Is it time for LL to change the way they develop?

Is it time for Linden Labs to change the way they develop the SL viewer?

The frequent, small updates have, over time, left the SL viewer riddled with bugs. Each new update seems to introduce more bugs than fixes or features. New features are wonderful when they work, but after the most recent updates I can't think of a new feature off the top of my head, I can however list several new bugs that have had a very significant and negative effect on my business.

With this in mind, is it time for LL to rethink their policy on updates in favor of less frequent updates with far more emphasis put on testing before updates go live?

Vale

I have often thought it would be a good idea to get a group to test up coming "updates" in world so they can preform all the items that need testing. The Beta Grid has too many limitations, all you can really do there is chat. I know a lot of people who would be more then willing to help work out bugs before it goes to the general population.

Get a good cross section of users so all parts of the view can be tested and fixed first.

as 1000 of fellow Sl members. the problem is not the frequency of the update. But the introduction of new features. SL never had time to mature... LL is insiting on introducing new features instead of FIXING the bugs. Right new the client shoudl be called early alpha. as the number bugs in SL is so great that any other reponcible software company whould not have put out yet.

Seem LL need to be forced to fix the bugs and stop introducing new features until the client is stable. What good is voice support or sculpted prims if SL is not even capable of the most basic 3d function (like applying correctly a simple textere on a flat surface).

Please LL listing to your consumers and don't add any new festures until the software is stable.

In the past LL update in much bigger intervals, and the update where much larger. The reason they changed, was because the large changes often has such a big unexpected effect that SL was almost unusable for a week or 2, until biggest bugs where fixed. Having smaller updates results in smaller bugs and less code to go through to figure out where it went wrong.

Fixes and new features are often being worked on by different teams, research in software development have showed that throwing more people on a problem often doesn't result in quicker fixes. More people means more time wasted on management and communication. Best thing to do is put the people with the most experienced with that piece of code on the problem.

But we aren't happy with that solution either.
To all the bugs you have to add the 5 hours downtime once or twice a month.

There must be an intermediary solution, like leaving the main grid/viewer to 1.15 while developing the beta grid/viewer to 1.20, only then updating the main grid/viewer.
But for this to go smooth we need more beta testers, many more.

How about making smaller updates less frequently... say, only as often as the previous updates had had thier bugs worked thru?

Why this pushpoush PUSH forward when so much behind them is still so unreliable?

When does innovation finally bow to stability and a reliable product?

*cough*
1.2 < 1.15
I think you meant 1.20
Software version numbers are not subject to the rules of decimals we learned in math. 1.2 != 1.20

Huh ? ah yes yes yes, I forgot that abbreviating it would make it conflict.