A client of mine has been using a product to do task, project, client
and billing management, AllocPSA. For various reasons they are
considering moving away from that to something more open and widely used
My recommendation, for task tracking, is Request Tracker: I have used it
with success for years, and I am very comfortable that it can work well
for managing job assignment, completion and history.
There are a couple of features that AllocPSA, being a "one stop shop"
provides that don’t come out of the box with RT and which I wondered
about what experience / advice y’all had for integration:
My client currently has per-project billing rates for their staff, so a
single engineer might be working on one job at $50 per hour, another at
$75 and so forth.
This is, currently, displayed on the tasks; my view is that this can be
extracted in an overlay to the ticket display and added, as a read-only
field, to one of the standard display boxes.
(Alternately, it can be set and maintained as a custom field on the
tickets, but I don’t want them to be mutable and probably do want them
more highlighted than a standard CF would be.)
They also have some CRM style features: the software maintains a
hierarchy of [ client 1 -> * project 1 -> * task ] in the system, so
that a single client can have multiple projects and a project multiple
It also provides contact management: a client has contact records, and
so forth, which are then used once you get down to the project or task
It would be good to keep this sort of integration and, from experience,
the user management part of RT really isn’t targeted at this space: it
doesn’t seem to provide a lot of “client” focused searching or display,
The two approaches I can see to this are:
Deploy some other CRM style application and encourage people to use
that, optionally using the command line tools to sync changes from
the external product into the RT client/user database.
Modify RT to use an external data source for client/user information,
pulling that from the external tool directly.
The second is more desirable, to my mind, since it avoids duplicating
the information – but I am not sure what sort of workload it would be
to incorporate that into RT?
I welcome your advice,