[oe] {RFC] console-image.bb

Koen Kooi k.kooi at student.utwente.nl
Wed Feb 17 19:12:12 UTC 2010


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 17-02-10 19:58, Frans Meulenbroeks wrote:
> 2010/2/17 Koen Kooi <k.kooi at student.utwente.nl>:
> On 30-01-10 11:47, Rolf Leggewie wrote:
>>>> Frans Meulenbroeks wrote:
>>>>> Wouldn't it be cleaner to rename ANGSTROM_EXTRA_INSTALL to e.g.
>>>>> DISTRO_EXTRA_INSTALL ?
>>>>
>>>> This has been discussed before, but not changed.  I think it made sense
>>>> back then and I think it makes sense now.
>>>>
>>>> I believe there is even a better name than DISTRO_EXTRA_INSTALL to be
>>>> found for the variable.  Maybe IMAGE_EXTRA_INSTALL?
> 
> of all the ways to solve it, you just had to go out of your way to annoy
> people.
> 
>> I BEG YOUR PARDON
> 
>> I did not make the change, actually I have not even seen it. I think
>> you owe me an apology for this.
>> I also would like to get the opinion of the TSC on the above accusation.

Please pay attention, I responded to rolfs mail, not yours

> 
> A simple IMAGE_EXTRA_INSTALL += ANGSTROM_EXTRA_INSTALL would have done
> the job, but no, lets create more work for people, break existing docs
> and especially ignore what the maintainer of the recipes said.
> 
> Way to go
> 
>> FYI: I popped a question three weeks ago.
>> Today Rolf asked me about the status in #oe.
>> As you can see above 4 people (Rolf, Mickey, Phil and me were in
>> favour of the proposal. In irc mckoan also backed the proposal.
>> There was one person opposed to it: you.

And funnily enough I'm the maintainer, so I get the final say.

>> Rolf asked whether he could make the change. As I was about to leave I
>> agreed that he made the change.

IRC isn't worth anything when it comes to decision making.

>> If you want it to be an angstrom recipe, please name it as such (or,
>> even better: start your own angstrom tree or branch).

There are multiple things wrong with the commit:

1) it silently changes the outcome of 'bitbake console-image' if you
have ANGSTROM_EXTRA_INSTALL set

2) It doesn't mention in the commit message that 1) happens

3) It invalidates docs without any hint

4) It wasn't discussed on the mailinglist, but on a medium where I'm not
present

5) the maintainer of the recipe wasn't informed of the change

With this change it seems angstrom developers have to update docs, deal
with user complaints, update sample config files, etc, etc, just because
Rolf jumped the gun. I find that unacceptable.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD4DBQFLfD+MMkyGM64RGpERAsHcAJ0TvgFEzIULXdGV3vTl1MdduSDDugCY3VgD
Ut2LwB0i1AH9GMz7YedxjA==
=F5i2
-----END PGP SIGNATURE-----





More information about the Openembedded-devel mailing list