data.hr - All Comments

re: SQL Server 2008 RTM'd

vmihalj — Thu, 07 Aug 2008 12:22:35 GMT

Wait a minute, why are you downloading it whan you're off to seaside...? I know, you'll be testing it there, to see how it behaves in extreme conditions: sun, sea and a bunch of girls in (topless?) swimsuits :-)

re: SQL Server 2008 RTM'd

anonymous — Thu, 07 Aug 2008 08:00:17 GMT

Well, nice to hear that *something" is finally reelased. But, I still do remeber a nasty little "trick" MS pulled when MSSQL2005 was released: database mirroring, so much talked about feature, was disabled by default, and was not officially supported nor recommedned until SP1.

Hope there will be nothing tricky with this one :-)

re: Our Lady of the Lake

anonymous — Tue, 24 Jun 2008 11:46:06 GMT

Hats off, salute :-)

re: Our Lady of the Lake

vmihalj — Tue, 24 Jun 2008 07:45:46 GMT

Hats off, chaps!

re: DLinqy-thingy, Part One

anonymous — Fri, 22 Feb 2008 20:51:33 GMT

That sort of users would abuse/and are abusing all sort of things anyway :-).

re: DLinqy-thingy, Part One

dvitner — Fri, 22 Feb 2008 19:30:49 GMT

Miha, thanks for the comment.

I am aware that the default behaviour of Linq2SQL could be tweaked, and even that stored procedures could be used. However, right now I'm more interested in "out-of-the-box" experience, something I'm afraid is the way this thing will be (mis)used by the vast majority of ordinary users.

And yes, I'm biased :)

re: DLinqy-thingy, Part One

anonymous — Fri, 22 Feb 2008 16:45:21 GMT

Dean, you can modify optimistic concurrency checking. Try reading TFM ;) "Optimistic Concurrency Overview (LINQ to SQL)"

About fetching all fields. Well, this is normal in ORM world because you are working with entities usually, not with their subsets. But you can also delay load (or not load at all <- not supported in LINQ to SQL I think) fields you wish. Or, you can use a projection. But then you can't save the data that easy.

re: I unikifikacija ima granice..

vmihalj — Fri, 22 Feb 2008 15:50:05 GMT

Ovo je bila fora u ispitima za MSSQL 2000 na koje nas je upozorio Damir Bočkal, akoja je bila nepoznata gomili novopečenih sikvelaša :)

Anyway, to su bila vremena kad je Damir još predavao, a MOC-ovi kod njega su bili mjesto na kojima si mogao *naučiti* raditi, plus gomila vlastitih primjera s kojima se bio susreo i koje je koristio kako bi demonstrirao point. Današnji MOC-ov za 2005... Nula u odnosu na stari dobri MOC 2073.

re: Lublana ma živalski vrt

anonymous — Fri, 26 Oct 2007 20:36:24 GMT

jako mi je drago da je bilo super, idući put nas ide više, pa makar sjedili u kafiću pokraj :)

re: Lublana ma živalski vrt

anonymous — Thu, 25 Oct 2007 15:31:51 GMT

Svaka čast, Dean & Dejan, nadamo se redeliveryu u dogledno vrijeme u Zagrebu :)

Potrudit ćemo se da nabavimo "normalan" mikrofon. Možda se povežemo s Ljubljanom, Osijekom i Splitom (i "kafićem pokraj" :) preko Office Communications Servera...

re: Lublana ma živalski vrt

anonymous — Mon, 22 Oct 2007 09:11:16 GMT

Bilo je odlično.

re: Lublana ma živalski vrt

tbronzin — Sat, 20 Oct 2007 19:31:14 GMT

Svaka čast i drago mi je da se suradnja nastavlja i dalje nakon uspješnog CodeCamp-a gdje su Dejan i Miha bili naši dragi gosti, ops, njih već gledamo kao da su domaći :-)

re: Kriza identiteta

dvitner — Thu, 27 Sep 2007 20:47:15 GMT

Naravno da gaps nisu bitni, nisam to ni rekao. Ono što je bitno je ovo:

create table bloggers (

bloggerid int not null identity (1, 1) primary key,

name varchar(30) not null,

email_address varchar(255) not null,

jmbg char(13)

);

insert bloggers (name, email_address, jmbg) values ('Dean Vitner', 'dvitner@gmail.com', '2002965302603');

select * from bloggers;

Ti kažeš da je ovo sasvim OK, i da tu imamo dva različita retka u tablici; ja kažem da si sfušao dizajn.

IDENTITY je štaka. Slažem se da je možda nekad nužno upotrijebiti ga kao PK, ali to bi morala biti dobro promišljena odluka, nikako ne default.

Šteta što se nećemo vidjeti u Varaždinu na bootcampu - tamo ćemo se otprilike pola dana baviti ovakvim nekakvim stvarima. Možda bi bilo jasnije :)

re: Myths&legends

petar.repac — Wed, 26 Sep 2007 03:21:17 GMT

Da, SQL Server ne podržava "deferred constraints"

re: Myths&legends

dvitner — Tue, 25 Sep 2007 21:08:28 GMT

Nazalost, constrainti se provjeravaju za svaki pojedini statement, što je ustvari jedan od problema, IMO:

use tempdb;

create table parent (c1 int primary key, c2 varchar(10) not null);

create table child (c1 int primary key, c2 int not null constraint fk_parent foreign key references parent(c1));

begin tran

insert child values (1, 1);

insert parent values (1, 'qwre');

commit

A inace, slažemo se da baza može biti stvarno konzistentna samo ako su svi constrainti izraženi deklarativno :)