At Thu, 23 Aug 2007 23:12:37 +0200,
=?UTF-8?Q?Tobias_Rundstr=C3=B6m?= wrote:
'morning all,
> How do we want to profile XMMS2? I think that one of the major pro's
> for XMMS2 is that we don't have to profile it with one face because
> of the abstraction
The abstraction just means that we have to profile it at a higher
level than mere arguments such as "it looks like iTunes" (Rhythmbox),
"you can read slashdot with it" (Songbird) or "it has a 3D cover
browser" (iTunes). However, we must be careful to provide _concrete_
[GitML(TM)] arguments and not abstract ones for users. In particular,
the argument Peter gave:
Peter> It's all about [quality (elegant, robust, efficient) and]
Peter> client/server for me.
Peter>
Peter> I'll take API over GUI any day.
are totally irrelevant to users. They are very strong arguments for
developers, but it's another side of it, which we can discuss in
another thread. Users don't care about the internals.
More concrete points for users:
XMMS2 ...
... is light enough to play on low-end systems (try running Amarok on
your old P90 or your new N800).
... can support your super-large music collection without dying (try
loading 100GB of music in Rhythmbox).
... can be controlled remotely (use a pretty GUI on your desktop to
play music on the headless server in the kitchen).
... plays almost all your crazy file formats (most other players do,
but not all, e.g. mpd).
... can stream your music to your boyfriend's computer (Icecast) or
your father's hifi (AirPort).
... hosts your music library, as well as your preferences (tags,
playlists, collections, listening habits), which can be accessed
from a great variety of interfaces (CLI, Qt, GTK, Evas - pick your
favourites), even simultaneously.
... allows you to search, browse and organize your music library using
groups, tags, playlists, dynamic party shuffles, etc.
... does or will support external services such as last.fm, podcasts,
ipod/rhythmbox, covers, etc.
Can you think of anything else?
Naturally, we should dump a long list like that, but refine, group and
clarify it in very simple and concrete terms/concepts. For example, I
think Songbird presents its features in a rather nice way (icon +
little description), but they still have too many items to be
readable:
http://www.songbirdnest.com/features
But I really agree with tru that USP are important to explain to users
why they might be interested in XMMS2.
> My view of XMMS2 is the following: I do it because it's fun!
I mainly contribute to XMMS2 to implement new features in order to
simultaneously make people excited and annoy puzzles.
> I would like to profile XMMS2 like the "opensource foobar2k".
I think the idea is good, yes. In short, it would mean targetting
users who _care_ about their music player. It doesn't mean they're
hardcore hackers and will write their own client in Prolog; it just
means that they go beyond using the pre-installed music player and
look for something that suits their needs and taste, be it a simple
GUI or a vis-enabled Qt4/Plasma imaginary client.
Personally, I think that Foobar2k is too elitist in that they make no
effort to make the customization process easy; in XMMS2 terms, though
less drastic, one step of customization is choosing a client, and that
just take one command using your package manager (ideally). In other
words, XMMS2 might be accessible to less computer-savvy users than
Foobar2k, but the spirit of wanting something suited to their personal
preferences remains, including a greater sense of ownership/intimacy
(like people who call GNU/Emacs their best friend, crazy nerds).
Somehow it's not too far from what Peter said about the
API/abstraction, i.e. that XMMS2 is a platform. But in user terms, it
amounts to presenting XMMS2 as a "world" in which all sorts of custom
players live. And you get to pick yours (your own, your precious).
--
Sébastien Cevey (theefer) - http://inso.cc/
"GNU/Emacs is my best friend."
--
_______________________________________________
Xmms2-devel mailing list
[email protected]
http://lists.xmms.se/cgi-bin/mailman/listinfo/xmms2-devel
