Andy Patterson <[email protected]> skribis:
> On Mon, 05 Dec 2016 21:48:36 +0100
> [email protected] (Ludovic Courtès) wrote:
>
>> > +(define-public sane-backends+hpaio
>>
>> Could you add a comment and a synopsis explaining what’s this is?
>>
>
> Sure, is the following patch ok?
Yup!
>> > + (package
>> > + (inherit sane-backends)
>> > + (name "sane-backends+hpaio")
>> > + (inputs
>> > + `(("hplip" ,(@ (gnu packages cups) hplip))
>> > + ,@(package-inputs sane-backends)))
>>
>> The closure size of this is 290 MiB whereas it’s 87 MiB for
>> ‘sane-backends’. I suppose that’s the reason to keep’em separated?
>>
>
> Yeah, and there's also the fact that hplip depends on sane-backends.
> But this is the reason I didn't change sane-backends into a private
> variable, and then make this the new sane-backends.
Oh right.
One last thing: should we renaming “sane-backends” to
“sane-backends-minimal”, and “sane-backends+hpaio” to “sane-backends”?
That way all scanners would work out of the box, at the expense of extra
disk usage, but that is roughly what patch 2/2 does anyway.
Thanks!
Ludo’.