Andre Poenitz wrote:
But that does not solve my current problem of having to call a base class'
handlers manually in a derived class as far as I can see. Or does it?
I don't understand your problems, but maybe they are local to mathed
and it's therefore I don't understand them. Normal insets
On Fri, Aug 16, 2002 at 10:39:31AM +0200, Juergen Vigna wrote:
Would you care to explain inside which step you have a problem?
It was a problem concerning inheritance, not nesting. And outer world
example would be CommandInset and LabelInset I believe.
With Jean-Marc's suggestion to call the
Andre Poenitz wrote:
> But that does not solve my current problem of having to call a base class'
> handlers manually in a derived class as far as I can see. Or does it?
I don't understand your problems, but maybe they are local to mathed
and it's therefore I don't understand them. Normal insets
On Fri, Aug 16, 2002 at 10:39:31AM +0200, Juergen Vigna wrote:
> Would you care to explain inside which step you have a problem?
It was a problem concerning inheritance, not nesting. And outer world
example would be CommandInset and LabelInset I believe.
With Jean-Marc's suggestion to call the
John == John Levon [EMAIL PROTECTED] writes:
John On Wed, Aug 14, 2002 at 07:10:34PM +0200, Lars Gullik Bjønnes
John wrote:
| approach with a silly patch a while ago ?
Yes, did I shoot it down?
John I think we all just um'd and ah'd for a bit ...
I for one liked it, except of the concern
On Thu, Aug 15, 2002 at 10:17:55AM +0200, Jean-Marc Lasgouttes wrote:
John I think we all just um'd and ah'd for a bit ...
I for one liked it, except of the concern about performance.
Performance is not critical unless proven otherwise ;-)
Anyway, if anything from that patch should survive
> "John" == John Levon <[EMAIL PROTECTED]> writes:
John> On Wed, Aug 14, 2002 at 07:10:34PM +0200, Lars Gullik Bjønnes
John> wrote:
>> | approach with a silly patch a while ago ?
>>
>> Yes, did I shoot it down?
John> I think we all just um'd and ah'd for a bit ...
I for one liked it,
On Thu, Aug 15, 2002 at 10:17:55AM +0200, Jean-Marc Lasgouttes wrote:
> John> I think we all just um'd and ah'd for a bit ...
>
> I for one liked it, except of the concern about performance.
Performance is not critical unless proven otherwise ;-)
Anyway, if anything from that patch should
I just wondered: Why is it sensible to have the dispatch switches in each
inset?
As far as I can tell, the same functionality can achieved by virtual
functions for individual LFUNs and a single big switch in the inset base
class.
The only disadvantage I see is that we'll have 100+ additional
Andre Poenitz [EMAIL PROTECTED] writes:
| I just wondered: Why is it sensible to have the dispatch switches in each
| inset?
| As far as I can tell, the same functionality can achieved by virtual
| functions for individual LFUNs and a single big switch in the inset base
| class.
| The only
On Wed, Aug 14, 2002 at 04:12:20PM +0200, Lars Gullik Bjønnes wrote:
| The only disadvantage I see is that we'll have 100+ additional member
| functions in the base class (which is certainly not Nice), but the total
| effort should be about the same.
Do we really want any of those approces?
Andre == Andre Poenitz [EMAIL PROTECTED] writes:
Andre But that does not solve my current problem of having to call a
Andre base class' handlers manually in a derived class as far as I
Andre can see. Or does it?
What you should do, I think is to have every InsetFoo::dispatch(...)
method call
On Wed, Aug 14, 2002 at 04:46:29PM +0200, Jean-Marc Lasgouttes wrote:
What you should do, I think is to have every InsetFoo::dispatch(...)
method call its parent dispatch on the 'default:' of the switch(). Do
you have to know explicitely the parent name for that in C++
Yes. But that's no
Andre == Andre Poenitz [EMAIL PROTECTED] writes:
Andre Anyway, the default: should solve my problem.
Except that it will cause problems every time the inheritence chain is
changed and one forgets to update dispatch().
JMarc
On Wed, Aug 14, 2002 at 04:57:19PM +0200, Jean-Marc Lasgouttes wrote:
Andre Anyway, the default: should solve my problem.
Except that it will cause problems every time the inheritence chain is
changed and one forgets to update dispatch().
Well... this should happen considerably less often
On Wed, Aug 14, 2002 at 04:12:20PM +0200, Lars Gullik Bjønnes wrote:
Wouldn't a system where the insets register LFUNS with the
dispatcher (which we would only need one of) be better?
Has time started running backwards ? Didn't I illustrate this exact
approach with a silly patch a while ago ?
John Levon [EMAIL PROTECTED] writes:
| On Wed, Aug 14, 2002 at 04:12:20PM +0200, Lars Gullik Bjønnes wrote:
|
| Wouldn't a system where the insets register LFUNS with the
| dispatcher (which we would only need one of) be better?
|
| Has time started running backwards ? Didn't I illustrate
On Wed, Aug 14, 2002 at 07:10:34PM +0200, Lars Gullik Bjønnes wrote:
| approach with a silly patch a while ago ?
Yes, did I shoot it down?
I think we all just um'd and ah'd for a bit ...
regards
john
--
It is unbecoming for young men to utter maxims.
- Aristotle
I just wondered: Why is it sensible to have the dispatch switches in each
inset?
As far as I can tell, the same functionality can achieved by virtual
functions for individual LFUNs and a single big switch in the inset base
class.
The only disadvantage I see is that we'll have 100+ additional
Andre Poenitz <[EMAIL PROTECTED]> writes:
| I just wondered: Why is it sensible to have the dispatch switches in each
| inset?
>
| As far as I can tell, the same functionality can achieved by virtual
| functions for individual LFUNs and a single big switch in the inset base
| class.
>
| The only
On Wed, Aug 14, 2002 at 04:12:20PM +0200, Lars Gullik Bjønnes wrote:
> | The only disadvantage I see is that we'll have 100+ additional member
> | functions in the base class (which is certainly not Nice), but the total
> | effort should be about the same.
>
> Do we really want any of those
> "Andre" == Andre Poenitz <[EMAIL PROTECTED]> writes:
Andre> But that does not solve my current problem of having to call a
Andre> base class' handlers manually in a derived class as far as I
Andre> can see. Or does it?
What you should do, I think is to have every InsetFoo::dispatch(...)
On Wed, Aug 14, 2002 at 04:46:29PM +0200, Jean-Marc Lasgouttes wrote:
> What you should do, I think is to have every InsetFoo::dispatch(...)
> method call its parent dispatch on the 'default:' of the switch(). Do
> you have to know explicitely the parent name for that in C++
Yes. But that's no
> "Andre" == Andre Poenitz <[EMAIL PROTECTED]> writes:
Andre> Anyway, the default: should solve my problem.
Except that it will cause problems every time the inheritence chain is
changed and one forgets to update dispatch().
JMarc
On Wed, Aug 14, 2002 at 04:57:19PM +0200, Jean-Marc Lasgouttes wrote:
> Andre> Anyway, the default: should solve my problem.
>
> Except that it will cause problems every time the inheritence chain is
> changed and one forgets to update dispatch().
Well... this should happen considerably less
On Wed, Aug 14, 2002 at 04:12:20PM +0200, Lars Gullik Bjønnes wrote:
> Wouldn't a system where the "insets" register LFUNS with the
> dispatcher (which we would only need one of) be better?
Has time started running backwards ? Didn't I illustrate this exact
approach with a silly patch a while
John Levon <[EMAIL PROTECTED]> writes:
| On Wed, Aug 14, 2002 at 04:12:20PM +0200, Lars Gullik Bjønnes wrote:
|
| > Wouldn't a system where the "insets" register LFUNS with the
| > dispatcher (which we would only need one of) be better?
|
| Has time started running backwards ? Didn't I
On Wed, Aug 14, 2002 at 07:10:34PM +0200, Lars Gullik Bjønnes wrote:
> | approach with a silly patch a while ago ?
>
> Yes, did I shoot it down?
I think we all just um'd and ah'd for a bit ...
regards
john
--
"It is unbecoming for young men to utter maxims."
- Aristotle
28 matches
Mail list logo
