It's in 4.0.2 and 3.2.2 as far as I know.  I've seen it's fixed in 4.0.3 and 3.2.3 and that's been awhile ago so it should probably be getting close to being released I would think but I've got no idea how they do their releases here for the core.  I assume it's being tested too.

The error is most definitely in third party modules also.  Those will probably need to be fixed by whoever writes the modules or if you have the source it could be checked now.

It's not really a logic error just more of a coding style that normally probably wouldn't effect anything at all but in this case there is just a tiny sliver of time where is possible for an object to be null and cause a section of code to be run that generates the error if it's not acting on something.

I made the changes in 3.2.2 for myself but since I don't run anything live on .Net 2.0 yet I haven't done anything with it yet.  You'll probably be safe until the official release is out.

Cheers,

Hi eveyone

I'm expriencing this problem too with DataCache.GetCache(). Okay, so reorganizing the code, removes some errors due to time problems. But I think the problem is beyond only reorganizing code.

I beleive that DataCache.GetCache() and SetCache() are the most called methods in the whole Core. But, strangely, these are not multi-threading safe. Check this code:

.... DataCache.SetCache(key, obj);

Object obj2 = DataCache.GetCache(key);

return obj2; // Most of the time, this is null

....

So what is happening is the SetCache method is not finishing synchrounosly and safely, when GetCache() tries to get that object , and whoooops no object !!!

So what I'm suggesting is to make the DataCache multi-thread safe.

Nice work guys