this seems to be a very important thread ...and so is the Pedi Project. Hopefully everyone follows the Pedi Project sufficiently or follows every trail on this list and therefore participates....but it might be good to start a more obvious threadname at some point.
I would like to ask Rick and others:
1. How has the Hui made visible the copyrightable portions of the 2003 release of HuiOpenVistA?
Hui certainly has a formal policy of technology transfer. Has there been a declared method of showing both the pruned omissions from FOIA and the grafted additions to FOIA to clearly show what is the supplemental part of the derivative work. Is visibility limited to a special marker/comment to identify these in the source code or are there other documents that accomplish this better?
2. What are the distinctions between Hui OpenVistA released 2003 and any WV OpenVistA releases since 2003?
I am asking basically if any of current WV OpenVistA is already a mosaic of some Hui OpenVista plus various patches to each FOIA that is reviewed and incorporated (WorldVista was said to be a collaborator even if Hui did the heavy lifting) If so perhaps there already is a license applicable, although that doesnt preclude negotiating something else.
3. I would like to hear examples of the typical recurring tasks that occur in joining new FOIA releases into each successive OpenVistA (whatever the regularity/predictability of WorldVista meetings and solitary contributions)
What are some of the classic patches that are going to have to be done again and again because of evolution in certain VistA modules?
Just hearing some recurring tasks described helps to formulate a view of appropriate license for this ongoing product. ( this disregards the work that would be required to review major new contributions that are not FOIA)
Naturally questions spawn other questions
Rusty Maynard
Frederick D. S. Marshall wrote:
Dear friends,
Nancy's question about how best to license the Pediatrics work applies equally well to OpenVistA as a whole. We know we need some kind of open-source license for it--after all, the CHCS experience helped teach us all what can happen to unprotected software--but WorldVistA has not yet decided on which license to use.
On this coming Friday's VistA Community Call, the main topic will be this very question, so in preparation many of us are studying the available licenses and seeking guidance.
Several VistA community businesses have expressed passionate concern about the GPL, describing it as viral and asserting that it would threaten their ability to do business. These are serious concerns, and I must discover if they are well-founded. We will not pick a license that injures our vendors.
On the other hand, it is possible that their concerns are misplaced, that the GPL would not interfere with their business strategies. I honestly do not know yet and am studying the matter.
I do know that where VistA is concerned this is a nontrivial subject. What constitutes a "program" in VistA is not the same as most other
programming systems. Certainly a routine is not a program, only a unit
of software storage. A VistA "program" is closer to a package, with
corresponding routines, globals, remote procedures, templates, tasks,
and so on--dozens of different kinds of software elements--parts of
which are stored in routines but much of which are stored in globals,
thoroughly blurring the distinction between code and data. Von Neumann
would be proud! Since VistA package elements are split up by type and
stored along with the corresponding elements from other packages, the
boundaries between one VistA package and another must be defined very
differently than for traditional systems.
Our system is both integrated and intricately modular. VistA software packages are independent enough to be installed and updated without directly affecting the rest of the system. Yet, they are so integrated with the rest of VistA that their connections with other packages are numerous and complex. The articles I have read so far about open-source licenses discuss such questions as whether one program uses parameter passing or APIs to interact with another--yet VistA packages although independent nevertheless interact with other packages in dozens of ways. By any traditional definition, to any mainstream programmer, VistA packages would have to be seen as not independent, as part of a single integrated system, yet the truth is more complex and subtle.
One very important question this raises is how exactly a "derivative work" should be defined with respect to VistA, given its unusual software structure. Until we are clear about that, we are not clear about how the GPL would impact the vendors like DSS, Informatix, Medsphere, and others who develop and market VistA extensions. This specific issue, which based on my reading to date is central to the question of whether the GPL would hurt the vendors, has been publicly raised to date neither by the vendors likely to be affected nor by the VistA-community advocates of the GPL.
However we may feel about the pros and cons of the various open-source licenses on their own merits, we dare not take a position as to which is the right choice for OpenVistA without understanding this issue. For example, we may need to explicitly spell out what is or is not a derivative VistA work to ensure the license we choose has only the effect we want--VistA is different enough that it may not be reasonable for us to expect lawyers and courts to understand how to apply a given license to VistA correctly.
I'm sure there are a few other technical issues like this one that will define for us the impact each family of licenses would have on our VistA vendors. We need to identify and master those issues.
Likewise, we need to get real clear about the results we want to see from adopting an open-source license. There may be a few individuals or organizations in our community for whom the most important part of this decision is that we be able to use the phrase "open source" to describe our software so we can benefit from the good will or perceived cachet of the phrase, but the rest of us have more tangible goals we need to achieve and we will not accept any license that improves over public domain in name alone. Anyone still of the opinion that public domain is adequate for our work should ask for an explanation of why most of us oppose releasing it with no protection at all. The question of how best to protect the work ties directly to what we are trying to achieve, how we want our work used, and how we want our work to interact with the work of others. I suspect most of us, vendors and nonvendors alike, agree on more than we suspect about the results we want to see. Few are interested in passing off the community's work as their own. Most just want to be able to pursue their programming work--for profit, for charity, or for government purposes--in a way that helps the results work together harmoniously.
Once we have some clarity about the VistA-specific issues involved in making this decision, as well as about what kind of protection we need, then we will be in a responsible enough position to choose (hopefully) or create (if necessary) the right open-source license.
Meanwhile, I am much appreciating the references to articles and explanations that people have been sending to this list. As a community, we have some studying to do to ensure we make a wise choice.
Yours truly, Rick
On Sat, 2004-11-13 at 15:45, Greg Kreis wrote:
Here are some articles that might help.
Court Evaluates Meaning of "Derivative Work" in an Open Source License. http://articles.corporate.findlaw.com/articles/file/00050/008924
General Public License, Explained http://www.sitepoint.com/article/public-license-explained
GPL: Viral Infection or Just Your Imagination? http://www.linuxinsider.com/story/33968.html
Perhaps someone familiar with how SAIC created the CHCS product could speak to this? Am I correct in understanding they started with VistA (DHCP at the time)?
Cameron Schlehuber wrote:
Bill, I meant no sarcasm in my response. I really did think you had some great questions and responses.
And in thinking further about my responses regarding traditional copyrights, they really didn't get at the very different twist that GPL introduces if applied to public domain works, which I think is what you're trying to get at. That is, can a copyright cause all derivative works to essentially remain in the public domain? And, is the GPL such a copyright?
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Nancy E. Anthracite Sent: Saturday, November 13, 2004 4:14 PM To: [EMAIL PROTECTED] Subject: Re: [Hardhats-members] Protecting Pedi Project IP [add] and SCO
Knowing Cameron, I would say that although he spoke of it in simple terms,
for the likes of me, it is not lightly considered and he probably knows as well as anybody what he is talking about as he, too, has been in the "business" a
while himself, albeit while employed by the US government, but he is very much a Hardhat and in favor of anything being open that can be.
It may be that getting a definitive answer for this will be impossible if
the GPL has never faced a court challenge.
However, that may change as I guess good old SCO is planning to do just that.
http://www.it-director.com/article.php?articleid=11807.
On Saturday 13 November 2004 04:34 pm, Bill Walton wrote:
Golly, Cameron. That's just *brilliant*. So whenever someone wants to know about VistA and IP, we just sing them a chorus or two. That never would have occurred to me, for sure. ;-)
I've been in the business world far too long not to know that if were as simple as you'd apparently like to believe, there'd be no lawyers.
My question stands. Or perhaps I should break it in to two questions.
1) Has WorldVistA (or anyone else) already investigated these issues?
2) If so, is there a document somewhere that provides the answers?
A simple yes / no response, sans sarcasm, to the first question would be appreciated.
Thanks, Bill
----- Original Message ----- From: "Cameron Schlehuber" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Saturday, November 13, 2004 1:47 PM Subject: RE: [Hardhats-members] Protecting Pedi Project IP
Great questions and great responses! I'm no lawyer either, but that doesn't mean that truth can't be easily determined. Let's try for some plain and simple examples from everyday life.
The poem "Mary Had a Little Lamb" is in the public domain in the same way
that VistA is in the public domain. Yet, there are numerous books with
that poem that are copyrighted in the traditional manner. It's the entire
work that is copyrighted ... the artwork, introductory remarks, perhaps
there are variations on the "Little Lamb" theme. Are there parts of the
original mixed in with the new parts? Yes. Are there such works with
more
than one kind of license? Yes. Look at most song books and you'll see a
mix of copyrights with different songs and hymns holding different kinds
of
copyrights (and some with none). Are there more "restrictive" copyrights as well as less? Yes. Can public domain works be more restrictively licensed? Yes. Can anyone other than the original author set new license terms? Yes.
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Bill Walton Sent: Saturday, November 13, 2004 12:04 PM To: [EMAIL PROTECTED] Subject: Re: [Hardhats-members] Protecting Pedi Project IP
K.S. Bhaskar wrote: /snip/
In the case of the pedi project, I recommend
the GPL.
As I understand the architecture of VistA, new functionality will tend to include / result from modifications to existing code rather than simply additions of new code modules that are completely seperate from the existing code. To the extent that this is true it seems to me that there may be some non-trivial questions that need to be addressed. For example...
Is is possible for a single piece of source code to have more than one license?
If the code is a mix of FOIA software and new additions, can the additions be licensed at all? For example, "lines 1,2,3 and 7 are FOIA but lines 4,5, and 6 are additions subject to the GPL." I've never heard of a situation like this.
Is it possible to release FOIA software under a more restrictive license?
In general, all the OSS licenses I'm aware of preclude the licensee from releasing derivative code under a more restrictive license. The licensee must give recipients at least the same rights he was given. FOIA software contains no license restrictions. OSS licenses do. So it would seem that there may be an issue attendant with the re-release of FOIA software under an OSS license.
Is it even possible to for anyone other than the original author to set
the
license terms?
I'm not aware of a situation where someone other than the original author(s) has (have) set the license terms. Is there any legal precedent for this?
Assuming that FOIA software *can* be re-released by someone other than the original author(s) under an OSS license, what steps have to be taken?
In general, for a license to be valid, its terms must be presented and accepted. Most software presents you with a license agreement upon installation. Are there accepted alternatives? Or does this capability have to be built into WorldVistA software? Also, for example, many OSS licenses require the terms to be included in the source code. Anybody sized this effort?
Even if the new code (e.g., pedi) is completely seperate from the existing code, if the FOIA portion can't be re-released under an OSS license, is there any precedent in the market for a mixed-license model?
I'm no lawyer, but I'd want the answers to these and other questions
before
I contributed code. If WorldVistA (or anyone else) has already investigated these issues, is there a document somewhere that provides the answers?
Thanks, Bill
------------------------------------------------------- This SF.Net email is sponsored by: InterSystems CACHE FREE OODBMS DOWNLOAD - A multidimensional database that combines robust object and relational technologies, making it a perfect match for Java, C++,COM, XML, ODBC and JDBC. www.intersystems.com/match8 _______________________________________________ Hardhats-members mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/hardhats-members
------------------------------------------------------- This SF.Net email is sponsored by: InterSystems CACHE FREE OODBMS DOWNLOAD - A multidimensional database that combines robust object and relational technologies, making it a perfect match for Java, C++,COM, XML, ODBC and JDBC. www.intersystems.com/match8 _______________________________________________ Hardhats-members mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/hardhats-members
