No More Waits
-
At least no more 5 year waits for SQL Server according to Mark Souza. It's going to be every two years, mostly because of software assurance, which is how I read it. Lots of customers got caught purchasing Software Assurance, expecting their upgrade cost to rolled up in it and they got a service pack or two and that's it.
And they would have gotten those for free!!!!
Sure some maintenance was included, which might have saved you a few $249 calls to PSS, but it really wasn't worth it, IMHO if you bought in 2000, 2001, or potentially in 2002. When it was released, it seemed at first glance to be a good idea, but as I thought about it more and dug in, I wasn't sure it was the best thing.
A couple years ago at PASS, Andy, Brian, and myself sat down with a couple product people from Microsoft over dinner and talked about the next version of SQL Server. At the time it was 3 years and they were worried because they were nowhere near being done. My opinion was that 4 or 5 years wasn't a bad time frame because most of us valued stability and didn't really want to upgrade any more often. Heck, there are still people pinging me about 6.5 at times!
I think a rapid release cycle, just to generate revenue, is a bad idea. It hurts the customers, puts pressure on developers to shortcut their work, and means that when IT people just get comfortable with a product, they must learn another. Already I see people skipping 2003 MCSE certification and waiting for Longhorn because some of them just finished a 2000 certification two years ago.
Personally, I think their plan is to get 45% the people to upgrade every other cycle, so every 2 years, 45% of the people will upgrade, having skipped the previous release. I know you're thinking that my math is challenged, but I think there are 5% that will upgrade every two years to be early adopters, and 5% that will upgrade once every 6 years or so.
This two year cycle, kind of an agile/XP/Scrum/whatever methodology can work. But it sounds suspiciously like the once a quarter cycle for service packs. Something that failed miserably because Microsoft didn't deliver it. If they can stick to a cycle, and still get service packs out, and still provide good support for 2 versions back, then I'm ok with it.
Or maybe they should just sell annual maintenance at a reasonable rate to fund support and some development and go to a point release every year that comes for free. That's what I'd prefer.
Steve Jones
-
We'd really know if the market wanted new features if they still sold old versions of software alongside the new ones, at reduced cost. Of course, few vendors will do this because of the increased support costs, but also because the take up of the new versions would be far reduced.
Today I'll be developing with SQL 2000 and Crystal Reports 8.5, and doing other bits of my job with Windows 2000, Office 2000 and some scarily aged Unix stuff. Sure, there are a few things in the later versions of each product I'd like to have, but could I justify the costs to an accountant?
Bill.
-
The cost of upgrading is enormous, far beyond the license cost. While some technologies (i.e. wireless) are changing so fast that upgrades are necessary, databases are pretty much a known entity. Keeping existing product running is far more cost effective than the overall costs of changing systems. Microsoft's planned obsolescence (with concurrent support dropoff) is the last thing we need.
...
-- FORTRAN manual for Xerox Computers --
-
Personally, I think their plan is to get 45% the people to upgrade every other cycle, so every 2 years, 45% of the people will upgrade, having skipped the previous release.
Does this mean that we should upgrade from 6.5 now?
-
Our company went from Win3.1 to Win98 to WinXP. And Office2 to Off97 to OffXP. But we went from Oracle 8i to SQL7 to SQL2K.
We generally skip the in betweens on the desktop because of the cost. And finally converted officeXP over the vet's day weekend. We will probably hit the SQL2K5 when SP1 comes out. But the end-user cost of upgrade is ridiculous and I think SA is a rip off.
----------------
Jim P.A little bit of this and a little byte of that can cause bloatware.
-
It's all about revenue from MS standpoint. From ours it's primarily stability then support followed by features. 5 year cycles are just fine (and probably more realistic). 2 year cycles are OK too (but we're not into the vaporware zone). Now whether the cycle is 2 years or 5 years is all up to MS - they just need to get marketing, sales, development and testing all to agree on the timelines and features. If that were done, then maybe the political 'face saving' gesture of the 2 year release plan would not be needed.
As for SA, well it allowed us to cash in (get a refund on) 90% our used case hours with MS PSS under Premier Support. Those refunded hours are being used for training and site inspections and such. SA may be a waste for some but not us. Just a note, we have SA for all MS products, not just SQL Server.
RegardsRudy KomacsarSenior Database Administrator"Ave Caesar! - Morituri te salutamus."
-
Yes, upgrade from 6.5!!!!!!!!!!!! No more questions, my memory can't handle it!!!!
I know it's all about revenue, and I do think we should support future development. But it has to be a fair profit and a good development effort. I'm kind of liking two year cycles with me upgrading every 4 years
Course the selling older version for less money is an interesting idea.
-
We are still on NT so fat chance we will upgrade anytime soon from SQL2K. The main reason? Our very expensive application software runs on these versions just fine and the software manufacturer has no reason to upgrade just because there is a new SQL version. Instead, they have created a new line of software product (but will still support both) but the newer version won't support our needs and our customizations we did to the source code. The improvements they made really are not related to SQL or the operating system so the latest version of either is just not needed. I think your 5% number might be low, at least initially. We are planning to upgrade the OS for other reasons but not SQL until SQL2K10 or so.
-
I prefer a lesser changes between versions and lesser development cycle.
Yukon - have amazing features.
But now adoption our system requires a lot of time.
And then upgrade on clients' servers - ... rrrrr!
-
In general, upgrading injects a lot of chaos in the enterprise and we try to avoid it. When MS added Reporting services to the tool set for SQL Server that was great. MS would be better of incrementally upgrading is various components over time, instead of all at once. Maybe releasing a new version of SSIS one year and then releasing a SQL engine another year would smooth out he change.
Viewing 10 posts - 1 through 9 (of 9 total)
You must be logged in to reply to this topic. Login to reply