opencog-dev team mailing list archive

[Cassio Pennachin <cassio@xxxxxxxxxxxxx>]

Trent,

> You admit that staging will accept checkins that break stuff.

No, I don't, at least to the extent it is possible to prevent it (see  
below).  I've said so many times today.

> So you're saying that Linas should just put up with his stuff being
> broken daily.. what do you want him to do?  Complain to whoever
> checked in the code that broke it and then just wait?   No.  Clearly
> you want him to fix the problems introduced by others instead of
> focusing on his work.

I have no idea where you got this conclusion from.  I've said the  
opposite of your interpretation many times today as well.

> Why is it so hard to ensure the unit tests are working before you
> check in?

It doesn't sound hard to me, which is why I think they should work  
before people commit and if they don't their commits can be rolled  
back.  This gets us close to "software level" working, although things  
can still be broken at the "AI level" of working, which can't be  
tested with just unit tests.

Cassio