or
Joe ObbishJoe Obbish
James JenkinsJames Jenkins
Jack DouglasJack Douglas
Erik DarlingErik Darling
Joe ObbishJoe Obbish
sql-server-2017
AndrewAndrew
sql-server-2016
VéraceVérace
mysql-5.6 dbfiddle
searle1986searle1986
sql-server-2019
searle1986searle1986
sql-server-2016
QuadeQuade
sql-server-2017
The Heap
Josh Darnell
Ah, okay.  No worries about eXtreme details.  I was just curious if you remembered off the top of your noggin.
Paul White
That might have been part of the 4199-covered fix but honestly I don't recall that level of detail
Paul White replying to Josh Darnell
As I recall it was back-ported
Josh Darnell
Do you know if the article is wrong, or the calculator was backported (or something else)?
Josh Darnell
@Paul Regarding [this article](https://www.sqlshack.com/join-estimation-internals/), it says the `CSelCalcSimpleJoinWithUpperBound ` calculator was introduced in SQL Server 2016.  I repro'd the problem (and solution) on 2014, since that was the OPs version.
Lamak
lol
GeorgePalacios
Pretty impressive.
Paul White
yowza
GeorgePalacios
Microsoft Sequel Server (MySQL Query Language)
GeorgePalacios
One line under skills reads
GeorgePalacios
Going through CV's today for a new DBA starting
GeorgePalacios
Had to share this.
Paul White
Someone v. reasonable proposed it
Josh Darnell replying to Paul White
Aha, that sounds like a v. reasonable change.
Paul White
Implied a rather higher level of structure and organization than was justified
Paul White replying to Josh Darnell
The Heap 1.0 was originally called The Clustered Index
Lamak replying to Zane
hey!, good to hear from you
Jack Douglas replying to Josh Darnell
well Joe is a fellow Oracle man ;)
Josh Darnell
Is that you, Joe?
Jack Douglas
heaps rule
Josh Darnell replying to PeterVandivier
O_O
PeterVandivier
  
> clustering is a one-time operation
PeterVandivier
be ye afeared
PeterVandivier replying to Josh Darnell
https://www.postgresql.org/docs/current/sql-cluster.html
Josh Darnell replying to PeterVandivier
Postgres doesn't have b-tree indexes?!
PeterVandivier replying to Josh Darnell
postgres backend demands respect
Josh Darnell
That's definitely been a popular decision, and one of the larger departures from how SE does things.
Zane
I like the bit where it's just always chat on the side. Turning comments into conversations.
Josh Darnell
I wonder why Jack didn't name it "The B-Tree"
Josh Darnell
Well welcome to the upgraded Heap.
Zane
That is correct. I did live and continue to do so!
Josh Darnell
I had to edit in some social distancing.
Josh Darnell replying to Zane
.  
.  
.  
He's alive!
Zane
Testing. 
PeterVandivier replying to Jack Douglas
i figured you might be, but then i thought it would sound snooty to ask if that's what it was if it was a bona fide outage
PeterVandivier
:-D
Jack Douglas
probably should put all that in a blog
Jack Douglas
the biggest change is that backends mostly put a hard limit on disk space per fiddle now
Jack Douglas
you might notice things being faster as I've fixed a few bugs along the way
Jack Douglas
some backends will be unavailable for a while
Jack Douglas
I need to move the physical hardware from one rack to another in the DC before we get total lockdown here so it's bee a bit of a rush
Jack Douglas replying to PeterVandivier
just doing some unannounced maintenance while America is asleep :)
PeterVandivier
> _shudder_
PeterVandivier
having to resort to... sql fiddle...
PeterVandivier
@Jack is db<>fiddle okay? seems unreachable on my end atm
Andriy M replying to Jack Douglas
Not as clever as the solution itself! Thanks though
Jack Douglas
that's some very clever tweaking ^^^
Jack Douglas
::: quote 856 32208 33 209,232,209 37,71,37
Andriy M *— 3 hours ago* *in [Join by overlap](/codegolf?room=856#c32208)*  
>Hey @Jack, I managed to squeeze your solution down to 364 bytes after finding a few more redundant spaces and observing that table aliases in column references could be removed by renaming columns of `z`. Then I noticed that the nested `with` could be rewritten relatively compactly as a single `select`. The result was 320 bytes. Somewhat annoying that I could manage to get it below 300. Oh well. Here's the result: https://dbfiddle.uk/?rdbms=postgres_12&fiddle=f04c882d7c13e752566490642fef8c8c Feel free to use it in your answer!
:::
Paul White
Morning
Andriy M
Morning
swasheck
stay safe everyone
swasheck
still 
swasheck
denver
Lamak
were are you now?
Lamak
that sucks
swasheck
my daughter likely has flu, and my wife is beginning to flu symptoms
swasheck
3/5 of my household is healthy
swasheck
roughly the same. we've been in a state of quarantine for two weeks now. 
Lamak
how are you?
Lamak replying to swasheck
hi!...I'm quarantined, but healthy
Jack Douglas replying to swasheck
my pleasure, literally in this case because I've found it really useful :)
swasheck
how's everyone doing? 
swasheck
@jack thanks for getting screen clipping up and running
swasheck
howdy, all
Paul White
Also clearing the last notification doesn't take one back to chat if the notification is a system one (I got one from CodeGolf)
Paul White replying to Jack Douglas
Quite impressed you managed that in SQL at all
Jack Douglas
(once we've finished messing around with db<>fiddle!)
Jack Douglas
yes that is something we need to fix
Jack Douglas
I need to read better
Jack Douglas
oh wait
Jack Douglas
does that not happen for you?
Jack Douglas replying to Paul White
no, clicking on reply should take you into the chat pane
Paul White
Is it a known restriction that *replying to* links don't work in the Notifications pane?
Jack Douglas
@@@ answer 991
Jack Douglas
I didn't realise code golf can be so addictive
Paul White
eeek
Jack Douglas
had to write a bit of C code either way which I find painful
Jack Douglas replying to Paul White
I eventually settled on using xfs quotas, after getting a more fragile implementation working with loop mounted files
Paul White
How did you do it for Postgres?
Jack Douglas
and about a million percent easier to implement then the postgres variety
Paul White
I forget things from the future sometimes
Jack Douglas replying to Paul White
actually that is pre-release as well
Paul White
I don't abuse db<>fiddle clearly :)
Paul White
I had forgotten how it worked already for SQL Server.
Paul White
Oh I see. Well that makes sense.
Jack Douglas
![blob](/image?hash=4c7a5c476d130cf67fd9650328360c8c2a94c8e6e1155c1d1197a1df0b146527)
Jack Douglas replying to Paul White
I guess I'll do that if it becomes an issue — but it will need custom code for different backends which I like to avoid.
Paul White
I was wondering if you could catch the error and print something more helpful instead
Jack Douglas replying to Paul White
true but I think I'll wait and see on that one as the limit is high enough to only catch people who probably already know they are testing the limits
Paul White
i.e. it could generate support requests
Paul White
That could cause people to think db<>fiddle has run out of disk space, rather than their individual fiddle being limited?
Jack Douglas
not released yet
Jack Douglas
![blob](/image?hash=c734ab6e24a2abdd2d0cc98584b1f72119e86feebd59e40dc35fef154f0c7909)
Jack Douglas
I think I have finally figured out how to impose disk usage quotas on postgres fiddles
Jack Douglas
morning
Monica
Here they just expect us to find out via the media, it seems.  Which I guess works because of social media, but still... somebody pushes amber alerts and storm warnings to my phone but a stay-at-home order doesn't rate?  Governments are weird.
PeterVandivier
![9E527581-1F84-4283-AF34-5062CECF68B7.jpeg](/image?hash=f9e99a22a653b232919bfc76a6840721f012c942015ccf6227fc9649605a237e)
PeterVandivier replying to Jack Douglas
just got my copy of that
Lamak
I'm quite impressed that the quarantine un the UK includes that you can exercise once a day
Jack Douglas replying to Tom V
![blob](/image?hash=52430ea4fd43456764cb259311e7e7d295f5c0d28b2f942f15b8271b7c658f38)