The CTE suggestion is really no different than our existing temp table technique. The problem is it still has to calculate the ranked metric for the entire dataset and...

select T.*

from yourTable T

join yourTable T2 on T.user_id = T2.user_id

and dateadd( hour, 1, T.punch_out ) = T2.punch_in

also consider eliminating the month(), date(), year() comparisons

SELECT COUNT(tblConferences.Division) AS Total,

tblConferences.Subdivision, tblConfAtt.ConfDate,tblConfAtt.CustomerID

FROM tblConfAtt

INNER JOIN tblConferences

ON tblConfAtt.Conference = tblConferences.Room...

also take a look at sybase powerdesigner.

try rounding both datetimes.

select '11:29',

dateadd(minute,

round(cast(datediff(minute, 0, '11:29') as float) / 60,0) * 60, 0)

Select ...

,(SELECT DISTINCT b1.FirstName + ' ' + b1.LastName FROM dbo.consultant

LEFT OUTER JOIN dbo.uvwConsultantDownLine AS B1 ON D.SponsorID = B1.ConsultantID) AS SponsorName

,D.SponsorID

...

from #Downline D with (nolock)

LEFT OUTER JOIN uvw_DownlineOrder O...

the left outer...

Yeah, agreed, except that I don't want it to 'try to figure out'. Thanks, guys, for the verification it's not a bug. I still don't like it though. I'm just...

Hey, SSCrazy. That sounds good except that now I have to question the design. You're right that the alias makes it fail correctly ( isn't that ironic? You have...

thanks again todd. i'm somewhat skeptical about what naysayers that never use a feature say about a feature.

them: "it's awful, i never use it for...

hey todd. thanks for the feedback. when i give a suggestion i don't always put on my qa hat as i just want to help someone down the road...

just convert getdate() to an int and leave run_date alone:

cast(convert(char(8),getdate(),112) as int)

sorry the comparison should be TimeStamp >= date expression

run these statements:

select dateadd(month, -12,

dateadd(day,1 - day(getdate()),

...

try this instead:

where TimeDateStamp <=

dateadd(month, -12,

dateadd(day,1 - day(getdate()),

convert(char(10),getdate(),120)

...

i've read that the drawback of row_number() is that it must be calculated for the entire partition of the over() clause and thus usually results in higher io. anyone...