Re: AutoPortage

Desktop Linux growth does look stagnant at the moment and more development
interest is focusing on web and mobile development but I think real
opportunities exist.  They are just not being exploited.

First, software applications for Linux are being sold either as part of
Linux and thus not full products in and of themselves.  A company that
provides each application in autopackage format with a book, interactive
tutorial, and online support network (wiki, discussion, chat channel, and
add-ons) could make a difference.  The software alone, might be made
available for free--but the value worth buying is in the full product.

As for distributions, I personally think we need at least one that focuses
on being a platform just as each application should be productized as a real
product and not a Linux add-on.  But this isn't really feasible until there
is a critical mass of autopackages.  Once done, I'd build this platform as a
basis for customized derivatives.  For example:

Home Server -- One server to control devices throughout your house, such as
thin clients in the rooms (for work, games, Internet, VoIP, music, movies,
etc.) and study areas, recipes in the kitchen, movies in the living room,
scheduled lights, heater, coffee, etc.

Small Business Server -- One server to serve DHCP, files, email, calendar,
applications and a content manager pre-configured as a portal and online
store.  This would support thin clients (Linux (autopackage apps) / Windows
via Remote Desktop from a virtual machine), Windows, Macintosh, printers and
other systems.  That is, everything you need as IT infrastructure for a
small business of 100 to 200 employees or so.

If software applications are treated as independent software packages and
the platform can focus on being a platform with derivatives focusing on
their customer bases' needs then I think desktop Linux will fit very well
into the next generation of computing and electronics.

What are the trends booming today, even in this economy?  The answer is (a)
Virtualization and (b) Collaboration.  The web 2.0 and mobile device
revolutions are really pushing the need for a return to time sharing systems
in various ways.  That's a strength of Linux.  And, it applies to desktops.

Matthew

On Tue, Dec 30, 2008 at 10:49 AM, Mike Hearn <[email protected]> wrote:

> It's definitely true that autopackage hasn't become a big success, so
> in that sense you can justify trying nearly anything. However I'd go
> further and say it's [desktop] Linux as a whole that has not been
> successful, and that's probably impossible to change. The only people
> using Linux are those who accept or have even decided they prefer
> having all their software vetted and provided by a third party, it's
> sort of the iPhone AppStore model in extremis.
>
> > As much as I regret saying this, I don't use Autopackage for own software
>
> Well, autopackage isn't really targetted at server software. Most of
> the things it provides aren't relevant there, or would just annoy the
> technical users who know and like The Unix Way. So that isn't
> surprising.
>
> > 1. The Linux hype is long over. The Linux user base is still growing
>
> Actually it's not. User-agents as seen by google.com show Linux being
> stagnant at slightly less than 1% of all visits, and that has been
> flat over a number of years. Apple is growing slowly and Vista is
> growing fastest of all.
>
> Basically, I think desktop Linux is a dead end - it's (barely) adding
> enough users to keep up with the general growth of the desktop
> computer market. It never managed to grow a commercial developer
> ecosystem, which means it's very susceptible to trends and fashion for
> survival ..... it seems most developers these days are not excited by
> desktop development anymore and prefer to focus on web and mobile
> development. Really who can blame them. The platforms are much nicer
> to work with, and there are well understood revenue models there.
>
> > 2. I'm noticing that pretty much everybody these days *expects* that
> > software is installed through their native package managers. I have to
> > conclude that there is no problem as far as most existing Linux users are
> > concerned.
>
> I think it's indeed less of a problem than when we started - the
> distro market has consolidated a lot lately, there's basically Ubuntu,
> Fedora and SuSE with Ubuntu (apparently) vastly in the lead market
> share wise. Of course the current setup is still  unworkable for any
> developer who actually cares what bits their users run, ie, all
> commercial developers, and any open source guys who have a strong
> desire for product quality.
>
> Something that interests me currently (as the owner of a new G1 Dev
> Phone) is ensuring Android doesn't end up in the same mess with
> respect to compatibility and app distribution. It's off to a pretty
> good start, although the whole Market concept suffers the same problem
> as Synaptic does in terms of crappy UI. But then, the solutions to
> that problem are well known and not hard to implement.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: autopackage-dev-unsubscribe-OfajU3CKLf1/[email protected]
> For additional commands, e-mail: autopackage-dev-help-OfajU3CKLf1/[email protected]
>
>