[oe] [PATCH 1/1] angstrom-2008.1.conf: add PREFERRED_PROVIDER_virtual/psplash -silence warning during parsing

[Otavio Salvador]

Koen Kooi <k.kooi at student.utwente.nl> writes:

> On 30-03-09 13:47, Otavio Salvador wrote:
>> Koen Kooi<k.kooi at student.utwente.nl>  writes:
>>
>>> On 26-03-09 01:10, Andrea Adami wrote:
>>>
>>>> +#Silence warning during parsing
>>>> +PREFERRED_PROVIDER_virtual/psplash ?= "psplash-angstrom"
>>>
>>> NACK, the 'psplash' package should get deleted and replaced with
>>> psplash-oe' or something. It was left in place to give people time to
>>> switch to the new way of installing branded psplash images.
>>
>> Hello Koen but I belive those are two different issues.
>
>> I added the virtual/psplash package to allow others to properly depends
>> on an alternative. Currently the only recipe using it is
>> initramfs-module-psplash but others may come.
>
> I don't think virtual/psplash is the way to go, but I don't have a
> better suggestion.

So what we should do? IMO if there's no better solution we use this
until we find one.

>> I fully agree that we should drop plain psplash recipe and have
>> people to use the branded one but I also think angstrom needs to
>> properly set the preferred version since it will then allow that
>> its users to use it without much hassle.
>
> Unfortunately this will end up with people spreading FUD like "can't
> change psplash logo in angstrom" just because they see that line in
> angstrom.conf.

Well, we could add a comment for that in the angstrom.conf file.

Something like:

# This enforces a specific brand to be used, in our case the angstrom
# one. In case you want to change the logo or something like that we
# suggest you to create another recipe (e.g: psplash-<brand>) for that
# and change this line according.
PREFERRED_VERSION_...

> But then again, people are still claiming angstrom is bloated and
> minimal isn't when minimal is currently angstrom.conf with sed
> s/angstrom/minimal/ applied to it...

I think this is another problem; even minimal isn't in a useful state
and it does need to be fixed, this doesn't fix the virtual dependency
problem.

So?

-- 
Otavio Salvador                  O.S. Systems
E-mail: otavio at ossystems.com.br  http://www.ossystems.com.br
Mobile: +55 53 9981-7854         http://projetos.ossystems.com.br