It can be confusing reading Fred’s blog but in case it is not clear - PatientOS is not a fork of anything, it is a unique application in every way - architecture, user interface, data model etc and workflow perspective (an older document is described here).

There are no plans to fork PatientOS (hopefully Fred will not do that despite his history as a major forker) at any time.

I am currently supporting one open source service company install the first site and I am more than happy supporting any number of small or large businesses looking to meet their clients needs.

You do not need to fork the code to customize the system - all of the UI is defined in the database, there is a lot Javascript scripting to customize interactions, there is a rules engine, Mirth (which I love), and much more.

With the next release the capacity to write plugins to overhaul the UI or workflow is being added.

Ultimately people will be able to write whole modules as plugins so as not to bloat the code base and to increase integration opportunities with existing systems.

I understand that PatientOS is new development, but Fred's point seems to be that the big issue is splintering development effort and that your project is attracting Java developers who might instead be working on Tolven.

My point is that it seems kind of like a "pot, kettle, black" situation.

It can be confusing reading Fred's blog but in case it is not clear - PatientOS is not a fork of anything

You are right. That is not clear and I will fix this.

it is a unique application in every way - architecture, user interface, data model etc and workflow perspective (an older document is described here).

This is where we fundamentally disagree, in fact the only thing that is demonstrably unique about your approach is that you insist on using a thick-client. Everything else is just standard Java best practices, plus the latest-greatest tools. In short the "architecture" has no significant improvement on any other Java project.

There are no plans to fork PatientOS (hopefully Fred will not do that despite his history as a major forker) at any time.

Just to note, I have forked a project only once, away from FreeMED. (Which is, BTW, the most forked away from EHR project ever...) MirrorMed is not a fork of ClearHealth, rather it is merely my trademark for the same codebase. Ironically, when I published FreeB, it became the most adopted cross-project codebase ever (I am happy to say that Mirth will soon win that distinction from me). So most of my programming efforts have been towards projects where I am not the project manager. My money (and code and time) is where my mouth is. This is another issue that I will have to make clear in my article.

-FT

• Re: Outing Patient OS by Greg Caulton on Monday December 03, 2007 @ 11:46 AM

  I think you are confusing software architecture with technologies - just because two projects use PostgreSQL, Hibernate and JBoss - that does not mean they have the same software architecture. Software architecture is far more complex - it is the relationship between the packages, systems, layers. It is the function of the core classes that define how the system works. It is the design patterns that control the inner workings. It is often influenced by the relationships of the tables. The software architecture significantly affects how the system can handle change especially with complex workflow, internationalization, localization, translation to multi-byte languages etc. Unless you look at the code, learn how the applications work, and effectively learn the system you cannot compare and contrast. Now briefly looking at OpenMRS, I can see it has a professional architecture, a strong structure, strong data model that is in line with the excellent reputation of Regenstreif. Through the architecture PatientOS dynamically generates all of the Swing code in real time based upon database settings. You can use the GUI to change any form - even the tools themselves (the recursive effect is interesting). Thus how we manage and interact with the system is fundamentally different. For example I just read this post on adding a tab while in OpenMRS you edit files, in PatientOS (hey they rhyme) you use the Dialog Builder - a rich client GUI to navigate the dialog hierarchy and add a new panel of type tab. You can customize menus, toolbars, controls on a form, add scripting behind controls, all through the GUI. All of that one could add to OpenMRS no doubt - but the design patterns are very different (and I am not saying one is better than the other - impossible to say while the products are so young). OpenMRS and other web applications are going to shine brighter than PatientOS in developing world as the client (browser) requirements are a tiny fraction of the processing power of the PatientOS Swing client. Now shiny ball in the form of Ultimate EMR is looking slick for an Open Source EMR - but the license link is not working for me...

  [ Reply to this ]

• confusing architecture and technology by Fred Trotter on Wednesday December 05, 2007 @ 12:00 AM

  It is certainly possible to have different architectures based on different technologies. However, there are limitations to this. Building an application that uses a hierarchical database is easier in MUMPS than in Java. Building a web-based interface is easier in php, etc etc. Your points about architecture are both valid and interesting, but they do not justify your actions, which is to attempt to draw developers to your "design" away from the Tolven/OpenMRS "designs". Both of those "designs" are mature, live systems. Your "design" is an experiment. If you were saying "hey I have this alpha system that has an interesting design, come help me with it" AND you mentioned that both Tolven and OpenMRS both probably justify at least 1.0 status, then I would have no problem with what you are doing. This is where the disconnect seems to happen. I am trying to enhance the communities ability to move forward, and I am grudgingly willing to accept second-rate designs to make that happen. You care about your personal preferences regarding design so much that you are willing to cripple the communities ability to move forward. So I will repeat what I told you over the phone regarding this issue. I hereby concede all design points to you. For the sake of this argument I bow down (virtually speaking) to your infinitely superior software architecture skills. From that position... you are still wrong to do what you are doing. There MUST be a way for you to get what you want regarding design, without starting yet-another-alpha-EHR, which will starve for resources, even as it steals resources from already solid projects. -FT

  [ Reply to this ]

• Re: confusing architecture and technology by Take It All Back Already on Wednesday December 05, 2007 @ 10:08 PM

  It is dreadful to die of thirst in the sea. Do you have to salt your truth so much that it can no longer even ‑ quench thirst? -Nietzsche PatientOS may never be what you don't want it to become, but Caulton is certainly not stealing. For you to take the position that he is stealing, well, you are wrong to do what you are doing. If you have such evidence, present it. Otherwise, give it up. Consider, am I "stealing resources" if... I buy a single copy of a CD and copy some or all of it to my computer and my iPod? I take my son to play football at the park, and we don't join the other guys who are already there, with whom we play sometimes, and who want extra players this particular day? I start a new church even though there are five established churches of very similar faiths within a mile radius that I could be attending instead? I have work done on my lighthouse using contractors from across town, knowing full well that my 4 immediate neighbors, with whom I am friends, consist of a professional plumber, painter, mason, and carpenter who could really use the work? I open a new gas station at an intersection that already has gas stations on two other corners? I drive the city streets instead of the tollway? I spend a hundred thousand dollars on cancer treatment for my wife, rather than donating it to the American Cancer Society? I give thousands of dollars to Salvation Army each year, and won't give a single hour of my time to the shelter that is in my immediate neighborhood? I spend money running for President of the U.S. with absolutely no chance of winning, when I could have donated the money to the candidate I would most prefer to win? I get my own hotel room while on the road with other employees, in violation of the company policy to bunk up in pairs to save on costs? I decline being a backup singer in Faith Hill's band, only to end up singing a hit song on American Idol, and within the next year beat Faith Hill out for every single award category she had been winning the last four years running? I give money to my brother-in-law who is mentally ill as a result of too much drug use in the past, but won't give a cent, not even toward treatment, to my own brother who is still on drugs? I buy discounted, round-trip airline tickets to Hawaii for two on a honeymoon special, with full intentions of flying out for business with my (non-spouse) business partner? I join the Peace Corp and serve in Somalia, consciously opting not to join the much more organized local Guard unit to assist during natural disasters here in the States or the military to go fight in Iraq? I do something with eager confidence, simply because I can and I choose to?

  [ Reply to this ]