« first day (4154 days earlier)      last day (37 days later) » 

5:20 AM
A chairde - Morning all!
 
 
1 hour later…
6:20 AM
@McNets Similar situation as the person who ordered his bat medium-rare
mOnring
 
 
4 hours later…
10:39 AM
@Vérace-СлаваУкраїні I can't answer my own posts - but just to let you all know that the word "monkies" kind of "grated" with me, so I checked. The correct plural of "monkey" in English is "monkeys" (and the plural of donkey is donkeys). So, these two animal nouns break the normal rule in English that the plural of a word ending in "y" is "ies" - body, bodies - duty, duties, - industry, industries. My apologies for misleading any non-English speakers in the group.
I shall be wearing my chain cilice for the next week as penance for this egregious error on my part!
 
11:35 AM
So, today I learned the programme manager is the Head of Change's girlfriend. Nepotism at its best, folks.
Needless to say, the project is a clusterfuck.
Business-led IT my arse.
 
12:17 PM
@Vérace-СлаваУкраїні the normal rule is "y" becomes "ies" in nouns ending in consonant+y. So no rule is broken
day - days
boy - boys
key - keys
Maybe you were thinking of "monies" which does break the rule ;)
 
σας ευχαριστώ - it's frequently non-(native)-language speakers who know the rules better than "from birth" speakers. No, I wasn't thinking of "monies" because I don't have any! :-)
 
de nada senior ;)
I have a question for you, did you grow up learning 2 languages from birth?
 
For a fun read, take a look here. Sometimes, errors in some of these really grate with me!
@ypercubeᵀᴹ No - I would have started to learn Irish from an early age, and then I learnt French at school, but learnt to speak it properly (to the point that I could spend hours with native speakers who couldn't tell I was a foreigner) from the age of 21.
English being my from-birth language.
 
12:51 PM
@Vérace-СлаваУкраїні not too many to be honest.
it says 100 but looks like about 60 or more fall into patterns (-sis -> -ses for Greek origin words, -us -> -i, -um -> -a for Latin origin words, etc)
 
@ypercubeᵀᴹ But one doesn't say stadia, but rather stadiums. "Stadia" would be considered extremely pedantic, and even wrong. I agree that there is a regularity in the irregularity - fractal English perhaps (as opposed to fractured English! :-) )
 
@Vérace-СлаваУкраїні well, you just said it ;)
 
1:53 PM
Hi. I wonder what is the best way to make DB backups.

Let's assume I have backups once per day and also I have DB mirror.

If I accidentally delete production DB, then I also delete mirror DB, so I can't recover from mirror. I can recover from this once per day backup, but it will be outdated.

I wonder if there is a way to have backups so it's possible to recover exactly to the point before DB deletion (so no data loss will occur). Can you point me where should I look?
 
@KarolZlot Transaction log backups. This is the Way.
It won't take you to the exact point before deletion of the database, but you shouldn't be allowing anyone that privilege on a day-to-day basis anyway.
 
2:19 PM
The details may differ from DBMS to DBMS but the idea is the same. Apply the latest full backup and then on top of it the transaction log backups (from the date that backup was taken till the disaster).
 
@HannahVernon Sounds good. I'm reading about this but I can't find why it can't restore to exact point before deletion of the database. What do you mean?
 
2:54 PM
@KarolZlot Probably meant the recovery point is limited to the frequency of log backups, without special effort. If you take a log backup every 5 minutes, you would lose up to 5 minutes' changes by applying all logs except the one where the damage occurred.
Of course, with some effort, and assuming the RDBMS in question supports point-in-time restore from logs, you could indeed get right back to the moment before disaster.
 
@PaulWhite Ok this is what I needed. Thank you. I see that Postgres can do it, so I will try to learn details.
 
No worries
 
3:27 PM
So apparently monkey (monkie) pox can be transmitted through respiratory droplets and contaminated materials such as bedding (WHO).
 
@PaulWhite I suppose one of the tricky parts would be to find which point that is.
 
@ypercubeᵀᴹ Yes. Sometimes you're lucky enough to know the approximate time, or maybe even used transaction markers. In SQL Server, I seem to recall restoring logs WITH STANDBY, which makes it possible to undo applying a log.
It's a long time since I've had to do this in practice though.
20 monkie cases in UK now, and the first one in Australia
Personally I think the pox is being spread by NFTs
 
4:05 PM
@PaulWhite yeah, hopefully NFTs and this monkey thing die a sudden death
 
so bored with apes
3
 
lol
@PaulWhite there are quite some layers there
 
4:20 PM
worked hard on that one
 
 
4 hours later…
8:46 PM
@ypercubeᵀᴹ - can you give me a dig out here? I'm trying to solve this in PostgreSQL, but I've arrived here but am stumped as to why I'm getting the ERROR: syntax error at or near "WITH" LINE 18: UNNEST(STRING_TO_ARRAY(x.sp, ' ')) WITH ORDINALITY a(elem,... error - any ideas?
 
There was a line break, which broke the markdown formatting. Took me a while to find it.
 
Yuor wleocme
 
@Vérace-СлаваУкраїні dbfiddle.uk/…
 
9:04 PM
@ypercubeᵀᴹ - excellent stuff - thanks for that. You've sort of crystallised for me the difference between the LEFT JOIN LATERAL and the implicit CROSS JOIN LATERAL as explained by EB here. I think that further study tomorrow (am knackered and going to bed - up at 06:00 this morning!) will prove profitable! Oíche mhaith a chara!
 
@Vérace-СлаваУкраїні enjoy!
the problem was that WITH ORDINALITY the unnest() produces two columns, so you can't use it in the select clause. WIthout it, you could.
 

« first day (4154 days earlier)      last day (37 days later) »