July 21, 2010 at 5:52 pm
Alvin Ramard (7/21/2010)
Grant Fritchey (7/21/2010)
rjohal-500813 (7/21/2010)
This chap ought to update his cv and brush up on his interview skills. Not sure I'd want to work there. This situation could only have arisen as a result of a whole series of wtfs.Yep, I think I'd probably walk out under my own power at this point.
I was going to recommend DBCC TimeWarp, but how can you backup the DB if it's gone? :w00t:
That's the beauty of Timewarp - we can do it together and all go back in time to before the DB was dropped.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
July 21, 2010 at 7:06 pm
Getting back to Joe Celko ..
A rather simple question posted to a forum yesterday evening with what I thought was a simple answer:
http://www.sqlservercentral.com/Forums/Topic956023-392-1.aspx#bm956671
Today along comes Joe Celko, posting an answer that would not have properly answered the OPs question, and which had a simple syntax error.
No further posting by Joe, but let me tell you, made me feel good that I could answer the OPs question, agreed a simple question but in this one case I beat out Joe .... lord have mercy on my soul.
July 21, 2010 at 7:48 pm
I've long had a theory that you can either be an *** or you can be wrong, but nobody can get away with being both.
I know some people here are friends with him, and he may be a nice guy, but so much of what he writes is just garbage on so many levels. Even when he is correct, telling someone that doesn't know how to write a simple select to go follow some hundred page ISO document is just absurd.
On the positive side, he makes the rest of us look good.
July 22, 2010 at 7:48 am
bitbucket-25253 (7/21/2010)
Getting back to Joe Celko ...
The man knows his business and I respect him for that. However, I have found that remembering he is looking at SQL through portability-tinted glasses makes dealing with his posts a lot easier.
He knows a lot about SQL, a lot more than I know, I just don't agree with his stance that everything written should be portable. Different database systems have different strengths and weaknesses; use the strengths, avoid the weaknesses as much as possible. And if you're going from one database system to another, there are going to be more changes that merely detaching a database and re-attaching it somewhere else, why not use what strengths the database system has? And by changes, I mean that minimually someone is going to go "Hey, since were changing systems, can we also change..." giving the devs/DBAs something else to do besides make sure their code works on the new system.
Joe Celko does know his business, and his business is portable SQL code. He's also on planet Earth and as such is subject to making mistakes, just like the rest of us. For myself, I can be tolerant of his foibles, wish he would be more forgiving of those who don't have the advantage of his experience and study and remember that he writes SQL code through portability-tinted glasses. I don't agree with that stance, but I respect it.
edit: to fix HTML tags
-- Kit
July 22, 2010 at 7:48 am
CirquedeSQLeil (7/21/2010)
Alvin Ramard (7/21/2010)
Grant Fritchey (7/21/2010)
rjohal-500813 (7/21/2010)
This chap ought to update his cv and brush up on his interview skills. Not sure I'd want to work there. This situation could only have arisen as a result of a whole series of wtfs.Yep, I think I'd probably walk out under my own power at this point.
I was going to recommend DBCC TimeWarp, but how can you backup the DB if it's gone? :w00t:
That's the beauty of Timewarp - we can do it together and all go back in time to before the DB was dropped.
I can never remember which parameter it is, and sometimes when I do get the right one, I find I've gone back too far.
Is there a parm to create trans logs to roll forward to a poijnt in time from there?
Oh for some better documentation.
Greg E
July 22, 2010 at 8:37 am
Oh for some better documentation.
Greg E
I heard there was a recently deciphered passage from the Dead Sea Scrolls that discussed such a parameter, but the fragment containing the ARID (alternate reality ID) was missing, so there is no way to be sure that it would even be applicable to the reality context you are currently operating under, Greg.
__________________________________________________
Against stupidity the gods themselves contend in vain. -- Friedrich Schiller
Stop, children, what's that sound? Everybody look what's going down. -- Stephen Stills
July 22, 2010 at 9:06 am
The Dixie Flatline (7/22/2010)
Oh for some better documentation.
Greg E
I heard there was recently deciphered passage from the Dead Sea Scrolls that discussed such a parameter, but the fragment containing the ARID (alternate reality ID) was missing, so there is no way to be sure that it would even be applicable to the reality context you are currently operating under, Greg.
I've been thinking my wife's uncle might be able to help me. The 60's was a lost decade for him, so he might have some clues.
And since he is retired, and every day seems to be a Saturday, he lives in a different reality context.
Greg E
July 22, 2010 at 10:39 am
I've got a QotD question for you'll...
Is it sufficient to base the answer off of BOL?
I'm getting a lot of grief today over a few index types that people are perceiving to be sub-types, and shouldn't be counted. (And I had issues with my last QotD on index structures.)
I say: if you disagree with BOL, submit a connect item to get it fixed or clarified. But a BOL reference should stand.
Wayne
Microsoft Certified Master: SQL Server 2008
Author - SQL Server T-SQL Recipes
July 22, 2010 at 10:49 am
BOL should be able to stand. The problem is that BOL is not always correct. I don't see a problem with submitting a connect ticket on it though.
Jason...AKA CirqueDeSQLeil
_______________________________________________
I have given a name to my pain...MCM SQL Server, MVP
SQL RNNR
Posting Performance Based Questions - Gail Shaw[/url]
Learn Extended Events
July 22, 2010 at 10:51 am
BOL should be a standard reference. You can add "community content" to items as well as submit a Connect Item if you wish.
July 22, 2010 at 10:57 am
WayneS (7/22/2010)
I've got a QotD question for you'll...Is it sufficient to base the answer off of BOL?
I'm getting a lot of grief today over a few index types that people are perceiving to be sub-types, and shouldn't be counted. (And I had issues with my last QotD on index structures.)
I say: if you disagree with BOL, submit a connect item to get it fixed or clarified. But a BOL reference should stand.
Agree BOL should be the standard. I got the answer right because of Grant's recent SQL University blog post[/url] which also lists 8 and references BOL.
Jack Corbett
Consultant - Straight Path Solutions
Check out these links on how to get faster and more accurate answers:
Forum Etiquette: How to post data/code on a forum to get the best help
Need an Answer? Actually, No ... You Need a Question
July 22, 2010 at 11:00 am
I added a note in the QOD as well, since I don't view BOL as "true", but as a reference.
July 22, 2010 at 11:15 am
Jack Corbett (7/22/2010)
I got the answer right because of Grant's recent SQL University blog post[/url] which also lists 8 and references BOL.
I must admit, I don't agree with that. I don't view unique indexes or indexes with include columns as separate types. Unique is just a property of a clustered or nonclustered index, it's not an alternative index type. If you say CREATE UNIQUE INDEX, you're getting a nonclustered index, not a separate type that's at the same level as clustered/nonclustered
As for indexes with include columns considered separate types, well only nonclustered indexes can have include columns, so how can that be considered a 'type' equivalent to clustered or nonclustered. Same with filtered, only nonclustered indexes can be filtered and you don't say CREATE FILTERED INDEX. you say CREATE NONCLUSTERED INDEX ... WHERE ...
And if we want to reference BoL, from the page on sys.indexes:
type_desc
Description of index type:
* HEAP
* CLUSTERED
* NONCLUSTERED
* XML
* SPATIAL
is_unique and has_filter are properties of the index, not types of indexes.
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
July 22, 2010 at 11:21 am
GilaMonster (7/22/2010)
Jack Corbett (7/22/2010)
I got the answer right because of Grant's recent SQL University blog post[/url] which also lists 8 and references BOL.I must admit, I don't agree with that. I don't view unique indexes or indexes with include columns as separate types. Unique is just a property of a clustered or nonclustered index, it's not an alternative index type. If you say CREATE UNIQUE INDEX, you're getting a nonclustered index, not a separate type that's at the same level as clustered/nonclustered
As for indexes with include columns considered separate types, well only nonclustered indexes can have include columns, so how can that be considered a 'type' equivalent to clustered or nonclustered. Same with filtered, only nonclustered indexes can be filtered and you don't say CREATE FILTERED INDEX. you say CREATE NONCLUSTERED INDEX ... WHERE ...
And if we want to reference BoL, from the page on sys.indexes:
type_desc
Description of index type:
* HEAP
* CLUSTERED
* NONCLUSTERED
* XML
* SPATIAL
is_unique and has_filter are properties of the index, not types of indexes.
Just so you know, I agree. I don't think unique or include columns are seperate types either, but I didn't want to get into a personal p*****g match with anyone on the topic. So I used the BOL reference and pointed out that honest people can disagree (and do) with that list. Fact is, your break down from sys.indexes is probably preferred. I should have used that.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
July 22, 2010 at 11:24 am
I'm gonna concur with Grant and Gail.....but just because I picked 5 since it was the next smallest number over 2, and I figured there couldn't really be that many.....:-P:-D
good learning tool though, I tend to shoot for something without too much research a lot of the time just to find the answer and learn from it at QotD. (lazy way to save time)
---------------------------------------------------------
How best to post your question[/url]
How to post performance problems[/url]
Tally Table:What it is and how it replaces a loop[/url]
"stewsterl 80804 (10/16/2009)I guess when you stop and try to understand the solution provided you not only learn, but save yourself some headaches when you need to make any slight changes."
Viewing 15 posts - 16,621 through 16,635 (of 66,712 total)
You must be logged in to reply to this topic. Login to reply