best practice to get outand back to the original stack/scripts?


Hmm… in that scenario you might need to type `delete stack 
"d2dc63f5-efeb-412c-a6ae-741d60b569de”` in the message box. Definitely the 
cache is not being cleared/delted when it should be here.


Cheers

Monte
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your 
subscription preferences:

http://lists.runrev.com/mailman/listinfo/use-livecode





___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Monte Goulding via use-livecode]

> On 16 Sep 2018, at 10:36 am, Sannyasin Brahmanathaswami via use-livecode 
>  wrote:
> 
> No doubt Monte is going to say :-) "this is not problem, best you ignore
> this" ha!~

This seems a bit unfair to me. I don’t think I’ve ever said that. Clearly if 
you are having a problem then there’s a problem. 

Your last email, however, it sounded much more like you were having a problem 
with operations outside of LiveCode (git commands and text editing outside the 
IDE) than within LiveCode. It’s easy to make a mistake there because the IDE 
does not reload stacks if the underlying file is changed.

As always the simpler the recipe the easier it is for us to figure out what the 
issue is and resolve it.

Cheers

Monte
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Monte Goulding via use-livecode]

> On 16 Sep 2018, at 1:47 am, Sannyasin Brahmanathaswami via use-livecode 
>  wrote:
> 
> OK, given that we understand the problem, to be solved one day.
> 
> Until then, how do we work around it?
> 
> Give an error, and the Script Editor "becomes"
> 
> stack "d2dc63f5-efeb-412c-a6ae-741d60b569de"
> card id 1002
> 
> What is best practice to get outand back to the original stack/scripts?

Hmm… in that scenario you might need to type `delete stack 
"d2dc63f5-efeb-412c-a6ae-741d60b569de”` in the message box. Definitely the 
cache is not being cleared/delted when it should be here.

Cheers

Monte
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Sannyasin Brahmanathaswami via use-livecode]

Well I play with it a while, I don't have  a perfect recipe yet...

1) debug error show up. Do edit the script. You check the error and
close the script editor. So far I think the IDE has no cache of it.

2) then on the phone, you go to App Info, Uninstall.

So far I didn't get a cached script. But if you "touch" the script. In
some way, even to change a space. Then even after you uninstall the app
on the phone and close the editor, then open it again make some changes,
and you change some code and (verified save text file of a behavior by
BBedit)

 Then, magically then old code may appear!

then carefully, close the SE, and uninstall the app, appears go away..

No doubt Monte is going to say :-) "this is not problem, best you ignore
this" ha!~

BR


On 9/15/18 6:39 AM, J. Landman Gay via use-livecode wrote:
> I asked this before and the fix is to quit the app, if I remember right. 
> The problem on Android is that apps never really quit, they remain in the 
> background until the OS needs the RAM. When the OS wipes the app, we don't 
> receive any messages so the debugger won't know when the socket closes. The 
> same thing happens when the user hits the Home or Recent buttons to leave 
> the app, which also does not notify us. We need messages for these events. 
> Android does send the app a notification but LC hasn't implemented it.
>
> The result is that LC Android apps usually never know what's going on. It 
> interferes with the ability to save the app state as well.
>
> You might try adding a Quit button to see if that works. The quit command 
> was updated a bit recently so that might work, even if you just use it 
> temporarily for debugging.

-- 
Svasti Astu, Be Well!
Brahmanathaswami
 
Get the SivaSiva app, it's free:
https://www.himalayanacademy.com/apps/sivasiva


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[J. Landman Gay via use-livecode]

I asked this before and the fix is to quit the app, if I remember right. 
The problem on Android is that apps never really quit, they remain in the 
background until the OS needs the RAM. When the OS wipes the app, we don't 
receive any messages so the debugger won't know when the socket closes. The 
same thing happens when the user hits the Home or Recent buttons to leave 
the app, which also does not notify us. We need messages for these events. 
Android does send the app a notification but LC hasn't implemented it.


The result is that LC Android apps usually never know what's going on. It 
interferes with the ability to save the app state as well.


You might try adding a Quit button to see if that works. The quit command 
was updated a bit recently so that might work, even if you just use it 
temporarily for debugging.


--
Jacqueline Landman Gay | jac...@hyperactivesw.com
HyperActive Software | http://www.hyperactivesw.com
On September 15, 2018 10:49:53 AM Sannyasin Brahmanathaswami via 
use-livecode  wrote:



On 9/13/18 10:47 AM, Monte Goulding via use-livecode wrote:
OK that definitely gives me an idea. The cache is meant to be cleaned up 
when the socket closes so should be happening between launches of the app 
but we can also force the debug session to end during the deploy of the new 
version which will hopefully ensure things are cleaned up.


OK, given that we understand the problem, to be solved one day.

Until then, how do we work around it?

Give an error, and the Script Editor "becomes"

stack "d2dc63f5-efeb-412c-a6ae-741d60b569de"
card id 1002

What is best practice to get outand back to the original stack/scripts?

BR



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your 
subscription preferences:

http://lists.runrev.com/mailman/listinfo/use-livecode





___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Sannyasin Brahmanathaswami via use-livecode]

On 9/13/18 10:47 AM, Monte Goulding via use-livecode wrote:
> OK that definitely gives me an idea. The cache is meant to be cleaned up when 
> the socket closes so should be happening between launches of the app but we 
> can also force the debug session to end during the deploy of the new version 
> which will hopefully ensure things are cleaned up.

OK, given that we understand the problem, to be solved one day.

Until then, how do we work around it?

Give an error, and the Script Editor "becomes"

stack "d2dc63f5-efeb-412c-a6ae-741d60b569de"
card id 1002

What is best practice to get outand back to the original stack/scripts?

BR



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Sannyasin Brahmanathaswami via use-livecode]

I made no changes in the second branch, so I deleted it and made a new
one from nightly.


But,

a) since I received a stack from Elanor
b) never have opened before
c) edit the stack script
d) see the stack script of a debug session from Elanor,

(Stack name is a usual long ID number)

It makes you wonder where did it come from?

BR


On 9/13/18 10:44 AM, Monte Goulding via use-livecode wrote:
> t does not sound like the remote debugger was involved at all in this story 
> so perhaps look elsewhere for the problem.
>
>> On 14 Sep 2018, at 6:12 am, Sannyasin Brahmanathaswami via use-livecode 
>>  wrote:
>>
>> Well this is somewhat "urgent" due to a scenario I just witness.
>>
>> It GIT -->  branch -- nightly # scripts updated.
>>
>> committed
>>
>> switch to branch "fixIosBrowser"
>>
>> git merge nightly
>>
>> # get the usual conflict of binary stacks..
>>
>> git checkout --theirs "view/view_SivaSivaBrowser.livecode"
>>
>> #presumably up to do date.
>>
>> reboot LC, edit script of stack "view/view_SivaSivaBrowser.livecode"
>>
>> And it open a script in debug mode, with breakpoint that I was working
>> on yesterday, in branch


-- 
Svasti Astu, Be Well!
Brahmanathaswami
 
Get the SivaSiva app, it's free:
https://www.himalayanacademy.com/apps/sivasiva


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Monte Goulding via use-livecode]

> On 13 Sep 2018, at 11:42 pm, Andrew Bell via use-livecode 
>  wrote:
> 
>> - The stack that caches object scripts is not being cleaned up properly 
>> between debugging sessions. This would mean that you would see the old 
>> script but what would be executing would be the new script. One way to test 
>> this might be to move a breakpoint statement from one line to another and 
>> see what line is highlighted in the debugger.
> 
> This is my vote for what is happening. I have noticed numerous times that 
> stepping through lines I could see things happening in the standalone that 
> were NOT the highlited line of code in the script editor at the time. For 
> instance, a blank line (for readability of code) would execute some other 
> line of code that was now 2 lines down due to a recent change in the script.

OK that definitely gives me an idea. The cache is meant to be cleaned up when 
the socket closes so should be happening between launches of the app but we can 
also force the debug session to end during the deploy of the new version which 
will hopefully ensure things are cleaned up.

Cheers

Monte
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Monte Goulding via use-livecode]

It does not sound like the remote debugger was involved at all in this story so 
perhaps look elsewhere for the problem.

> On 14 Sep 2018, at 6:12 am, Sannyasin Brahmanathaswami via use-livecode 
>  wrote:
> 
> Well this is somewhat "urgent" due to a scenario I just witness.
> 
> It GIT -->  branch -- nightly # scripts updated.
> 
> committed
> 
> switch to branch "fixIosBrowser"
> 
> git merge nightly
> 
> # get the usual conflict of binary stacks..
> 
> git checkout --theirs "view/view_SivaSivaBrowser.livecode"
> 
> #presumably up to do date.
> 
> reboot LC, edit script of stack "view/view_SivaSivaBrowser.livecode"
> 
> And it open a script in debug mode, with breakpoint that I was working
> on yesterday, in branch
> fixIosBrowser! and the script that is nightly are not same!
> 
> git status
> Already up to date
> 
> but the scripts are different? Oh boy! now I can't depend on GIT and
> *something* is caching scripts.
> 
> Look like I have to make of branch fixIosBrowser switch night, make a
> copy and the do a level diff in bbedit between the two, Yikes!
> 
> Maybe is best to remove that ext Remote Debugging (doesnt work on iOS
> anyway) and turn of script Debug with building a standalone... only safe
> thing to do.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Sannyasin Brahmanathaswami via use-livecode]

Well this is somewhat "urgent" due to a scenario I just witness.

It GIT -->  branch -- nightly # scripts updated.

committed

switch to branch "fixIosBrowser"

git merge nightly

# get the usual conflict of binary stacks..

 git checkout --theirs "view/view_SivaSivaBrowser.livecode"

#presumably up to do date.

reboot LC, edit script of stack "view/view_SivaSivaBrowser.livecode"

And it open a script in debug mode, with breakpoint that I was working
on yesterday, in branch
fixIosBrowser! and the script that is nightly are not same!

git status
Already up to date

but the scripts are different? Oh boy! now I can't depend on GIT and
*something* is caching scripts.

Look like I have to make of branch fixIosBrowser switch night, make a
copy and the do a level diff in bbedit between the two, Yikes!

Maybe is best to remove that ext Remote Debugging (doesnt work on iOS
anyway) and turn of script Debug with building a standalone... only safe
thing to do.

 
On 9/13/18 8:30 AM, J. Landman Gay via use-livecode wrote:
> I think so too. I've seen updated card layouts appear, but the debug 
> script that was displayed didn't include additions that managed the 
> layout changes.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[J. Landman Gay via use-livecode]

On 9/13/18 8:42 AM, Andrew Bell via use-livecode wrote:
- The stack that caches object scripts is not being cleaned up 
properly between debugging sessions. This would mean that you would 
see the old script but what would be executing would be the new 
script. One way to test this might be to move a breakpoint statement 
from one line to another and see what line is highlighted in the 
debugger.


This is my vote for what is happening. I have noticed numerous times 
that stepping through lines I could see things happening in the 
standalone that were NOT the highlited line of code in the script editor 
at the time. For instance, a blank line (for readability of code) would 
execute some other line of code that was now 2 lines down due to a 
recent change in the script.


I think so too. I've seen updated card layouts appear, but the debug 
script that was displayed didn't include additions that managed the 
layout changes.


I think I asked here some time ago about how to purge the stored 
scripts. I saw some oddly-named stacks in the App Browser that seemed 
relevant so I deleted those, and the next time I tried a test the whole 
IDE went down. So that wasn't it. :)



--
Jacqueline Landman Gay | jac...@hyperactivesw.com
HyperActive Software   | http://www.hyperactivesw.com

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Andrew Bell via use-livecode]

- The stack that caches object scripts is not being cleaned up  
properly between debugging sessions. This would mean that you would  
see the old script but what would be executing would be the new  
script. One way to test this might be to move a breakpoint statement  
from one line to another and see what line is highlighted in the  
debugger.


This is my vote for what is happening. I have noticed numerous times  
that stepping through lines I could see things happening in the  
standalone that were NOT the highlited line of code in the script  
editor at the time. For instance, a blank line (for readability of  
code) would execute some other line of code that was now 2 lines down  
due to a recent change in the script.


--Andrew Bell


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Monte Goulding via use-livecode]

> On 13 Sep 2018, at 5:08 am, J. Landman Gay via use-livecode 
>  wrote:
> 
> I've seen the same thing repeatedly. I wasn't able to work around the 
> problem. I'd like a way to manually force the debugger to unload itself and 
> reload the new script.

I’d really like to get to the bottom of this. FWIW the script that you see in 
the debugger actually comes over the socket from the app running on the device. 
As a result there’s really only two things I can think can be happening:

- When the deploy library (Test button) installs the app on the device it is 
silently failing and then launching the older version. We could rule this out 
if those of you that see this regularly had some way to know for sure that the 
new build actually made it to the device. Perhaps a different color background 
or something?

- The stack that caches object scripts is not being cleaned up properly between 
debugging sessions. This would mean that you would see the old script but what 
would be executing would be the new script. One way to test this might be to 
move a breakpoint statement from one line to another and see what line is 
highlighted in the debugger.

Cheers

Monte
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[J. Landman Gay via use-livecode]

I've seen the same thing repeatedly. I wasn't able to work around the 
problem. I'd like a way to manually force the debugger to unload itself 
and reload the new script.


There is a bug report you can add to:


On 9/12/18 8:40 AM, Sannyasin Brahmanathaswami via use-livecode wrote:

-- Shut down LC
-- Android phone is connected  by USB
-- install in app on Android
-- fix a script (change a typo)
-- save the script; save the stack
-- From the test menu, choose Android
-- build the app; install on device
-- the debugged breaks on the script that I just saved, and it shows the
script at it was *before* change the typo!
-- Leave debug mode, try to change the script again,
-- try to save LC give an error message so on cannot file the path to
save the script

The above is just one scenario: after running debug on Android, you have
no clue whether you are editing a "phantom script" or the "real library"
on in your hard drive. Or you are ending a script the  lies in "depths
of the standalone maker" even though the app is installed on the phone.

In some scenarios, like the above, you feel *certain* you are editing
the "real" script. (you can tell by opening BBEdit) You delete the app
from the device, run the test again: the debugger shows that error that
you just fixed!

It is getting tiring to have to shut down LC and reboot everytime you
test. But that seem to be only way that there is not "latent" memory of
the script, as it was, broken, in the bowels of LC, even you have fixed
in it disk.

Can any provide a rock solid procedure to be *sure* that  SA *always*
get a corrected version of any script?

BR



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode




--
Jacqueline Landman Gay | jac...@hyperactivesw.com
HyperActive Software   | http://www.hyperactivesw.com

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote Debugger Anomalies

[Sannyasin Brahmanathaswami via use-livecode]

Hmm I just go a test stack from Elanor... never open before

It has the debug script when she ran it on android saved into SA builder!

I went to edit it and get unusual ID

536434756567-36465 (etc.)

?




On 9/12/18 3:40 AM, Sannyasin Brahmanathaswami via use-livecode wrote:
> -- Shut down LC
> -- Android phone is connected  by USB
> -- install in app on Android
> -- fix a script (change a typo)
> -- save the script; save the stack
> -- From the test menu, choose Android
> -- build the app; install on device
> -- the debugged breaks on the script that I just saved, and it shows the
> script at it was *before* change the typo!
> -- Leave debug mode, try to change the script again,
> -- try to save LC give an error message so on cannot file the path to
> save the script
>
> The above is just one scenario: after running debug on Android, you have
> no clue whether you are editing a "phantom script" or the "real library"
> on in your hard drive. Or you are ending a script the  lies in "depths
> of the standalone maker" even though the app is installed on the phone.
>
> In some scenarios, like the above, you feel *certain* you are editing
> the "real" script. (you can tell by opening BBEdit) You delete the app
> from the device, run the test again: the debugger shows that error that
> you just fixed!
>
> It is getting tiring to have to shut down LC and reboot everytime you
> test. But that seem to be only way that there is not "latent" memory of
> the script, as it was, broken, in the bowels of LC, even you have fixed
> in it disk.
>
> Can any provide a rock solid procedure to be *sure* that  SA *always*
> get a corrected version of any script?
>
> BR
>

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Remote Debugger Anomalies

[Sannyasin Brahmanathaswami via use-livecode]

-- Shut down LC
-- Android phone is connected  by USB
-- install in app on Android
-- fix a script (change a typo)
-- save the script; save the stack
-- From the test menu, choose Android
-- build the app; install on device
-- the debugged breaks on the script that I just saved, and it shows the
script at it was *before* change the typo!
-- Leave debug mode, try to change the script again,
-- try to save LC give an error message so on cannot file the path to
save the script

The above is just one scenario: after running debug on Android, you have
no clue whether you are editing a "phantom script" or the "real library"
on in your hard drive. Or you are ending a script the  lies in "depths
of the standalone maker" even though the app is installed on the phone.

In some scenarios, like the above, you feel *certain* you are editing
the "real" script. (you can tell by opening BBEdit) You delete the app
from the device, run the test again: the debugger shows that error that
you just fixed!

It is getting tiring to have to shut down LC and reboot everytime you
test. But that seem to be only way that there is not "latent" memory of
the script, as it was, broken, in the bowels of LC, even you have fixed
in it disk.

Can any provide a rock solid procedure to be *sure* that  SA *always*
get a corrected version of any script?

BR



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode