July 26, 2012 at 7:57 pm
SQLRNNR (7/26/2012)
Lynn Pettis (7/26/2012)
Brandie Tarvin (7/26/2012)
WayneS (7/26/2012)
You'll might find this interesting: http://www.sqlperformance.com/2012/07/t-sql-queries/split-stringsVery interesting indeed. Poor Jeff. Poor Tally Tables.
Did you take a close look at how Jeff's split routine was modified? His routine is optimized for varchar(8000) or nvarchar(4000) depending on which routine you use.
I take that testing with a grain of salt or two.
+1
Aaron is really good at TSQL too though.
I guess I need to get around to writing a > 8K splitter. I've been putting it off because a couple of folks I know have been working on one of their own.
--Jeff Moden
Change is inevitable... Change for the better is not.
July 26, 2012 at 8:09 pm
capn.hector (7/26/2012)
so now that jeff has posted his response shall the SSC Dog Pile commence??
Good idea. I was utterly dumbstruck by his performance ranking of the rCTE solution to the point of disbelief.
Let the debunking commence.
My thought question: Have you ever been told that your query runs too fast?
My advice:
INDEXing a poor-performing query is like putting sugar on cat food. Yeah, it probably tastes better but are you sure you want to eat it?
The path of least resistance can be a slippery slope. Take care that fixing your fixes of fixes doesn't snowball and end up costing you more than fixing the root cause would have in the first place.
Need to UNPIVOT? Why not CROSS APPLY VALUES instead?[/url]
Since random numbers are too important to be left to chance, let's generate some![/url]
Learn to understand recursive CTEs by example.[/url]
[url url=http://www.sqlservercentral.com/articles/St
July 26, 2012 at 8:16 pm
so now that jeff has posted his response shall the SSC Dog Pile commence??
Thanks for the support, folks. Being human and since I've had more than one code based run-in with him, I have to admit that thought is enticing especially since I've proven him wrong at every turn... but it wouldn't be the right thing to do. Give him the opportunity to fail or succeed by giving him a chance to respond to my post. After that, if you do still decide to respond, let's keep it on the up-n-up. Show him and all that read your responses that SSC is a haven for professionals, not professional dog-pilers.
--Jeff Moden
Change is inevitable... Change for the better is not.
July 26, 2012 at 8:19 pm
Well :blush:, I meant dog-piling in a professional sort of way of course.
My thought question: Have you ever been told that your query runs too fast?
My advice:
INDEXing a poor-performing query is like putting sugar on cat food. Yeah, it probably tastes better but are you sure you want to eat it?
The path of least resistance can be a slippery slope. Take care that fixing your fixes of fixes doesn't snowball and end up costing you more than fixing the root cause would have in the first place.
Need to UNPIVOT? Why not CROSS APPLY VALUES instead?[/url]
Since random numbers are too important to be left to chance, let's generate some![/url]
Learn to understand recursive CTEs by example.[/url]
[url url=http://www.sqlservercentral.com/articles/St
July 26, 2012 at 8:20 pm
dwain.c (7/26/2012)
capn.hector (7/26/2012)
so now that jeff has posted his response shall the SSC Dog Pile commence??
Good idea. I was utterly dumbstruck by his performance ranking of the rCTE solution to the point of disbelief.
Let the debunking commence.
THAT you can do. I just ask that you do it in a fashion that you'll be proud of the next day. That means let the code do the talking and no ad hominem attacks.
--Jeff Moden
Change is inevitable... Change for the better is not.
July 26, 2012 at 8:23 pm
dwain.c (7/26/2012)
Well :blush:, I meant dog-piling in a professional sort of way of course.
You know I had to make sure. ๐
--Jeff Moden
Change is inevitable... Change for the better is not.
July 26, 2012 at 8:25 pm
Jeff Moden (7/26/2012)
dwain.c (7/26/2012)
capn.hector (7/26/2012)
so now that jeff has posted his response shall the SSC Dog Pile commence??
Good idea. I was utterly dumbstruck by his performance ranking of the rCTE solution to the point of disbelief.
Let the debunking commence.
THAT you can do. I just ask that you do it in a fashion that you'll be proud of the next day. That means let the code do the talking and no ad hominem attacks.
ill be working on some code to possibly show how it is wrong and if not i am not one to break godwin's law. your response was the best there is with out code examples. i just know with all the coders here we should have a decent code pile to refute his claims.
For performance Issues see how we like them posted here: How to Post Performance Problems - Gail Shaw[/url]
Need to Split some strings? Jeff Moden's DelimitedSplit8K[/url]
Jeff Moden's Cross tab and Pivots Part 1[/url]
Jeff Moden's Cross tab and Pivots Part 2[/url]
July 26, 2012 at 8:35 pm
capn.hector (7/26/2012)
Jeff Moden (7/26/2012)
dwain.c (7/26/2012)
capn.hector (7/26/2012)
so now that jeff has posted his response shall the SSC Dog Pile commence??
Good idea. I was utterly dumbstruck by his performance ranking of the rCTE solution to the point of disbelief.
Let the debunking commence.
THAT you can do. I just ask that you do it in a fashion that you'll be proud of the next day. That means let the code do the talking and no ad hominem attacks.
ill be working on some code to possibly show how it is wrong and if not i am not one to break godwin's law. your response was the best there is with out code examples. i just know with all the coders here we should have a decent code pile to refute his claims.
Because he cited me in his article, I'm both deeply humbled and honored that you would take the time to do such a thing. I was going to do similar but it would be my honor to watch you good folks in action.
--Jeff Moden
Change is inevitable... Change for the better is not.
July 26, 2012 at 8:40 pm
Jeff Moden (7/26/2012)
dwain.c (7/26/2012)
Well :blush:, I meant dog-piling in a professional sort of way of course.You know I had to make sure. ๐
I am not "he would should not be named" (the one whose recent post was deleted). ๐
I would help to support Capn Hector if I could but I'm afraid I might not find the time this weekend.
My thought question: Have you ever been told that your query runs too fast?
My advice:
INDEXing a poor-performing query is like putting sugar on cat food. Yeah, it probably tastes better but are you sure you want to eat it?
The path of least resistance can be a slippery slope. Take care that fixing your fixes of fixes doesn't snowball and end up costing you more than fixing the root cause would have in the first place.
Need to UNPIVOT? Why not CROSS APPLY VALUES instead?[/url]
Since random numbers are too important to be left to chance, let's generate some![/url]
Learn to understand recursive CTEs by example.[/url]
[url url=http://www.sqlservercentral.com/articles/St
July 26, 2012 at 8:55 pm
capn.hector (7/26/2012)
...since i reread the article a couple times a month (any time i stumble onto it in my google searches) as it is a very good description of your (Jeff's) thought process and each time i read it i learn something new.
Shifting gears, thank you good Sir! You made my day... maybe the whole month. Although I had a great deal of fun with the dust bunny, beer popsicle, and other metaphors, I'm glad there's people that can see beyond those.
And, yes, I really do have the "outside the box" artwork I spoke of. It's one of my prized possessions. Hmmmm... I wonder if I can take a picture of it without taking the glass out of the frame so I can post it.
--Jeff Moden
Change is inevitable... Change for the better is not.
July 26, 2012 at 9:03 pm
Hate to say it, but I posted a reply. When I started it, Arron had not yet replied, but when I posted it, he had. I hope what I said sits okay with everyone, I tried to be polite yet not very happy about the article.
I did include that I hoped people that read his srticle take the time to read Jeff's article and the discussion that followed.
July 26, 2012 at 9:04 pm
Stefan Krzywicki (7/26/2012)
Yeah, CLR and XML have come a long way!
SNORRRTT!!! BWAAAA-HAAAAA!!!!:-D:-P:hehe: ROFLMAO!!!! ๐ Ya made me blow a porkchop bone clean out of my nose!!!! I can't stop laughing! ๐
--Jeff Moden
Change is inevitable... Change for the better is not.
July 26, 2012 at 9:08 pm
Lynn Pettis (7/26/2012)
SQLRNNR (7/26/2012)
Lynn Pettis (7/26/2012)
Brandie Tarvin (7/26/2012)
WayneS (7/26/2012)
You'll might find this interesting: http://www.sqlperformance.com/2012/07/t-sql-queries/split-stringsVery interesting indeed. Poor Jeff. Poor Tally Tables.
Did you take a close look at how Jeff's split routine was modified? His routine is optimized for varchar(8000) or nvarchar(4000) depending on which routine you use.
I take that testing with a grain of salt or two.
+1
Aaron is really good at TSQL too though.
That may be, but after reading that post and the Tally Oh article, I have a problem with the comparision done and the changes made to Jeff's routine. It takes a routine designed to work with strings upto 8000 bytes and takes it into an area we know it may not be a performant. Looks like apples to oranges to me.
Ah, my ol' friend from the desert. Thank you for your support and the absolutely correct observation.
--Jeff Moden
Change is inevitable... Change for the better is not.
July 26, 2012 at 9:39 pm
NEWS FLASH -/-
The SQL-verse was set aflutter today by a purported attempt to debunk the myth perpetrated by Jeff Moden that his DelimitedSplit8K was the fastest 8K SQL-based string splitter that the known SQL-verse had ever seen.
Reports are coming in that the debunkers of the debunkers are planning a mass rally in Jeffโs favor, and that the code churning machines have been turned up to full volume production levels. SQL Server performance across the known SQL-verse has now dropped to all-time lows as these unprecedented levels of code flurry, expend CPU cycles attempting to make sense out of these unverified reports.
The victor in these skirmishes will no doubt rise to the top of the string-splitting throne, with scepter in hand, to declare โall those that have come before me shall finally bow to the prowess of this solution.โ
Stay tuned to BBSQL for breaking news updates as this drama unfolds.
My thought question: Have you ever been told that your query runs too fast?
My advice:
INDEXing a poor-performing query is like putting sugar on cat food. Yeah, it probably tastes better but are you sure you want to eat it?
The path of least resistance can be a slippery slope. Take care that fixing your fixes of fixes doesn't snowball and end up costing you more than fixing the root cause would have in the first place.
Need to UNPIVOT? Why not CROSS APPLY VALUES instead?[/url]
Since random numbers are too important to be left to chance, let's generate some![/url]
Learn to understand recursive CTEs by example.[/url]
[url url=http://www.sqlservercentral.com/articles/St
July 26, 2012 at 9:47 pm
dwain.c (7/26/2012)
NEWS FLASH -/-The SQL-verse was set aflutter today by a purported attempt to debunk the myth perpetrated by Jeff Moden that his DelimitedSplit8K was the fastest 8K SQL-based string splitter that the known SQL-verse had ever seen.
Reports are coming in that the debunkers of the debunkers are planning a mass rally in Jeffโs favor, and that the code churning machines have been turned up to full volume production levels. SQL Server performance across the known SQL-verse has now dropped to all-time lows as these unprecedented levels of code flurry, expend CPU cycles attempting to make sense out of these unverified reports.
The victor in these skirmishes will no doubt rise to the top of the string-splitting throne, with scepter in hand, to declare โall those that have come before me shall finally bow to the prowess of this solution.โ
Stay tuned to BBSQL for breaking news updates as this drama unfolds.
Easy, big guy. First, it's not my splitter. It's our splitter. A lot of good folks have put some good time and effort into it and 2 folks in the discussion of the "Tally Oh!" article actually took my code and, in the SSC spirit, made it even better.
Although Aaron was apparently unaware of the affect that simply changing the input variable to a MAX datatype would have and he certainly went about it all the wrong way, his goal wasn't to challenge N4K or 8K splitters. Despite my personal disappointment in how he went about it, his goal was to find out what the best MAX splitter(s) were.
--Jeff Moden
Change is inevitable... Change for the better is not.
Viewing 15 posts - 37,081 through 37,095 (of 66,712 total)
You must be logged in to reply to this topic. Login to reply