Industry Zero: One man’s thought on software, business and life

Just downloaded JavaFX and played with the demo.  I am sorry to say that by looking at the codes of the demo, it seems too complicated to do some simple thing.   How come Sun doesn’t learn from some of the mistakes in Java? First, why create  a new scripting language, why not just use Ruby, Python or any existing scripting language.  If you are going to create something new, make it simple and take advantage of scripting language feature like dynamic typing. Why still does static typing?   Crazy!!!

NPR has a radio segment on Personal Health Record. Revolution Health’s PHR and trackers were mentioned in the radio segment. Since I am the Engineering/Dev Manager for the both of the products. I am very happy about it.

Let’s face it. We all want our code to be solid and be able to last long.  We code it such a way that it can  be extended in the future and we thought of all the scenarios that the requirements can change.  But then it always happened like this:  Business has changed, technologies has changed, it is better to rewrite the code from scratch then to maintain it.  Your old code is gone.  So why spend all the time thinking about the future.   Code it now.  Change is going to happen.  Change is going to be a pain. But that is the nature of the game.

There are a few open PHR systems that will allow third party to integrate with them. One is Google Health PHR which will open up their API soon. The other is an open source project called Indivo which anyone can use and allow import and export from any third party systems that will create an interface to it. The third one is Microsoft Health Vault which also allow third party to connect to it. I am trying to figure out which system is better to integrate with. So far the only system that I can take a look under the hood is Indivo. I wish Google and Microsoft will open up their API soon. Whoever can get the most doctors integrated with their system will win. There are a lot of other opportunities to provide add on services on the top the PHR system. For players who doesn’t want to fight a battle with Google or Microsoft. They should concentrate on the add-ons instead.

Personal Health Record battle is heating up. From Microsoft to Google
, and also RevolutionHealth (disclosure: I am the engineering manager for the PHR in RevolutionHealth), all trying to figure out a way to create a electronic health record system that can be managed and used by consumer, physicians and hospitals. After working on this product for awhile, I am not sure if we get to the point that we can change consumer behaviors yet. No matter how easy we can make PHR for users to use, it is still extra time that consumers needs to spend on managing it. Consumers need to see  great benefits before they are willing to manage their health records online. In fact, consumers probably don’t want to spend time to manage it. They just want to have access to it online. They will not want to import or export data between service providers. Service providers should do that for them.

I recently found this website CrimeReports.com. This website contains crime data overlaying on top of a Google map. So you can see what crimes happened at where in your neighborhood. What a great site! The best part is that their business model is not ad driven. Police department pay them to publish the data. What a cool concept!

Why does it always be so hard to upgrade to a new version of anything? There is always something that will break. I recently did an upgrade rails from 1.1.6 to 1.2.6. for a simple app. Lot of stuffs broke and it took many hours to fix them. Crazy!

1. Improve search by using delicious. You are not going to beat Google search doing the same thing that Google does. Do something different. Use users submitted link as part of the ranking algorithm. delicious contains tons of links. Find an algorithm that can take advantage that.
2. Integrate Flicker to Microsoft desktop.
3. Combine MSN and Yahoo into single one portal. Kill MSN.
4. Integrate Yahoo Answer to Microsoft desktop
5. Integrate Yahoo Music and come up with an iTune killer
6. Create SSO platform that will work for all websites
7. Buy ebay.

One of my good developers resigned to day.  I asked her about the reason of leaving.  She told me that she wanted a more stable company and the company that we are working for is high stake, high risk and she doesn’t really know if it will last.  I asked her if there is any other reason such as work environment, job satisfaction and salary.  She told me no.  As a manager, I can do what I can to improve the work environment, job satisfaction and salary.  There is one thing that I can’t promise is if a startup will success or not.  So I can’t really do something for her.   Software development is all about getting and keeping good developers.  A lot of people want to put emphasis on process.  However, software development is not a factory process even a lot of people want it to be.  It is a creative, research and learning process.  It needs people who are creative, able to learn and explore and attach new problems.  My question for the day is how do you keep good people?:

There were many efforts to create portable electronic health record and turn them into personal health record that can be controlled by users.  For example, Waltmart and a few companies are doing this thing called: Dossia.  I have been working on the similar product: https://www.revolutionhealth.com/my-health-portfolio/.  Almost everyone has this grand idea that if we can just store the health record electronically and allow them to transferred between providers and consumers, we will reduce health cost by reducing duplication of lab tests and reducing medical errors.

Sound great in theory.  The problem is how to actually convince consumers and doctors to use it.  What is the incentive for doctors to spend ton of money to put their patients record online?  So that they can easily send their patients’ records to other doctors or their patients… don’t think so.  Will electronic health records save doctor money?  Not with the existing insurance system.  Doctor would get penalize for duplicated tests.  Insurance pay for it.   Medical errors?   Insurance pay for it in case lawsuit.  How about patients?  Won’t they care?  Yes, they care about medical errors but not cost because insurance pay for it.  Furthermore, patients are not going to be able to force their doctor put their records electronically.  Nor will patient want to enter those data in themselves.

So the problem is not able providing a secure platform to store electronic health record.  The problem is that there is no business case for it under the current insurance pay system.  Unless all the insurance companies get together and say all doctors might use electronic health record by some dates in the future or the government mandate it.  We are not going to see wide use of electronic health record anytime soon.

There is another business case for electronic health record that should worth explore.  Providing health assessment, diagnosis and treatment guide based on a person’s health record.  However, how to build a business model on it is another tough problem to solve.  The key is still how you can convince doctors to spend money on it.