Addressing more sysvinit bugs
Jesse Smith
jessefrgsmith at yahoo.ca
Sat Nov 3 00:30:23 GMT 2018
On 11/2/18 9:19 PM, Adam Borowski wrote:
> On Fri, Nov 02, 2018 at 05:58:07PM -0300, Jesse Smith wrote:
>>> This appears to be a bit late for Debian Buster, at least when I look at
>>> the freeze timeline¹.
>>
>> Agreed, 2.92 probably won't make it into Buster. However, I am hopeful
>> that 2.91 will, which upstreams several Debian patches and fixes several
>> bugs compared to the older sysvinit 2.88 version.
>>
>> Since almost everything in the 2.92 branch at the moment is a bug fix
>> for Debian, perhaps those fixes can be backported in the Debian 2.91
>> package? I'm not 100% sure of the Debian policy on backporting fixes.
>
> Do you have any disruptive or controversial changes already in to-be-2.92?
> If not, what about making a smaller release sooner? It's not like git tags
> or archive uploads are very costly...
>
I wouldn't say there is anything disruptive in the 2.92 branch. There
are some minor behaviour changes (disabling shutdown PC speaker beep,
allowing last to display long usernames, enabling a "quiet" mode for
shutdown). But nothing really big.
I'm not opposed to publishing a release little sooner to make the Buster
freeze as long as we can still fit in a couple of weeks of beta testing
to make sure nothing broke during the 2.92 development phase.
When would sysvinit-2.92 need to be released in order to be included in
Buster? I know Debian has its own sequence to go through to get a new
package from Unstable into Testing.
- Jesse
More information about the Debian-init-diversity
mailing list