A big hello to everyone subscribed on this list.
:-)
I'm sharing my initial thoughts on a probable course of action for
the Froid project. I look forward to hear your opinion and
thoughts.
Our goal is to provide a middle-ware that will be able to handle
transactions of orders and results to and from laboratory analyzers.
The project aims to support both serial and ethernet connections.
Most of the manufacturers use the ASTM E-1381-95 [1] (
transport
level protocol) standard to transport the relevant records
across the analyzer. Although this standard is relatively "old" its
still being used, it have some revisions but the main concepts
remain. I have found some relevant documents online here and there,
but if anyone can pinpoint any resource, please feel free to do it.
The Froid project should not care -at least at the beginning- to
support the application level protocols that a analyzer uses. Every
manufacturer adds more or less some extensions to the application
level protocol and i cannot see a simple way to acquire all those
documents and to implement them. Example of
application level
protocols are the ASTM E-1238-94 and ASTM E-1394-97.
Instead, we will design a clean interface to the underlying
transport layer. Any application who implements and pushes orders
with this interface, will be able to reach the analyzer, and acquire
the results through the same interface. My initial though on this
was to use redis as a queuing subsystem in which various clients
will push orders and take back the results. The orders/results will
be defined as .json or .yml files.
Having
froid.io registered there are ongoing conversations to
set up a discourse [2] community forum to handle all design -
development - testing cycles and communication. I think this will
also help to expand the project to other type of instruments.
Last but not least, we need a "manifesto", an ethical guide to
publish to the medical world. We need to state the reasons we are
developing the Froid project. Me as a purely technical guy do not
have any experience with medical laboratories so I'm assured that
the need for an open source middle-ware are better known to you
guys.
[1]
https://www.astm.org/DATABASE.CART/HISTORICAL/E1381-95.htm
[2]
https://github.com/discourse/discourse