> The final validation is quite simple: we freeze all files (tests, libs > and device files) and then check if it passes all tests. Yes, a lot of > work, but straight-forward and should not take too long. > If the /validated is for this, then you're right - we should not move > anything to this directory.
For now, for the very first release, yes, this is what I think. > But before the final validation, there is long process of creating and > improving files. If we assume that most of the changes are > improvements, I think it makes sense to move files linke blink - tests > for a specific chip - to /validated. It means 'it works, or at least, > it has worked'. "it has worked" also means "I don't know now if it still working" so it shouldn't be validated. This is mainly because device files are still changing a lot. For any other lib, IMHO *we* (that is more than one person) have to vote before moving things in validated. And BTW we have to *move* files, not *copy* them. > We could argue if the device file should be migrated too. I would not > do this since there is still work to do on those (i think, seeing they > are still updated regularly) and most of the features aren't tested > yet. The idea of validated is: if someone wants to work with tested file, he goes to validated. If he goes here, and there's just blink sample without any device files, it's not functional, he can't do anything with just a blink sample. What is in validated must be functional, thus device file should be put here too, BUT device files are still changing a lot, SO... it's too early for now IMHO > > This brings me to an other point: are we capable of supporting a > specific peripheral (like serial_hardware) on all pic's before we > concider it validated? I think this is a huge amount of work, and > with bugs for 18F chips* > > An option would be: > - blink goes to validated if it works > - generic libs (external and jal) go to validated if they work on at > least 2 different chips. > - periperal libs go to validated if they work on at least 2 different > chips. But if a periperal lib supports different versions of the > periperal, it should be tested on at least one version of each. > - device libs move to validated if they work with the relevant generic > and periperal libs. (Relevant means the lib should support the chip. > If serial_hardware has no 18F support, 18F device files can be moved > to validated). OK, I agree, nice criterias, we need to put this a the validation process wiki page. Cheers, Seb -- Sébastien LELONG http://www.sirloon.net http://sirbot.org --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "jallib" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/jallib?hl=en -~----------~----~----~----~------~----~------~--~---
