[Rebase review] Hardy netbook-lpia branch rebase to 2.6.24-28.70

[Ike Panhc]

On 06/11/2010 09:25 PM, Stefan Bader wrote:
> On 06/11/2010 02:30 PM, Andy Whitcroft wrote:
>> On Fri, Jun 11, 2010 at 09:59:29AM +0200, Stefan Bader wrote:
>>> On 06/10/2010 03:01 PM, Ike Panhc wrote:
>>>> Hi,
>>>>
>>>> Please review for the rebase on Hardy netbook-lpia branch including lbm/lrm/lum.
>>>> The git tree is at the following place:
>>>>
>>>> git://kernel.ubuntu.com/ikepanhc/hardy-kernel/.git branch netbook-lpia and
>>>> tag NBK-Ubuntu-2.6.24-28.70netbook01
>>>>
>>>> git://kernel.ubuntu.com/ikepanhc/hardy-lbm/.git branch netbook-lpia and
>>>> tag NBK-Ubuntu-2.6.24-28.70netbook01
>>>>
>>>> git://kernel.ubuntu.com/ikepanhc/hardy-lrm/.git branch netbook-lpia and
>>>> tag NBK-Ubuntu-2.6.24.25-28.1netbook01
>>>>
>>>> git://kernel.ubuntu.com/ikepanhc/hardy-lum/.git branch netbook-lpia and
>>>> tag NBK-Ubuntu-2.6.24-28.70netbook01
>>>>
>>>>
>>>> What has been done list below.
>>>>
>>>> For Hardy kernel:
>>>>  * Rebase to tag Ubuntu-2.6.24-28.70
>>>>  * Disable LGUEST_GUEST for two reason: we dont need it and build failed on it.
>>>>  * Remove debian/control and debian/control.stub: Those are auto-generated files
>>>
>>> The last step is not wrong but unusual compared to the master tree. Usually we
>>> carry those files there and update them whenever we make an abi bump.
>>> You can go forward and remove them, but just need to keep in mind that you will
>>> need to call debian/rules clean now before you create source packages.
>>>
>>> I try to get you a better review, but it might get tomorrow while I am on the train.
>>
>> I think I suggested he rip out the built files as he had been bitten by
>> not updating them.  As we have ripped them out of all the other trees
>> there seems little chance of confusion.  Perhaps we should do the same
>> control-sectomy on the master branch too.
>>
>> -apw
> 
> I see that this makes the behaviour more like later releases. And maybe it works
> just ok (but then you should rip out kernel-versions next). But it feels for an
> unexplainable reason uneasy to have some parts removed to give similar behaviour
> and possibly have other pitfalls. And maybe confuse people being used to
> re-commit those files for Hardy. But that might be overcautious.
> 
> Stefan
> 

I check for which file will be touched after 'fdr clean'. They are control
control.stub and kernel-version.

If those file is removed, who want to build hardy master/netbook branch have to
'fdr clean' before building.

For a developing tree, I think it will be good to remove them. But since Hardy
is a maintaining tree. I will vote for keep it as usual (not to remove those
three files).

--
Ike Panhc