README.hardware

README.hardware

[Gary Thomas]

This file seems quite out of date, in fact it's not been updated
for 9 months :-(

I don't know enough details to suggest a patch, but I think it
would be useful to update it to reflect "today's Poky/Yocto" world,
especially with the new visibility of the project.

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

Re: README.hardware

[Joshua Lock]

On Wed, 2010-12-08 at 06:47 -0700, Gary Thomas wrote:
> This file seems quite out of date, in fact it's not been updated
> for 9 months :-(
> 
> I don't know enough details to suggest a patch, but I think it
> would be useful to update it to reflect "today's Poky/Yocto" world,
> especially with the new visibility of the project.
> 

I agree, although I'm also wondering if it would be better to have this
information on the wiki where it can be more easily maintained by the
community.

Or a balance? any machine in meta/conf/machine documented in
README.hardware and anything else on the wiki.

Thoughts?

Cheers,
Joshua
-- 
Joshua Lock
        Intel Open Source Technology Centre

Re: README.hardware

[Gary Thomas]

On 12/09/2010 12:08 PM, Joshua Lock wrote:
> On Wed, 2010-12-08 at 06:47 -0700, Gary Thomas wrote:
>> This file seems quite out of date, in fact it's not been updated
>> for 9 months :-(
>>
>> I don't know enough details to suggest a patch, but I think it
>> would be useful to update it to reflect "today's Poky/Yocto" world,
>> especially with the new visibility of the project.
>>
>
> I agree, although I'm also wondering if it would be better to have this
> information on the wiki where it can be more easily maintained by the
> community.
>
> Or a balance? any machine in meta/conf/machine documented in
> README.hardware and anything else on the wiki.
>

Wikis are great, but I tend to think of the Poky "checkout" tree
as self-sufficient.  It has the full manual, etc, already so forcing
the user to have to go to the Wiki for such information (which is
really only overview anyway) would not be ideal.  A combination
of hard files and Wiki would be OK as long as the README.hardware
at least mentions the "well sanctioned" ports, which right now
it does not.

I also tend to err on the side of my [sadly] many customers
that may have to work in non-internet-connected environments,
so making them have to get to a Wiki to find such answers is
not really acceptable.

-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

Re: README.hardware

[Tom Zanussi]

On Thu, 2010-12-09 at 11:18 -0800, Gary Thomas wrote:
> On 12/09/2010 12:08 PM, Joshua Lock wrote:
> > On Wed, 2010-12-08 at 06:47 -0700, Gary Thomas wrote:
> >> This file seems quite out of date, in fact it's not been updated
> >> for 9 months :-(
> >>
> >> I don't know enough details to suggest a patch, but I think it
> >> would be useful to update it to reflect "today's Poky/Yocto" world,
> >> especially with the new visibility of the project.
> >>
> >
> > I agree, although I'm also wondering if it would be better to have this
> > information on the wiki where it can be more easily maintained by the
> > community.
> >
> > Or a balance? any machine in meta/conf/machine documented in
> > README.hardware and anything else on the wiki.
> >
> 
> Wikis are great, but I tend to think of the Poky "checkout" tree
> as self-sufficient.  It has the full manual, etc, already so forcing
> the user to have to go to the Wiki for such information (which is
> really only overview anyway) would not be ideal.  A combination
> of hard files and Wiki would be OK as long as the README.hardware
> at least mentions the "well sanctioned" ports, which right now
> it does not.
> 
> I also tend to err on the side of my [sadly] many customers
> that may have to work in non-internet-connected environments,
> so making them have to get to a Wiki to find such answers is
> not really acceptable.
> 

I touched a bit on the current README.hardware in this posting, which
suggests converting each machine into a BSP layer:

https://lists.yoctoproject.org/pipermail/poky/2010-December/001104.html

In that scheme, the individual sections in README.hardware would be
moved into their own machine-specific READMEs for each BSP, keeping
everything together.  The individual READMEs could also be made
separately available on the 'BSP downloads' page and/or wiki for
convenience.  Does that sound reasonable?

Thanks,

Tom

Re: README.hardware

[Joshua Lock]

On Thu, 2010-12-09 at 13:35 -0600, Tom Zanussi wrote:
> On Thu, 2010-12-09 at 11:18 -0800, Gary Thomas wrote:
> > On 12/09/2010 12:08 PM, Joshua Lock wrote:
> > > On Wed, 2010-12-08 at 06:47 -0700, Gary Thomas wrote:
> > >> This file seems quite out of date, in fact it's not been updated
> > >> for 9 months :-(
> > >>
> > >> I don't know enough details to suggest a patch, but I think it
> > >> would be useful to update it to reflect "today's Poky/Yocto" world,
> > >> especially with the new visibility of the project.
> > >>
> > >
> > > I agree, although I'm also wondering if it would be better to have this
> > > information on the wiki where it can be more easily maintained by the
> > > community.
> > >
> > > Or a balance? any machine in meta/conf/machine documented in
> > > README.hardware and anything else on the wiki.
> > >
> > 
> > Wikis are great, but I tend to think of the Poky "checkout" tree
> > as self-sufficient.  It has the full manual, etc, already so forcing
> > the user to have to go to the Wiki for such information (which is
> > really only overview anyway) would not be ideal.  A combination
> > of hard files and Wiki would be OK as long as the README.hardware
> > at least mentions the "well sanctioned" ports, which right now
> > it does not.
> > 
> > I also tend to err on the side of my [sadly] many customers
> > that may have to work in non-internet-connected environments,
> > so making them have to get to a Wiki to find such answers is
> > not really acceptable.
> > 
> 
> I touched a bit on the current README.hardware in this posting, which
> suggests converting each machine into a BSP layer:
> 
> https://lists.yoctoproject.org/pipermail/poky/2010-December/001104.html
> 
> In that scheme, the individual sections in README.hardware would be
> moved into their own machine-specific READMEs for each BSP, keeping
> everything together.  The individual READMEs could also be made
> separately available on the 'BSP downloads' page and/or wiki for
> convenience.  Does that sound reasonable?

This sounds reasonable to me.

Cheers,
Joshua
-- 
Joshua Lock
        Intel Open Source Technology Centre

README.hardware

[James Abernathy]

[-- Attachment #1: Type: text/plain, Size: 395 bytes --]

Question on the subject document:

In the atom-pc platform section, the Hard Disk option mentions a bitbake
target of 'core-image-minimal-directdisk'.  This target is not listed as an
option for bitbake after 'source oe-init-build-env' and if you try to use
it, the bake will fail at some point due to invalid target.

Am I missing something or is this a documentation problem.

Jim A

[-- Attachment #2: Type: text/html, Size: 490 bytes --]

Re: README.hardware

[Paul Eggleton]

On Wednesday 09 November 2011 09:34:40 James Abernathy wrote:
> In the atom-pc platform section, the Hard Disk option mentions a bitbake
> target of 'core-image-minimal-directdisk'.  This target is not listed as an
> option for bitbake after 'source oe-init-build-env' and if you try to use
> it, the bake will fail at some point due to invalid target.
> 
> Am I missing something or is this a documentation problem.

This is a documentation problem - for 1.1 there are no -directdisk or -live 
images; instead this is done by adding "live" to IMAGE_FSTYPES which is done 
automatically if MACHINE is set to atom-pc, so all you need to do is build the 
image you want (e.g. core-image-minimal) and you will also get a corresponding 
live image.

I noticed that we missed updating this document the other day and am putting 
together a patch for README.hardware to fix it.

Cheers,
Paul

-- 

Paul Eggleton
Intel Open Source Technology Centre