The Heap™ – Consultancy ©®

1:30 AM

@ErikDarling I understood that reference

4 hours later…

5:35 AM

Morning

5:59 AM

Morning

Vérace

6:13 AM

A chairde - Morning all!

morning

6:52 AM

One in the edit review queue...

dba.stackexchange.com/review/suggested-edits/187786

Any programmers in here? Is the edit reasonable?

> Edit summary: The point is to set the maximum length, not to use legacy non-Unicode encoding, so the example should not override nvarchar with varchar.

@JoshDarnell Might know

trino seems to be a new thing.

@JohnK.N. It's possible the varchar part of the answer relates to the OP's comment:

I know it does hurt performance. When I define my own tables with SQL I use varchar(n). My question was more about how much it hurts performance(though I realize as posted that wasn't explicitly clear). — Nate Aug 21, 2013 at 14:30

The answer seems consistent, without a strong reason to prefer the suggested change. I'd probably reject it. I'm prepared to be corrected by anyone who knows more about these things though.

7:34 AM

Me three

7:57 AM

Last day in my current job

Angels ~~singing~~ rejoicing in the background: Hal-lelujah, hal-lelujah, hallelujah, hallejujah, hal--leee-luuuu-jahhhhhh

8:14 AM

Now, I've heard there was a secret chord.

8:27 AM

When you debug something with a Microsoft Debugger, you require to have the Windows OS ..... and preferably the product ....

I can't remember the correct word. Not the signatures or libraries. The other thingy-ma-jig.

symbols?

do you mean loading the symbols?

8:54 AM

@Zikato yes, I think so. But you have to have them before you can load them.

I think only some are public. But I'm out of my depth here

9:59 AM

good morning

do you see any reason why Azure Synapse dedicated SQL pools are created with compatibility level 130 by default? why not 150?

also, what the gymnastics group thing is about? I've missed something, that much I get

Dec 24, 2020 at 22:06, by Philᵀᴹ

Just wanted to take a minute to wish you and fam a very merry Christmas and a happy year, most of all good health!
These days people don't spend much time or thought on some personal words to their friends and family, they just copy and paste some random rubbish and send it on.
So, after all we've been though together this year I want to thank you for your friendship and support and wish you a happy 2021!
You are the best gymnastics group anyone could ask for.
Best wishes,
Phil

I've missed quite a lot, based on that date
and thank you very much indeed

haha, perfection.

10:17 AM

Only 13 days for the Fanatic gold badge

I wonder if visits to the site while suspended count toward that

*sigh*. So close

10:38 AM

@dezso I know next to nothing about Synapse, but doesn't it use DW compatibility level instead?

docs.microsoft.com/en-us/sql/t-sql/statements/…

1 hour later…

11:44 AM

One more thing about this approach (as it is): it works if the IDs are consecutive and from 1 to N. If they are not (e.g. 2, 3, 7, 9) that will be updated to 4,3,2,1 (and not to 9,7,3,2). Not sure what the OP wants to do in that case, perhaps it's ok. — ypercubeᵀᴹ 43 secs ago

12:30 PM

@PaulWhite I'd like to see an answer with that "CTE (or subquery) update without updating the ID column and without a self-join"

Josh Darnell

@PaulWhite I would reject it too - and have! cc: @JohnK.N.

Seems weird to call varchar "legacy."

Sean Gallardy - Mostly Retired

Is varchar(24) also legacy?

1:00 PM

@Zikato (⓿_⓿)

Is SQL Server 2022 going to support nth_value() window function ?

1:25 PM

dbfiddle.uk/WnG9XbSD for Paul, when he comes online

@JoshDarnell thanks

@ypercubeᵀᴹ I'm no good at creating dummy data for a PostgreSQL table. Would creating a 80 Million row / 8000 byte wide table required a lot of resources? And querying that table with a count(*) take a long time to run, say on db<>fiddle?

@PaulWhite that's an interesting question (haven't heard about this other compatibility level yet), however, the same page tells me 130 is not a valid option (only 10, 20, and 9000 are available)

@JohnK.N. I wouldn't throw that much into dbfiddle

I could do it on a test instance at work.

the count would be ok but inserting so much might be a problem. They run as VMs as far as I know

1:34 PM

but at the same time ALTER DATABASE test SET compatibility_level = 150; says Incorrect syntax near '150'.

Also consider that if you publish the link, every reuse will try to redo all the work.

@ypercubeᵀᴹ no

@ypercubeᵀᴹ I'll plug it into a test database at work using How can I generate big data sample for Postgresql using generate_series and random?

@dezso yes it's a different thing, a replacement if you will

@ypercubeᵀᴹ crikey I'm going to need to focus a bit on that one

@JoshDarnell thanks

Erik Darling

1:50 PM

hello and good tidings

@PaulWhite pity.

Postgres has the nth_value() but the update statement doesn't work the same, so we do need the join on PK in the final cte/query.

2:08 PM

@ypercubeᵀᴹ yes

@ErikDarling bon soir, mon ami

@ypercubeᵀᴹ well I just broke (as in filled up) my test instance with 80 Mio rows of random 4000 bytes per row of data.

anyway, Azure Data Studio tells me when opening the DB that it's at CL 130
probably I am not alone with being confused :D

ERROR:  could not extend file "base/17667/17669.8": No space left on device
HINT:  Check free disk space.

@dezso Yes. With the caveat I have no idea about this, it seems to me the DW setting replaces db compat. level, or rather supersedes it. You can't set db CL, only the DW version. Synapse is built on a core shared with SQL Server, so bits show through. That's my opinion.

So, if you change DW, I would expect db CL to also change. 9000 => 160 perhaps.

But it's all a horrible mess, so I wish you the best of luck in your endeavours

2:26 PM

thanks
TBH the docs (I mean, learns) are also terribly confusing, with about four slightly different products all documented in a single place. It's probably time to start from scratch, and call it something like msdntechnetbooksdoconline.microsoft.com

It doesn't help there are two 'sections' for essentially the same thing

and that, indeed
and some links will navigate you from one to the other silently

@dezso Check out this "explanation" What's the difference between Azure Synapse (formerly SQL DW) and Azure Synapse Analytics Workspace 🤡

my former boss was always complaining about AWS UX and the docs
I still should suggest him trying to work with Azure for a week or two, to raise his satisfaction level

Erik Darling

3:05 PM

what is everyone doing this weekend

aside from being great

thought I might go shopping for potassium iodide

stock up on canned food etc.

Erik Darling

3:22 PM

aw, prepper paul

are you putin on a bbq