Skip to Content.
Sympa Menu

sm-discuss - Re: [SM-Discuss] Future plans on the progress

sm-discuss AT lists.ibiblio.org

Subject: Public SourceMage Discussion List

List archive

Chronological Thread  
  • From: Jeremy Blosser <jblosser-smgl AT firinn.org>
  • To: sm-discuss AT lists.ibiblio.org
  • Subject: Re: [SM-Discuss] Future plans on the progress
  • Date: Thu, 29 Jan 2015 11:38:09 -0600

On Jan 29, Ismael Luceno [ismael.luceno AT gmail.com] wrote:
> On Wed, 28 Jan 2015 17:56:08 -0500
> Sukneet Basuta <sukneet AT gmail.com> wrote:
> > On Wed, Jan 28, 2015 at 4:34 PM, Ismael Luceno
> > <ismael.luceno AT gmail.com> wrote:
> >
> > > We're very far feature-wise from becoming popular. Pre-configuration
> > > and sane defaults for unattended usage, and binary packages are a
> > > must.
> > >
> >
> > Not having that stuff is what I like about Source Mage. There are
> > enough distros out there that offer pre-configuration and 'sane'
> > defaults. I like that Source Mage has per package build options and
> > crazy customization. That's a huge factor to it's appeal IMO. If I
> > wanted binary packages with sane defaults, I'd use a binary distro.
> <...>
>
> It doesn't mean diminishing the customization potential in any way.
>
> A lot of software doesn't work right out of the box, or has security
> issues. We could at least make sure those services don't get enabled by
> default, provide warnings and instructions, and perhaps configuration
> examples.
>
> None of those would go against our current policies.

Our policy since inception basically has been to promise admins that if
they install a spell they will get the same thing they would have gotten if
they'd done 'download; extract; configure; make; make install', unless they
actively tell us they want something different or it's an FHS change.

This is only partly about "customization potential", it is even more about
"principle of least surprise". I switched to SMGL predominantly because I
was sick and tired of spending time debugging system issues only to find
the problem was some !@%@^& patch some other distro had applied to "fix"
something that wasn't broken if some packager's assumption wasn't true.

Packages that don't work when run this way have bugs that should be filed
against upstream. Minimal patches can be applied on our end to fix it, but
only for very critical things that don't work anywhere and don't make
assumptions about the state of the system that aren't universal or
confirmed.

> That said, if we want to appeal anyone besides hobbyists, we should
> consider what the different user profiles expect, and deliver.

First, the distro appeals just fine to professionals, thanks. SMGL has
been the distro facilitating at least millions of dollars of credit card
transactions for some pretty large companies whose names you would
certainly recognize, some of them "too big to fail".

Really, almost any serious admin who sees it gets it, the problem is just
adoption where they work if they can't spend the time to get critical mass
of their colleagues interested. Every project with low name ID has this
problem and it's never an easy one to fix, it involves a lot of chickens
and eggs.

Second, the goal of this distro has always been to first meet the needs of
those that develop it. Finding ways to get the word out to like-minded
people so they'll join us as developers or users is great, changing what
we're doing to appeal to people who aren't us is not.







Archive powered by MHonArc 2.6.24.

Top of Page