The Heap™ – Consultancy ©®

8:10 AM

Morning

8:39 AM

Morning

9:20 AM

Morning all

9:44 AM

Seen on the interwebs today ...

4

10:52 AM

@JackDouglas what's the best way to forward you some background info? dbfiddle twitter DMs okay?

@PeterVandivier dbfiddle as a messaging platform? That's interesting.

XP

@ConcernedOfTunbridgeWells the tallest blade of grass is the first to get cut

the early bird gets the worm. the early worm gets eaten

What are you trying to say? The tallest mountain gets the most motivated people's corpses?

manifestly so!

11:13 AM

I somehow ended up in a game "scrum poker"

I have no idea what this is and wtf they are doing now, I'm supposed to give points to something without knowing what they represent

brilliant

lol - i actually love scrum points for that. it helped me explain "query bucks" as another thing that only has value in relation to itself

unless there's secretly an exchange rate scrum points : query bucks ...?

11:39 AM

Gamification, that's what DevOps was missing all along. Now I get it.

4

i always terminate the last VM to finish configuring as a lesson to the others

@TomV-TeamMonica It's a game intended to bring peer pressure into the estimation process. You're supposed to come to a consensus on the pointing. The idea is that everybody is now invested in the estimates and is thus motivated to feel personally attacked by the party tasked with implementation argues the toss when it turns out to be a can of worms. If you've ever seen Private Benjamin or Full Metal jacket, you can then punish the team if any of the estimates are wrong.

sounds like you've been in some pretty toxic scrums, eh?

@PeterVandivier Never been on a non-toxic agile project.

System Of A Down - Toxicity (Official Video)

well i was listening to Frank Sinatra, but this warrants a mood change

►

11:53 AM

To do is to be - Focault
To be is to do - Sartre
Do be do be do - Sinatra

2

The first story we needed to score was something for which the architecture isn't decided yet, so I said "40" (whatever that means, it was my first time doing this, and I just pulled up the wikipedia link so I could pretend I knew what I was doing)

@TomV-TeamMonica What colour did the scrum master turn?

Seriously - "Scrum Master?" What year is it - 1996?

@ConcernedOfTunbridgeWells I don't know, is that score extremely high?

presumably the color of segfault, 40 isn't a fibonacci number

@TomV-TeamMonica Poker cards would be 1,2,3,5,8,13,20

11:57 AM

nah... ours go on from there

ours had 1,2,3,5,8,13,20,40 and 100

max you can point any 1 story is 100

ah... "simplified fibonacci" 🤮

Roughly translating to days or half days of effort, depending on how you calibrate the process. Although technically it's not supposed to be a concrete time estimate, that's how it tends to get treated.

@TomV-TeamMonica do you also have the zero, infinity, question mark, and coffee cards?

no

I would have used infinity for the one where we didn't even know the architecture for the application the feature was supposed to be added to

I mean, how can you estimate a "story" when you don't even know the app you should add the story to

11:59 AM

I've seen a theory on this to the effect that you should always go with the highest estimate because the party coming up with that is the most likely to have put some thought into what's involved in the implementation.

that's actually appropriate though... the idea is if you throw infinite or question mark, then that's an indicator that you haven't defined the project well enough to commit to a timeline

But since there wasn't a question mark i just said a really high number

they argued "lets just assume the architecture is going to be x, and re-vote"

bollocks.gif

As long as you get the scrum master to document the assumption.

@TomV-TeamMonica boom - that's your out. that's why you define acceptance criteria

GIVEN x WHEN y THEN z

you're assuming x

if x changes, then you close the ticket since the criteria have changed

and you re-define what you're willing to commit to

Also, there is a server component and a native mobile app, there were stories for the backend and stories for the frontend, why would you allow a backend dev to estimate a frontend change and the other way round?

12:03 PM

@PeterVandivier In other words, learn how to manage your audit trail.

@ConcernedOfTunbridgeWells yuuuuuuuuup.gif

@TomV-TeamMonica Because in Agile, there is no 'back end' and 'front end'. Everybody is supposed to be interchangeable. Good luck with that.

@ConcernedOfTunbridgeWells what could go wrong, sigh

@TomV-TeamMonica I've seen things you wouldn't believe.

Actually, scratch that. You probably would believe them.

attack ships on fire off the shoulder of orion?

12:05 PM

@PeterVandivier Far worse.

c beams glittering in the dark near the Tannhäuser Gate?!

Even worse than that.

One side effect of agile is that I'm more in meetings and calls than actually doing something useful. So the time "saved" by not doing a basic analysis upfront is spent discussing as you go

6

I've seen project management in the financial services industry.

😂

12:07 PM

Cargo cult Agile

When I get out of a call with subcontractors and consultants like that I can't help but think "Well, that's 3K well spent"

Agile has about the same effect on project delivery as any management structure that puts politicians in charge of technical implementation decisions.

3

one of the ways i've seen the time drain addressed in the past is to chargeback meeting time to managers. you take the average hourly salary in IT and bill managers' expense reports per man-hour in meetings they call

Scrum treats developers as irresponsible code monkeys, assumes a 1:1 mapping between requirements and assumes the value of analysis and design is essentially zero because refactoring. It puts all of the decision making in the hands of the scrum master and whoever is responsible for grooming the backlog. There is a sham process where the business assigns priorities and this is presented as a no-cost decision.

lol, i've clearly had much better experiences with it than you have 😆

12:15 PM

@PeterVandivier I guess you probably have; I've never seen an agile project where the process wasn't being manipulated in a profoundly self-serving way by someone in a position of authority.

i've seen shitty management, but i don't blame that on inherent structural failings in agile

Maybe it works if you don't have toxic narcissists running the show, although (a) this applies to most endeavours and (b) there's something about agile dogma that seems to bring them out of the woodwork.

> if you don't have toxic narcissists running the show

☝️ that's the ticket

Right. My mission is now to find an agile project that isn't being run by bellends.

i think it probably exacerbates certain weaknesses of people who already suck at managing, but i've not worked enough places where it wasn't used to have a basis for comparison

12:19 PM

@PeterVandivier One does get around as a contractor, so I've seen quite a few projects now - by my back-of-fag-packet count around 40. Certain patterns do emerge.

I think agile dogma is probably more toxic than the process itself. It's really the dogmatic zealots who won't flex when it's obviously needed that make trouble.

Sometimes you really do need to sit down and spend a month or two - or 6 - on a complex discovery process.

it's not hard to imagine why someone who hired a contractor would be resistant to the truth that a 6 month discovery process is needed 😂

I've also got a thing for decent written specs. The format doesn't necessarily matter, but a written spec document turns a M:M communication graph into 1:M

@PeterVandivier It's more likely to be a function of them having already promised to management that it's not needed because Agile.

Story refinement is essentially writing specifications but I've seen way too many folks come out with knee-jerk resistance to the notion.

well i mean... a story you're given and can't change isn't a "story"

the whole point of the agile manifesto iirc is that it's "self-organizing"

if you're given a task list... that's just not agile on the face of it

We are progressing quickly with our sprints, lets just hope the goal doesn't turn out to be Amsterdam as we are sprinting in the direction of Paris for now

depends what direction you're coming from whether that's a good or bad thing ;)

for example, if you're coming from london, neither of those is a good option

:p

McNets

1:26 PM

Morning

That's a new one

1:57 PM

@PaulWhitesaysGoFundMonica Hey, why are the links not underscored? Custom CSS?

Q: Removing underline style for Stack Exchange links without flicker

0

As part of the rollout of new network site themes, many of the Stack Exchange sites I visit regularly now have links in posts and comments underlined. Preferring the non-underlined look, and since I primarily use Chrome (68.0.3440.106 (Official Build) (64-bit)) and Edge (42.17692.1004.0), which ...

Thanks

2:11 PM

Underlines all over the page drive me bonkers

2:24 PM

Yet for some reason someone thought it would be blinking brilliant

Good thing they didn't decide to bring back blinking

<blink>just checking</blink>

2:37 PM

4

Josh Darnell

LOL

@AndriyM ha ha ha

2:58 PM

<marquee>That's enough of that silly rubbish you lads</marquee>

Mathematics

3:32 PM

Hi, is it possible to generate database diagram of all dbs in a server

3:42 PM

I have to apologize upfront because I'm that crappy situation where I'm not even sure what the right question to ask is, so sorry if I'm blithering. I've got two big, nasty stored procedures, both of which involve dozens of tables. One is blocking the other - I've got a whole host of evidence to show that. So my question is, now what? Is there any way to narrow down which tables are involved here or do I just have to compare the contents of the two procedures manually?

whoisactive.com

Run that while you think you have blocking and it will show what queries are fighting over a resource

(yes, you can write custom queries against sys.dmsomething_lock and join that to other tables and what not or just use the handy dandy procedure, after installing)

@Mathematics SQL Server, Oracle, Informix, DB2, MySQL, Postgres, DBase, Access, something else?

@billinkc this looks helpful, if nothing else it will keep me busy for a while

brentozar.com/archive/2016/07/…

Mathematics

@billinkc SQL :)

@MikeTheLiar It can handy to poll your workload and dump to a table and then go check the entrails from time to time

3:50 PM

@billinkc this is good general career advice. If your workload is onerous, dump to a table, check the entrails. You will likely find your workload to decrease substantially.

I'm guessing what you'll see is a combo of transactions being taken early/held too long and inefficient queries

Had a client last week that stumbled upon the pattern of opening a transaction for the duration of every procedure... and trigger. And those triggers had all this workflow in them, not the least of which was sending email. In said transaction and do you know what, it was slow and there was lots of blocking

4:12 PM

If the transaction rolled back, was the email recalled?

dezso

@ConcernedOfTunbridgeWells discovering how the lock works?

@PeterVandivier actually, I saw this working well. Very small case (in the end, I was the only one actively coding), with a background that made initial and most later assumptions obsolete or at least half-true quite fast. Still, the output was fairly decent (and not only because I am a brilliant Java developer, which I am totally not)

but all participants wanted to work together nicely and achieve a goal

@PaulWhitesaysGoFundMonica They'd never know as they never checked commit status. But, from my own experience with SSIS and distributed transactions, yes, the emails would "send" and then the rollback happens and why am I not getting error emails!

Brillant

2

The words literally went out of my mouth: The email works when the package succeeds. Profiler shows the sendmail waking up for the failures but it's like the transaction's...getting...rolledbackI'manidiot

The Brillant ~Paula Bean~ Paul White

2

4:22 PM

Wow, blast from past there Mr Liar

Those were....days, huh

14 years... ugh

thedailywtf.com/articles/The_Brillant_Paula_Bean

@MikeTheLiar yes exactly

The spelling was quite deliberate

@billinkc Oh no. Ugh.

ypercubeᵀᴹ

5:31 PM

@Mathematics that's not an answer

Well technically it is, just not a very helpful one

ypercubeᵀᴹ

I expected better from someone with the alias Mathematics!

5:54 PM

Universal set

In set theory, a universal set is a set which contains all objects, including itself. In set theory as usually formulated, the conception of a universal set leads to Russell's paradox and is consequently not allowed. However, some non-standard variants of set theory include a universal set. == Notation == There is no standard notation for the universal set of a given set theory. Common symbols include V, U and ξ. == Reasons for nonexistence == Zermelo–Fraenkel set theory and related set theories, which are based on the idea of the cumulative hierarchy, do not allow for the existence of ...

hth

it has been mathematically proved that it is impossible to have all databases diagrammed

Even with that explanation I feel you still need to hope that helps