On 8 March 2017 at 08:52, Philip Prindeville
<philipp_s...@redfish-solutions.com> wrote:
>
>> On Mar 5, 2017, at 8:06 AM, Matthias Schiffer
>> <mschif...@universe-factory.net> wrote:
>>
>> This could also avoid adding host UCI support to the build system, making
>> the whole patchset a lot smaller.
>
>
> Host support for UCI is useful to have anyway, for other reasons… not
> something to avoid doing (not clear if you were against the principle of
> having UCI be on the host, or just trying to keep this particular patchset
> down to the bare minimum).
>
> For instance, you could have platform-specific scripts that dynamically
> generate the initial configuration in the target image by doing UCI
> batch/add/commit sequences…
>
> -Philip
>
The usefulness of such availability only makes sense when it is going
to at least have an actual user. Feel free to bring it up in another
occasion if you find it necessary to achieve your goals there ;)
We always have a choice to introduce new features when there is no
other way around. But the addition will add to the maintenance
burden, and they will only be removed or replaced if we decide to make
a hard choice to abandon backward-compatibility and this is really
bad...
yousong
_______________________________________________
Lede-dev mailing list
Lede-dev@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/lede-dev
