DML

[Doug Easterbrook]

hi Randy:

just remember that you asked the question

>> Why do developers using DML feel like they have to eventually migrate to SQL?  I have been using DML for over 25 years and it is rock solid.  I’d be interested in knowing how many developers are using DML.  Going forward, can DML and fat client take advantage of JS, web and mobile app technology?


the opinion, which I do share, is that if you are not hurting, DO NOT migrate to SQL.

but those that eventually migrate from DML to SQL ARE HURTING for some reason.  if you don’t fit that category or hit the various pain points that make you want to migrate, you have your life in a good place.

sooner or later, DML will eventually kill a developer.  One of the key points I mentioned in a prior email are reasons to jump.   if they don’t apply, you are a happy person and there is no reason to re-write code that doesn’t need rewritten.

but you did ask the question:   why do developers feel like they have to migrate from DML
- speed/performance/scalability
- reliability
- acid compliance
- technology obsolescence
- size and depth of application
- size of database
- ability to integrate at the data level with other applications
- 24x7 operation
- live/checkpoint/full integrity backups while system running
- replication of databases for master/slave capabilities
- ability to run on industrial strength hardware using some sort of *nix  (aka NOT windows)
- ability to run background database threads and improve the interface of omnis ad get rid of the spinning beachball for some synchronous tasks.

and I could go on.  if you don’t need any of the above, don’t migrate.


but the one thing you should never do is feel that you are 100% safe with DML as a long term strategy.    We thought we could be, until
- we started getting clients that pushed us to 50 to 100 gig databases.
- people wanted to operate their web site 24x7
- they stated to get frustrated because of an omnis bug in DML with small block sized databases.
- they couldn’t ever be down to do a database reorganize/optimize
- upgrades to data and schema were killing them when we introduced new features as innovation is  absolutely required in our market segment

and I could go on.


the point is… stay where you are if you are comfortable.  Look to trends to see if they apply to you, and ignore them after due consideration.  Sometimes trends are faddish and bad, sometimes they are the way to go forward.

The people that hurt are those that don’t even see change on the horizon or have a plan to deal with whether it will affect them or not.

 notable examples.  motorola/nokia and the smartphone.  sony and the walkman.   kodak and 35 mm film/pictures, xerox and copiers.  fax machines.  CPM (was an excellent os),   Commodore.


pick what works for you.  just note that DML as a technological choice has well known long term issues that may or may not bite.  Thats why people are tending to migrate.


Doug Easterbrook
Arts Management Systems Ltd.
mailto:doug at artsman.com
http://www.artsman.com
Phone (403) 650-1978