[Libosinfo] [osinfo-db][PATCH] opensuse, install-script: Skip second stage

[Fabiano Fidêncio]

On Thu, May 31, 2018 at 11:11 AM, Christophe Fergeau
<cfergeau at redhat.com> wrote:
> On Sat, May 26, 2018 at 10:29:38PM +0200, Fabiano Fidêncio wrote:
>> As we don't do anything in the second stage, we can just skip it
>> entirely without any issues.
>>
>> Signed-off-by: Fabiano Fidêncio <fabiano at fidencio.org>
>> ---
>>  data/install-script/opensuse.org/opensuse-autoyast-desktop.xml.in | 1 +
>>  1 file changed, 1 insertion(+)
>>
>> diff --git a/data/install-script/opensuse.org/opensuse-autoyast-desktop.xml.in b/data/install-script/opensuse.org/opensuse-autoyast-desktop.xml.in
>> index 6554eb6..537c266 100644
>> --- a/data/install-script/opensuse.org/opensuse-autoyast-desktop.xml.in
>> +++ b/data/install-script/opensuse.org/opensuse-autoyast-desktop.xml.in
>> @@ -144,6 +144,7 @@
>>                <ask-list config:type="list"/>
>>                <mode>
>>                  <confirm config:type="boolean">false</confirm>
>> +                <second_stage config:type="boolean">false</second_stage>
>
> Are these 'config:type' parameters something which is exposed to the
> library user, right? Or Are they something purley internal?

Those are public. They're used for specialized scripts, they're still
supported we just never used it.
It's going to work without issues with the Leap series. For the older
ones, we don't even support express installation.

> If they are public, what happens in libosinfo if we try to set a
> non-existent parameter? In other word, should we consider this an
> API/ABI break, or are we fine with dropping it?

We're totally fine dropping it.
If we try to set a non-existent parameter (or pattern) autoyast is
smart enough to not break the installation because of that.

>
> Christophe



-- 
Fabiano Fidêncio