Why not leave it as it is? Thinking about it, I don't find it confusing. Other languages have similar concepts and have lived with it. Thoughts anyone? ...
... Second option: convert IoMyCPP.c to .cpp and downgrade the error. ... I'm sure you will... it can be tricky, but you have to find out how to do it only...
... Unfortunately, that would mean editing the Queue_inline.h file as well, since it has conversions from (void *) to (void **) and such... breaking the ...
... Success! I have one question though, the main.cpp file you have in that folder which has the following: #include "MyCPP.h" int main(int argc, char **argv)...
... WTG! Glad you did it! ... (...) ... You're right, the main.cpp file was only intended to test the plain C++ class, before writing the bindings. I didn't...
... Well I'm back with a little roadblock. I am trying to move the code found in your main.c into a C++ class, ala CIoScriptHost, so my engine can deal less...
... List_sizeTo_ ... Ok, this was easy to fix, I just ended up enclosing any #includes of Ioxxx.h files with extern "C" { ... } Links happily now :)...
1766
Daniel A. Koepke
dkoepke@...
Aug 3, 2003 12:53 am
... I think there are well-defined ways to interconnect C and C++ code that are a bit less error-prone. Given this class: class MyCPP { public: int getInt(); ...
... Hi Dustin, I'm and OSX user but there are folks on the Io mailing list that might be able to provide a precompiled windows binary. I'm forwarding your ...
Another possible solution: - don't set the proto slot on locals object - set the locals "setSlot" slot to Object setSlot (so assignment works) - change the...
... I forgot that that breaks iterators and some other stuff. The other option is to set the locals proto to some base Object, but that prevents overrides of...
... That worked. It seems to work for "make test" and is just as fast. Would anyone like to beta test this feature before I release it? (I'll also wait a few...
... (...) ... I agree with you that it's better, but perhaps a little less straightforward because of the struct MyCPP hack on the .h file. Anyway, yours is a...
I'll get it ready soon enough... if you want just the VM, I can compile it and make it available for download (I'm taking longer because of IoServer and...
... My personal preference is the circumfix |...| form, but maybe that just shows my Smalltalk bias. However, I prefer it as follows: Account |balance| = 10 ...
... Hi Steve That sounds promising. However, I don't quite understand the implications of this change. Slots are still being created automatically instead of...
... Hi Roberto I agree with you that the *concept* is not confusing. However, in practice the automatic creation of variables is a VERY error prone mechanism....
I was thinking support for := operators could be added by changing the parser to treat symbols that begin with a : as operators instead of messages. I don't...
... Hi Steve, While we're on the subject, I've been wanting to raise the following question. In Smalltalk a distinction is made between the following ...
... You can test for identity like this: if ( x uniqueId == y uniqueId ) then ( "they are identical" print ) Cheers, Steve OSX freeware and shareware:...
... Sorry, my mistake, I forgot that Io uses == for comparison and not = (it's a bit late at night for me). Thanks for the above information, it's actually...
... So I got := working. The only (minor?) problem is that you must use a space before the := to avoid the : being taken to be part of the message (see below)....
... I think it is safe too, as I have also never seen selectors beginning with ":". Cheers, -- Sébastien -- «Too old to be alternative, too alternative to be...
... The problem I see here is that we have a special syntax for a special operator, which once again breaks the overall consistency. Maybe we could consider...
... I'm open to that, but haven't seen any good proposals. Do you have any? Cheers, Steve OSX freeware and shareware: http://www.dekorte.com/downloads.html...
... I must say that I only quickly read the messages regarding the need for an ":=" operator, but what I understoof is that we need a way to declare that we...
... This would also require a special change to the parser. Right now the only symbols that act like operators are super and return. Cheers, Steve OSX freeware...
