The Heap™ – Consultancy ©® - 2020-01-24

« first day (3307 days earlier) ← previous day next day → last day (1547 days later) »

12:01 AM

Fucking Microsoft. excel.uservoice.com/forums/274580-excel-online/suggestions/…

If someone has an Excel worksheet and a formula uses CELL("filename") to get the name of a tab in the formula, if I pull it down through Sharepoint and then use PowerShell to import it into SQL Server, because Excel for the web doesn't support CELL("filename") all those cells are #VALUE!

Even though no one is really opening it using Excel for the web.

Because Excel is sharing our live edits through Teams and Sharepoint Online.

And after several years there is still no workaround for CELL("filename") except to maybe put a cell in the worksheet and use a fixed cell reference or something.

2 hours later…

1:50 AM

@CadeRoux that's why genetic analysis on person-by-person basis is becoming increasingly important, Inatead of generic wide swaths of the population beimg lumped in one or the other category where maybe they won't get the treatment that is effective for them.

2:12 AM

And SharePoint sucks, too!

2

We're seeing over 2 million logins from our UAT SharePoint farm every day.

5 hours later…

6:47 AM

@PaulWhiteReinstateMonica Watching the game?

7:05 AM

Happy Australia Day long weekend everyone. Go throw a prawn on the barbie!

@MaxVernon seriously? To no database?

@MaxVernon But if I understand correctly what Cade has said, it's "being lumped in one or the other category" correctly (i.e. according to your physiology) that makes it possible for you to get the treatment that's effective for you. My understanding is, sex at birth "affects whether a particular part of the heart or vessel is on the small or large side for a person", which determines the right treatment. But perhaps I misunderstood something

Morning

1 hour later…

8:31 AM

@TomV-trytopanswers.xyz Why buy them if you're not going to read them?

@Johnakahot2use Ditto.

Tom V - try topanswers.xyz

8:51 AM

@FaheemMitha I was partly joking, but I think everybody buys a book they plan to read sometimes and then never get around to actually reading it, even though I've read most of what I bought

3

@TomV-trytopanswers.xyz I have about 20 of them, unread I mean or started and haven't gone past the 1st chapter. More if I count my wife's books.

@FaheemMitha also some books are only references, like lexicons. Nobody actually reads them from start to finish.

I have Knuth's TAOCP for example. I can joke sometimes that I read it during breakfast (who knows, perhaps Bill Gates will hire me ;) but it is more of a reference. And certainly not a start-to-end read.

JEAGL:

> Covers of the third edition of Volume 1 quote Bill Gates as saying, "If you think you're a really good programmer… read (Knuth's) Art of Computer Programming… You should definitely send me a résumé if you can read the whole thing."

Tom V - try topanswers.xyz

@ypercubeᵀᴹ I do have some science fiction I'm really going to read at some point. I just have a tendency to fall in to the Amazon suggestions trap when ordering a book and end up buying 10 at a time

2

The Art of Computer Programming

The Art of Computer Programming (TAOCP) is a comprehensive monograph written by computer scientist Donald Knuth that covers many kinds of programming algorithms and their analysis. Knuth began the project, originally conceived as a single book with twelve chapters, in 1962. The first three volumes of what was then expected to be a seven-volume set were published in 1968, 1969, and 1973. Work began in earnest on Volume 4 in 1973, but was suspended in 1977 for work on typesetting. Writing of the final copy of Volume 4A began in longhand in 2001, and the first online pre-fascicle, 2A, appeared later...

10:02 AM

@TomV-trytopanswers.xyz that's called tsundoku :)

3

bbc.com/news/world-44981013

Tom V - try topanswers.xyz

10:14 AM

Aren't we generous with stars today

Perhaps we are. Or perhaps we are making up for the times when we aren't

@PaulWhiteReinstateMonica That was a quite interesting game, wasn't it?

Peter Vandivier

10:50 AM

so i've just noticed when i kill the postgres service from session A, my psql client in session B drops me into a !> prompt (with a little coaxing) rather than booting me back out into the parent shell

can you actually do anything from this state? or are you just meant to cross your fingers and hope for a reconnect

or just manually ctrl+d ?

John aka hot2use

Morning

2 hours later…

12:23 PM

@PeterVandivier Is ctrl+d the only way you exit psql or does it have its own exit command?

Peter Vandivier

@AndriyM i usually ctrl+d out, but there's also "\q" and "exit"

@PeterVandivier Do those work when psql loses the connection?

Peter Vandivier

🤔 not sure, let me spin up a VM real quick

i assume so though

since you're still, you know, in the psql client

I mean, if they do then there's your answer ;)

Peter Vandivier

i mean... can you do anything productive

12:28 PM

Yeah, I was afraid that was what you meant. Because in that case I don't have an answer, sorry

Peter Vandivier

like... i was expecting to be booted out of the psql client when i killed the service, like a broken pipe ssh message

@AndriyM lol no worries

interesting (not really). \q works but exit fails

postgres=# select 1;
FATAL:  terminating connection due to administrator command
server closed the connection unexpectedly
        This probably means the server terminated abnormally
        before or while processing the request.
The connection to the server was lost. Attempting reset: Failed.
!> exit
->
-> ;
You are currently not connected to a database.
!>

Tom V - try topanswers.xyz

@PeterVandivier can you \c postgres for example?

Peter Vandivier

12:48 PM

The connection to the server was lost. Attempting reset: Failed.
!> \c
All connection parameters must be supplied because no database connection exists
!>

can't seem to figure the format it wants to get further with it

!> \c postgres
All connection parameters must be supplied because no database connection exists
!> \c -D postgres
All connection parameters must be supplied because no database connection exists
!> \c -D /data/base -l /data/log/log -o "--config-file=/data/conf/postgresql.conf"
invalid port number: "/data/log/log"
!> \c -D /data/base 5432 -l /data/log/log -o "--config-file=/data/conf/postgresql.conf"
invalid port number: "-l"
!> \c -D /data/base 5432
All connection parameters must be supplied because no database connection exists

in case you're interested

got it

1 sec

well sort of

!> \c -D /data/base 127.0.0.1 5432
could not connect to server: Connection refused
        Is the server running on host "127.0.0.1" and accepting
        TCP/IP connections on port 5432?
!> \c -D /data/base 127.0.0.1 5432
FATAL:  role "/data/base" does not exist

looks like the syntax to confirm the service is down only works when the service is down and you need something else to confirm the service is up

all told i'm feeling like the answer is "no, you cannot do anything productive from this state"

@PeterVandivier I think you can

just try to\c again

Peter Vandivier

gimme dem deets

FATAL:  role "/data/base" does not exist
!> \c
All connection parameters must be supplied because no database connection exists
!>

Yeah, just tested it

Peter Vandivier

is this worth a question on main?

feels a bit whiffy

!> \c pantelis pantelis 127.0.0.1 5432
Password for user pantelis:
SSL connection (protocol: TLSv1.3, cipher: TLS_AES_256_GCM_SHA384, bits: 256, compression: off)
You are now connected to database "pantelis" as user "pantelis" on host "127.0.0.1" at port "5432".

Peter Vandivier

1:02 PM

yup

\c <db> <role> <ip> <port>

You can do that any time, even if you are connected. It will try to change to new connection (could even be different server). If it fails, you stay in the current connection

Peter Vandivier

yea, but \c always kills the current pipe and re-establishes a connection anyway, right?

even on a healthy connection

@PeterVandivier it says "kept" but not sure

pantelis=# \c x x 127.0.0.1 5433
could not connect to server: Connection refused
	Is the server running on host "127.0.0.1" and accepting
	TCP/IP connections on port 5433?
Previous connection kept

Peter Vandivier

mucho strange-o

I have some 500 tabs open... A youtube video started and took me 30 sec to find out where the hell it was. Luckily I had the headphones on and didn't wake up the whole floor.

Peter Vandivier

1:17 PM

😂

that's too many tabs

It was the video that @TomV-trytopanswers.xyz posted yesterday, with the hysterical laugh !

Tom V - try topanswers.xyz

@ypercubeᵀᴹ haha, that would have been nice if it was playing for the whole office

fireefox does have a speaker icon on the tab that is playing sound, and you can click it to mute the tab without opening it

Peter Vandivier

Five. Hundred. Taaabbbbbbbzzzzzzzzzz.

@TomV-trytopanswers.xyz I think I misclicked on the tab without realizing and then clicked somewhere else and the first click reloaded the page or started the video, not sure. Then I was lost for a while, trying to figure what happened, while a colleague had just stopped by to ask me something ;)

1:34 PM

@ypercubeᵀᴹ well, as in All Things SQL Server™, it depends of course. But I fail to see, for instance, the importance of have a column that differentiates between male or female on my drivers license. It has a picture for identification, my legal name, my date of birth, etc. What benefit does having the gender marker actually confer? None, that I can think of; in fact it probably is only useful to discriminate in various ways against me.

And, as for SQL Server statistics and indexes, a single column that has only two possible values is of minimal use to the query optimizer.

As for what @Cade was talking about earlier in regards to medical databases, I see gender as only a way to broadly categorize people in ways that are probably not scientifically relevant for a statistically significant number of people. I certainly want the best care I can get, and if that means a karyotype marker that distinguishes certain characteristics such as ventricle size, then I'm all for that.

There are a lot of different karyotypes in nature, and they are probably far more important than knowing that I identify as something other than my birth "sex".

Of course, I'm fairly pragmatic, and I "get" that having a gender marker to differentiate between heart attack treatments for women versus men is probably better than just having a single treatment type for everyone, but in my opinion it's problematic in enough cases to make the non-judicious use of a gender marker simply a way to screw up someone's life.

Anyway, if someone has a good reason to store gender in a database, I'm all for hearing it.

Sports databases?

As a for-instance, there is no valid reason for TicketMaster to know my gender.

@ypercubeᵀᴹ you have teams, and you have members of those teams. Why does that need a gender marker?

in the database, I mean

@MaxVernon because events are divided to "only men", "only women", "mixed", "2 male team", "2 women team", "mixed team".

@ypercubeᵀᴹ right but, the people admitting members to the team would presumably be able to distinguish that if possible. Don't get me wrong, I have no problem with the sport being "for men" or "for women". Although that in itself is riddled with assumptions.

What harm is there if the db stores which of the two persons in the 2007 mixed team that won the Wimbledon was the male and who the female?

1:40 PM

@ypercubeᵀᴹ what benefit does it confer, though?

I hate to be so reductionist, but it's in my nature.

it's valuable for marketing youtube.com/watch?v=cjEK7qQKRDY

and also for stirring up shit with statisitcs

I mean if saying "Roger Federer" won isn't enough, I don't know what is.

@MaxVernon ok, you know Roger. Do you know Chris Evans?

@James but marketing can suck it quite frankly. That's the perfect example of allowing someone to put themselves in a category that's only going to limit sales.

Does everyone know who that is?

1:43 PM

@ypercubeᵀᴹ no, the point is why does it matter if Chris is female or male, or some other karyotype? How does that matter?

it only matters to prejudice

Tom V - try topanswers.xyz

@MaxVernon It would be useful to compare results and see if it counts as a record for example

I don't care for karyotype. Health databases might.
But I might care if he was playing in men's or women's tournaments.

Tom V - try topanswers.xyz

You can argue that it shouldn't matter, but as it is now in reality it does, so at some point a database does have to reflect that

or health databases or whatever

As long as the business or the organisation or whatever cares, the dba has to care at least from a technical perspective

I agree with Tom that most businesses don't ask it so they can provide you a better service. They are asking you because everybody asks, and because they can get extra value out of knowing it.

@TomV-trytopanswers.xyz no, I don't think so. I used to see that line of thinking as pertinent, but no longer. I mean, there are non-karyotipical women that compete at the Olympics level, that are not "transgender", they just aren't karytopically female, in that they may be XXY or something. Should the be excluded from competing?

Tom V - try topanswers.xyz

1:45 PM

you can argue the business or the organisation shouldn't care, but that's irrelevant to the discussion

*I agree with max

herpaderp

Tom V - try topanswers.xyz

The point isn't if you or I think it's relevant, what the consumer of the database thinks is relevant

if your customer or employer or whatever wants genders for something, you will have to think about how you design it, period

@MaxVernon that's a different problem though. Tough indeed, but different.

@TomV-trytopanswers.xyz actually, I don't think it is. I think its completely relevant - if no one cares about it, how will perceptions ever change?

@TomV-trytopanswers.xyz of course, but if you're in the position to be designing a database, I think you should think far more carefully than simply adding a boolean field for "male" versus "female".

Tom V - try topanswers.xyz

@MaxVernon You can try to change perceptions as a person, but as long as the client hasn't changed his perception at some point you have to design a database with the concept of gender

1:47 PM

it excludes a bunch of people who may no longer deal with you. and that may cut into profits.

Tom V - try topanswers.xyz

@MaxVernon Ah but that I agree on

@MaxVernon I don't buy that. Really.

Tom V - try topanswers.xyz

But again, it's up to the client to decide if they want male/female/unspecified or male/female/trans/queer/....

@TomV-trytopanswers.xyz that's when I'd add a column that could have a variety of genders. Why not?

I mean if there is exclusion comes from the other things, not from the db design

Tom V - try topanswers.xyz

1:49 PM

@ypercubeᵀᴹ yes

@ypercubeᵀᴹ really? I've personally experienced being excluded like that and it does make a difference to my perception of the company I'm dealing with.

@ypercubeᵀᴹ the database design facilitates those exclusions, though.

@MaxVernon gotta think about hard drive space, though...

@MaxVernon I don't mean that there is no exclusion in the world. And yes, the db design may reflect those. But it is not the cause.

@James yah, because 2 TB drives, even really really expensive ones are only a couple of thousand bucks.

;D

you've also got to consider that adding 72 items to the gender dropdown will also affect people's perception of the company

maybe positively, but maybe not, depending on who you're serving

and the CEO may not like it if you hijack his business for your own agenda

1:51 PM

@ypercubeᵀᴹ of course it's not the cause, but it certainly facilitates exclusion, and as such we should be aware of that, and work to prevent that. That is unless your client is the Nazis, in that case, you won't have much choice.

(unless you are the CEO)

@James hah, I'm certainly not saying you should make those decisions by yourself without consulting stake-holders (fuck I hate that word, btw). But, in many cases, it just takes a simple discussion to win them over to the side of making more profit.

Another example, not about gender/sex. Several questionnaires in the UK ask for ethnic origin. Options are: English, Irish, Other White, Roma, Black, Indian, Other Asian, ... etc (I may not remember correctly the actual options). Should I be offended that I would have to select "Other white"? Is that exclusion?

oh no, I swore. Watch out for incoming mods.

@ypercubeᵀᴹ again though, why is it important? Only for discrimination. Fight me.

:-)

Actually these questionnaires are supposedly to fight exclusion.

1:54 PM

to be clear, I know there are questionnaires that provide under privileged groups with better funding.

But if I have 49.9% of my heritage from what could be considered an "under-privileged" group, where do I fit? Who makes that decision?

Well, the Saxon half of your whitness is oppressed and deserves money, and your norman half is the oppressor and owes money, so it cancels out

3

me? I self-identify on the questionnaire, and who changes it? In other words, if I'm under-privileged, it should be based on things such as my income level.

@James that made me lol in my cubicle, I might add.

=]

What is not lost on me at this point is we're a bunch of privileged white males (ok mostly) talking about others.

@MaxVernon spoken by Norman Max. Now let the other self talk again ;)

2:02 PM

who do you love more your dad, your mom, your two dads, your two moms, none of the above?

@ypercubeᵀᴹ ok. As an example of why caring about gender is exclusionary, check this out: twitter.com/CateSpice/status/1220598778333261824?s=20

@McNets probably, you hate your parents like any sensible person would.

:-)

/sarcasm of course

@MaxVernon this is it!!! ;)

@MaxVernon yeah, there is racism and sexism in most societies.

Tom V - try topanswers.xyz

An example of some database I had to create some time ago

@ypercubeᵀᴹ which I think is honestly why, as caring people, we need to passionately work against that.

2:06 PM

gender is an attribute that slowly but steadily will tend to disappear from all forms

Tom V - try topanswers.xyz

You know those magazines that cater to people over 50 or 60 or whatever? so they want age stored in the database and filter specifically on age. Do I think it's stupid? yes I do. Why the cutoff at some arbitrary number? However they have based their entire business model around finding advertisers who want to advertise in that segmented space, writing content they think is for that specific age group and marketing to those groups

will some people feel excluded? sure, will some people be included but think they shouldn't? sure. do I think it's stupid? absolutely, do I have anything to say as a developer? Not unless I want to walk away from my job

not in my company, of course, they still live in the 19th century

and you've got to know your customers. For example, that gillette add that lost them 9 billion dollars

Age, Sex, Race, Ethnicity - I have to store all for BI for making cohorts. Age is most important. There are certain FDA procedures only approved for people over a certain age because younger people have not had enough clinical trials to deem them safe, while older, more at-risk patients are OK because it has enough improved outcomes to be justified.

sorry, work calls, be back shortly

2:10 PM

Younger people can still have full open heart surgery, while olds get the high-tech!

i guess that means full open heart is the future, right?

You don't want to get a procedure/device and then find out the expectancy means you need to get it replaced several timed before you die. Old people will die before they need a replacement.

Some of these things just haven't had the long history to be able to be studied adequately.

ahh, there's a star trek episode about that

There is a question when you enable this sorting and searching via technology whether it can be misused. That was the root of how the Nazis did it. With the punch cards and the census and then deportation and labor camps and eventually genocide. It was all about what they could fit about people into the IBM 80-columns (they weren't called IBM yet, but it was their subsidiary Dehomag selling the cards and consultants).

I don't have a lot of confidence that our PII and PHI is protected from the power of the legislature to change things to say that the government has the right to suck it all in, when they are already regularly giving rights to private industries over the objections of the public.

@James ah, nice reminder!

Is Picard episode out?

2:24 PM

Must check for Picard on Plex...

I think it's on Amazon prime

I think it's only on CBS All Access. Must bug my Plex guy to get it loaded.

It is out, but I heard you might not like it

George.Palacios

@MaxVernon The problem with inclusivity is that eventually you have to start excluding people, to create an "inclusive" environment.

Star Trek Picard Leaks | My God It's Full Of Suck! (SPOILERS!)

Novalium Company

2:38 PM

What's this chatroom for?

chatting

Novalium Company

Wooooaaah amazing.

about data

and whatever else

Novalium Company

Oks

I'm joining

user image

This is the hole where Picard should be.

George.Palacios

2:40 PM

General on- and off-site discussion for dba.stackexchange.com. Jokes explained at great length (JEAGL) please. We are using SQL

Very important rule that JEAGL

Novalium Company

Alright

Someone interested in AI?

2:54 PM

Define "interested"

George.Palacios

@NovaliumCompany I've written some little "AI's"

John aka hot2use

<tl;dr>

Well, I did actually.

@CadeRoux about data and Data ;)

@ypercubeᵀᴹ I saw it yesterday

Not in the official site...

Novalium Company

@mustaccio "showing curiosity or concern about something or someone; having a feeling of interest." You can use google to check it out. As most people do, I'm using the most common meaning of words.

@mustaccio I understand that you are trying to appear philosophical but I'm here to have conversations that are based on the notion of accepting the most common meaning of words.

3:08 PM

I've written some AI's, too. Do case statements count? I think they should.

I'm currently listening to an update on our machine learning product, it's a long hard process - it's been in development since before I started at this company. To be fair, I don't think the ML part is the hard part.

@NovaliumCompany Why do you need to know if someone is interested in AI?

Novalium Company

@CadeRoux Cool, what's your company?

@mustaccio It's a polite way to start a conversation about a topic I'm interested in.

Ascend Health IT. We specialize in technology surrounding cardiovascular reporting. Where reporting in this context is signed medical reports that a cardiologist prepares. and where the ML is being used to help tag and categorize the images that come in from the devices. When imaging sends that data in, they aren't really categorize or labeled as to what part of the heart they are looking at or from what angle. So the cardiologist has to do that basic sorting and naming.

Tom V - try topanswers.xyz

@NovaliumCompany It wasn't a joke on mustaccio's part but I'll explain at great length anyway. What mustaccio probably meant was that there are different gradations of interest. One can be interested "oh yea, I read most of the headlines when it hits the news" or one can be interested "Oh yes, I wrote 3 gradient descent algoritms".

So Mustaccios reply should be interpreted as "Yes I'm interested in a broad sense, but depending on what your next message will be I will determine if i'm interested in the sub-topic you want to talk about or not"

3:18 PM

My particular side, I have 3 products: analytics - where the structured data in the studies is shredded into a relational DB for analysis, a user usage analysis product to try to help them optimize their processes/knowledgebase and a data mart which makes it easier to use the analytics data in PowerBI/Tableau.

Novalium Company

Oh, alright. Then I'm talking about AI only to people interested in the scale of 1 hamburger to 10 Liberian bananas.

Tom V - try topanswers.xyz

weird

Sounds like an artificial scale to me

George.Palacios

@NovaliumCompany I have done a fair amount of this in a semi-enterprise environment. Used regression to predict estate data growth, created a tool to automatically allocate tasks using a cost-based optimiser

Amongst other bits and bobs

Novalium Company

3:34 PM

Cool

@NovaliumCompany so what is your interest in AI?

George.Palacios

I'm also engaged with trying to create an Enterprise framework for use of AI within the enterprise

Novalium Company

@ypercubeᵀᴹ 10 Liberian bananas.

is that suposed to be funny?

George.Palacios

What a surreal conversation lol

3:36 PM

i didn't think there were any drugs for AI's yet

Novalium Company

@ypercubeᵀᴹ it's supposed to be whatever you find it to be.

George.Palacios

I feel like this isn't really a conversation about AI... What's your agenda?

Novalium Company

I don't particularly understand what this chatroom is about. What topics am I supposed to discuss here?

George.Palacios

@NovaliumCompany on and off-site discussion. IE Pretty much whatever you want

it's primarily for moderation of the DBA forum

3:38 PM

whatever you and others in the room like but yeah, occasionally that ^

Novalium Company

Oks.

George.Palacios

Like, I'm totally happy to talk to you about AI, but you don't actually seem that interested? Is there some subtext I'm missing here or something?

3:51 PM

Some of my very first programs were on playing games.

Life, Othello and Sprouts

Othello was not too bad for its era, it could beat my friends, running on a 386, with 2 or 3 moves look ahead.

@James I just finished ep 1. I like it

cool, I'll give it a watch, I suppose

...when I can figure out how

@James But there are some that make one's intelligence a bit artificial, temporarily

Novalium Company

@George.Palacios Nevermind :)

4:29 PM

Well shit, my Plex guy's dad died and he's out of town.

5:26 PM

@George.Palacios in my experience, those people exclude themselves after a while.

4 hours later…

9:17 PM

If linear regression is AI, does that mean I’m a data scientist when I click the button in Excel that does it?

Applied statistics/mathematics

Every regression you run in Excel brings our civilisation one step closer to the singularity.

The problem with AI is that nobody can define what intelligence is.

9:34 PM

My understanding is that the current "AI" are the same exact neural networks we had in the early 90s but that we can actually just run much larger ones (more neurons) with GPUs and things.

And there are standard frameworks like Tensorflow that people use for it.

@CadeRoux That is true to a large extent. The same can be said about Excel.

2 hours later…

11:11 PM

@MaxVernon ,Ypercube, Tom V, James (no order) thanks for the discussion regarding the gender table/row question, it got me to update myself and even have a laugh, it really brought forward much background which I was not expecting, this gave me good points to think about moving forward. Thanks and Cheers!

« first day (3307 days earlier) ← previous day next day → last day (1547 days later) »

Transcript for

Jan '2024

The Heap™ – Consultancy ©®

General on- and off-site discussion for dba.stackexchange.com....

frogs mysql nosql oracle postgresql sqlite sql-server

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile