[Toaster] [review-request] adamian/20150707_bugs

[Barros Pena, Belen]

On 07/07/2015 17:27, "Damian, Alexandru" <alexandru.damian at intel.com>
wrote:

>Hi,
>
>
>I have some comments below,
>
>
>Alex
>
>On Tue, Jul 7, 2015 at 3:54 PM, Barros Pena, Belen
><belen.barros.pena at intel.com> wrote:
>
>
>
>On 07/07/2015 15:10, "Damian, Alexandru" <alexandru.damian at intel.com>
>wrote:
>
>>Hello,
>>
>>
>>I'm pushing two patches for review
>>
>>
>>- one is fixing 7955
>
>Thanks for the quick fix, Alex! I've tested this on master, and I've
>noticed a couple of things:
>
>1. The invalid data from the layer index causes warnings when importing
>the information. This might be because the debug mode is enabled, though,
>but I thought I'd bringing it up just in case.
>
>​This happens because data doesn't match what Toaster expect - Toaster
>has a bit stricter ​requirements than Layer Index. I would expect that
>the warning messages are helpful to the user, and they should not be
>obscured. If a particular user wishes to not see some of the messages,
>they can set up the debug level to something higher in settings.py. Or we
>can ship with a higher debug level by default, but I don't think we
>should silently ignore bad data.

Sure: I can see your point. I was only bringing it up, but I am not sure
what's best, to be honest: showing them or not. As you said, falling
silent doesn't sound right; on the other hand, the messages look a bit
alarming.


>
>2. I can see at least one recipe in the 'all recipes' table without a
>name. This particular one is provided by meta-ivi. You can add the layer
>and you get a build button, which you can click, although when you do so
>the build does not seem to start. I think we need to hide any recipes that
>do not have a name from the list. They are invalid, and should not be
>exposed to  users.
>
>​We can add such a check, of course, on imported data. I would say this
>is the object of a different bug report, though.

Done! https://bugzilla.yoctoproject.org/show_bug.cgi?id=7969

>
>Also, I think we are going to need to back port the fix to Fido.

Is back porting ok? Should we track this somehow?

Thanks!

Belén

>
>Cheers
>
>Belén
>
>>- one is fixing various issues highlighted by pylint
>>
>>
>>https://git.yoctoproject.org/cgit/cgit.cgi/poky-contrib/log/?h=adamian/20
>>1
>>50707_bugs
>>
>>
>>
>>​Can you please review ?
>>
>>
>>Cheers,
>>Alex​
>>
>>
>>
>>
>>--
>>Alex Damian
>>Yocto Project
>>
>>SSG / OTC
>>
>>
>>
>
>
>
>
>
>
>
>
>
>
>-- 
>Alex Damian
>Yocto Project
>
>SSG / OTC 
>
>
>
>