Re: is it not a *requirement* that layers start with "meta-"?

["Robert P. J. Day" <rpjday@crashcourse.ca>]

On Thu, 8 Jan 2015, micke.prag@telldus.se wrote:

>
> Message: 6
> Date: Thu, 8 Jan 2015 08:31:58 -0500 (EST)
> From: "Robert P. J. Day" <rpjday@crashcourse.ca>
> To: Anders Darander <anders@chargestorm.se>
> Cc: Yocto discussion list <yocto@yoctoproject.org>,
> 	"peterengcomau001@adam.com.au" <peterengcomau001@adam.com.au>
> Subject: Re: [yocto] is it not a *requirement* that layers start with
> 	"meta-"?
> Message-ID: <alpine.LFD.2.11.1501080831200.31772@localhost>
> Content-Type: TEXT/PLAIN; charset=US-ASCII
>
> On Thu, 8 Jan 2015, Anders Darander wrote:
>             * Robert P. J. Day <rpjday@crashcourse.ca> [150105 18:09]:
>                   On Mon, 5 Jan 2015, Rifenbark, Scott M wrote:
>
>                         If anyone can locate some scripts or something that
>                         actually does require "meta-" to be prepended please let
>                         me know. That changes how we describe this.
>
>                   as an admittedly simple example, i ran across this in oe-core's
>                   scripts/contrib/documentation-audit.sh:
>
>                   for pkg in `bitbake -s | awk '{ print \$1 }'`; do
>
>                   ... snip ...
>
>                   if [[ "$pkg" =~ ^meta- || "$pkg" =~ ^packagegroup- || "$pkg" =~
>                   -image ]]; then # Skip meta, task and image recipes continue fi
>
>                   checking that the package name starts with "meta-" seems to pretty
>                   clearly suggest that it will be treated differently than if it
>                   doesn't. i'm sure no one is terribly worried about that particular
>                   script, but it does seem to show an example of this.
>
>             But that's checking *recipe* names, not layer names... Thus, you'll have to look
>             for another example. ;)
>
>   ah, quite so, so perhaps the "meta-" prefix is not a real
> requirement, just convention.
>
> rday
>
> --
>
> The layer "openembedded-core" does not start with "meta-".

  well, yes, but that's kind of a special case as it's the only
*required* layer. in any event, i apologize if i made a bigger deal
out of this than was warranted, i just thought the rules for layer
naming were a bit ambiguous and i wanted some clarification.

rday

-- 

========================================================================
Robert P. J. Day                                 Ottawa, Ontario, CANADA
                        http://crashcourse.ca

Twitter:                                       http://twitter.com/rpjday
LinkedIn:                               http://ca.linkedin.com/in/rpjday
========================================================================