[Gnumed-devel] Re: Draft outline for some local programming help I should try to find

[Syan Tan]

On Sun Oct 2 21:52 , J Busser sent:

> Hi Syan
>
> Could you have a look at what I have written below as an outline of 
> some help (work) that might be useful for me to have done locally? 
> The open source projects EGADDS and TAPAS are being done in Postgres 
> with Java, so these people are potentially available to do a little 
> extra work toward that anticoagulation project for which I have some 
> limited funds.
>
> Would you suggest any major revision to the parts I *think* I am looking for?
>
> Also, I am still finding my way uncomfortably through this issue of 
> having a bit of funds available. I imagine most gnumedders have 
> adequate base salary/income, and may not want the responsibility of 
> personally providing code which becomes any obligation, and if they 
> did I may not afford their rates (:-)) but if you think it anything 
> worth discussing, please do.
>
> ============================
> Part One: GNUmed database setup
>
> - setting up / configuring PostgreSQL on a debian sarge server that I 
> have in my basement
> - installing and configuring GNUmed databases onto it
>    - 0.1 schema version
>    - development (CVS) version
> - documenting the procedure and maybe improving existing documentation
>
> --- I don't know whether the above could (e.g.) be done in a day, 
> 7-10 hours. Postgres is I am sure available as a debian "deb" binary, 
> and the gnumed server code is approaching also being available as a 
> binary (or doable from CVS). There is also some documentation but it 
> may need improvement.

 

 - Can Carlos help ? He has experience in settiing up remote ssh accounts,

  so that people can do remote work on servers. Perhaps he can describe it as a walkthrough.

 

-  I can  post a walkthrough of my recent experience with installing a debian system

upto setting up a gnumed_v2  database , using the alternative of loading a partial global dump and gnumed dump

from the salaam postgres instance, if this is any help.

 

 - where is the url for info about EGADDS and TAPAS ?

 

 

 

 

============================
Part Two: PathNet lab results import interface

- examine the OSCAR code rewrite of the Andromedia module
(hopefully it will soon be done, at least the connection portion)
- estimate the work required to modify in Java or Python to
- - permit it to connect from Linux (assumes current is Windows-specific)
- - import into GNUmed schema

--- I believe the Andromedia module was in Java, as I presume is the
rewrite. Although GNUmed is primarily in in Python, it may be
workable to have a Java module doing the lab importing, and that may
be the easiest first step if that is the main expertise of those
locally available. However, if the logic has been figured out for
PathNet making it easy for a person to redo in Python, that too would
be of interest. The GNUmed team could certainly help with the
Postgres part though if your Postgres person were available that
could also be available. Now I believe it took the Andromedia
consultant a full week to do the original work. Maybe it is illusory
to think it could take less but an examination and estimate could be
the first step.

 possibly could help; you sent  test data and adromedia code in java, which

was translated to python, and a simple parser that converted the hl7 v2 data in the test data

into  print key:value pairs. Perhaps  a parser could create python map objects representing a result element,

 order request , and message subject headers, and these more conveniently inserted into gnumed schema.

what would be needed would by identities in a gnumed database with names, date of birth, or pupics , that match test data message subjects.

The link mechanism would also need to be specified ( e.g. ftp, sftp, email ,  periodic polling ; 

the recent ozdocit.org   wagmail  is convenient, as it can poll email accounts and do some verification/decryption

before dumping email with attachments to  a  configured data directory.

a modification here might be use different dump directories depending on which verified email source address,

so that email from verified path source ,  will be directed a directory which is read by another process

which imports files from that directory using the python port of the adromedia parser  , into gnumed results

schema ). 

=============================
Part Three ???

Availability on some kind of low level basis, helping me to maintain
it, install updates & patches etc. Also I understand Postgres
databases benefit from some maintenance like "vacuuming". And maybe
troubleshooting. Since mine would not at this stage be a live /
production installation nor a primary test unit it can run crippled,
if that should happen, for a week or two at a time by which time.
Maybe 2 hours a month?

if Horst and Carlos can give you tips on how to setup and manage remote access ssh accounts,

I think there will  be plenty of people who could help.