VistA Internationalization...lost horizon?
32 views
Skip to first unread message
PeterBodtke
Aug 27, 2012, 10:18:24 PM8/27/12
to hard...@googlegroups.com
Chris, Kevin and Sam,
Writing this note to you because you have done work to enhance VistA toward multiple language support, other are welcome to comment as well. Kevin's recent decision to stop supporting for multiple languages in TMG CPRS has give me great pause. I certainly understand he reasons and do not fault him one bit. Several requests were posted ask Kevin to continue, but I didn't see any offers to fund the effort. The whole Internationalization concept seems like a chicken or egg problem, if the work was done there would be a huge market, without the work already done no one is seriously interested in adoption...so either happens.
I am 8 days away from my next trip through Central and South America, 6 months overland, touching every country in Central and South America. The last time I travel in Latin America (2008-09) I had meetings with government offices (Belize, Costa Rica, El Salvador and Columbia) and universities (Guatemala, El Salvador, Costa Rica and Panama) where VistA was demonstrated to great interest, but no traction gained. The biggest stumbling block, in my mind, was the lack of Spanish support (or any non-English language...)
Here are the questions that keep me up at night:
Best regards,
Peter Bodtke
Writing this note to you because you have done work to enhance VistA toward multiple language support, other are welcome to comment as well. Kevin's recent decision to stop supporting for multiple languages in TMG CPRS has give me great pause. I certainly understand he reasons and do not fault him one bit. Several requests were posted ask Kevin to continue, but I didn't see any offers to fund the effort. The whole Internationalization concept seems like a chicken or egg problem, if the work was done there would be a huge market, without the work already done no one is seriously interested in adoption...so either happens.
I am 8 days away from my next trip through Central and South America, 6 months overland, touching every country in Central and South America. The last time I travel in Latin America (2008-09) I had meetings with government offices (Belize, Costa Rica, El Salvador and Columbia) and universities (Guatemala, El Salvador, Costa Rica and Panama) where VistA was demonstrated to great interest, but no traction gained. The biggest stumbling block, in my mind, was the lack of Spanish support (or any non-English language...)
Here are the questions that keep me up at night:
- Is it possible and practical to create an Internationalization process for FIOA VistA (including available clients)?
- What languages would a "conversion" process support in the deepest recesses of the server, i.e. Spanish? (extended character and all...)
- How large would the effort be in hours and dollars?
- Is there a sustainable road map to incorporate new enhancements in the foreseeable future?
- Is VistA too complex and too embedded with English for practical support of another language?
Best regards,
Peter Bodtke
Nancy Anthracite
Aug 27, 2012, 10:26:47 PM8/27/12
to hard...@googlegroups.com, PeterBodtke
Peter, I don't know if you saw the recent discussion on Hardhat where Larry
talked about what could be supported in terms of characters. The thread
started "VistA and UTF-8" which should help you find the emails.
Good to see you back on the mailing list!
--
Nancy Anthracite
On Monday, August 27, 2012, PeterBodtke wrote:
> Chris, Kevin and Sam,
>
> Writing this note to you because you have done work to enhance VistA toward
> multiple language support, other are welcome to comment as well. Kevin's
> recent decision to stop supporting for multiple languages in TMG CPRS has
> give me great pause. I certainly understand he reasons and do not fault him
> one bit. Several requests were posted ask Kevin to continue, but I didn't
> see any offers to fund the effort. The whole Internationalization concept
> seems like a chicken or egg problem, if the work was done there would be a
> huge market, without the work already done no one is seriously interested
> in adoption...so either happens.
>
> I am 8 days away from my next trip through Central and South America, 6
> months overland, touching every country in Central and South America. The
> last time I travel in Latin America (2008-09) I had meetings with
> government offices (Belize, Costa Rica, El Salvador and Columbia) and
> universities (Guatemala, El Salvador, Costa Rica and Panama) where VistA
> was demonstrated to great interest, but no traction gained. The biggest
> stumbling block, in my mind, was the lack of Spanish support (or any
> non-English language...)
>
> Here are the questions that keep me up at night:
>
> - Is it possible and practical to create an Internationalization process
> - Is there a sustainable road map to incorporate new enhancements in the
> foreseeable future?
> - Is VistA too complex and too embedded with English for practical
talked about what could be supported in terms of characters. The thread
started "VistA and UTF-8" which should help you find the emails.
Good to see you back on the mailing list!
--
Nancy Anthracite
On Monday, August 27, 2012, PeterBodtke wrote:
> Chris, Kevin and Sam,
>
> Writing this note to you because you have done work to enhance VistA toward
> multiple language support, other are welcome to comment as well. Kevin's
> recent decision to stop supporting for multiple languages in TMG CPRS has
> give me great pause. I certainly understand he reasons and do not fault him
> one bit. Several requests were posted ask Kevin to continue, but I didn't
> see any offers to fund the effort. The whole Internationalization concept
> seems like a chicken or egg problem, if the work was done there would be a
> huge market, without the work already done no one is seriously interested
> in adoption...so either happens.
>
> I am 8 days away from my next trip through Central and South America, 6
> months overland, touching every country in Central and South America. The
> last time I travel in Latin America (2008-09) I had meetings with
> government offices (Belize, Costa Rica, El Salvador and Columbia) and
> universities (Guatemala, El Salvador, Costa Rica and Panama) where VistA
> was demonstrated to great interest, but no traction gained. The biggest
> stumbling block, in my mind, was the lack of Spanish support (or any
> non-English language...)
>
> Here are the questions that keep me up at night:
>
> for FIOA VistA (including available clients)?
> - What languages would a "conversion" process support in the deepest
> recesses of the server, i.e. Spanish? (extended character and all...)
> - How large would the effort be in hours and dollars?
> - Is there a sustainable road map to incorporate new enhancements in the
> foreseeable future?
> - Is VistA too complex and too embedded with English for practical
Kevin Toppenberg
Aug 27, 2012, 10:38:11 PM8/27/12
to hard...@googlegroups.com
Peter,
I agree that it is good to have you back around. Are you travelling by motorcycle again?
I don't have all the answers to you questions, but I can just say that it is too big for me. I think it may be easier to teach them the bare English they need to know to run the menus than it would be to try to safely convert all of CPRS (and server code) to another language--and stay in the patch stream.
Best wishes in your travels.
Kevin
Peter Bodtke
Aug 27, 2012, 11:32:51 PM8/27/12
to hard...@googlegroups.com
Nancy, Thanks for pointing out the thread. I spoke to Larry several weeks back and character set support was an open questions. I never left, just lurking for sometime. I will track down the thread mentioned and hope it
is digestible by my semi technical brain. Perhaps a call to Larry will
be the fastest route to understanding.
Kevin, I wrestle with the "teach'um English" approach, but don't it works in third world countries were nurses do a lot and those with higher skill sets get hired away. Better to provide a system in the native language. VistA is huge, clearly bigger that any person or small group of people for that matter. Testing is critical to ensure safety in a translated version...QA is something that the open source model doesn't typically attract. I know, I spent a few years as a manual and automated regression analyst, testing is mind numbing. Still struggling to arrive at a solution for poor counties and wondering if a simpler system is better than the best...
Yes, a new motorcycle adventure is in store. Going all they way to Terra del Fuego this time. Details offline.
cheers,
Peter Bodtke
--
Peter Bodtke
Kevin, I wrestle with the "teach'um English" approach, but don't it works in third world countries were nurses do a lot and those with higher skill sets get hired away. Better to provide a system in the native language. VistA is huge, clearly bigger that any person or small group of people for that matter. Testing is critical to ensure safety in a translated version...QA is something that the open source model doesn't typically attract. I know, I spent a few years as a manual and automated regression analyst, testing is mind numbing. Still struggling to arrive at a solution for poor counties and wondering if a simpler system is better than the best...
Yes, a new motorcycle adventure is in store. Going all they way to Terra del Fuego this time. Details offline.
cheers,
Peter Bodtke
--
http://groups.google.com/group/Hardhats
To unsubscribe, send email to Hardhats+u...@googlegroups.com
--
Peter Bodtke
r...@rcresearch.us
Aug 28, 2012, 3:05:00 AM8/28/12
to hard...@googlegroups.com, PeterBodtke
Peter;
This has been a set of questions I had been asking since 2002 when we
started WorldVistA. Everybody wants it but nobody wants to invest in
it. I worked up a methodology to extract english strings from VistA and
replace them with instrumentation to automate the internationalization
to the point where the language supported for a user was selectable by
the user. The process instruments all the routines in VistA in about 45
minutes on a 800 MHZ laptop and databases the results in the DIALOG
file. The population of the Dialog File goes up from 1900 entries to
over 165,000 entries in that time. I also have a routine that will load
the DIALOG English phrases into spreadsheets ready for the local
population to enter their translation of the English terms into their
native language. An additional routine loads those new terms as
translations for that language and that version of the English phrase.
In this way, we can load multiple languages into the DIALOG file and
have the proper language presented when that application is accessed
with the language selected by the specific user language. I am now
asking overseas to see if we can get support there for the work I
started in 2005. This is a means of keeping your VistA in the Patch
Stream. We only need to keep a version that is kept up to date, and
running the conversion routine can be run on the patched version when
done patching and replace the old routines. The old DIALOG file can be
used and the IENs for the old codes should match the new IENs.
There is more and the way that CPRS was done, is a nightmare. It is
possible that the EWD solution could be supported as well. The lack of
support in this country has forced me to consider taking it to Europe
where there is research money for a transmittable international
solution.
Best of luck; Chris
This has been a set of questions I had been asking since 2002 when we
started WorldVistA. Everybody wants it but nobody wants to invest in
it. I worked up a methodology to extract english strings from VistA and
replace them with instrumentation to automate the internationalization
to the point where the language supported for a user was selectable by
the user. The process instruments all the routines in VistA in about 45
minutes on a 800 MHZ laptop and databases the results in the DIALOG
file. The population of the Dialog File goes up from 1900 entries to
over 165,000 entries in that time. I also have a routine that will load
the DIALOG English phrases into spreadsheets ready for the local
population to enter their translation of the English terms into their
native language. An additional routine loads those new terms as
translations for that language and that version of the English phrase.
In this way, we can load multiple languages into the DIALOG file and
have the proper language presented when that application is accessed
with the language selected by the specific user language. I am now
asking overseas to see if we can get support there for the work I
started in 2005. This is a means of keeping your VistA in the Patch
Stream. We only need to keep a version that is kept up to date, and
running the conversion routine can be run on the patched version when
done patching and replace the old routines. The old DIALOG file can be
used and the IENs for the old codes should match the new IENs.
There is more and the way that CPRS was done, is a nightmare. It is
possible that the EWD solution could be supported as well. The lack of
support in this country has forced me to consider taking it to Europe
where there is research money for a transmittable international
solution.
Best of luck; Chris
rtweed
Aug 28, 2012, 6:47:04 AM8/28/12
to Hardhats
Just an FYI - EWD has a very comprehensive and sophisticated automated
multi-lingual capability built-in. It was originally developed as a
manageable and scalable solution for Xerox across Europe to support
their managed print services across 16 languages. I can provide all
the details if anyone is interested.
Whilst Xerox used translation service companies to do their phrase
translations for them, the guys at Quest Diagnostics did some tests
with it some years ago using the Google Translate APIs: they
automatically translated their entire EWD-based EzOrder application to
Spanish in about 5 minutes flat. OK some of the translations might
have been iffy, but then EWD provides all the mechanisms needed to
review on-screen and modify in-situ any translations. EWD also
provides a mechanism for managing the translation of text that is
generated at the back-end at run-time.
I've not really promoted the multi-lingual capabilities of EWD for
VistA, but it's all there under the covers and I can push out the
documentation as required
Rob
multi-lingual capability built-in. It was originally developed as a
manageable and scalable solution for Xerox across Europe to support
their managed print services across 16 languages. I can provide all
the details if anyone is interested.
Whilst Xerox used translation service companies to do their phrase
translations for them, the guys at Quest Diagnostics did some tests
with it some years ago using the Google Translate APIs: they
automatically translated their entire EWD-based EzOrder application to
Spanish in about 5 minutes flat. OK some of the translations might
have been iffy, but then EWD provides all the mechanisms needed to
review on-screen and modify in-situ any translations. EWD also
provides a mechanism for managing the translation of text that is
generated at the back-end at run-time.
I've not really promoted the multi-lingual capabilities of EWD for
VistA, but it's all there under the covers and I can push out the
documentation as required
Rob
Sam Habiel
Aug 28, 2012, 12:08:06 PM8/28/12
to hard...@googlegroups.com
Peter,
My opinion on this matter, having lived this, is that many people think of using VISTA in any place outside of the US as a problem solved using translation of strings. Not true! There are a myriad of issues; my opinion is that there needs to be massive changes to VISTA to accommodate various issues I outlined in a presentation I gave at the VISTA Expo last year. I am attaching that for your reference.
One of the issues I don't mention in the presentation is ICD-10 which runs all over the world but not in the US.
For a budgetary number, I would estimate that adapting VISTA for international use would cost something in the order of 100 million dollars over 10 years; each hospital implementation would cost about 5 million dollars as a base line. In order to fund development, I would say that each hospital will need to really pay 5 million dollars more to continously fund development. Remember that Fileman and the Kernel would by and large work okay in different places around the world; the issue is mainly the clinical applications.
Sam
My opinion on this matter, having lived this, is that many people think of using VISTA in any place outside of the US as a problem solved using translation of strings. Not true! There are a myriad of issues; my opinion is that there needs to be massive changes to VISTA to accommodate various issues I outlined in a presentation I gave at the VISTA Expo last year. I am attaching that for your reference.
One of the issues I don't mention in the presentation is ICD-10 which runs all over the world but not in the US.
For a budgetary number, I would estimate that adapting VISTA for international use would cost something in the order of 100 million dollars over 10 years; each hospital implementation would cost about 5 million dollars as a base line. In order to fund development, I would say that each hospital will need to really pay 5 million dollars more to continously fund development. Remember that Fileman and the Kernel would by and large work okay in different places around the world; the issue is mainly the clinical applications.
Sam
Wolfgang Giere
Aug 28, 2012, 1:01:26 PM8/28/12
to hard...@googlegroups.com
Dear all,
We are, again, trying to get an interested hospital to become a testbed plus a stable organization to develop and support a German version of WorldVistA. We do have some hope that this time the effort might be successful.
For Germany some of the problems listed by Sam Habiel do not exist. Others are solvable, partly with the help of the Xerox tools integrated in EWD (Thanks, Rob!). But it would be a huge project which would need your help!
My personal background is in thesauri, automated translation of narrative text and classification: My babies are the ICD-10 thesaurus (used in Germany, Austria and Switzerland) and half of the official German OPS procedure codes.
I hope we can join efforts to provide an English-English and a German version of World VistA. We are actually trying to get funding. How ist the situation in GB?
Btw: Are there installations in Israel or a Jewish hopital?
Best regards
Wolfgang
We are, again, trying to get an interested hospital to become a testbed plus a stable organization to develop and support a German version of WorldVistA. We do have some hope that this time the effort might be successful.
For Germany some of the problems listed by Sam Habiel do not exist. Others are solvable, partly with the help of the Xerox tools integrated in EWD (Thanks, Rob!). But it would be a huge project which would need your help!
My personal background is in thesauri, automated translation of narrative text and classification: My babies are the ICD-10 thesaurus (used in Germany, Austria and Switzerland) and half of the official German OPS procedure codes.
I hope we can join efforts to provide an English-English and a German version of World VistA. We are actually trying to get funding. How ist the situation in GB?
Btw: Are there installations in Israel or a Jewish hopital?
Best regards
Wolfgang
Reply all
Reply to author
Forward
0 new messages