Summary/Minutes from today's FESCo Meeting (2016-07-15)

[Josh Boyer]

===================================
#fedora-meeting: FESCO (2016-07-15)
===================================


Meeting started by jwb at 16:00:06 UTC. The full logs are available at
https://meetbot.fedoraproject.org/fedora-meeting/2016-07-15/fesco.2016-07...
.



Meeting summary
---------------
* init process  (jwb, 16:00:06)

* #1593 cryptkeeper package is unmaintained / as well as upstream  (jwb,
  16:02:15)
  * LINK: https://fedorahosted.org/fesco/ticket/1593   (jwb, 16:02:15)
  * AGREED: Retire cryptkeeper (+8, 0, 0)  (jwb, 16:05:41)
  * jwb to retire cryptkeeper  (jwb, 16:05:54)

* #1568 F25 Self Contained Changes  (jwb, 16:06:14)
  * LINK: https://fedorahosted.org/fesco/ticket/1568   (jwb, 16:06:14)
  * AGREED: All Self-Contained Changes approved (+8, 0, 0)  (jwb,
    16:12:15)
  * golang may need to be a system-wide change in the future  (jwb,
    16:14:02)

* #1594 F25 System Wide Change: IBus Emoji Typing  (jwb, 16:14:19)
  * LINK: https://fedorahosted.org/fesco/ticket/1594   (jwb, 16:14:20)
  * AGREED: Defer approval to allow more discussion on implementation
    and dependencies (+8, 0, 0)  (jwb, 16:20:25)

* #1595 F25 System Wide Change: Automatic Provides for Python RPM
  Packages  (jwb, 16:21:01)
  * LINK: https://fedorahosted.org/fesco/ticket/1595   (jwb, 16:21:01)
  * AGREED: Approved Automatic Provides for Python RPM Packages (+8, 0,
    0)  (jwb, 16:24:15)
  * follow up on packaging guidelines draft is needed  (jwb, 16:24:45)

* #1597 F25 System Wide Change: Ruby on Rails 5.0  (jwb, 16:25:28)
  * LINK: https://fedorahosted.org/fesco/ticket/1597   (jwb, 16:25:28)
  * AGREED: Approved Ruby on Rails 5.0 (+8, 0 ,0)  (jwb, 16:27:22)

* #1598 F25 System Wide Change: SSSD fast cache for local users  (jwb,
  16:27:33)
  * LINK: https://fedorahosted.org/fesco/ticket/1598   (jwb, 16:27:34)
  * AGREED: Approved SSSD fast cache for local users (+7, 0, 0)  (jwb,
    16:30:21)

* #1599 F25 System Wide Change: Unicode 9.0 support  (jwb, 16:30:36)
  * LINK: https://fedorahosted.org/fesco/ticket/1599   (jwb, 16:30:36)
  * AGREED: Approved Unicode 9.0 support (+7, 0, 0)  (jwb, 16:35:22)

* #1590 F25 System Wide Change: GHC 7.10  (jwb, 16:35:53)
  * LINK: https://fedorahosted.org/fesco/ticket/1590   (jwb, 16:35:54)
  * AGREED: Approved GHC 7.10 (+8, 0, 0)  (jwb, 16:38:02)

* #1601 F25 System Wide Change: Fedora Media Writer as Primary
  Downloadable  (jwb, 16:38:18)
  * LINK: https://fedorahosted.org/fesco/ticket/1601   (jwb, 16:38:18)
  * AGREED: Approved Fedora Media Writer as Primary Downloadable (+8, 0,
    0)  (jwb, 16:44:54)

* #1600 F25 System Wide Change: KillUserProcesses=yes by default  (jwb,
  16:45:20)
  * LINK: https://fedorahosted.org/fesco/ticket/1600   (jwb, 16:45:20)
  * LINK: https://fedorahosted.org/fesco/ticket/1600#comment:4
    (sgallagh, 16:46:26)
  * AGREED: FESCo approves KillUserProcess=yes by default with the steps
    sgallagh has proposed in the ticket (+7, 0, 0)  (jwb, 17:04:58)

* Next Week's chair  (jwb, 17:06:44)
  * maxamillion to chair next week  (jwb, 17:07:11)

* Open Floor  (jwb, 17:07:19)

Meeting ended at 17:17:52 UTC.




Action Items
------------





Action Items, by person
-----------------------
* **UNASSIGNED**
  * (none)




People Present (lines said)
---------------------------
* jwb (121)
* sgallagh (69)
* maxamillion (42)
* number80 (37)
* nirik (32)
* zodbot (23)
* paragan (19)
* kalev (15)
* jsmith (11)
* zbyszek (10)
* tibbs (2)
* smooge (2)
* pjones (1)
* dgilmore (0)




Generated by `MeetBot`_ 0.1.4

.. _`MeetBot`: http://wiki.debian.org/MeetBot

[Jóhann B. Guðmundsson]

On 07/15/2016 05:19 PM, Josh Boyer wrote:

>     * AGREED: FESCo approves KillUserProcess=yes by default with the
> steps
>       sgallagh has proposed in the ticket (+7, 0, 0)  (jwb, 17:04:58) 
"Tier 1 packages must be ported to support operation under 
KillUserProcesses=yes"

Is it safe to assume that FESCo members will do all the porting since 
they where the ones that proposed, decided and agreed upon that condition?

JBG

[Björn Persson]

Jóhann B. Guðmundsson wrote:
>  On 07/15/2016 05:19 PM, Josh Boyer wrote:
>  
>  >    * AGREED: FESCo approves KillUserProcess=yes by default with the
>  > steps sgallagh has proposed in the ticket (+7, 0, 0)  (jwb,
>  > 17:04:58)  
>  
>  "Tier 1 packages must be ported to support operation under 
>  KillUserProcesses=yes"
>  
>  Is it safe to assume that FESCo members will do all the porting since 
>  they where the ones that proposed, decided and agreed upon that
>  condition? 
Don't be silly. Naturally those who want to see KillUserProcesses happen
will have to do the work to make it happen (or pay someone to do the
work). That's how it always works in Fedora. Fesco merely decided that
at least some programs are important enough that SystemD isn't allowed
to nonchalantly pull the rug out from under them. If some people want to
change the contract between user programs and the operating system,
then the work it takes to adapt the user programs is part of the work
it takes to make that change. That much really should have been obvious
from the beginning.

Björn Persson

[Stephen Gallagher]

On 07/16/2016 05:33 AM, Björn Persson wrote:
>  Jóhann B. Guðmundsson wrote:
> > On 07/15/2016 05:19 PM, Josh Boyer wrote:
> >
> >>    * AGREED: FESCo approves KillUserProcess=yes by default with the
> >> steps sgallagh has proposed in the ticket (+7, 0, 0)  (jwb,
> >> 17:04:58)  
> >
> > "Tier 1 packages must be ported to support operation under 
> > KillUserProcesses=yes"
> >
> > Is it safe to assume that FESCo members will do all the porting since 
> > they where the ones that proposed, decided and agreed upon that
> > condition?
>  
>  Don't be silly. Naturally those who want to see KillUserProcesses happen
>  will have to do the work to make it happen (or pay someone to do the
>  work). That's how it always works in Fedora. Fesco merely decided that
>  at least some programs are important enough that SystemD isn't allowed
>  to nonchalantly pull the rug out from under them. If some people want to
>  change the contract between user programs and the operating system,
>  then the work it takes to adapt the user programs is part of the work
>  it takes to make that change. That much really should have been obvious
>  from the beginning.
>   
This is a piece of it, yes. There are certainly some very common work patterns
that would be disrupted here. This change is highly polarizing and while it does
have some clear benefits, if it were turned on right now, FESCo believes that
the consequences would outweigh them.

Rather than forbidding the Change outright, we settled on putting together a set
of conditions that would have to be satisfied in order to end up with a solution
for which the benefits would outweigh the costs. In this case, we decided that
FESCo would rule on a certain set of important packages for which not having a
transparent solution for our users would be treated as a blocker to this effort
(Tier 1). We will also maintain a second set of packages (Tier 2) that we
recognize may be affected but are not critical and can be fixed (or not) at a
later time.

In the end, we hope this will lead to a minimum impact on our users while
gaining the maximum benefit from KillUserProcesses.