« first day (1903 days earlier)      last day (2959 days later) » 
01:00 - 19:0020:00 - 23:00

1:21 AM
@AndriyM Yes it is expected. From the SE FAQ:
> On Stack Overflow, migration requires a super-majority when closing, which means that it will only be migrated automatically if four users vote to migrate it to the same site, on other sites, three votes for the same target suffice.
74
A: What is migration and how does it work?

Kyle CroninWhat is migration? Migration allows an off-topic question to be gracefully moved to another site in the Stack Exchange network. It preserves the current revision of the question, all its answers, any comments on any post, as well as most of the votes. Side effects of migration Down votes are ...

@AndriyM Of course you're right: we should not throw rubbish to any other site, in the same way we don't want rubbish from them.
 
1:50 AM
@PaulWhite Thanks! I knew there had to be some fine points about migration I wasn't aware of, but that FAQ article has proved to me there are some that are important to always keep in mind.
 
2:02 AM
Really, unless the question is high quality and clearly on-topic, it's usually better just to vote to close (not migrate). The asker can normally delete here and re-ask on the target site anyway. Flag it if you see this sort of thing again.
 
2:14 AM
@AndriyM Edited.
@dezso Because there's a difference between being regular and being good :)
3
 
 
1 hour later…
3:30 AM
About temporal tables in SQL Server 2016: no real surprises. They present as two tables (current value and history values) twinned through meta data, with enough syntactic sugar and sensible defaults to make it easy. There are a few obvious restrictions on what can be done, and a few less obvious ones that feel like v1 software. A nice addition to the portfolio but nothing revolutionary.
In the execution plan the "current" and "history" tables show up separately. There's a Concatenation operator if the predicates would touch both.
 
Yep that sounds right. I haven't paid much attention partly because there's not much (anything?) you couldn't do yourself, and it is all a bit V1.
Saves a bit of repetitive effort though I guess.
 
 
1 hour later…
4:43 AM
@JamesLupolt sorry. just getting this. what's up?
 
 
2 hours later…
6:29 AM
@PaulWhite And gone :)
 
@AndriyM Yeah, well I tried. Maybe he learnt something about formatting and attribution.
 
And, perhaps, answering questions that might not be his cup of tea.
 
@PaulWhite obviously got the peer pressure badge
 
@dezso Not this time, a bit earlier, but yeah, he's definitely got it :)
 
 
1 hour later…
8:00 AM
@PaulWhite @AndriyM problem solved. Question closed as unclear (migration rejected). stackoverflow.com/questions/36120369/mongo-in-php-is-deprecated
The original is on our site, still: dba.stackexchange.com/questions/132800/…
 
Cool (well not really, it wasted a bunch of time, but still).
 
@ypercubeᵀᴹ I guess I'll restore my original comment then.
 
@PaulWhite Well, I'm regular nowadays. The Lactulose really worked a treat.
 
(If I can find it)
 
Your question is unclear. Please add more details. — Andriy M 10 hours ago
@AndriyM That one?
 
8:06 AM
@PaulWhite The version I wrote initially just before casting my close vote. It didn't make sense once I realised the question was migrated, so I edited it.
 
Ah.
 
8:19 AM
Had to adjust it a bit, but basically this is what it was like initially:
Your question is unclear. It is also likely to be off-topic here while it may be on-topic on Stack Overflow. The community did attempt to migrate your question over there but it was closed as unclear and so the migration was rejected. If you are positive that yours is a database issue, please provide more details to make that point clear and allow your question to be re-opened. Otherwise, if I were you, I'd start afresh by posting the question on Stack Overflow (and deleting this one here) and see how it goes. You can always flag your question for moderator attention and request migration, if necessary. — Andriy M 2 mins ago
 
8:31 AM
And TIL: rejected migrations don't seem to be too much of a mess any more. Previously, when rejected on the target site, they used to remain locked on the originating site. Looks like that has been fixed.
 
@AndriyM There's plenty of scope for rejected migrations to stuff up. Don't be fooled by the simplest possible case "going well" :)
There's nothing good about a rejected migration.
 
All right then
 
8:53 AM
MVCC somehow says that they can proceed without locking. They say that Version control systems do the same. I wonder how do they accomplish check-in without locking. Moreover, I revision control seems to relay upon the locking. Is it different from version control?
 
@ValentinTihomirov I think "revision control" and "version control" are 2 names for the same thing
 
9:26 AM
Anybody still knows concorde XAL? :)
it's going to be one of those days
 
9:43 AM
I do not get why 3-phase commit is more reliable. They say that 2-phase commit cannot stand the coordinator failure. But how the 3rd phase of the commit protocol can solve this problem?
 
@ValentinTihomirov Who are they and where do they say that?
You know that commenting on random stuff you read somewhere does not mean that others know what you are talking about.
3
Please provide at least a link, if you want others to have a clue of what you are talking.
 
Sorry, I believed that it is a standard treatment of 3PC.
Reading more carefully, 2-phase cannot recover if both cohort member and coordinator fail. You do not know if cohort member has received the commit message or not. But, I do not see how pre-commit helps here.
 
10:08 AM
@ValentinTihomirov There is also a possibility that what you are talking about may not be trivial or common knowledge enough for people here to be able to quickly understand the issue, much less to have something to say in response.
 
I do not expect answers form people who are unaware of what is 3-phase commit.
 
That's understandable.
 
@TomV What on earth is that? I hope it's supersonic at least :)
 
@PaulWhite the help about says copyright 1990-1998
it's about 20 years old and AX's predecessor
the customer has a license issue, apparently the thing is licensed to have a database of a certain size, but you cannot get new licenses anymore (for obvious reasons)
 
50 mins ago, by Tom V
it's going to be one of those days
Seems like.
 
10:18 AM
@ValentinTihomirov Only I'm a little at a loss if this MVCC thing is indeed something not uncommon for database people to know about in some detail or if you are just taking your chances asking these questions here. (Don't get me wrong, please, I don't mind at all, just a little concerned for you, because your questions on this subject seem to have been getting little useful response.)
 
@PaulWhite The solution "for now" 5 years ago was removing data (old inventory transactions and stuff)
it has been the solution every year since then
 
Nothing's more permanent than temporary.
4
 
Yep.
 
They now are using 1185 mb from the 1220 in their license
 
10:42 AM
I can't see anyone ever needing 1220MB of storage.
 
@PaulWhite Yes, I edited that immediately
 
@TomV Huh for some reason it didn't refresh on my screen until I hit F5.
 
11:23 AM
Ever worked with adhoc reports anyone ?
this is the only subject I struggle to find help on internet for :/
 
12:28 PM
@Mathematics No movies about it?
3
 
12:46 PM
@TomV not that I am aware of :(
 
What version of report builder?
 
1:00 PM
I designed standard reports using SSRS, and showed them to clients using report viewer in asp.net, but now requirements are to add adhoc reporting
but, I can't use SSRS report builder
because of 2 reasons,

1 - we want a nice looking interface
2 - our web app is multi tenant app (so security will a massive issue and emerging report builder with our app is aweful as welll)
 
That's a bit broad for a chat don't you think?
 
1:32 PM
pretty sure people had developed it before me, just struggling to find keywords to search for it
should I call it reporting engine
it will look like this
 
@Mathematics Are you looking for products that have such functionality?
 
@ypercubeᵀᴹ nope, as then I will have to integrate them to our web application
I want to design it myself, infact design is already done
I have all of entityframework set up as well
just want to find out how people save these sort of information as templates in database
 
2:03 PM
7
Q: Is it ever ok to not test a feature?

ZX9Is there any point at which you become so familiar with your language/database/system that there is no need to test a new feature/configuration/query/etc. by contained/simulated testing before implementing it in your system (especially concerning a feature that modifies data)? Or is it always ess...

@MaxVernon @ypercubeᵀᴹ @KennethFisher Are you guys still of the opinion that question should be reopened?
@MaxVernon Posting a comment meme was not the response I had in mind :)
 
@PaulWhite I'm not certain there is any point at this time? I'd love to know how you calculate risk on work, like the OP wants.
@PaulWhite lol....ok
 
@MaxVernon Well there's a couple of points: it narrowly failed vote to reopen twice, has seven up votes and no downvotes, an OP that is still active on the site, and at least one comment that suggests someone would have answered had it not been closed.
Getting a reopen vote on dba is quite an achievement. Especially from users whose opinion I trust.
 
@PaulWhite hmmm in that case, I'd have no problem reopening it. It's not a particularly bad question. I expect the VtC was around "don't be stupid", without a lot of actual thought.
 
@MaxVernon You voted to reopen :)
I guess I just wanted to check with everyone instead of actually being decisive.
 
@PaulWhite I thought that was why you asked me about it above. I have no recollection of actually voting to reopen it, or what my particular thought was about why.
I suspect my reason was to allow Kenneth to answer.
 
2:15 PM
Maybe yes.
 
I'll hit the switch again
done
 
Ooo nice one. OK.
 
I liked your answer here - catching the mismatched types is pretty darn important. I should have tweaked on that when I saw the datetime2 type; at least enough to check that. ;-)
 
@MaxVernon It's a topic close to my heart :) It's the a shame that OP asks so many questions and doesn't always come back to older ones. I'm still hopeful he will say something new on that page.
He's after the Socratic gold badge of course.
 
Even then you can't deny him being good at generating questions that qualify for the badge
 
2:26 PM
Sure. It's a gold badge for a reason. We like people to ask good questions.
 
@TomV I am not ignoring you, I will reply as soon as I can — marcello miorelli Oct 6 '15 at 16:05
I suppose he still hasn't had the chance :)
 
Ah. Yes. Many questions = many jobs to do.
 
why no one wants to speak about reporting here
 
@Mathematics it's not very cool
 
@PaulWhite He works at a customer of ours
 
2:30 PM
@TomV Does he?! Small world!
 
But he doesn't know :)
 
Secrets are cool.
 
Not that I ever met him, it's a multinational and he must be in some other office, but I saw it in a screenshot of his
 
@PaulWhite way more so than SSRS
 
@TomV But he might as soon as he sees your face? :)
Oh, you didn't meet him
 
2:32 PM
@MaxVernon and MVCC :)
 
@PaulWhite amen to that!
 
I like Fast Reports, though.
 
@TomV So tempted to superping him here so he discovers your secret (not really).
 
Not going to talk about it anyway :)
 
@AndriyM I like fast cars.
And pretty execution plans.
 
2:34 PM
@Mathematics no one wants to talk about programming one
 
@PaulWhite It's no secret, I just didn't feel like going "hey dude, I notice you work there, I work with one of your vendors" in the comments
 
@PaulWhite I like both of those too. Not very good at getting either, however.
 
@TomV Well that's no fun.
 
@PaulWhite just to clarify, are you're talking about model-view-controller and the code it generates, or read-committed-snapshot?
 
But we can pretend it's sooooper secret if you like
 
2:36 PM
@MaxVernon Three phase MVCC.
 
@PaulWhite ahhh
 
@TomV Well it's all a bit late now, you've already ruined it! :)
@AndriyM Me either, but it's good to have goals.
 
@Mathematics OK I'll bite (waiting for a restore anyway) Why would you want to design your own ad-hoc working tool? It'll pretty much suck
If you take into account the number of man-hours put into the existing ones and most of them suck too
 
@TomV pretty sure no one here would want to do that either.
 
@TomV Can confirm. Did one. Sucked
 
2:38 PM
@MaxVernon (That link for MVCC context).
 
Well..not personally. We built ReportBuilder into an internal application
 
@TomV I shouldn't get into it if I were you
 
@PaulWhite yah, cool. thanks!
 
@Mathematics use excel pivot tables or an existing tool
 
2:40 PM
did someone say pivot? :P
2
 
unpivot
 
Good grief it's nothing short of spooky.
Might be a bot! Let's ask it a clever question.
@bluefeet Are you real?
 
@PaulWhite Did I have the opinion to reopen that? I must have been drunk ;)
 
@PaulWhite I don't understand that question. :)
 
@bluefeet Ha ha ha awesome!
@ypercubeᵀᴹ Yep
 
2:43 PM
@PaulWhite There's no indication there he was drunk
 
@AndriyM You were expecting a little beer or wine glass icon? That would be a cool feature.
 
lol... I'm a scholar, but not scholarly enough for an upvote.
 
Or blood alcohol level
 
@PaulWhite Perhaps I was thinking that Kenneth wanted to answer. I can't remember now
 
31 mins ago, by Max Vernon
I suspect my reason was to allow Kenneth to answer.
 
2:50 PM
Ha I see Erik lost his beard
 
lol, thanks @PaulWhite
 
Nagging As A Service.
@TomV Has he?
 
@PaulWhite loooool. and now he's accepted.
 
Oh! Or perhaps he'd just shaved more recently?
@Erik
 
2:56 PM
Bam, I got told!
I deleted the "old" dll I found. It doesn't matter what ddl it is it's a custom dll for SSIS script task. There is a standard way of using a DLL for a script task. Which is why I am asking if there is another place where the dll can be. Since there seems to be a standard way of loading custom dlls in script task SSIS — Ya Wang 14 mins ago
 
Ya Wang?
 
@TomV Can't have to integrate it into our web applications
 
@billinkc Perfect!
 
if anyone interested
0
Q: How to store tables, columns and filters as templates

MathematicsI am able to create standard reports using SSRS which work perfectly, but now I need to create ADHOC reporting engine/solution, we can't use standard SSRS report builder because of following reasons, It looks awful and is very difficult to integrate it with our web application Our web app is mu...

 
2:58 PM
@billinkc I should start using this when someone mentions pivot
 
@bluefeet you should!
 
@Mathematics the questions says C# and asp.net, but you list "very difficult to integrate it with our web application" as one of the reasons to not use SSRS. Correct me if I'm wrong, but isn't it very simple to integrate them in that environment?, it's one of the "pro"s of it
 
Yes, but not if I am doing adhoc reporting like in the picture :)
 
@Lamak It might be that their web app makes it exceptionally difficult to integrate
 
it works perfectly for standard reports
 
3:02 PM
@Mathematics We are usually not interested in SO questions. If you cared to migrate it to DBA.SE ...
 
@billinkc it might
@ypercubeᵀᴹ which would be then closed as off topic I assume?
 
Quite likely.
 
@Lamak Why? How to store metadata? It's not very clear, but I'm sure it might be a valid db-design question, even if it needs some polishing
 
(crying quietly)
 
@ypercubeᵀᴹ Are you drunk again?
3
I thought that might earn a star :)
 
3:05 PM
@ypercubeᵀᴹ well, yeah, you are right. Didn't read the whole question actually
 
@PaulWhite The first was from me ;)
 
@ypercubeᵀᴹ I kinda knew that.
 
I'm happy I found a bug that has cost me like 3 days to identify. And another 5 days to the developer that wrote (the code with the bug). So, drunk with relief.
 
1
A: SQL Server on SAN, same LUN: one logical drive vs multiple

Max VernonIn your question you say "one LUN assigned to a SQL Server VM" - what you want is one LUN assigned to the host VMware machine that can have multiple VMDKs for the SQL Server stored on it. This allows the VM to use a much larger number of threads to access the disk. In order to get the best queu...

Max gets my last vote of the day. Congrats.
 
@ypercubeᵀᴹ is it possible for you to migrate it or should I flag it myself ?
thanks
 
3:16 PM
@Mathematics We can vote. That'll need 5 votes. Your flagging might be faster.
 
thanks, let me try
 
@PaulWhite wooohoooo!
 
@PaulWhite I decided I should look a bit more professional, so I shaved and updated my picture.
 
Vertica, boom!
 
lol
 
3:19 PM
I heard my name
Boom
 
it's too bad Chat Rooms don't have a sound effect as someone enters the room.
 
Does Database Administrators want this?
1
Q: vertica Received fatal signal SIGSEGV

Sergii SuchkovOne of my vertica cluster node down, and in log file a found : 2016-03-19 20:01:55.263 Init Session:0x7f3420013970-b00000000a70ab [EE] <INFO> SortManager found maxMerges 7 too small(38 MB Assigned). 2016-03-19 20:01:55.263 Init Session:0x7f3420013970-b00000000a70ab [EE] <INFO> After disabling op...

 
customizable for each person of course
 
@MaxVernon i'd prefer a ping to my cell phone
 
3:20 PM
Even when on an airplane..
 
@TomV Yeah it didn't cause quite the splash as when @MaxVernon updated his but I'm ok with that. :)
 
7
Q: PyGreSQL vs psycopg2

maxpWhat is the difference between these two apis? Which one faster, reliable using Python DB API? Upd: I see two psql drivers for Django. The first one is psycopg2. What is the second one? pygresql?

Anyone?
 
@Erik I sure wasn't expecting anyone to notice. The response really amazed me.
@Kermit seems a bit thin on details
 
@MaxVernon per usual
 
no doubt
 
3:24 PM
@MaxVernon You're a popular and highly visible guy with a distinctive pic. I noticed the change before the meta post.
 
@Erik so many things have changed since High School.
 
lol ain't that the truth!
luckily in my case the changes are almost all for the better
 
@Erik And I got your beard anyway, just in case :)
 
@Erik Still the old picture on the SE network, or should I blame caching? ;)
 
@AndriyM Thanks. I'll give you a holler when I want to look like a vagrant again :)
 
3:30 PM
@Erik Ha, sure! :)
 
@PaulWhite Yeah I think it is a caching thing. I replaced it last week and updated my profile blurb too. I clicked the button to propagate the changes. I'd think that the cache would be invalidated by now though...
 
@ypercubeᵀᴹ SO has an epic flag queue.
 
I'll check some of my other SE profiles to see if the changes are there.
 
@Erik That's true.
Huh. No 1-box. edit: Finally!
 
Indeed, my purple triangles and diamonds are still there in the SE profile too.
 
3:32 PM
@PaulWhite Thanks for the heads up. I just manually updated it based on my dba.SE profile
 
@Erik That desire to look professional fades away usually
 
@Erik Confirmed beardless for me now too, network-wide.
 
Huh? And the link just oneboxed
 
@TomV yeah when I was in the Marine Corps I looked sharp everyday. As you can see that didn't stick for the long term... I signed up for LinkedIn too. So far the longest I've ever stayed on social media is 6 months. We'll see if that holds true on LinkedIn too.
 
@AndriyM I fudged it manually. That's flair.
 
3:35 PM
@PaulWhite Ah, I can see now!
 
@AndriyM It looks like you and @TomV 's profile images are new too. I think you guys both had default images last time I was on chat
 
@Erik We did.
 
Both of your images look much nicer
 
Paul coerced us suggested we change them
 
3:38 PM
lol
 
Very few defaults in the recent user list now.
 
Like most of his suggestions I think it was the right call
 
@Erik And thanks!
 
@AndriyM I believe he was trolling :)
 
@Erik MOST?!
3
(lol)
 
3:41 PM
No I think it is nice to connect names with faces. You are legit people now instead of bots that Paul White created to get a more votes in a single day :)
@PaulWhite I haven't seen it yet but I'm sure its possible... :)
 
did i miss something?
 
@PaulWhite so i missed nothing
 
@swasheck You missed Ya Wang.
Scroll back. Everyone's been unusually entertaining today.
 
(sorry. deleting)
 
3:49 PM
@swasheck See if you'd posted that at the time, you would have gotten 100 stars.
That's what you get for neglecting The Heap.
 
or being in a different timezone
 
Excuses, excuses
 
@AndriyM i didnt realize your face wasnt purple and white
 
If I can be here at 4:50am I'm damn sure you can make an effort.
 
@swasheck Yeah, my real one is less professional-looking
 
3:51 PM
@PaulWhite you're superhuman. i'm subhuman
 
And on that note, I hand over to @swasheck to keep order. Back at the UTC rollover.
 
@PaulWhite ... cats and dogs living together ...
 
heh
 
wtf, apostrophe?
 
@swasheck Nothing urgent. Just wanted to talk about some ideas for what stats to collect for my metrics DB
 
4:06 PM
is there something I'm missing here? I can't get SQL Server to understand what he's talking about.
0
Q: NULL-coalesce a Boolean expression to FALSE

HeinziI have a T-SQL Boolean expression exp which can evaluate to TRUE, FALSE or (due to SQL's ternary logic) NULL.¹ Is there a generic method to null-coalesce such an expression to FALSE? (Obviously, COALESCE(...some Boolean expression..., FALSE) won't work, since T-SQL does not have a FALSE literal...

I'm tempted to VtC as "unclear", but maybe I'm just being dumb
 
@AndriyM Now I gotta update my picture of you
At least I hadn't finished my @TomV masterpiece before these new avatars rolled in
 
@MaxVernon See my answer. Might make it clearer.
 
@billinkc Looking forward to it!
 
@AndriyM I'm not :)
 
@JamesLupolt all of them
 
4:13 PM
@TomV I'm sure it'll be something nice. I was a king last time :)
 
@PaulWhite VtC'd as Off Topic - too localized
I liked your answer though.
and I'm downvoted. Nice. Guess he must think I'm not providing the answer he wants.
@PaulWhite I +1'd it since hopefully that answers his question!
I deleted my answer since nothing about it makes sense to me. Perhaps I just don't understand his logic.
 
EXISTS is interesting. It can only return true or false, never unknown.
 
@swasheck What metrics would you use to watch for anomalies in non-normal data?
 
4:28 PM
@JamesLupolt depends on your architecture. i watch OS VM metrics (for VMs) ... a subset of Logical Disk counters ... Network Counters
Process Counters
 
@MaxVernon: I have taken your advise to "write a good question" and have greatly rewritten the question (made it less theoretical and more practical). I'd be glad for feedback. Is it easier to understand now? — Heinzi 1 min ago
@MaxVernon He is trying, at least.
 
Sorry, I mean what statistics would you use to tell whether a metric with a non-normal distribution was unusual
 
@PaulWhite true. I still don't get it, but I'll give it another go.
 
@JamesLupolt well first i'd try to figure out what type of distribution it is
bimodal or some other polymodal?
if you really want to geek out, then perhaps try to curve fit (R is great for this sort of thing) and then you can evaluate actuals vs. expected
 
@TomV What are you wearing in your photo? Does that coat have a hazmat hood attached to it?
 
4:33 PM
I wonder what the Apple event will bring us. Starts in 1/2 hour
 
@Phil Apple products
NAILED IT!
 
@swasheck Yeah I guess that's what I need to do. Figure out 1) how normal the data is and 2) for data that isn't normal, what distribution it resembles.
 
@Phil Hep-C
 
@PaulWhite ok, I do get it. And your answer appears to be exactly what he wants!
 
This presentation has some ideas.
 
4:33 PM
@billinkc +1 Insightful
 
@AndriyM There ought to be a shorter form using NULLIF or IIF if 2012 +
 
@Phil Even a blind squirrel finds a nut every now and then
 
@MaxVernon I think AndriyM's might be better. Too tired to really think.
 
@PaulWhite IIF will certainly make it more concise. NULLIF – can't see it yet.
 
The (shorter) SQL standard would be WHERE c1 IS DISTINCT FROM 1
 
4:41 PM
One of those questions that gets more interesting the more you think about it.
 
@JamesLupolt seems like you'd need to also calculate a seasonality (reading through the deck ... EMA is where i'm at and it seems like a good start). i used to know how to calculate that.
 
@swasheck Dejan's chapter in TSQL Querying talks about how to do EMA in TSQL. And there is this paper about how to update it incrementally if you can't keep all the measurements.
 
@ypercubeᵀᴹ If you remove ELSE 0, then you are back to ternary logic, no?
 
@JamesLupolt is EMA a core calculation for seasonality?
 
@swasheck No idea. I'm ignoring seasonality right now.
 
4:44 PM
@AndriyM no, You either have 1 or null as a result
 
I'm only comparing each 15 minute segment on a given day of the week (e.g. for each metric, comparing every instance of 23:45:00-23:59:59 on a Sunday).
 
This might turn out to be a terrible idea
 
@ypercubeᵀᴹ Comparing it to 1 you get TRUE or UNKNOWN. Negating that, you get FALSE or UNKNOWN.
 
@AndriyM Your first query would work ok. The second, no, you are right.
 
4:46 PM
2 hours ago, by Paul White
@ypercubeᵀᴹ Are you drunk again?
 
At least that's where I stopped when I was considering omitting the ELSE.
 
@JamesLupolt that would be a good way to predict ... is this "normal" for Monday at 1330?
 
@PaulWhite "active" since 2007. Gotta love Connect.
 
@PaulWhite Added the IIF option. Still can't see a NULLIF solution (not a concise one anyway)
 
@MaxVernon It was going to be fixed in SQL Server 2008. Then they seem to have forgotten about it...
 
4:58 PM
@AndriyM Ok cool.
 
And my reference has been invalidated.
 
A_V
Hi
Has anyone here been granted access to SQLserver on linux preview ?
 
@PaulWhite I loved the miraculously appeared (y) ;)
 
@ypercubeᵀᴹ It was z when I wrote it first, but x (z) seemed wrong.
 
Found in the wild the ever elusive RIGHT JOIN
 
5:10 PM
@swasheck Yeah, that's what I was thinking. I think it will help in situations where metric x is being blamed for new problem y, but the values of metric x haven't really changed much.
 
22
A: Does anyone use Right Outer Joins?

Martin SmithTo give one example where a RIGHT JOIN may be useful. Suppose that there are three tables for People, Pets, and Pet Accessories. People may optionally have pets and these pets may optionally have accessories CREATE TABLE Persons ( PersonName VARCHAR(10) PRIMARY KEY ); INSERT INTO Pers...

 
@A_V I don't think many (any?) people who've applied through public sources have been given access.
 
@ypercubeᵀᴹ Did your cat get on your keyboard?
 
@PaulWhite afraid of asking again if he's drunk?
 
5:12 PM
yes, a human cat
 
@Lamak A bit.
 
A_V
@JamesLupolt What's the cool people's way of getting access ?
 
@JamesLupolt the first step in any good statistical analysis is discovery ... i'd recommend a plot (using R, python, Excel, whatever) to look the trend of the particular metric you're analyzing
@A_V we're not cool enough to know
 
The first rule of...
 
i thought you were sleeping?
@RoboPaulWhite
 
5:14 PM
@Zane thanx for reminding me. I keep procrastrinating righting an answer to a recent question (involving RIGHT joins ;)
 
@ypercubeᵀᴹ i know, RIGHT?
 
@swasheck Oh yes, I forgot. Bye!
 
(hey all. dad's gone. let's play with matches and kerosene)
 
@A_V No idea. Probably become an MVP, get a job at Microsoft, or work for a company that wants to dump Oracle and give MS a lot of money.
 
I've leave you all trying to right the right right join, right?
 
5:15 PM
@swasheck if it wasn't a good joke, I'd have flagged it - for improper English :)
 
@ypercubeᵀᴹ if it was improper english then i wouldn't have typed it
 
1
A: Is it ever ok to not test a feature?

Kenneth FisherI want to start by saying everything I do is SQL Server so those are the examples I give. In general however this applies to any form of code regardless of system. Let's start by breaking this down a bit. Upgrades You have a system and are about to upgrade some or all of it. For example upgr...

 
@A_V I received an email last friday cc @JamesLupolt
> Thank you for signing up to receive more information about our plans for running SQL Server on Linux. We look forward to sharing updates with you periodically. We are currently in the process of reviewing applicants and will begin sending out invitations to join the preview over the next few weeks, gradually increasing the number of participants added each week.
@billinkc It's my roll-up trollface
 
<3
 
No it's just a real warm coat with a hoodie, it was cold that day
 
A_V
5:25 PM
@TomV yeah got the same thing, was just wondering if some people were luckier or cooler than me on here :)
 
@A_V I don't think an MVP would need to apply, they could just ask
@A_V You want to use linux to be able to patch without rebooting?
/JK
 
A_V
ahue
Nah, I'd like to try it out just for laughs
So I can be some kind of sqlserver hipster and say I tried the very first version before everyone else
 
Do you have *nix experience?
 
A_V
on oracle only
It'll probably be the same MDF/LDF setup but it'll run on unix servers so it shouldn't be too hard to learn.
still it feels like a dumb setup to run oracle on windows and sqlserver on unix
 
It's only dumb when it's mandatory. Otherwise it's experimental.
 
A_V
5:41 PM
The question is, at which point a server architect decides it'll be wiser to run either of those setups
 
@AndriyM Indeed, I just would like to see how quickly it breaks (if it does), I have no intent of running it in production
And I run the OS on anything but the work laptop anyway so testing it shouldn't be much more work than spinning up a docker container
From the last slide of a AX version++ session
When you join the beta but are high as f*ck
 
@ypercubeᵀᴹ he's been on fire lately
 
/out, playing cards tonight
 
@TomV is it similar to bridge?
 
A_V
5:57 PM
@TomV To be more precise, I've mostly used solaris and redhat
 
6:18 PM
Is it right that with atomicity and isolation I reach the consistency? I mean that if I isolate transactions and perform their action atomically then I always have system in a consistent state, assuming that transactions do not violate the constraints.
 
01:00 - 19:0020:00 - 23:00

« first day (1903 days earlier)      last day (2959 days later) »