A Plea for Sanity

7 views
Skip to first unread message

K.S. Bhaskar

unread,
Jan 29, 2009, 7:16:42 AM1/29/09
to Hardhats
The intense discussion has convinced me that it is way to premature to
create standards for VistA implementation directories. Artificial
times and pressure for immediate responses are incompatible with good,
thoughtful standards. Ergo, this is a plea for sanity to go about
standards making in a deliberative and thoughtful way.

Ignacio, I realize that you are trying to set up an implementation.
Make what seems to you to be the best choice and proceed, as so many
others have before you. GT.M and Linux are flexible enough that you
won't be painting yourself into a corner no matter how you set it up.

If you are determined to set up a vote, please understand that it is
really a vote for what you think is the best way you should proceed in
your setup. I would encourage you to include a "None of the above" or
"I want to think about it some more" option. Instead of a vote, you
may as well make your choice and proceed. That is my counsel.

For anyone who wants to implement VistA, please consider placing an /
etc/<vistaflavor>_<identifier> file or directory that can be sourced
to provide the environment variables needed to define your VistA
implementation. Alternatively provide a /usr/local/bin/
<vistaflavor>_<identifier> shell script, or directory that contains a
run shell script.

Regards
-- Bhaskar

Brian Lord

unread,
Jan 29, 2009, 7:51:30 AM1/29/09
to Hard...@googlegroups.com
Bhaskar, I've kept quiet on this point even though as most know we have
adopted standards for our deployments. Your post however I think is a
perfect example of what the role of WorldVistA is perceived as and what
the members of World VistA need.

I agree with you a fast answer is not always a good answer, but no
answer is what people are dealing with. Yes VistA and Linux are
flexible, but the confusion about how to setup a system is visible on
the hardhats list every week.

The members and other interested parties need guidance. If WorldVistA
were to create a standard, and then realized that something about there
standard caused problems they could then correct that and everyone would
know that the standard was updated and people should follow this. To use
your own words "Release Often" would correct any problems.

However to have no standard at this time when once again eyes are
turning back to VistA gives the perception that WorldVistA EHR is still
in the wild west days and people are looking for a solution that they
can trust. At this point whether it is by consensus or by a group of
folks that put together the first WorldVistA EHR standard setup
protocols it really doesn't matter. Peer review will correct problems
and WorldVistA will be looked at as taking the lead in sorting out the
confusion.

I know that there are people on this list that have participated in
dozens of implementations. You yourself have a standard that you created
for your VistA appliance. Putting out one of these standard
implementations would be the first step in formalizing this process
which could then be more and more refined. However until someone from
WorldVistA publishes a standard there is going to be a desperate need
among the community for someone to help make sense of all the pieces.
Pieces, that if there were a standard, implementors would never again
need to worry about unless they generated a problem, which as you
pointed out, most would never be a problem at all.

I agree we need some sanity, my suggestion is to take a WorldVistA
certified implementation that is currently in long term use, (we have
several and David Whitten and Larry Landis are very familiar with them)
start with that and get it put out to the group.

Changing something that is wrong is much easier than trying to draft
something from scratch.

Just my attempt at sanity, and you must forgive me I'm not very good at
sanity these days.

"A wise man learns to understand, and obey nature. Therefore all
innovation shall come from unwise men." Abba Eban

fred trotter

unread,
Jan 29, 2009, 9:13:45 AM1/29/09
to Hard...@googlegroups.com
> The intense discussion has convinced me that it is way to premature to
> create standards for VistA implementation directories. Artificial
> times and pressure for immediate responses are incompatible with good,
> thoughtful standards.

I think this is, to a great extent, a false dichotomy.

We (the community) have been discussing the need for a standard
installation pattern for years, perhaps even decades. So slow,
deliberate, repeated discussions actually do not work either.

The problem here is that it is entirely possible to fully understand
and respect an opponents opinion on VistA layout and still disagree.
As a result, this issue has been in continuous deadlock. Each
appliance or installer attempt has gone their own way.

Debian can afford to be differently laid out than Fedora, because both
projects have enormous resources and happily work in parallel. Not
being sure of "where things go" is an impediment to us collaborating
together, and while the WorldVistA community is pretty large it is not
so large that you can afford to have issues like this continue to
stand in the way of collaboration.

Ignacio must do something, he has something to build which relies on
this. He feels, even if the community does not, that it needs to be
built soon. I, and apparently Brian, agree on that point.

So far, repeated discussions have failed. We have tried the slow,
deliberate way and no good, thoughtful standards have appeared.

I think Ignacio can tolerate a Vista Standards Layout version .9 so
that we can officially acknowledge that the discussion is not over.
But we should continue to move at this pace. The choice is not between
slow and fast, it is between fast and not happening at all.

-FT


--
Fred Trotter
http://www.fredtrotter.com

r...@rcresearch.us

unread,
Jan 29, 2009, 9:15:28 AM1/29/09
to Hard...@googlegroups.com
Fred and assembled multitude;

I agree with you, Fred. We need a mark in the sand and try it. If it
works, we can build a trench. If it doesn't, then we move the mark.
This is in the time-honored tradition of the VA, "Ready, Fire!, Aim".
In these situations where we are majoring in the minors, we need to
just do it. Establish a standard, and sit back and collect the
impressions people have. Any really good reason may be enough to
change our minds, (but it has to be a good reason). We can always
modify the standard later. This is needed so that we can start
building tools that know where to get information about the
environment. Without the guide-lines, these tools become impossible to
build.

Larry, you have the Chief Technical Officer role. Make a decision and
publish. You now have input from a bunch of strong willed people. You
cannot please everyone, but you please most of them by making a
decision and publishing. Let's move on this. If we can stay
professional about this, we can move to a situation where everyone is
served by making the move.

Best wishes; Chris

David Whitten

unread,
Jan 29, 2009, 11:15:10 AM1/29/09
to Hard...@googlegroups.com
On Thu, Jan 29, 2009 at 8:15 AM, <r...@rcresearch.us> wrote:

Fred and assembled multitude;

  I agree with you, Fred.  We need a mark in the sand and try it.  If it
works, we can build a trench.  If it doesn't, then we move the mark.
This is in the time-honored tradition of the VA, "Ready, Fire!, Aim".
In these situations where we are majoring in the minors, we need to
just do it.  Establish a standard, and sit back and collect the
impressions people have.  Any really good reason may be enough to
change our minds, (but it has to be a good reason).  We can always
modify the standard later.  This is needed so that we can start
building tools that know where to get information about the
environment.  Without the guide-lines, these tools become impossible to
build.

  Larry, you have the Chief Technical Officer role.  Make a decision and
publish.  You now have input from a bunch of strong willed people.  You
cannot please everyone, but you please most of them by making a
decision and publishing.  Let's move on this.  If we can stay
professional about this, we can move to a situation where everyone is
served by making the move.
  Best wishes;  Chris

A minor point, I (David Whitten) have the Chief Technical Officer (CTO) role in
WorldVistA. Larry Landis is the Chief Information Officer (CIO).
That being said, I think that roles are not the real issue here,
but rather, community agreement is the issue.  Simply declaring an
answer by Fiat is not going to make sure that good input such as from
Bill Ackerman, Bhaskar, and Jonathan Tai is heard by the community.
Several us setup a "de facto" standard as long ago as the 1st VistA
Community Meeting (VCM) in 2001.  I think it is appropriate for
everyone's voice to be heard.

Dave

David


ival...@gmail.com

unread,
Jan 29, 2009, 11:36:05 AM1/29/09
to Hardhats
Perhaps the difference is that this time:

1) There is specific objectives.
2) All 140+ postings, several VM appliances, Nancy's setup and your
proposal have been read, considered, thought about sensitively, and
empathic-ally and incorporated into the proposal wherever possible.
3) Consensus from this difficult but necessary process appears to be
emerging.

-- IV

On Jan 29, 6:16 am, "K.S. Bhaskar" <ksbhas...@gmail.com> wrote:
> The intense discussion has convinced me that it is way to premature to
> create standards for VistA implementation directories.  Artificial
> times and pressure for immediate responses are incompatible with good,
> thoughtful standards.  Ergo, this is a plea for sanity to go about
> standards making in a deliberative and thoughtful way.
>
> Ignacio, I realize that you are trying to set up an implementation.
> Make what seems to you to be the best choice and proceed, as so many
> others have before you.  GT.M and Linux are flexible enough that you
> won't be painting yourself into a corner no matter how you set it up.
>
> If you are determined to set up a vote, please understand that it is
> really a vote for what you think is the best way you should proceed in
> your setup.  I would encourage you to include a "None of the above" or
> "I want to think about it some more" option.  Instead of a vote, you
> may as well make your choice and proceed.  That is my counsel.
>

Actually it is a vote for what the community thinks is best as the
proposal contains elements from multiple sources. The specific vote on
the preferred <base-dir> will be because there is contention over it.
Like you, my bias is for /opt but I can live with /opt, /var/opt, /
home/<userid> and even /opt/mumps as that is far more specific than
what is there now which is nothing.

> For anyone who wants to implement VistA, please consider placing an /
> etc/<vistaflavor>_<identifier> file or directory that can be sourced
> to provide the environment variables needed to define your VistA
> implementation.  Alternatively provide a /usr/local/bin/
> <vistaflavor>_<identifier> shell script, or directory that contains a
> run shell script.

That seems to already be in the VSB proposal:

/<base-dir>/<branding>vista/<instance>/etc # Version, release
and other configuration information.
/<base-dir>/<branding>vista/<instance>/etc/vista.conf # contains
version, release and other config information.
/<base-dir>/<branding>vista/<instance>/etc/patch_listing.txt #
Optional listing of patches for this instance.
/<base-dir>/<branding>vista/<instance>/etc/env # for gtm/VistA
environment variables.

/<base-dir>/<branding>vista/<instance>/bin # for server controller
type software.
/<base-dir>/<branding>vista/<instance>/bin/run # Bhaskar'sconvention,
run an instance.
/<base-dir>/<branding>vista/<instance>/bin/vistastart # Bhaskar's
convention.
/<base-dir>/<branding>vista/<instance>/bin/vistastop # Bhaskar's
convention.

-- IV

Nancy Anthracite

unread,
Jan 29, 2009, 12:58:09 PM1/29/09
to Hard...@googlegroups.com
Larry is the CIO, not the CTO.
--
Nancy Anthracite

r...@rcresearch.us

unread,
Jan 29, 2009, 3:38:34 PM1/29/09
to Hard...@googlegroups.com
David;

You are correct. You are the CTO. As such, it is more than appropriate
for you to publish an RFC so that there is a firm document that people
can argue over. Right now, we are all shooting at ideas floating is
soup. That gets nothing done except a lot of heat by stirring the soup.
Withan RFC out there for people to all point to and say, I like this,
and I don't like that, we can at least get the issues and the reasons in
a form that others can study and make comment on. You can republish the
RFC any time you feel that the model has changed (or should be changed).
This at least moves us from soup to thick oatmeal that can finally
reflect all of the issues and provide the basis for a real consensus.
Keeping everyone in solution is not getting us anywhere but around and
around the bowl.

Bon Appetite; Chris

fred trotter

unread,
Jan 29, 2009, 3:54:26 PM1/29/09
to Hard...@googlegroups.com
I would like to point out that Ignacio is not merely deciding here....
The last changes were specifically designed to address the points that
Bhaskar has been making technically.

We are reaching a consensus here.

Bhaskar, does the recent split between the "Enterprise" version of the
VSB and the small office setup make you more inclined to support the
overall design?

ival...@gmail.com

unread,
Jan 29, 2009, 4:01:33 PM1/29/09
to Hardhats
False. This is progressing very rapidly with a steady forward linear
progression and in an organized fashion that will be voted on by the
community very soon. The document is being organized by me with
massive public input from multiple sources like LSB, reviews of all
available appliances and much community generated documentation and
proposals. It is now in its 10th public draft form. I am not wishing
to say that I know everything, impress or boast, but to impress upon
this community that I have every qualification to do this: I am a
worldvista member, I have a BS, MS in computer science, MD and years
of experience as a software engineer for IBM, Compaq and others. I
have used VistA in the VA. I have personally, successfully, but
incompletely as yet implemented VistA in a real private hospital
system. While I would like to have more experience with VistA, I think
I have enough experience with it and Linux as well as the motivation
to get this part done and it is getting done. This is a small but
essential part of much more work to do.

-- IV

fred trotter

unread,
Jan 29, 2009, 4:15:38 PM1/29/09
to Hard...@googlegroups.com
Perhaps we can tentatively agree that the VSB that Ignacio is
publishing as .9 will be automatically considered as DRAFT 1.0 with
RFC. This way we can both slow down and speed up.

I have done some work revising documents for CACert.org and that
experience has taught me that a formal tool for taking document
comments is a good idea. I recommend that we use co-ment.

http://www.co-ment.net/

At this stage worldvista should -not- host an instance, but knowing
that we someday could makes the software safe to for us to use.

At this stage the ratio of comments about the process of taking
comments are greatly out-numbering the comments about the actual
proposal.

Can I please propose that we accept the compromise of a slower process
for the "1.0" version and lets get back to actually discussing the
merit of Ignacio's most recent proposal?

David, while I would not think that you should in any way squash
community comment, an endorsement/decision from you about how to move
forward collecting those comments for the "next" version might calm
some of this argument down... so that we can actually focus on which
blessed directories that VistA actually belongs in?

Does any one have a significant problem with my basic proposal? Here
it is in short:

* Allow Ignacio to move forward by momentarily shifting focus to
actually discussing the contents of his proposal.
* Once this process is complete, make that document both the official
.9 version and the 1.0 DRAFT. (I think what Ignacio is looking for is
-broad- insights on what might be missing from his proposal)
* Setup a mechanism for formally collecting and commenting on very
specific portions of the DRAFT text, using stet, co-ment or equivalent
Document commenting tool.

Is that a passable way to address the concerns of both those who want
to finish soon and those who want to further discuss? This same basic
structure would work for any WorldVistA document written in the
future, and is essentially the same way that FSF writes its new
licenses.


-FT

K.S. Bhaskar

unread,
Jan 29, 2009, 4:25:54 PM1/29/09
to Hard...@googlegroups.com
Fred --

I have expended as much time and energy on this topic as I can spare at
this time. I respectfully withdraw from the discussion. I will update
my recommendations in the future based on my understanding of best
practices and I certainly will be looking for good ideas wherever they
may come from.

But for now, on this topic, this horse's bolt is shot.

Regards
-- Bhaskar

_____________

The information contained in this message is proprietary and/or confidential. If you are not the
intended recipient, please: (i) delete the message and all copies; (ii) do not disclose,
distribute or use the message in any manner; and (iii) notify the sender immediately. In addition,
please be aware that any message addressed to our domain is subject to archiving and review by
persons other than the intended recipient. Thank you.
_____________

ival...@gmail.com

unread,
Jan 29, 2009, 4:53:28 PM1/29/09
to Hardhats
Hang in their cowboy! I think we are 2 votes away from gettin 'er
done. I just put the call out for the vote on <base-dir> We are nearly
there. -- IV

Jonathan Tai

unread,
Jan 29, 2009, 6:44:29 PM1/29/09
to hard...@googlegroups.com
On Thu, 2009-01-29 at 13:15 -0800, fred trotter wrote:
> Does any one have a significant problem with my basic proposal? Here
> it is in short:
>
> * Allow Ignacio to move forward by momentarily shifting focus to
> actually discussing the contents of his proposal.
> * Once this process is complete, make that document both the official
> .9 version and the 1.0 DRAFT. (I think what Ignacio is looking for is
> -broad- insights on what might be missing from his proposal)
> * Setup a mechanism for formally collecting and commenting on very
> specific portions of the DRAFT text, using stet, co-ment or equivalent
> Document commenting tool.
>
> Is that a passable way to address the concerns of both those who want
> to finish soon and those who want to further discuss?

I've given all the technical input that I can at the current time. As
with many standards, you won't know the problems with the standard until
you try implementing it. I'd suggest Ignacio move forward with the
tools he wants to write based on the input he's received on this list
and see where it goes. If the tools are compelling, people will have a
reason to follow the standard.

- Jon

signature.asc
Reply all
Reply to author
Forward
0 new messages