Erik's site is finally back online.

The backup was from Saturday morning. So any comments / etc. since then will have been lost, unfortunately.

Did you forget the tail of the log 😀

Good job, anyway

@PaulWhite Apparently 😅

It had too many VLFs.

Gave me an opportunity to fix the typo in my comment anyway

I tried restoring the one from this morning, but it failed for unknown reasons.

Is it MySQL under the hood?

Yeah.

Seems right.

Good to see a professional web dev involved at last

Who is that? 🤣

ha ha well it's all relative of course

Erik's had some bad luck

Or it's perfectly normal

Yeah, today's debacle is a real doozy.

I suppose it's cute the way we still expect IT people to be vaguely competent

Given the evidence we see every day

And tbqh I don't mind the way the site looks rn at all

I didn't think the new look was so much better anyway

But then what would I know about it really

I don't have a great eye for design either. I did like the new design though.

It would have been better with another month's worth of posts included.

and some plugins

Right!

I suppose I'm also factoring in the thousands of dollars

How many thousands of dollars have you spent on your site design?

Well quite

Of course, my site could barely be described as having a design.

v clean

@JoshDarnell Yours is all static generated using some tool isn't it? Self-hosted?

Jekyll or sthg?

Mentally I associate your blog with Kendra's

Zikato uses static pages maybe I'm thinking of that

joshthecoder.com ftr

@PaulWhite Right, Jekyll. It's on GoDaddy shared hosting.

It is statically generated yeah. I write the posts in markdown.

Me too

I occasionally think about switching to GitHub Pages.

But that would require me to, like, do stuff.

The payoff, though, being that it's free hosting. I think.

Yeah doing stuff can be v time-consuming

@ErikDarling blogs are a mistake

Read 331 messages

Morning

BTW, where is Evan EVAN THE GREAT?

Morning

@JohnK.N. Sulking

I've finally finished the Spider-Man game, so I can catch up with life

@PaulWhite I use Hugo - same as Kendra. Her blog post motivated me to start, and I've used the same theme as well. I have several custom integrations "borrowed" from the community.

Can anyone confirm: the weird ODBC format {d '2022-08-15'} and {ts '2022-08-15 11:00:00} seem to be dateformat safe also?

@Charlieface you deserve a medal for that recent interaction where you flagged a comment

There should be a badge for putting up with a barage of comments while maintaining civility beyond the call of duty.

@HannahVernon Is there anything we can do about that particular poster?They seem to have a penchant for doing some mild reading on a particular subject, then blasting out a ton of basic questions on it, most of which could be satisfied with reading the docs or "trying it out".

@Charlieface what do you mean by dateformat safe?

@HannahVernon I mean that it returns the same thing always, no matter what language or set dateformat you have. See dba.stackexchange.com/questions/166771/…

@Charlieface we have communicated our distaste for their antics in the past. I just sent them another message about not over commenting.

@Charlieface I believe the odbc functions are subject to the same rules as other date functions; in other words they are sensitive to localizations such as set dateformat. Perhaps that should be asked on the main site.

@HannahVernon dbfiddle.uk/… Doesn't seem like it to me

Well, it's useful to me. I'm learning new stuff.

@Zikato I learn something new every day. And it seems like every day, I make a totally wrong assumption. Perhaps that's the thing I'll learn today, to not make assumptions. Probably not today, but maybe.

@Charlieface you'd think I'd remember this, but alas my brain is old and apparently decrepit. How does this syntax work? {fn CurDate()} or {fn Now()} etc

what the HELL

> The single-threaded CPU performance of your processor is usually the final performance bottleneck in a well-tuned system that does not have any other substantial bottlenecks. - Is the CPU Important for SQL Server?

Based on that, it seems like the answer to the blog title is "no" 😀

so if all your reads are unblocked index seeks from pages in memory...

Exactly.

I get that everyone has their pet peeves, and one of Glenn's is when people don't know their CPU's special name. But just because something bugs you doesn't mean it's important.

*fewer

one of my pet peeves is when people don't know their CPUs are offline because their 20 person dba team installed the cal bits on every server under the roof

that's probably more important than a cumulative gigahert

but i have no sql skills or authority

Hahaha

That does seem like a slightly bigger problem.

Your Intel Space Pond Gen 8 Series X isn't do much for you when SQL Server can't use it.

muppet lake

@ErikDarling is that a Twitter Hot Take™?

sometimes, you just need to close your browser

~17GB freed up

I just got to install SQL Server 2005 lol

"got to"

:o

Did you mean "forced to?"

lol, not really - I needed to answer a question from 2012; I already had SQL Server 2000 installed, but wanted to see if the problem had persisted into 2005. It had not as it turns out.

10 year old question answered on Internet

Actually I wasn't so concerned about answering the question as i was about removing the comments on the question.

the comments contained answers

That is some serious dedication to comment deletion.

comments were a mistake, as Paul is so fond of saying.

Anyone got any idea on this one? How is it possible for a foreign-key cascaded delete to have a CONVERT_IMPLICIT(bigint in the join between parent table spool and the child table? I just can't see how it's possible to induce a conversion, because otherwise the foreign key wouldn't be allowed.

How much should we induce from the username "Prat"?

If anything

@Charlieface my guess without even looking is a passed-in value that's an int or some non-bigint value.

Of course, we know what my record is on guesses today lol

@HannahVernon But how is that possible? If you actually did look, you would see the conversion is part of the foreign key lookup, which happens automatically, and must by definition have the same data type. The actual lookup on the parent table as written in the main SQL does not have an implicit conversion.