English - United States  


Usenet Posting #29 - Trigeminal Software, Inc. (English)

Subject: INFO: Are my replication methods the best?
(Originally posted 09/13/01)

This post has to do with the way I managed replication projects (it also applies to the way I handle globalization projects, incidentally!). I work by VERY ironclad principles and I do not make exceptions to them. I expound on those principles here often -- because when contrasted with the problems people report, they work.

Can I claim that my ways are better? Yes, I can. Because my replicas *work* and theirs do not. "My children are better than yours, nya, nya, nya!" (heh heh heh, sorry I could not resist).

Can I claim that my ways are the best? Not really, especially since they have changed over time to reflect many real world issues. They have not had to change in quite some time, but that does not mean that they cannot be modified again if a technique proven to be superior comes forward.

Can I claim that every technique I have developed or help developed was done with strict adherenece to the scientific method? Nope. In the most extreme case, as I have told many people, I first came up with the concept of replica farms while I was drunk, almost falling on the floor, in a pub in London. Hardly scientific. But after careful sober consideration it was realized that not only was the suggestion consistent with the replication model but that it also fixed many often-reported bugs!

Can I claim that I understand every side effect of everything I do not recommend? Hell no! Because I do not USE those techniques. How on earth could I fully explore the consequences of them?

Now, I am not entirely oblivious -- I have the experiences of many other developers to draw on, I have the knowledge of all of the specific bugs that were found and fixed in various schemes as well as the behavior that used to occur. At one point I was working for both Access and SQL Server, and I owned three different wizards across three different products for more than two versions -- and often had reason to investigate specific bugs and issues in Jet. Now I do not do that sort of thing any longer, but I do remember many of the design principles. And I do have the code that I myself wrote and was given permission to redistribute, which included many utilities that call Jet APIs directly, using that knowledge.

But I want to make it clear that even though there are many things I know, I am hardly infallible, because once I learn that a particular road is bad to travel upon, I do not travel upon it. If someone finds out that my initial reports of glass on the road did not mention the fact that it has a 50 foot hole in it, then one can assume that I really should not be held responsible, even in a philosophical sense.

In the end I am just a software design engineer. A developer. A geek who makes code, a man whose life is often both way more interesting and way more boring than anyone could imagine. I am very cocky about a very small focal point of issues that I know a lot about -- and pretty silent on anything I do not. Further respondent sayeth naught.


Problems with this site? Please contact the webmaster@trigeminal.com
with your comments, questions, or suggestions.