On 7 March 2013 15:55, Rob Weir <[email protected]> wrote:

> On Thu, Mar 7, 2013 at 8:56 AM, janI <[email protected]> wrote:
> > On 7 March 2013 09:53, Andrea Pescetti <[email protected]> wrote:
> >
> >> Rob Weir wrote:
> >>
> >>> On Wed, Mar 6, 2013 at 5:46 PM, Kay Schenk wrote:
> >>>
> >>>> "Sustainability concerns due to our use of unsupported applications
> (from
> >>>> Apache Infra perspective), including phpBB and MWiki and reliance on a
> >>>> very
> >>>> small number of system admins." ...
> >>>> https://cwiki.apache.org/**confluence/display/OOOUSERS/**
> >>>> Website+Strategic+Plan<
> https://cwiki.apache.org/confluence/display/OOOUSERS/Website+Strategic+Plan
> >
> >>>>
> >>> And let's not forget alternative ways of addressing this concern:
> >>> 1) Work with Infra to make MWiki be officially supported.
> >>> or
> >>> 2) Form our own admin group
> >>>
> >>
> >> Indeed, I see no reasons to revisit, especially at the current stage,
> the
> >> MWiki vs [any other wiki] issue.
> >>
> >> MWiki is the biggest source of information about the project; while its
> >> content is often outdated, it's still very useful. Existing links to the
> >> MWiki pages are in the thousands (or more). It has been updated with a
> huge
> >> effort lead by Jan, and it is now stable and reliable.
> >>
> >
> > I totally agree with this point of view, but when a PMC raises doubt
> about
> > using Mwiki, it is something we all have to listen carefully to, afterall
> > one of the points in being PMC is to secure the long term stability of
> the
> > community and product.
> >
> > This is also the reason I researched cwiki and moin...not that I
> personally
> > would like to change, but because I read the mail from one PMC and the
> > reactions from others.
> >
>
> But please note that I never suggested moving off of MWiki.
>
> >
> >>
> >> If our problem is to enlarge the administrators group for MWiki or to
> get
> >> it officially supported, I prefer to explore these options first. We
> >> already have three committers who can administer MWiki on the system
> >> administration (shell access) side, right? (jani, imacat, rbircher). How
> >> many more do we need? I've also just asked Infra for clarifications
> about
> >> the "supported applications" issue.
> >>
> >
> > At the moment we have no-one with "just" shell access, all 3 have (or can
> > have) root access. Furthermore rjung (infra) pratically maintains httpd,
> > gmcdonald (infra-root) helps with more or less everything and Clayton
> helps
> > with mwiki setup (without access). BIG Thanks to all three for their
> great
> > help !
> >
> > After the upgrade we have had one incident (which rjung handled), one
> > upgrade proposition (which I handled wrong) and 2 request for change (one
> > pending and one I have handled). At the same time, nothing have been done
> > "inside" mwiki regarding old pages, strange categories, spammed paged,
> > misleading information.
> >
> > At the moment 2/3 of the time I use on mwiki goes to logistic and
> > coordinating people, if the group is expanded my experience is that the
> > overhead grows exponentially. I respect the wish of the community and
> when
> > the group is expanded, I will withdraw (after a handover), I am here to
> get
> > things done and not to coordinate people.
> >
> > I fully understand we need many sysop, and the idea of having a mail list
> > (or just a wiki-page, with sysop access level) is a good idea. Our
> biggest
> > job at the moment is not sys-admin but normal sysop work.
> >
>
> This is what I see:
>
> 1) We brought over MWiki and had a single person (Terry) who really
> knew what was going on.  We may have had others who had permissions,
> but they were not (IMHO) able to keep the wiki stable.
>
> 2) Terry left, and the wiki was not maintained as well.  Eventually it
> fell due to massive spam.  No one of the existing admins stepped up to
> fix the problem.  Whether this was from skill, permissions, time,
> inclination, I don't know and I don't judge.  But the actions were to
> disable new account creation and treat that as a new way of life.
>
> 3) You stepped up and took the lead on getting MWiki to be properly
> maintained again.  If you had not done that I am pretty sure that we
> would not have the ability for new years to create accounts today.  We
> were already down on one knee when you offered to help.  I'm pretty
> sure without your help things would continue to degrade.
>
> 4) So what would things look like without you?  Sure we have others
> who have some permissions.  But is that enough?  (It wasn't enough
> before).  What do we need to do so we are not dependent on a single
> person?  Not just from a theoretical standpoint (X and Y have
> permissions) but from a practical skill and knowledge level as well.
>

I think there are too much focus on the sysadmin part, which isnt really a
problem..mwiki is very stable, it needs an upgrade now and then (2 times a
year in average). Our httpd, ats, firewall, ubuntu needs a lot more regular
maintenance, and thats where people like rjung and gmcdonald are of great
help.

I know "sidestepping". I can see your point of having multiple active
skilled persons on the team (even though I dont agree with it), but getting
more volunteers without the skills/motivation is simply not helping !
Before I jumped in there has been plenty of time for other volunteers to
step up alongside with the 2 who already have access. If someone comes
along today, I will happily tell them how the current setup (and now it is
even documented in svn).

As I wrote I respect whatever decision is made, but please also respect my
wish. I do want to use my limited time to bring our community forward and
not discussing, which is going to happen with a larger number of sysadmins.


>
> 5) Note the same applies to phpBB as well.
>
> From a high-level perspective, one way is to think of it like this:
> As a large project with diverse technical infrastructure we make high
> demands on Infra, hardware, bandwidth, sys admin time, etc.  Since we
> have such a dependency on this technical infrastructure, more so than
> other Apache projects which are less end-user facing, this starts us
> off with greater risks.  Our use of non-standard services like MWiki
> and phpBB increases the risk.  Fine. So how do we balance those risks?
>

Simple answer is "we dont", we have the people, we have the needed help
from infra. I stepped in when needed, when I am not around another person
steps in.

I wish we had the same attention on our the software...how many people do
we have the really understands the build procedure or for that matter the
l10n process ? How do we handle a situation deep in the software. I think
the support for our software is pretty much like mwiki, in many cases only
0 or 1 who knows the code, and a couple who knows in generally.

But I agree in a strategy paper anything can be written, and if our main
focus shall be on non-essential parts like mwiki etc. so be it.

rgds
Jan I.


> > I honestly think we should consider more how to stabilize the "inside" of
> > mwiki instead of focussing on the sysadmin part, but that is just my
> > opinion.
> >
>
> These are not mutually exclusive options.  One of the luxuries of a
> strategic plan is I can suggest more than one priority.
>
> Regards,
>
> -Rob
>
> > rgds
> > Jan I.
> >
> >
> >
> >
> >
> >>
> >> Regards,
> >>   Andrea.
> >>
> >>
> >>
> ------------------------------**------------------------------**---------
> >> To unsubscribe, e-mail: dev-unsubscribe@openoffice.**apache.org<
> [email protected]>
> >> For additional commands, e-mail: [email protected]
> >>
> >>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>

Reply via email to