Experimental Toaster - WorldVistA EHR on GT.M in UTF-8 mode

I have created an *experimental* Toaster (software appliance, virtual machine) of WorldVistA EHR VOE/ 1.0 on GT.M V5.3-003 on Ubuntu Jeos 8.04 - *running in UTF-8 (Unicode) mode*. See http://www.fidelityinfoservices.com/user_documentation/html/rn_tb/GTM_Unicode_Support.html for information about GT.M Unicode support. Instructions for use are the same as for the VistA Tour CD (http://downloads.sourceforge.net/fis-gtm/VistATour20081118.iso) except that the file name of the virtual disk image differs.

To use a Toaster requires Qemu (on the VistA Tour CD and elsewhere). On Linux on CPUs with hardware virtualization, kvm can also be used to run it. I am deliberately not posting detailed instructions here because I want only those who already know how to run Toasters to try this.

Execution of MUMPS code compiled for UTF-8 mode is guaranteed to be slower than for M mode. This is because operations such as $Piece() must actually parse the string, and cannot optimize based on a character being the same as a byte. I don't know whether this performance penalty in practice is vanishingly few microbleems, or whether it is perceptible. Please give me your feedback.

For some reason, CPRS connection attempts fail. I don't yet understand why. But the aulde roll and scroll works.

Just in case I forgot to mention it before, this Toaster is *experimental*. Please use it accordingly.

Regards
-- Bhaskar