Like Malone,
Smith
[
Smith Collective
]
proposes a model of collaboration with several dimensions/parameters:
Information type built by the group:
intangible private,
ephemeral,
tangible instrumental,
tangible target.
Time over which the collaboration occurs:
few seconds, several minutes,
15 minutes,
several hours,
several days,
several weeks,
several months,
several years,
life's work.
Space over which collaboration occurs:
office,
cluster,
floor,
building,
site,
1-hour travel for access,
1-day travel for access.
Agents collaborating:
people,
informal coalitions,
teams,
collection of teams,
group as a whole.
Processes involved:
cognitive,
metacognitive,
social interaction,
mediated,
organizational,
collective/distributed,
algorithmic.
Collaborative tools can be classified according to which points alongs these
dimensions they support.
For instance,
CES, which does checkpointing, can support
collaborations spanning all durations
whereas a tool such as RTCAL supports collaborations of less duration
than the time between computer failures.
(We need
John's help in defining the points along these dimensions and
using them to classify our tools.)