On 05.05.2009 17:45, Steve Borho wrote:
> On Tue, May 5, 2009 at 10:30 AM, Adrian Buehlmann <[email protected]> wrote:
>> On 05.05.2009 16:54, Steve Borho wrote:
>>> On Tue, May 5, 2009 at 1:56 AM, Adrian Buehlmann <[email protected]> wrote:
>>>> On 05.05.2009 08:04, Steve Borho wrote:
>>>>> -b1 includes the python2x.dll in the library.zip file.  There's a 50%
>>>>> hg and GTK will be ok with that configuration, but issue #67 is mostly
>>>>> about removing python out of the shell extensions.  Moving the menu
>>>>> code to the RPC server means that python is not being loaded into
>>>>> every process that opens a file browser.
>>>> I am confused.
>>>>
>>>> Why is there any need to "move menu code" into a separate RPC server
>>>> process?
>>>>
>>>> TK's C++ shell extension work includes a context menu handler
>>>> programmed in C++.
>>>>
>>>> If we combine that with my C++ overlay code we have a complete
>>>> solution for context menu and overlay icons, all done in C++ in
>>>> a dll loaded into explorer (and loaded into applications, if they
>>>> use file open dialog).
>>>>
>>>> So there is no python left to be loaded into explorer
>>>> (or applications) any more.
>>> This gets back to Mark's original e-mail.  I don't want to have to
>>> maintain the context menus in C++ in perpetuity.  There's a lot of
>>> ugly details that are much easier to deal with in Python.  Having
>>> minimal builtin menus is fine for the short term, but eventually I'd
>>> like to have a Python process serve up the menus (over a pipe, so
>>> python doesn't infect the application).
>> This does not make much sense to me. I never heard of a software
>> that needs a server process just do something as simple as a
>> context menu.
>>
>> I don't think that maintaining/expanding C++ would be a problem here,
>> since adding/changing menu entries or icons would be trivial,
>> even for programmers that have never written a single line of
>> C++ before. I'm sure this could be organized as some script
>> like area in the C++ (or even moved into an include file).
> 
> There's more intelligence in our menu code than just building static
> menus.  We're adding/removing menu options based on the files selected
> and the state of the repository.  There's also i18n issues, and
> configurability (on crew you can pick which options are promoted to
> the top menu).

If I need a server process just for that, I can well waive such
features. And who said we can't configure the C++ shell extension?

But I'll stop banging that C++ drum now... (er, well we might even
read the menu strings from a resource-like file, but, whatever...)

>> But we will see how your "interim" step goes. I'm currently just
>> interested in that. Maybe I can then continue to use (fork?) it while
>> you take the-python-server-for-the-cmenu route later ;-)
> 
> Feel free to beat me to this interim step :)

We'll see.

------------------------------------------------------------------------------
The NEW KODAK i700 Series Scanners deliver under ANY circumstances! Your
production scanning environment may not be a perfect world - but thanks to
Kodak, there's a perfect scanner to get the job done! With the NEW KODAK i700
Series Scanner you'll get full speed at 300 dpi even with all image 
processing features enabled. http://p.sf.net/sfu/kodak-com
_______________________________________________
Tortoisehg-develop mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/tortoisehg-develop

Reply via email to