Re: [platform-debug-dev] Configuring the padded string in a memory rende

[Samantha Chan <chanskw@xxxxxxxxxx>]

Hi -

Thanks Mikhail for giving me the feedback.  I would work on providing a
global preference for the padded string.  If you rendering does not wish to
use the padded string from this preference, you can implement your
rendering to use a different string.  It is up to your implementation to
honour this setting.

As for memoryByte flags, although it makes sense to treat an invalid byte
as read-only, I am inclined to keep them separated.  I do not want to
enforce such implication on the API.  I rather have clients explicitly tell
me if a byte is writable or not.  Some clients may complain in the future
if they come across with a platform that allows them to modify
unretrievable memory. Although it may seem unlikely, but I would like to
keep that option opened. Otherwise, we will have to break API compatibility
if we want to allow invalid bytes to be writable in the future.

I am also looking at your proposal about changing IMemoryBlockExtension's
#getBytesFromAddress and #getBytesFromOffset.  I will keep you posted on
that.

Thanks
Samantha




                                                                           
             "Mikhail                                                      
             Khodjaiants"                                                  
             <mikhailk@xxxxxxx                                          To 
             >                         <platform-debug-dev@xxxxxxxxxxx>    
             Sent by:                                                   cc 
             platform-debug-de                                             
             v-admin@eclipse.o                                     Subject 
             rg                        Re: [platform-debug-dev]            
                                       Configuring the padded string in a  
                                       memory rendering                    
             01/12/2005 02:50                                              
             PM                                                            
                                                                           
                                                                           
             Please respond to                                             
             platform-debug-de                                             
                     v                                                     
                                                                           
                                                                           




Hi Samantha,

I would definitely prefer to have one global preference, but I can't rule
out the possiblity that some renderings would need a padding string
different than default. So far I haven't ran into this problem, but I
haven't started to work on the disassembly rendering yet.

Mikhail

BTW I have a question related to the MemoryByte flags. If byte is not
VALID,
shouldn't it been treated as READONLY by the Memory view?

----- Original Message -----
From: "Samantha Chan" <chanskw@xxxxxxxxxx>
To: <platform-debug-dev@xxxxxxxxxxx>
Sent: Tuesday, January 11, 2005 3:16 PM
Subject: [platform-debug-dev] Configuring the padded string in a memory
rendering


>
>
>
>
> Hi -
>
> There is a method called getPaddedString in
> IMemoryBlockExtensionRetrieval.
> The string returned by this method call is used for rendering memory
bytes
> that are not retrievable for display.   As noted in previous discussions,
> this method should be part of Debug UI instead of Debug Core.
>
> However, looking at this method again. we feel that replacing this method
> with an user preference is more appropriate.  So, the proposal is to
> remove
> this method from IMemoryBlockExtensionRetrieval. Instead,  Debug platform
> will provide a global user preference to allow user to configure this
> setting.  The setting is defaulted to "??".  Clients who implements
> IMemoryRendering will have to consult this preference when rendering a
> memory block.  All of the bytes that are marked as invalid will be
> represented by this string specified in the user preference.
>
> So, my questions are:
> * Are you ok with having just one global padded string preference for all
> rendering types?
> * or Would you prefer to configure this setting programmatically via the
> model presentation or some other interface?
>
> Please let me know if you have any comments / questions to this proposal.
>
> Thanks
>
> Samantha
>
>
>
> _______________________________________________
> platform-debug-dev mailing list
> platform-debug-dev@xxxxxxxxxxx
> http://dev.eclipse.org/mailman/listinfo/platform-debug-dev
>

_______________________________________________
platform-debug-dev mailing list
platform-debug-dev@xxxxxxxxxxx
http://dev.eclipse.org/mailman/listinfo/platform-debug-dev