Re: Message forwarding...

[Barry McMullin]

At 15:15 19/02/97 -0800, Scott Christley <address@hidden> wrote:
>At 03:17 PM 2/19/97 +0000, Barry McMullin wrote:
[...]
>>I tried it and it works.  It's not a very severe test because
>>my forwarded messages all have trivial interfaces (at most an
>>int argument, and an int return), but it does prove the principle.
>
>Warning! Warning! Danger! Danger!  (Wave your arms around) :-)
>
>Forwarding is great an wonderful but there are definitely problems depending
>upon what version of the ObjC runtime that you have in your GCC.  The most
>recent snapshot that I released back in 960906 still has a problem, so try
>both of the programs to test; this only relates to i386 machines though as
>far as I know.  

Thanks to Scott for the extra info.  For myself, I had already decided to
revert to "explicit" forwarding (where each message to be
forwarded has a corresponding explicit method in the forwarding
object, which just relays it).  I reckon this is "guaranteed" to
work robustly, and also avoids a couple of ancillary difficulties
I was having (e.g. gcc complaining that all methods of an adopted
protocol were not being implemented...).

Anyway, thanks again for the advice,

Barry.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
| Barry McMullin, Autonomous Systems Group,  |    address@hidden |
| School of Electronic Engineering,          |  Voice: +353-1-704-5432 |
| Dublin City University, Dublin 9, IRELAND. |  FAX:   +353-1-704-5508 |
| http://www.eeng.dcu.ie/~mcmullin/home.html |                         |
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++