On 17 Feb 2009, at 18:18, Damien Katz wrote:
On Feb 17, 2009, at 6:53 AM, Jan Lehnardt wrote:
2.) Learn CouchDB [link to the docs].
3.) Use CouchDB [link to list of DBs].
Just as a general idea, details can vary. The idea is to show the
first most
likely actions a user wants to take and permanent links to the
documentation,
user mailing list, issue tracker (with BOLD LETTERS that for
general questions,
user@ should be used).
Thinking of it, the 'validate your installation' step should not
show up after
it has been run E.g. runtime config could set a new flag:
[futon]
tests_passed = true
Nah, no need to do that. Validating the install is something people
will want do at other times, as a sanity check.
I don't want to hide tests forever after the first run, but remove it
from the list
of things a user/dev can/should do "next" when coming to Futon. Sort
of a
guided tour (I didn't say "Wizard!") that doesn't do things twice. The
possibility
to run the test suite at any time should stay as is, just not placed
prominently on
the first page (except for the first run), but in the sidebar/menu as
it is now.
List of installed apps. With CouchApps shaping up, Futon could show
a list of
installed applications alongside created databases. The way to
detect CouchApps
is not yet defined, though there are sensible proposals.
I don't dislike this idea, but this seems more like an end user
feature? Like a start menu for CouchApps? I think most end users (as
opposed to devs and admins) should never need to use Futon.
This is meant for a dev managing multiple CouchApps, not an end user.
I agree
that Futon is not an end-user tool. Granted, the line with user and
dev is blurry
with CouchApps, but this would make a dev's live easier to get a list
of apps instead
of hunting attachments in design docs (3-4 clicks at least). Yes,
bookmarks..., but still :)
[...]
In my opinion, Futon should remain an admin and utility client for
admin and developers. End users really shouldn't be using Futon,
applications should provide all the UI necessary for users. Not just
to make it easier for users but so they don't go fiddling with the
wrong things.
Agreed.
Also, I worry a little that it will expand into some sort of big IDE
like environment, like a web-based Eclipse. Not that I think that's
a bad idea, I think it's a really interesting idea. But I think it
shouldn't be Futon, but rather a separate sub project. The Futon
core should remain small and focus on making common admin tasks
simple. Being able to use it from a wide range of browsers (like
from a phone) is a big benefit to admins.
And agreed again. I don't like to see a Kitchen-Sink-Futon*, but a
little
flexibility would be nice.
* This does not sound relaxing at all!
Cheers
Jan
--
