[oe] [oe-commits] org.oe.dev SlugOS: ixp4xx-kernel - added 2.6.23.8 recipe and supporting files. (Note that

[Rod Whitby]

Leon Woestenberg wrote:
> On Nov 23, 2007 7:24 PM, Mike (mwester) <mwester at dls.net> wrote:
>> The set of developers that will actually be using that recipe is
>> extraordinarily small (two at present), so the relative benefit of such
>>
> Sorry, this seems to be some misconception amongst the nslu2 
> developers. The use base for the mainline bleeding edge kernels is
> much larger*. I assume that every active board developer want to try
> and verify the new kernels against his hardware *and* is using
> OpenEmbedded to do so ;-)

No misconception at all.  In fact, that's the very reason why the
ixp4xx-kernel.inc file exists - to make it very easy to test the
absolute latest bleeding edge kernel.

For example, just by naming your .bb file ixp4xx-kernel_2.6.24-rc3-git1
(which is about as bleeding edge as you can get right now), you get
exactly that version built, without having to do anything else in the
SRC_URI field.  So tracking the latest -gitNN bleeding edge checkin by
Linus (or by Alan with the -ac series, or by Andrew with the -mm series)
is as simple as changing the name of the .bb file.

This is all clearly explained in the voluminous comments at the top of
the ixp4xx-kernel.inc file.

Of course Koen didn't write it, so instead of reading the comments to
understand what it does, he says the following:

Koen Kooi wrote:
> I would have fixed it myself, but the huge block of
> python magic in ixp4xx-kernel.inc scared me away.

Which is pretty much par for the course for Koen ... anything he didn't
write is by definition substandard and shouldn't be used.  He'll
probably even reply to this to explain why all this ixp4xx-kernel.inc
functionality has been implemented the wrong way and doesn't follow the
unwritten "spirit" of OE ...

-- Rod