-----Original Message-----
From: pschloeffel [mailto:[EMAIL PROTECTED]]
Sent: Sunday, 12 November 2000 9:46 AM
To: 'Horst Herb'
Cc: 'Beale_Tom'; 'Heard_Sam'; 'Rowed_David'
Subject: RE: license issue: GT.M and VISTA
Horst,
[Peter] As I said in my earlier reply to Andrew Ho on the list, I apologize if I
offended - I do not want to provoke an argument with you or anyone else in the
Open Health community but I acknowledge that I did "shoot from the hip". We had
a very productive meeting in LA last Monday night organised by the CPRI,
Internet2, and Joe DalMolin. Tim Cook, Andrew Ho, and a number of the other
Open Health contributors were there and Joe gave a very good talk on his
philosophy on Open Source and collaboration after Sam Heard and I had given
talks on GEHR. I won't repeat all of the material here which I put in my reply
to Andrew but will address a couple of your points specificly.
[HH] The point of open source development is
* to avoid vendor lock in
* to guarantee "survival" of a software project
* to allow as many developers as possible to contribute
[Peter] I agree with each of these points but I stand by my comments on use of
commercial tools when necessary as re-iterated in reply to Andrew.
[HH] By the choice of ISE Eiffel GEHR misses the point in all three criteria. I
still like the concept, I even like the Eiffel language, but I believe GEHR
would be thriving better and find more acceptance among the open source
community if the development would a least have taken place in SmallEiffel.
Hell, even our gnumed project would have happily embraced and supported
GEHR as our middleware if it would have been a true open source project.
[Peter] You may be right Horst. I would encourage you to perhaps do a
conversion of the kernel to SmallEiffel.
[HH] Quote from MSDJ online: "However, the folks from Redmond who bring you COM
have been burning the midnight oil, adding plenty of completely new stuff to the
once compact COM."
Do we really want / need this? OK, you are open for other interfaces. But the
initial choice of COM
[Peter] As I said in my original posting, the COM interface was a requirement of
the federally funded GPCG project because the GP vendors involved in the project
(MIMS, Locum, and Medical Spectrum - MD was invited but declined to participate)
all use VB or Delphi for their application software. We have a great deal of
interest in the US from the CORBAmed community in assisting with a CORBA
interface (the CORBAmed services are compatible but not all have yet been
implemented).
[HH] & the choice of ISE & the initial choice of the data base chosen for your
OSTORE
[Peter] Tom chose Matisse for the initial implementation for several reasons but
one very practical one was that we had very limited resources at that stage
(yes, we could have chosen to do an Open Source development from the ground up
like Littlefish and I presume Gnumed but we did not think this was feasible at
the time) and the fact that he had experience of using it in a previous project.
[HH] (might have changed by now) and ... - all together they are heading into
the wrong direction, away from the open source community.
Do you seriously think any large scale open source project will risk
invalidating its status for a "bullet proof middleware" depending on closed
source proprietary compilers?
[Peter] Yes, I do Horst - if the "closed source proprietary compilers" are
judged to be superior for the task at the time. Rightly or wrongly, we
certainly believed that in the case of ISE Eiffel. We may have to agree to
disagree on this point but I hope this will not "invalidate" us completely in
your eyes and the eyes of the rest of the Open Source community.
[HH] Look at all the "open source" projects developed with M$ VC++, Delphi,
etc. They never took off for that very reason, and many of them would have been
excellent projects otherwise.
[BTW, It has nothing to do with extremism. Same thing with open source as with
pregnancy: you can't get just "a little bit pregnant". It is either or.
[Peter] Again, I do not agree. There is already a wide spectrum within the Open
Source community as evidenced by those who think that GPL (or similar) is the
only valid form of Open Source licence and those who believe that Mozilla type
licences have a valid and important role to play within Open Source. You will
probably already know that our openEHR licence for the Ocean GEHR kernel (on
www.gehr.org) is essentially the Mozilla licence with required name and
jurisdictional changes. GEHR was always intended for use with both commercial
and Open Source end-user applications so the GPL type licences are of course not
applicable.
> How the use of "proprietary closed source" tools such as ISE Eiffel
invalidates > the legitimacy of the GEHR kernel as an Open Source product (ie
"the real > thing") is beyond my understanding.
[HH] In that case you should do some reading on the open source philosophy :-)).
Maybe you start at http://www.gnu.org
[Peter] I have in fact read all of the material on this site and a lot of the
linked and other Open Source material such as the Cathedral..., the Haloween
Documents, and the excellent OpenSources book published by O'Reilly.
Peter
