Falsetti: Had to break up your imaginary Risse at the right points

Botwin: Http://sqlfiddle.com/#!9/58e60

Adamson: That gives me “Table ‘db_9_58e60.infos’ doesn’t exist”

Swaim: Looks like a nice site, but its damn slow

Swanton: Worked for me. what do you want the query to return and why?

Mena: One moment, im getting it to work .

Blasing: Click build schema before run sql.

Schor: Http://sqlfiddle.com/#!9/85cfd/2

Krysiak: Every product has its sort defined in “infos”

Mcguirt: One moment, i still have it wrong.

Cafaro: Here we go: http://sqlfiddle.com/#!9/3b117/1

Handly: Water should be at position 2

Gartenhaus: Because there is no info for “water in drinks” so it should use the general position for water.

Chagnon: Not sure if there is an elegant way in sql to accomplish that. what are your thoughts?

Allgeier: So where do you want beer, the only one without and info?

Kalfa: Well, those could have a hardcoded sorting like 10 or something

Granstrom: ORDER BY coalesceI.sort,10

Willcox: Well, the categories have to be taken into account. let me make a better example.

Pusser: That tool would be awesome if it wasnt so slow.

Barut: Strangely, this magically gets it right:

Rayas: Http://sqlfiddle.com/#!9/34e0d/1

Guilfoos: Trying to build a version that shows the problem.

Montroy: Looks like it thinks NULL=anything but *prefers* a real match. is that do***ented behaviour?

Matz: Http://sqlfiddle.com/#!9/7207a/1

Kess: Water should be sorted at position 2

Subia: Is that possible in sql?

Lysher: The sort order of “water” should be 2. so it comes after soda.

Nowakowski: Wouldn’t cola be considered soda?

Bertoldo: Lets implement strong AI after we got the basic sorting right.

Deangelus: The only way I can think of to do it would be to ***ign weights or something where you could specify an order like you want to. I’m far from an SQL guru though so perhaps others have a better option.

Villaluazo: Or maybe position would be a better field name

Nevue: We need an sql guru here

Sitzman: I can use SELECT SUMf1 FROM t1 to get the sum of f1. but how do i get the sum of f1 and f2 from t1?

Weisenfluh: No_gravity: we need people asking questions to take less than 45mins to explain themself

Correl: Aias: SELECY SUMf1, SUMf2 FROM t1;

Correl: Wrksx: Be careful with that :

Correl: Wrksx: SUMf1+f2 can differ from SUMf1+SUMf2

Correl: Wrksx: Think about it and you will see why

Sitzler: I thought he needed the result for each row

Jaster: Therfore thought he didn’t need sum at all

Goethals: I need one result for all rows

Fulginiti: Aias, so salle answer is correct for ya as usual

Correl: Wrksx: Here we go: http://pastie.org/10428532

Prevette: Danblack: i think i have the sqlfiddle right now. this is the ouptut i am looking for: http://pastie.org/pastes/10428529/text

Correl: Aias: It will help if you explain what exactly you want to select

Kuhner: Salle, can’t see what the f is happening there

Exe: Seems so broken to me

Correl: Wrksx: Think : Hint: SUM is aggregate function

Correl: Wrksx: No this is correct behaviour

Hennard: Guessed it is, else you wouldn’t warn me that way

Correl: No_gravity: We have plenty of sql gurus here, but they are usually busy

Massar: Gotta love pastie’s legal page: “if anything bad happens or the stuff hits the fan, it’s your fault, not ours”