Do frequent versions raise red flags?
Tue 9 Nov 2004, 10:35 PMTweet
by Ben Langhinrichs
I was talking with a fellow ISV about new versions of his software, and he said that he thought that putting out too many new versions of software indicated instability, so he didn't think you should ever put out a new version, except maybe a critical patch, more than once every four months. My feelings on this are mixed. We certainly don't release Midas Rich Text LSX versions more often than that on a regular basis, but that product is seven years old, and it is understandable that the frequency of changes would be low. Similarly, we don't release new versions of CoexLinks that often, but the product is much more sharply defined, so in the absence of major new bugs, frequent versions would be surprising.
But I would expect frequent updates of the Midas Rich Text C++ API and @Midas Formulas. Those products are brand new, are being used by more people in more scenarios every day, and we keep finding small ways to make big differences. That seems fairly understandable in a new product... but understandable to me may not be the same thing as understandable to customers.
So, what do you think? Do you think frequent updates are a sign of poor testing or agile response to customer needs? Do you think infrequent updates are a sign of stability or lack of caring? Does it seem more understandable for a new product to have frequent updates, or do you measure any product using the same ruler?
Copyright © 2004 Genii Software Ltd.
What has been said:
229.1. Duffbert (11/10/2004 04:10 PM)
I don't think it's so much the number of releases but why they are coming out. Bug fixes are to be expected early on. But if you're up to 2.13.05 (let's say that covers 50 releases) and you're still fixing significant bugs, then there's an issue. But if you're up to that point and you're adding features based on feedback, I'm impressed.