Re: [openhealth] Open standards are meaningless.

[Thomas Beale]

Tim Churches wrote:
> Thomas Beale wrote:
>   
>
> Thomas,
> I provided fairly extensive, and I thought constructive, criticism on
> the licenses which you indicated (and which teh openEHR web site seems
> to suggest) apply to openEHR archetype definitions, which are not
> software. As far as I can divine, based on your previous statement and
> what it says on the Web site, openEHR archetypes are not licensed under
> the GPL/LGPL/Mozilla triple license, but rather under the
> openEHR-specific licenses discussed at length in previous posts to this
> list.
>   
You are right, sorry I skimmed some posts too fast. I don't have time to 
respond for a few days due to the MedInfo deadlines and other events. 
However, I would like people here to perhaps consider a repositioning of 
attitudes. I don't see the need for adversarial criticism; I do see a 
need for constructive criticism and collaboration. As far as archetypes 
go (for example) we probably do need a better license. My invitation to 
this community on this topic is as follows: have a think about what 
archetypes are useful for (see resources links below), and based on the 
assumption that the intent of any license for archetypes is of the 'open 
free use, but keeping original copyright intact' nature, make a 
suggestion for a license. It may be that Tim's post already solves the 
problem.

Archetype design activities are occurring all over the place, but a key 
international one is here: 
http://detailedclinicalmodels.org/wiki/index.php?title=Main_Page

The latest ADL workbench (reference parser and viewer) you can get it 
here: 
http://svn.openehr.org/ref_impl_eiffel/TRUNK/apps/doc/adl_workbench_help.htm
The Ocean Archetype Editor (VB.Net/Eiffel) is at 
http://oceaninformatics.biz/archetype_editor/ArchetypeEditor_download.html
The Linköping editor (Java) is at http://www.imt.liu.se/mi/ehr/

All of these are open source software.

- thomas beale

Re: Re: [openhealth] Open standards are meaningless.

[Tim Churches]

Tim Churches <[EMAIL PROTECTED]> wrote:
> 
> Thomas Beale wrote:
> > Tim Churches wrote:
> >> I think a more correct assertion would be that data without metadata 
> are
> >> not very useful, and that is analogous to saying that medical data
> >> stored in an openEHR system are not useful without access to the
> >> archetype definition used to store them. Which is why I have been 
> trying
> >> to elucidate the precise licensing of the openEHR archetype 
> definitions.
> >>
> >>   
> > The licensing on openEHR software is pretty clear and I imagine 
> > uncontroversial  - namely the Mozilla tri-license of GPL/LGPL/MPL. The 
> 
> > other licenses for use of documentary materials may or may not need 
> > revision, and of course we are happy to take advice on this. The 
> current 
> > versions (http://www.openehr.org/about_openehr/t_licensing.htm) were 
> > drafted by UCL's legal department, and possibly have an academic 
> > flavour. They are also some years old now. If you want to criticise 
> > these, go ahead, but please provide constructive criticism that we can 
> 
> > actually use.
> 
> Thomas,
> I provided fairly extensive, and I thought constructive, criticism on
> the licenses which you indicated (and which teh openEHR web site seems
> to suggest) apply to openEHR archetype definitions, which are not
> software. As far as I can divine, based on your previous statement and
> what it says on the Web site, openEHR archetypes are not licensed under
> the GPL/LGPL/Mozilla triple license, but rather under the
> openEHR-specific licenses discussed at length in previous posts to this
> list.
> 
> Is that correct or incorrect. It does actually matter.

Further to this, the use of the GPL, LGPL and Mozilla licenses to cover openEHR 
archetype definitions - if they are in fact the licenses which cover them, in 
contradiction to Thomas' previous statement - then such use becomes quite 
problematic because archetype definitions in the form of ADL files are clearly 
not software, and all three of those licenses are written to cover software, 
and thus their intrepretation becomes tricky with respect to archetype 
definition files. For example, all three licenses refer to "larger works", or 
"work as a whole" - what does this mean in the context of an openEHR archetype 
definition used (and needed) to store and query confidential medical data 
stored in an openEHR system?

I would submit that it is highly desirable that openEHR archetypes are released 
under a license which provides for strong "copyleft" protection i.e. others are 
free to copy, redistribute and modify the archetype definitions, but cannot 
re-issue them (or parts of them) under more restrictive licenses - thus 
preventing them from becoming "proprietary" and being used to lock-in anyone's 
data.

The GFDL (GNU Free Documatation License) - see 
http://www.gnu.org/licenses/fdl.txt - might be a better choice in order to 
provide such copyleft protection to archetype definitions, but even it is not 
quite right. As I said, I really think that the openEHR Foundation needs to 
engage a lawyer who is expert in open source licensing matters (I have 
suggested one) in order to resolve and clarify these archetype licensing 
issues, and as soon as possible, else it is quite possible that we will see 
openEHR archetypes popping up in closed source, proprietary openEHR-based 
systems , licensed in such a way that they lock in users' data. And that would 
be a shame.

Apologies in advance if Thomas or others consider s boring and/or 
non-constructive, but personally I think it matters a lot.

Tim C

Re: [openhealth] Open standards are meaningless.

[Tim Churches]

Thomas Beale wrote:
> Tim Churches wrote:
>> I think a more correct assertion would be that data without metadata are
>> not very useful, and that is analogous to saying that medical data
>> stored in an openEHR system are not useful without access to the
>> archetype definition used to store them. Which is why I have been trying
>> to elucidate the precise licensing of the openEHR archetype definitions.
>>
>>   
> The licensing on openEHR software is pretty clear and I imagine 
> uncontroversial  - namely the Mozilla tri-license of GPL/LGPL/MPL. The 
> other licenses for use of documentary materials may or may not need 
> revision, and of course we are happy to take advice on this. The current 
> versions (http://www.openehr.org/about_openehr/t_licensing.htm) were 
> drafted by UCL's legal department, and possibly have an academic 
> flavour. They are also some years old now. If you want to criticise 
> these, go ahead, but please provide constructive criticism that we can 
> actually use.

Thomas,
I provided fairly extensive, and I thought constructive, criticism on
the licenses which you indicated (and which teh openEHR web site seems
to suggest) apply to openEHR archetype definitions, which are not
software. As far as I can divine, based on your previous statement and
what it says on the Web site, openEHR archetypes are not licensed under
the GPL/LGPL/Mozilla triple license, but rather under the
openEHR-specific licenses discussed at length in previous posts to this
list.

Is that correct or incorrect. It does actually matter.

Tim C

Re: [openhealth] Open standards are meaningless.

[Thomas Beale]

Tim Churches wrote:
>
> I think a more correct assertion would be that data without metadata are
> not very useful, and that is analogous to saying that medical data
> stored in an openEHR system are not useful without access to the
> archetype definition used to store them. Which is why I have been trying
> to elucidate the precise licensing of the openEHR archetype definitions.
>
>   
The licensing on openEHR software is pretty clear and I imagine 
uncontroversial  - namely the Mozilla tri-license of GPL/LGPL/MPL. The 
other licenses for use of documentary materials may or may not need 
revision, and of course we are happy to take advice on this. The current 
versions (http://www.openehr.org/about_openehr/t_licensing.htm) were 
drafted by UCL's legal department, and possibly have an academic 
flavour. They are also some years old now. If you want to criticise 
these, go ahead, but please provide constructive criticism that we can 
actually use.

- thomas beale

Re: [openhealth] Open standards are meaningless.

[Thomas Beale]

David Forslund wrote:
> Just a comment:
>   I like OpenEHR, but it is erroneous to refer to it as an "open standard".
> The term "standard" is usually reserved for something certified as a 
> standard
> by an organization or that is a de facto standard by its widespread 
> use.  Neither
> of these apply to OpenEHR at this time.  It apparently has a published API
> but this doesn't make it a standard. 
>
>   
As I have said many times in the past, openEHR is not an official 
standard, apart from ADL (the archetype definition language) which is 
now European standard CEN EN13606-2, and which will soon have an ISO 
number as well. Usually I have used the term "open specifications" or 
"public specifications". Many other people refer to it as standards, and 
have ceased to correct them. From my point of view, the only standards 
worth the ink they are written with are ones with demonstrable 
implementations and tools. OMG standards rate pretty well by this 
criterion, as do W3C ones (which doesn't guarantee their fitness for 
purpose); most of the other standards relevant to health IT do not.

The openEHR Foundation is improving its process to be something fairly 
close to the OMG process, but with minimal bureacracy, mainly to make it 
clear how anyone could submit a new specification and have it processed.

- thomas beale

Re: [openhealth] Open standards are meaningless.

[David Forslund]

It certainly should be, but I've not had a chance to test this.
The standards OpenEMed are based on include support for the old
GEHR project which is now OpenEHR.I would like
to see this happen, but probably don't have the time to do
it myself.  We have a CTS implementation called OpenCTS
as part of OpenEMed that could be used in this context.  All
open source, of course.

Thanks,

Dave
Nandalal Gunaratne wrote:
>
>
> --- David Forslund <[EMAIL PROTECTED] > wrote:
> I think
> > EHR applications should be interoperable without
> > having to use the same
> > underlying code. Given some time and effort I
> > would like to show that
> > OpenEMed
> > can accommodate the OpenEHR specifications.
>
> Since the archetypes are central to the OpenEHR and
> these are in turn dependent on terminologies,
> ontologies and vocabualries, how does OpenEMed support
> this? Is the Terminology Query Service sufficient to
> support these?
>
> _
>

Re: [openhealth] Open standards are meaningless.

[Nandalal Gunaratne]

--- David Forslund <[EMAIL PROTECTED]> wrote:
 I think
> EHR applications should be interoperable without
> having to use the same
> underlying code.   Given some time and effort I
> would like to show that 
> OpenEMed
> can accommodate the OpenEHR specifications. 

Since the archetypes are central to the OpenEHR and
these are in turn dependent on terminologies,
ontologies and vocabualries, how does OpenEMed support
this? Is the Terminology Query Service sufficient to
support these?





 

Do you Yahoo!?
Everyone is raving about the all-new Yahoo! Mail beta.
http://new.mail.yahoo.com

Re: [openhealth] Open standards are meaningless.

[Gregory Woodhouse]

On Dec 2, 2006, at 12:51 PM, Fred Trotter wrote:

>
> Ignacio is saying that IP would be meaningless without an  
> implementation and
> he is correct. TCP/IP was a success because of the focun rough  
> consensus AND
> working code.
>
> -FT

At any rate, I'm probably not aware of the full context here.  I  
don't follow every thread here (Hardhats keeps me busy enough!) but  
this subject line got my attention. In my opinion, we ought to be  
fostering the development of open standards, not condemning them. I  
fully agree that standards must be implemented, even have multiple  
interoperable implementations from different code bases. But it would  
be illogical to toss out innovative work because  such  
implementations do not (yet) exist.

Gregory Woodhouse
[EMAIL PROTECTED]

"If everything seems under control,
you're just not going fast enough."
-- Mario Andretti




[Non-text portions of this message have been removed]

Re: [openhealth] Open standards are meaningless.

[Gregory Woodhouse]

On Dec 2, 2006, at 12:51 PM, Fred Trotter wrote:

> Ignacio is saying that IP would be meaningless without an  
> implementation and
> he is correct. TCP/IP was a success because of the focun rough  
> consensus AND
> working code.
>
> -FT

I know the slogan (and the paper) well. But the implementation still  
isn't the standard. I'm also well aware of the maturity levels in the  
IETF process.

Gregory Woodhouse
[EMAIL PROTECTED]

"Mathematics is the science of patterns."
--Lynn Arthur Steen, 1988





[Non-text portions of this message have been removed]

Re: [openhealth] Open standards are meaningless.

[Tim Churches]

ivhalpc wrote:
> Open standards alone are an artificial separation of code and data
> that is un-tenable. Source code without data and data without source
> code are not very useful. To be circular: a .odt (Open Document Text)
> file without OpenOffice.org is a .odt file without Openoffice.org.

Ignacio,

have you heard of Koffice? Runs unde KDE, very good, and it reads and
write .odt files too. So do other software products ege Google Docs and
Spreadsheets - there is a comprehensive list here:
http://opendocumentfellowship.org/applications

The whole point of the ODF specs is to remove dependence on a single
source of software, regardless of whether that source is open or closed
source.

I think a more correct assertion would be that data without metadata are
not very useful, and that is analogous to saying that medical data
stored in an openEHR system are not useful without access to the
archetype definition used to store them. Which is why I have been trying
to elucidate the precise licensing of the openEHR archetype definitions.

Tim C

Re: [openhealth] Open standards are meaningless.

[Fred Trotter]

>
> Nonsense. If open standards (such as the IP suite of protocols) were
> meaningless, we wouldn't have an Internet today.



Ignacio is saying that IP would be meaningless without an implementation and
he is correct. TCP/IP was a success because of the focun rough consensus AND
working code.

-FT




-- 
Fred Trotter
http://www.fredtrotter.com


[Non-text portions of this message have been removed]

Re: [openhealth] Open standards are meaningless.

[David Forslund]

Just a comment:
  I like OpenEHR, but it is erroneous to refer to it as an "open standard".
The term "standard" is usually reserved for something certified as a 
standard
by an organization or that is a de facto standard by its widespread 
use.  Neither
of these apply to OpenEHR at this time.  It apparently has a published API
but this doesn't make it a standard. 

An example of what I'm saying is that the API for OpenEMed is an "open
standard" developed by the OMG (and its member organizations).  OpenEMed
is an open source implementation of those open standards.   So the two
issues are orthogonal.   As an aside, I would like to see OpenEHR use
and open standard API (Perhaps something such as COAS or PIDS, or
an equivalent such as the upcoming EIS and RLUS specifications).  I think
EHR applications should be interoperable without having to use the same
underlying code.   Given some time and effort I would like to show that 
OpenEMed
can accommodate the OpenEHR specifications. 

Dave
ivhalpc wrote:
>
> Open standards alone are an artificial separation of code and data
> that is un-tenable. Source code without data and data without source
> code are not very useful. To be circular: a .odt (Open Document Text)
> file without OpenOffice.org is a .odt file without Openoffice.org.
> Major thinkers going back to Alan Turing have noted that the
> artificial separation of code and data is just that: artificial. It is
> like space-time. Time without space and space without time are
> meaningless. You can argue with me on this, but I don't think I can be
> convinced otherwise. Everytime I hear someone advocating open
> standards without or against open source I remember this.
>
> -- IV
>
> _

Re: [openhealth] Open standards are meaningless.

[Gregory Woodhouse]

On Dec 2, 2006, at 11:13 AM, ivhalpc wrote:

> Open standards alone are an artificial separation of code and data
> that is un-tenable. Source code without data and data without source
> code are not very useful. To be circular: a .odt (Open Document Text)
> file without OpenOffice.org is a .odt file without Openoffice.org.
> Major thinkers going back to Alan Turing have noted that the
> artificial separation of code and data is just that: artificial. It is
> like space-time. Time without space and space without time are
> meaningless. You can argue with me on this, but I don't think I can be
> convinced otherwise. Everytime I hear someone advocating open
> standards without or against open source I remember this.
>
> -- IV

Nonsense. If open standards (such as the IP suite of protocols) were  
meaningless, we wouldn't have an Internet today.

BTW, I agree with you 100% on the point that the separation between  
code and data is artificial, but that is not the point.

Gregory Woodhouse
[EMAIL PROTECTED]

"We may with advantage at times forget what we know."
--Publilius Cyrus, c. 100 B.C.





[Non-text portions of this message have been removed]