I finally found a bit of time to read through the whole discussion in one
piece, so I hope it's not too late to add my thoughts on the subject. Following
up on Andre's email summarizing main points made in this thread, let me list
some of the "ponies" that have been proposed, and the important hurdles, as I
see them, that Chandler faces to be a viable player in that niche. These are
either fundamental limitations imposed by existing design and technology, or
missing functionality that will probably take significant effort to implement
-- all the bug fixing and UI tweaking for improved usability is completely
glossed over. :-)
But before I begin, I have to say that *even though* I use Chandler by myself
and started using it because of its goals as a PIM, I found Phillip's arguments
more compelling. Looking at these ponies, it just seems like a) small-group
calendaring (or, more accurately, calendaring-plus-other-collaborative-stuff)
is more easily reachable; and b) there is room for Chandler to carve out its
own niche in this area.
Now here is the list:
"Cross-silo" collection of information and "the source of truth":
To really do this, Chandler would have to tightly integrate with the platform
and its basic sources of information: mail app, web browser, the filesystem.
This is possible, but unfortunately very platform specific, and so was pretty
much written off as a goal the second OSAF chose cross-platform capability and
wxWindows as its building blocks. (It might have been doable in a
cross-platform manner by joining the Firefox+Thunderbird ecosystem.)
Without this tight integration, what are we left with? Chandler can access IMAP
mail folders, to a certain extent, but it's not meant for use it as the main
mail client. Any other bit of information has to be manually copied into
Chandler, and if it exists anywhere else will need to be maintained in two
places. Chandler cannot refer to local files, and it doesn't know about URLs
and hyperlinks, so if I have references to such items in Chandler, I will need
to cut and paste them to another application to open them.
"How it's done": EccoPro can embed any file type, which can then be edited with
its application (probably other recent Windows PIMs can do the same stuff,
using OLE/DDE.); URL or mail address can be opened by double-clicking. iGTD
integrates pretty tightly with Mail and iCal on the Mac, and the OS lets you do
a lot with other apps. Devonthink can store and view anything, including web
page clips, add cross-references and structure, not to mention its searching
and classification capabilities.
Join an existing PIM "cult":
Chandler would have to offer either a faithful implementation of an existing
methodology in an attractive and easy to use package (like iGTD) or be flexible
and configurable enough that users can customize it to fit that methodology
(like Ecco or OmniOutliner with kGTD). I think we can all agree that it fits
neither of these two categories.
Chandler as a new kind of information and task management tool:
I think Chandler currently misses a few features that would are rather crucial
for effective task and information management. The foremost among those is
having a quick overview of current and upcoming tasks, i.e., sorting within
list views. Other important features would be, in no particular order:
filtering, customizable columns in list views, linking of related tasks (e.g.,
as outlines/hierarchies), custom attributes (e.g., as tags), better editing
capabilities within the notes field.
"How it's done": as Phillip noted, this is a very packed and long-existing
niche. Pick your own target for comparison, there are tons of them that do the
job quite well
Calendaring and collaboration:
I should note that really like Chandler's calendar view and spend most of the
time in it. This could be because of the limitations of the list view I
mentioned above, but at any rate it might colour my positive perception of it
as a calendaring tool.
Missing basic calendaring features: drop-down calendars for editing data
fields; full UI for event recurrences
Missing collaboration features (based on comments here and personal
observation, not my experience as I don't use Chandler for collaboration):
finer-grained access rights, smoother signing-up, more discoverable sync
conflicts, being able to mount Cosmo as a remote (webdav) folder, easy
definition of groups/teams of users, better sharing workflow when the both
parties are Cosmo users
some of the competitors: Zimbra (calendaring and messaging), Remember the Milk
(shareable task management), Goolgle Calendar
To sum up, it seems like the latter two areas (Chander as a new PIM and
small-group calendaring) are both reachable, but not concurrently within in the
available time and resource limits. It's ultimately OSAF's decision which one
to target, but I hope this email makes it a little easier to keep the
discussion going towards a happy resolution.
Happy New Year to everybody,
Davor
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Open Source Applications Foundation "Design" mailing list
http://lists.osafoundation.org/mailman/listinfo/design
