Hi,
Damianos Sidiropoulos wrote:
The biggest problem is the lack of an official desktop. It's hard to
try to pitch user experience without something a user can download and
install.
Unfortunately having just apps is not sufficient as that will look and
potentially behave differently depending on the environment in which
it runs.
well, depends.. also it is not the biggest problem. Those who think
GNUstep as a porting framework couldn't care less about an official desktop.
Also right now we have apps and separate desktop projects and given the
complexity behind them, it will be so for a while.
I also never remembered we discussed an official desktop except for the
fact that some people like a desktop! But this lead me to a very
interesting observation.
It is a novum, we usually discussed to death the concept of "reference
distribution".
Of course, a reference distribution could come with one or more desktops
chosable.
I might not have communicated it in the slides sufficiently so I will
try to clarify. The short term solution was to have GNUstep be a
dedicated developer website.
However once the desktop part of the equation is figured out, the
desktop solution would largely dictate how the desktop web presence
would be handled
Possible solutions and how they affect website strategies:
If the desktop is solved with an official desktop project developed
and maintained by GNUstep then integrate it into the main site with
subdomains or whatever.
If GNUstep chooses to have multiple official desktop projects then
that will involve further considerations.
If GNUstep chooses to instead just create an official relationship
with one or more 3rd party desktops like GSDE, then that would require
another approach.
I don't know if we should have an official desktop or even provide an
official desktop. There are different desktop implementation because
there are divergent ideas on how to implement it. Some cater the classic
look (e.g. GSDE, NEXTSPACE) up to the point of replacing or hacking
various applications to fit in.
Very hard is also the dependency on a specific Linux distribution.
Others are more generic (e.g. GAP, where I contribute) which try to stay
more agnostic on the underlying OS, user theming and do not tweak apps.
This also means postponing diffdicult choices.
But just a good observation which should make us all reflect a bit and
calm down.
The "framework" discussion camp is very active in stressing different
UXP support, theming, high Apple compatibility... you pointed it out on
your slides.
Looking at the "Desktop camp", we see a different trend. Both GSDE and
NEXTSPACE are very traditional in the look and experience, up to a
painstaking effort of NEXTSPACE.
GAP is agnostic - but to my part - tries to have an OPENSTEP like UXP
but modernized in things, not a "feature clone" and be mostly adaptable.
Etoile is innovative in terms of look, experience and technologies, but
it is more or less dead.
So if you look at the "desktop camp", most people are interested in NeXT
look, experience and probably care less in flexibility, and latest apple
features or windows UXP and portability.
To say, two camps, two very different "requirements" and "interests" in
one thing, the core system you want so much to highlight from GNUstep.
This goes back also to many discussions there were in the part : it
shows the interests of different people.
For today, I finish with this wisdom and insight which did not occur to
me before.
Riccardo
