« first day (4539 days earlier)      last day (336 days later) » 

12:27 AM
😒
 
1:04 AM
@PaulWhite lol
 
1:22 AM
So I know we're not moderating these days, but I thought Dan's edit on this one was funny: dba.stackexchange.com/q/328088/150011
 
 
2 hours later…
3:02 AM
@JoshDarnell I find these interesting. I've bookmarked the link. Recently I heard about bitslicing and bitweaving (pages.cs.wisc.edu/~jignesh/publ/BitWeaving.pdf) for the first time. I look forward to working my way through those, too.
 
 
3 hours later…
6:38 AM
Morning. Tricky Wordle today
 
6:59 AM
Very much so, yes.
Wordle 720 5/6*

⬛⬛⬛⬛⬛
⬛🟨⬛⬛⬛
🟨⬛⬛🟨⬛
⬛🟨🟨🟨⬛
🟩🟩🟩🟩🟩
 
 
2 hours later…
8:45 AM
@SeanGallardy surely they'll stop at 12 (or 24 or 60). 11 or 13 seasons would be blasphemous geometry
 
8:59 AM
Hello, esteemed experts.

I have SQL Server question, that I don't think deserves being posted as question. (But if you disagree, I will do so.)

Is my understanding correct, that that
1. clustered index implicitly INCLUDEs every other column of that table
2. non-clustered index implicitly INCLUDEs columns of clustered index (if it exists)
?
 
yeah, that sounds about right.
 
Wordle 720 4/6*

⬛⬛⬛⬛⬛
🟩⬛⬛⬛⬛
🟩🟨⬛⬛⬛
🟩🟩🟩🟩🟩
@Yano_of_Queenscastle Broadly, yes. For #2, the b-tree clustered index keys (including any uniqueifier) are present as extra keys (not includes) in the NC index if the b-tree nonclustered index is not declared unique.
Gosh it's hard to summarize in one sentence. The clustered index keys are always present as keys at the leaf level of the NC index. They're also keys at the non-leaf levels if the NC index isn't declared unique.
7
A: Why do root and intermediate levels of NON-unique NON-clustered indexes additionally store row ID?

Paul WhiteThe SQL Server storage engine requires every entry in an index be unique at all levels at all times. For a unique nonclustered index, only the nonclustered keys are required above the leaf because they're unique by definition. The full bookmark (aka row locator, RID) to the associated heap or clu...

@Yano_of_Queenscastle There is also a subtlety to point #1. Depending how you define "includes" the clustered index only 'includes' columns stored in row. Any data stored off-row (row overflow or large object data) isn't included by some definitions.
 
9:26 AM
@PaulWhite Interesting. I feel like I have more questions now than before :D Gotta read something about this.
 
@Yano_of_Queenscastle There are quite a few Q & A on the main site already. Have a look at the one I linked above then look around more as necessary. Your understanding is broadly right, there are just always more details if one is interested in them.
 
@PaulWhite I will.
Thank you both for your help.
 
10:31 AM
@ypercubeᵀᴹ 13 is a sacred number so it's all good
 
11:04 AM
Wordle 720 4/6*

🟨🟨⬛⬛⬛
🟨🟩⬛🟩⬛
⬛🟩🟩🟩🟩
🟩🟩🟩🟩🟩
What the h-e-double-hockey-sticks is that
I thought I had a good vocabulary, but I've never heard that before
Ok I'm an idiot, of course I've heard of it, you just have to say it the right way lol
 
Wow
Well you got there in the end
 
12:02 PM
Yah, felt kind silly when I realized what it was. Should have had it in three but I threw away the third guess by reusing a letter I knew wasn't the first letter.
I see the CMs have been doing a little cleanup
Quite surprised they've not removed the featured tag from the meta post
Seen on Discord:
> being confident matters more than being right
4
 
Yep
 
12:19 PM
@HannahVernon well we all know I didn't get my reputation by being right. 😉
 
12:31 PM
FGITW?
 
12:46 PM
Hmm?
 
Fastest Gun In The West
 
oOo. Yes, that helps lol
 
254
A: The Many Memes of Meta

Robert CartainoMeme: Fastest Gun in the West (FGITW) Originator: Omer van Kloeten First Heard: September 11th, 2008 Definition: A problem identified as a side-effect of sorting votes by descending score where first-posted answers are quickly up-voted, trumping vote opportunities for people who sit down and ...

 
Too many acronyms in this world.
 
TMAITW
2
 
1:03 PM
When I started working for the govt insurer here back in 2014, they gave me a list of the acronyms they use. It was well over 2,000 items and they stored it in a database with specific tooling just for looking up acronyms and their definitions.
 
@HannahVernon Lol sounds about right.
Gotta appreciate the organization at least.
 
I like ponies
 
@MichaelGreen Oh wow, that's cool too. Thanks for sharing.
 
1:21 PM
4 hours ago, by Yano_of_Queenscastle
Thank you both for your help.
Paul is now so helpful people think he’s two people
 
> ypercube's feelings
 
4 hours ago, by ypercubeᵀᴹ
yeah, that sounds about right.
I just repeated ypercubeᵀᴹ more verbosely
@JoshDarnell That's trademarked you know
 
@HannahVernon I briefly joined that Discord. It was a little overwhelming haha.
I departed the server after like an hour and a thousand messages.
 
Just a little lol
 
@PaulWhite Oops!
 
1:25 PM
About to get SUED
I always thought "discord" was a strange choice of name
Noun: discord (countable and uncountable, plural discords)
  1. Lack of concord, agreement or harmony.
  2. A false witness that speaketh lies, and he that soweth discord among brethren.
  3. 1775, Edmund Burke, Conciliation with America
  4. Peace to arise out of universal discord fomented in all parts of the empire.
  5. Tension or strife resulting from a lack of agreement; dissension.
(2 more not shown…)
 
Should name the "AI" chatbots that
 
> The name Discord was chosen because it "sounds cool and has to do with talking", was easy to say, spell, remember, and was available for trademark and website. In addition, "Discord in the gaming community" was the problem they wished to solve
 
@PaulWhite No, you got it right. More widely, this is exactly right:
> A false witness that speaketh lies, and he that soweth discord among brethren.
 
@Zikato I see
 
Finally ended our discord when choosing between Mumble, Ventrilo or TeamSpeak
 
1:27 PM
I used TeamSpeak for years, wrote many auto programs to integrate with its API
First one I ever used was Roger Wilco
 
It was a tough choice in times of high latency and low bandwith
 
Also someone posted an "AI" answer to a question they had and asked if it was correct. The question was how to lower SQL Server memory usage. The "AI" response was to exit the application.
 
Anyway, it's not a name that has aged well
 
Not wrong, that'll definitely lower the memory usage of the application...
 
@SeanGallardy Amazing.
 
1:34 PM
@JoshDarnell Especially when you figure in these are MS support people
 
I remember using Ventrilo back in the days of heavily playing Starcraft 2.
 
Imagine getting a response of "To use less memory, exit the application"
@JoshDarnell and Red Alert 2
 
@SeanGallardy Oof.
 
Perhaps AI is acquiring a sense of humour
 
Must be trained on your humor then
 
1:37 PM
No u
 
@HannahVernon Ah, is that who cleared all the spam from the front page? I wasn't sure if regular community folks had flagged things to death or not.
 
I looked once and "Charlotte" had handled some flags.
I think some or all of the spam was sufficiently flagged to death by ordinary users.
Oh, Catija too
 
1:53 PM
@SeanGallardy lol love it.
 
I have been more useless than usual
It’s been easy since I’m been on mobile only all week
Maybe the strike will end Monday
 
2:24 PM
Maybe
 
2:43 PM
I forget what the terms of the strike are
But I am enjoying my vacation
Maybe I’ll blog answers to questions instead of answering them 😤
 
Nothing wrong with asking or answering questions. We're just not moderating.
Up to you what you do or don't do ofc
 
Oh I thought it was a general activity strike
69
A: June 2023 Data Dump is missing

AMtwo DISCLAIMER: I was recently impacted by the Company's layoff. I am going to carefully respond in a way that ensures I don't reveal anything the Company may feel is confidential--particularly with regard to strategy, or future plans. Any knowledge I have on strategy or future plans is both dated a...

 
> Until this matter is resolved satisfactorily, we will be pausing activities including, but not limited to:

* Raising and handling flags.
* Running SmokeDetector, the anti-spam bot.
* Closing or voting to close posts.
* Deleting or voting to delete posts.
* Reviewing tasks in the various review queues.
* Running various other bots designed to assist in moderation, such as detecting plagiarism, low-quality answers, and rude comments.
> Until Stack Overflow, Inc. retracts this policy change to a degree that addresses the concerns of the moderators, and allows moderators to effectively enforce established policies against AI-generated answers, we are calling for a general moderation strike, as a last-resort effort to protect the Stack Exchange platform and users from a total loss in value.
 
Anyone know if SQL Server has a native function similar to IS_ROLEMEMBER() but for checking if a securable is mapped to a database role for a given permission?
 
2:59 PM
Ask ChatGPT.
😜
 
@JoshDarnell It told me to exit the app.
 
@J.D. I can't quite parse your question. Can you give a concrete example of the kind of test you want to perform?
 
Wordle 720 5/6

🟨⬜⬜🟨⬜
⬜🟨🟨🟩⬜
⬜🟩🟩🟩⬜
🟨🟩⬜🟩⬜
🟩🟩🟩🟩🟩
yeah, I know I fucked up
 
@PaulWhite I want to pass in the name of a Database Role and the name of an object, e.g. Table, and then get back a list of Permissions assigned to that Database Role for that Table.
 
@Lamak You might like to try "hard" mode
 
3:03 PM
Or furthermore, pass in the name of a permission as well, e.g. SELECT, and get back a value indicating that permission is provisioned or not.
 
@PaulWhite it's already too hard for me
 
It's easy enough to query for in the sys.database_permissions table. But I need to incorporate this check into some dynamic SQL, so hoping to simplify the code by calling a function that does the check.
 
@Lamak I personally find "hard" mode easier because it prevents silly mistakes
 
@PaulWhite ah, I thought you were messing with me...is there actually a hard mode?
 
@Lamak Yes, it's one of the options you can select
 
3:08 PM
what is different in hard mode?
 
> Any revealed hints must be used in subsequent guesses
 
@PaulWhite I see...that makes it seem like easier mode though
 
I agree
But that's how it works
@J.D. You could probably use HAS_PERMS_BY_NAME or sys.fn_my_permissions but you'd need to impersonate a user who is a member of only the target role with no other explicit permissions.
 
@PaulWhite hmm I'll check them out. For now I created a scalar function to refactor it for me. It's a small use case, so I'm not too concerned about it being scalar.
Thanks!
 
Why bother to ask if you don't care about the answer or have already solved it another way?
 
3:16 PM
I do care, as I noted I'll check out what you offered. I'm time limited right now unfortunately, so was coding a solution in tandem that'll do for now.
Always room for change. Each day better than the last, eh? :)
 
16 hours ago, by Erik Darling
@J.D. thanks for bringing it up 😃
 
Anytime! 🙃
But seriously, I do appreciate you answering me. Will actually look into that.
 
It's probably better to do it the way you're doing it. But, I do wonder why you're checking at runtime what permissions a ROLE you presumably have full control over has.
Normally, a role would encapsulate all the permissions a group of users need to perform some tasks. If you're implementing one of those tasks, it is sufficient to check if they are a member of that role, or even just assume things will be ok.
 
We replicate a few tables from a vendor system. Sometimes those tables get dropped from and re-replicated to the Subscriber as a result of upgrading that vendor app. On the subscriber side we have a user-defined database role that those tables are a part of. And essentially Domain Users are mapped to that database role, so that our custom apps can work.
 
Ah, replication
 
3:24 PM
Not a big deal for me to manually re-add the dropped and re-synchronized tables to the database role, but I was sleeping when the call came in this morning so I rather have a SQL job check and automate it for me.
Shame on me, I do this already for one of our other replicated vendor systems. Should've setup this up already for this vendor's replicated database.
 
> By default, replication does not execute GRANT statements on the subscription database. If you want the permissions on the subscription database to match those on the publication database, use one of the following methods:

* Execute GRANT statements at the subscription database directly.

* Use a post-snapshot script to execute the statements. For more information, see Execute Scripts Before and After the Snapshot Is Applied.

* Use the stored procedure sp_addscriptexec to execute the statements.
 
Yup yup.
 
@HannahVernon any 10 Gb Nics to recommend for home use?
 
This all horribly reminds me of when I used to have to copy a complicated replication setup to dev and test environments.
 
3:50 PM
🧊💊
 
It was a long time ago, I'm over it now
Mostly
eye twitch
 
Someone recently asked me to help on a Replication set up project that has a sensible use case. I’m waiting to hear back some more details.
 
@ErikDarling Interesting. Are you able to generically share the kind of use case it is?
@PaulWhite Word. Yea it's not exactly replications fault per se in this case. We generally truncate the tables, but because of certain schema changes the vendor made, it was just simpler to drop the tables on the Subscriber side and let a re-initialization re-add them.
 
Analytics people run shitty queries against transactional data and they want to get them off there, so the plan is to replicate tables over and add column store indexes to the subscriber side.
 
Filtered nonclustered columnstore on the main tables not an option?
 
4:05 PM
It’s an option of course but the queries are resource intensive and step all over the normal workload
And of course they’re all ad hoc whim and fancy so there’s no hope of properly doing anything
Well, the normal workload and other reporting queries, so let them figure that out somewhere else
Something like 80 hours of resource semaphore waits in 120 hours of uptime
Average wait of two minutes
 
Ah so they already have NCCS
 
so, is SO really not going to provide the data dump anymore?
 
seems like it
 
4:12 PM
jeezus, that's far worse than the stupid not-moderate-AI-answers thing
 
22
A: Is SE [going to be] selling our content for AI model training? And what exactly does "reinvest back into our communities" mean?

PhilippeI'm here to respond to some of the points raised in this post. No doubt I'll miss some, so I appreciate your forgiveness and understanding that I'm working to answer as many questions as I can. Some things I'm not competent to opine on (legal interpretations for instance), and I've passed off to ...

you've reached my caring limit on this issue though
 
Caring limits being what they are these days
 
Quite
 
mmmm not that I care more than that either...just reaaaaally don't like the decision
 
This is interesting:
> Chandrasekar [...] argues that [selling content to train LLMs] will also help future chatbots, which need "to be trained on something that's progressing knowledge forward. They need new knowledge to be created."
And at the same time they allow bot-generated content on the site, ultimately training bots on the content generated by bots.
I'm sure that will progress knowledge way faster than before
soon we'll be able to say: "knowledge being what it is these days"
 
4:31 PM
Maybe they’ll pay us for content now!
 
I can only imagine the circular downward spiral as these "AI" technologies don't actually create anything new. They don't actually think (as in problem solve in an abstract way) so I'm unsure how that's going to helpful to train bots on bot actions. In fact, taking a page from video games, it's quite terrible what ends up happening.
 
It’ll be like those old videos of Siri and Alexa talking to each other for three days.
 
Exactly.
 
@ErikDarling Yea custom indexing on the replicated data is a big + over AGs for example. :)
 
One might say.... the only plus
 
4:42 PM
Heh well it falls under the bigger umbrella of being able to customize the schema of the replicated data (related to what we discussed the other day).
I go as far as creating indexed views in our replicated databases, but I'm a madman.
Custom Row-Level Security, custom database security, etc.
 
5:01 PM
@J.D. Yep, threw out all of the usual options pretty quickly
 
@ErikDarling You mean they weren't jumping out of their seats to implement SSIS? 🙃
 
Thankfully they’d never heard of it
 
Worddd
Yea I raked the coals on the options a few years ago when I started here at Haugland. Potential data synchronization "solutions" I recall could be 1. Replication (inclusive of the various kinds), 2. AGs, 3. SSIS, 4. Triggers + Linked Server, 5. SQL Agent Job + Custom SQL. 6. Application code. Did I miss any? 👀
 
5:29 PM
An ocean of third party tools
 
True, fair enough.
 
If we’re counting AGs and snapshot replication then log shipping is in the picture
 
5:56 PM
@SeanGallardy I've only got 1GB lol
 
6:37 PM
@ErikDarling ah yes, I looked into that too at the time.
 
 
1 hour later…
8:03 PM
@ErikDarling Based on Change Data Capture/Change Tracking?
 
8:30 PM
@PaulWhite usually yes
Some claim to work off base tables with just a primary key available but who knows
 
8:41 PM
Hey, welcome to SO! You're new here, but someone should have told you: traditionally we used the Reddit dumpster fires as examples of what not to do. — Shog9 29 mins ago
 
8:54 PM
Hahahah good old Shog9
 
Jody is not any more popular internally
Though I suppose as people “leave” his popularity will rise
At least until more people need to leave who are diminishing his popularity
 
Big yikes.
I actually just got on Twitter for the first time in a week or so and I saw Kendra has parted ways with SO.
 
9:20 PM
Yeah, directly related to a lot of the nonsense you’re seeing publicly and a whole lot more
 
Jeez. That really sucks.
 
Might as well
176
Q: June 2023 Data Dump is missing

Data DudeThe data dump usually get refreshed the first weekend of the month, every 3 months. The current Data dump is still from March. Is there just a problem and it's delayed like in the past?

Oops
176
Q: June 2023 Data Dump is missing

Data DudeThe data dump usually get refreshed the first weekend of the month, every 3 months. The current Data dump is still from March. Is there just a problem and it's delayed like in the past?

Why isn’t my comment link working
 
“No downstream restrictions. You may not offer or impose any additional or different terms or conditions on, or apply any Effective Technological Measures to, the Licensed Material if doing so restricts exercise of the Licensed Rights by any recipient of the Licensed Material.” — Erik Darling 3 mins ago
Who knows.
 
Mobile was a mistake etc.
 
A chairde - Morning all!
Only 18 questions left till we break the 100K barrier!
 
9:35 PM
I’ve heard that line before
Perhaps with fewer moderations it will last longer
 
Feb 15 at 14:12, by Josh Darnell
Question 100,000.
Back in February!
 
9:52 PM
🎯🎯🎯
 

« first day (4539 days earlier)      last day (336 days later) »