Difference between revisions of "Internationalization"

Revision as of 16:38, 29 September 2009

INTERNATIONALIZATION of VISTA

The goal of this page is to collect information, plans and resources about internationalization of VistA and clients in one place. Feel free to add details to the descriptions below, insert links, note new stakeholders, share opinions, or anything you believe will advance the effort to support multiple languages in VistA.

Client Support for Spanish

There have been two translation effort for CPRS. In 2007, David Fonseca (University of Costa Rica) completed a manual Spanish translation of the field labels (the version of CPRS indeed quickly became outdated.) In 2008, Dr. Kevin Toppenberg's used a plug-in that copied and replaced the field labels with markers. The field label values were translated and made available in database, along with the original English labels. The language displayed in the field labels depends on a user selection. These translation projects did not address the significant translation development effort that remains within the VistA server.

A demonstration of enhancements delivered in TMG CPRS can be watched here: http://www.youtube.com/watch?v=ojcIoIy7HC4

The Astronaut WorldVistA and TMG CPRS installer can be downloaded here: http://sourceforge.net/projects/worldvistaautoi/files/

Kevin used the DKLang Localization Package to edit the language file: http://www.dk-soft.org/products/dklang/index.php?lang=en I believe Kevin also used another tool to automatically copy the field labels and insert code that would support the display manipulation. There is a caveat, the version of CPRS that Kevin enhanced is now one version old. There are many additional features that Kevin added to TMG CPRS that you will hopefully enjoy and find useful.

Medsphere's CIS

For web enabled interfaces a similar approach was taken by Medsphere. Their product, CIS, uses the open source development framework, Mono. Demonstration work has been done leverage Gettext routines to support display of multiple languages. It is not clear to me if the development work is limited to the CIS interface code with some changes to the OpenVistA or if the enhancements that support internationalization were extensively made to the OpenVistA server. I suspect the former.

Links describing Medsphere's work:

Internationalization (i18n) and Localization (L10n) http://medsphere.org/community/project/openvista-cis/blog/tags/translate

OpenVista CIS Features and Functions http://medsphere.org/docs/DOC-1202

Again, as far as I know, the translation process is only just starting with CPRS or CIS, as much more work needs to be done on the server side.

DSS Translation Effort

DSS attempted a full system translation into Spanish and Portuguese. The project was reported to have been put on the back burner by the winter of 2008.

MSC FileMan

George Timson's MSC FileMan (a.k.a. FileMan 23?) has internationalization capabilities which leverage the DIALOG files. It is assumed that an effort to fully leveraging MSC FileMan's internationalization potential would require considerable effort and resources, for integration coding, translation and quality assurance. Rick Marshall commented that such an effort would create fork like no other in VistA. I suspect that the extent of code modification would reach far beyond the realm of KIDS patches, but others would need to comment on this point. Integrating MSC FileMan would be very attractive as it would support additional languages for all of the 'roll-and-scroll' applications and display text parsed by GUI applications. [1] (updated 8/2009)

Automated Tools

Chris Richardson developed a tool which automates the selection and replacement of strings for translation purposes in VistA. Richardson's code will parse all of the VistA routines and populate the DIALOG file. Once this is accomplished, then the problem is generating spreadsheets that can be used to make the parallel translations possible. That code exists as well. Basically, the process identifies literal strings in English and makes sure that they are in the DIALOG file. Then the literal is replaced in parallel, generated instrumented code. The selection of the languages seen is selected by the user.

Rick Marshall has commented that the tool is useful, but significant additional development is required to complete the work. I'm not sure if Richardson's tool was ever released to Sourceforge or TRAC. I believe it was shared internally with some of the leadership, but I have lost track of it status.

Interested Parties

Mexico - two point of interest: IMSS experienced issues integrating translated catalogues and they leveraged middleware to support Spanish

America, South America and Spain - yes, there are many interested parties (government ministries, educational, and commercial), but none that have shared or indicated a commitment to developing VistA for use in these regions.

OHUM - has a long term objective to complete a translation effort, but it is not on their immediate schedule.

Jordan - has a goal to support Arabic in a limited number of user interfaces

Central American Universities that have shown interest in VistA: Universidad de Costa Rica (San Jose, Costa Rica) Instituto Tecnológico de Costa Rica (Cartago, Costa Rica) CONARE (San Jose, Costa Rica) Universidad Technologica de Panama (Panama City, Panama) Universidad del Valle de Guatemalla (Guatemalla City, Guatemalla) Universidad Centroamericana "Jose Simon Canas" (San Salvador, El Salvador)

Related Resources

Google Group "VistA Latino" http://groups.google.com/group/vista-latino?lnk=srg&hl=es This group is dedicated to discussing issues related to VistA in Latin America. Topics of interest include supporting Spanish and Portuguese in VistA.