« first day (4999 days earlier)      last day (31 days later) » 

12:15 AM
According to so republicans in Buttfuck, wherever, Joe Biden is actually dead and has been replacedvby a double! Hilarious!
12:59 AM
@Vérace expecting/waiting for an evisceration.
@HannahVernon Hoping? :-)
Most definitely hoping
@HannahVernon So far, damp squib!
1:27 AM
Harris disaapointing on what should have been her strong suit - women's rights!
Eating pets! Wow!,,
1:57 AM
Hate to say it but Trump is holding his own - Harris not aggressive enough on his lies!
> MariaDB seeks to eliminate the constraints
That's how you LLM the AI
@HannahVernon breaktime - your thoughts?
@JoshDarnell Let me guess they didn't use await either?? I mean, the docs clearly say use -1 for an infinite timeout #rtfm
Anyone any thoughts on this one?
I don't doubt Paul is right, but he doesn't say anything about having two indexes. My testing seems to show that it does infer uniqueness, see the final query where the Scan Count is still 1 on a non-unique index dbfiddle.uk/afPKcC0KCharlieface 5 mins ago
Case is a unique and a non-unique index on the same columns, does the optimizer infer uniqueness when seeking the non-unique index?
2:24 AM
Letting Trump speak too much!
2:58 AM
PBS panel (mix of d&r) thinks it's a win for Harris with undecideds! I'd say marginal at best...
3:53 AM
Quelle surprise - Fox has Trump winning by 94%! <snigger>
4:26 AM
12
A: Should I mark a composite index as unique if it contains the primary key?

Paul White Should I mark a composite index as unique if it already contains the primary key? Probably not. The optimizer can generally use information about the uniqueness of the contained key column anyway, so there's no real advantage. There is also an important consequence of marking an index unique on...

4:46 AM
 
2 hours later…
6:20 AM
A chairde - Morning all!
Wordle 1,180 2/6*

🟨⬛🟨⬛🟩
🟩🟩🟩🟩🟩
OMFG - talk about steeped!
 
1 hour later…
7:26 AM
RTÉ - Ireland's national broadcaster has an American expert on who thinks that the undecideds would go with Harris - she seems to have unsuprisingly widened the gender gap! Not stellar but solid. V.P. debate on 01/10!
 
4 hours later…
11:49 AM
@PaulWhite That's an interesting answer. But what is that fiddle supposed to tell me?
 
1 hour later…
12:50 PM
@Vérace nice!
1:00 PM
Wordle 1,180 3/6*

⬛⬛🟨⬛🟩
⬛🟩⬛⬛🟩
🟩🟩🟩🟩🟩
@Vérace Chump talks dumb and it it so hard to listen to. I think Harris is the far-and-away better candidate, I just wish she didn't appear as scared as she did.
Must have been a late night for you!
@Charlieface Heh, well. It was awaited. But then the task completed after 24.8 days and chaos ensued.
@JoshDarnell I'm not a .net person so legit question: Why is everything new in .net that I see always setup for async calls. Everything is async await, even trivial functions. Is this expected due to changes in .net design/paradigm or is it just people taking things too far? Or, I guess, a little of column a and b.
1:16 PM
@SeanGallardy Because people suddenly realized that most paradigms they had been using were fundamentally broken or difficult to use, and that await and "promises" (aka Task) was the answer. C# is not unique in this, JS has also been going there among others.
i've heard waiting is the hardest part
Also the requirements of high throughput and low memory usage of web apps and the cloud effectively mandated offloading waiting threads, because native OS threads take up too much memory (both kernel and usermode).
@Charlieface The answer in general or the answer with the .net side of things?
@Charlieface This makes total sense, thanks!
In general, at least when it comes to any app or service that interacts with others or the disk a lot.
CPU-heavy loads don't make sense for this though.
@HannahVernon didn't put the head down till ~ 04:30 my time!
1:20 PM
Think of it this way: you've sent a request to an IO device (disk, network, serial) or you want to wait for a timer. Why ask the kernel to freeze a whole thread just to do that, when you can just tell the kernel to call yo back when it's done.
@HannahVernon If he was standing in Ireland, he'd be lucky to get 2% of the vote - maybe 1%?
Eat my pet!
@Charlieface Depends on the end goal, I can see both sides of it.
This was always possible, but you had to write a Begin and End function for every time you do this, which was too much work for most people. await helps by splitting the function under the hood into separate pieces, and allowing those callbacks to happen but you can still code as if it's all one function.
Debugging is still a bit of pain, because the stack frames sometimes don't carry over properly.
@SeanGallardy What's the other side?
Fox blaming moderators!
the other side is you may want a dedicated thread handling those requests instead of waiting for .net threading (which would have to be implemented over native threads) to pick up the signal.
1:25 PM
@Charlieface Well said, I don't think I could add much to that, @Sean.
Awesome, thanks guys!
I think language semantics come into it to some extent too. Because once one thing is async, the whole call stack needs to be. So it's more "ergonomic" if developers to have the async option available.
@SeanGallardy That's what the threadpool is for, you can even implement your own threadpool if you so desire. It's just a bunch of threads waiting on kernel wait handles in a loop.
@JoshDarnell Yeah there's that. It's the "what colour is your function" problem, see eg journal.stuffwithstuff.com/2015/02/01/… and blainehansen.me/post/red-blue-functions-are-actually-good
C# hasn't solved this, and now there's this whole ecosystem of functions that use Task and ValueTask so the train has already run away, which is a shame. I wish there was a more simplified calling convention for async code.
1:47 PM
so what's everyone up to today
@Charlieface There's an extra Stream Aggregate for the DISTINCT when the data isn't known to be unique
@ErikDarling Testing crazy dynamic SQL schemes. You?
was thinking about trying buying some heroin and a guitar, seems to work for a lot of people
We might get there yet
> After adding just 1000 rows with some non-NULL values in c2, the issue went away entirely. In the production system we un-archived this tiny amount of old data, updated statistics and poof! Plans are stable and have seek+lookup with estimated number of rows around 1 as expected.
 
3 hours later…
4:47 PM
shia magic dot gif
 
1 hour later…
6:17 PM
The opaqueness of certain sentences in mathematics approaches infinity, at least for me. For instance, "In mathematics, an amenable group is a locally compact topological group G carrying a kind of averaging operation on bounded functions that is invariant under translation by group elements." What?
that sentence is harder than math
I literally had to look up everything. I understand what amenable means out of the gate, but what the hell is an amenable group? Is that what we are?
at times
> Amenability has many equivalent definitions.
Well, that solves that then
as long as they are equivalent, I mean
If I understand that original sentence, then the elements of a group form a likeable group if the average bounds of the group don't change. I am indubitably wrong.
what does that even mean though
6:25 PM
> ‘Sometimes one is forced to consider the possibility that affairs are being conducted in a manner which, all things being considered and making all possible allowances is, not to put too fine a point on it, perhaps not entirely straightforward.’
- Sir Humphrey Appleby, in 1986
yeah save it for livejournal
@ErikDarling precisely
I'm in this rabbit hole because I thought it might be nice to learn a little about Von Neumann. I was wrong.
> An intuitive way to understand this version is that the support of the regular representation is the whole space of irreducible representations.
to WHOM
the name should have been an appropriate warning
7:23 PM
@ErikDarling Oh how lovely :-/
wait till they start using stats stream
I regret answering sometimes. Well, kind of. The answer has merit for the future, even if the OP might not
are you saying the op has no merit or no future
7:56 PM
I think I was deliberately ambiguous
8:14 PM
@ErikDarling Sleeping! :-)
@HannahVernon Amenable groups hang around around the bar in Flatland, get drunk and have the odd fight over the cute Tangles that drop by after work!
lol nice
@HannahVernon Maybe you should master cryptography before tackling sentences like that one?
9:21 PM
@PaulWhite I really need to update my answer in lieu of all this new (to me) information.
Has anyone used software to make an animated avatar of themselves and then link that to your input video so that you're the animated character instead of your normal video? If so, which did you use?
I was looking at Adobe's but it seems... ridiculous
9:59 PM
Just use AI bro
(╯°□°)╯︵ ┻━┻
@JoshDarnell It's tough to cover everything first time. That's not a reason to avoid answering btw.
America's in a worse state than I thought - this woman, (who makes Marjorie Taylor Greene look like a moderate), while not winning elections, does well in Florida! <shakes head in pity and disgust...>
10:25 PM
@PaulWhite Thanks! Yeah. And I imagine there's a point of diminishing returns too haha.
I prefer to think of it as there always being more to learn docs.
You never quite know which bits of knowledge will be practically useful in future.
It's not like one can actually complete C# either.
Maybe there's a final SQL Server boss level where you have to take on the entire development and architecture teams in physical combat, idk.
10:40 PM
I’d pick that fight
@JoshDarnell Set a task to remind yourself to update it in 24.8 days
@ErikDarling I'd pay good money to watch that
You may get your wish!
I hope the PPV cut is fair
@PaulWhite hahahha
We'll await with interest
But I'd still leave early to watch Erik's cage match
So, expect a cancellation
As a token of my regard
11:05 PM
I hope I don’t have to get a Draft Kings tattoo
11:30 PM
Maybe just a temporary one
As all tattoos are, really
Some of mine are really starting to show their age
Might need a second coat

« first day (4999 days earlier)      last day (31 days later) »