The Heap™ – Consultancy ©®

« first day (2971 days earlier) ← previous day next day → last day (1896 days later) »

McNets

7:18 AM

Morning

hot2use

7:52 AM

Morning

Andriy M

8:17 AM

Morning

dezso

9:02 AM

@TomV the default destination is stderr, which is sorta easy to lose

postgresql.org/docs/11/… for an overview, or ask a specific question if you feel so

1 hour later…

Peter Vandivier

10:21 AM

Morning

2 hours later…

Philᵀᴹ

12:10 PM

@dezso 2>&1 is your friend

ConcernedOfTunbridgeWells

@MDCCL Elmasri and Navathe strike again.

A: Do candidate key concept exist only in theory?

I'll go out on a limb and guess you've done a course based on Elmasri and Navathe's textbook as they use this terminology. Many if not most questions like this seem to originate from students trying to make sense of this textbook. In data modelling, a candidate key is a possible key that you co...

dezso

@Philᵀᴹ no, my friend is /dev/null

ConcernedOfTunbridgeWells

12:35 PM

That textbook seems to have sown more confusion amongst students of database papers than any other thing I can think of. Maybe I should write a database text to save the I.T. industry from it.

Tom V

@dezso Thanks I'll have a look

ypercubeᵀᴹ

@ConcernedOfTunbridgeWells that would be great.

It's too big from what I remember. Now that I think of it, all database textbooks are too big.

ConcernedOfTunbridgeWells

At least.

Elmasri and Navathe - proudly sowing chaos in database papers for 20 years.

It would be an interesting challenge - write a database textbook suitable for 200-300 level papers and keep it to a reasonable page count.

I suppose 200 level would really be data modelling, SQL and some theory like transactions and relational algebra.

300 level could get into query optimisation (i.e. some view of how optimisers work), 'advanced' sql topics and physcial storage.

Some basic stuff like what a BTree is might go into the 200 level stuff.

Not sure whether relational synthesis is all that useful. E&N seem to love it, but I've never seen it used on an actual project. Maybe it was originally used in profiling data from flat file systems.

Maybe you could add more internals like how log files work into the 300 level stuff.

Maybe I would have to design some hypothetical 200 and 300 level papers and think of the topics to add to them.

Is 'principles of database administration' a sensible topic for a 300 level paper?

You could do a whole 'internals' paper picking PostgreSQL apart.

ypercubeᵀᴹ

12:58 PM

@ConcernedOfTunbridgeWells are you in the UK these days? We should organize something

ConcernedOfTunbridgeWells

@ypercubeᵀᴹ Yes. If that's a hint about drinkies It's on my hit list but I've had a nasty trapped nerve in my shoulder since december.

ypercubeᵀᴹ

just kidding. Yeah, it's a long time since we had any meeting.

dezso

1:22 PM

@ConcernedOfTunbridgeWells I hope it's only one shoulder, so that you still have a hand to lift a pint

but with children that should be more than inconvenient

ConcernedOfTunbridgeWells

1:45 PM

@dezso Fortunately it's my off hand.

1 hour later…

MDCCL

2:47 PM

@ConcernedOfTunbridgeWells My favourite text on data modelling is IDEF1X, it has about 150 pages if I recall correctly, but is rather clear IMO.

1 hour later…

Max Vernon

4:05 PM

@paj28 - that may be, however when you use an ORM to write SQL for you, you're making it almost impossible to troubleshoot performance later on, when it does matter. ORMs are a shortcut that have a tendency to shoot you in the foot when you least expect it. If you're at all concerned about writing great code, you write the SQL yourself. — Max Vernon 32 secs ago

2 hours later…

ConcernedOfTunbridgeWells

5:59 PM

Agile is so passive-aggressive.