[csw-pkgsubmissions] /newpkgs amanda

[Darin Perusich]

On 05/06/2010 06:41 PM, Philip Brown wrote:
> On Thu, May 6, 2010 at 6:39 AM, Darin Perusich
> <Darin.Perusich at cognigencorp.com> wrote:
>> Hi Phil,
>>
>> New amanda packages are in newpkgs for deployment.
> 
> hi again Darin,
> 
> I'm tying multiple threads together, by copying this to the
> "pkgsubmissions" mailing list.
> The issues involved, were discussed on the maintainers list, with subject
> "cswinetd and cswetcservices?"
> (We also exchanged a few emails privately, after that discussion)
> 
> Darin, I'm afraid that I'm not going to allow in your latest amanda
> packages for the following reasons. The summary being, that it goes
> against core principles of,
> "To provide a straightforward, easy-to-use experience for the user".

I think this is ridiculous. If I was going against the "core principles"
then why wasn't this package denied acceptance previously for that very
reason?

> You have chosen to deliver it in a way that makes the user
> unnecessarily jump through more hoops, if they do want it installed
> and running.

I'm continuing to release amanda no differently then it has been since
2006 when I began maintaining the package. If the manor in which I was
releasing the package was causing users to "jump through hoops" as you
state wouldn't it be reasonable to expect a bug/feature request to
change the behavior? If you review the bugtrack history for amanda
you'll see that there has *never* been such a ticket opened, so your
insistence that I'm forcing users to "jump through hoops" is baseless
and strictly your opinion.

> Right now, we have a straightforward, easy to use, *same for
> everyone*, framework, that allows our users to individually choose
> whether they want services to automatically be enabled at pkgadd time,
> or disabled.  Please use it.

This straightforward framework, which the standards pages says is "in
flux", is *only* detailed in the building standards, how is the end user
suppose to know of it's existence? Tell me, if this is so
"straightforward" and "easy to use" which package includes the
/opt/csw/etc/csw.conf and /etc/opt/csw/csw.conf files and where is the
end user documentation detailing the usage?

> From your posting on the maintainers list, it seems that you have
> already done the work to convert the package to use our standard
> csw----- classes.
> So this is not even a matter of asking you to do "more" work; only to
> simply release what you have already done in that regard.
>
> The argument of "well this is how it acted previously", is not enough
> to allow them through. If that were enough, then we would never make
> significant improvements to our packages.

Given this, should one conclude that  *all* new packages are being
evaluated to ensure they conform with these improvements? I expect not
and this just happened fresh in your mind based on recent discussion.

> if for some reason you decide to stick with postinstall scripts
> instead of our easy to use, standardized class action scripts, that is
> fine... but you will then need to do the extra work of making your
> postinstall scripts respect the autoenable_demons setting in csw.conf
> Also recognizing our documented standards, that if that setting is not
> present, it is understood to be defaulted to "yes, enable them".


-- 
Darin Perusich
Unix Systems Administrator
Cognigen Corporation
395 Youngs Rd.
Williamsville, NY 14221
Phone: 716-633-3463
Email: darinper at cognigencorp.com