Hacker News new | ask | show | jobs
by WesBrownSQL 3377 days ago
This is a completely anecdotal analysis. Every time this comes up it makes me cringe. Yes, I use MSSQL. Yes, I really like MSSQL. I know loads more about MSSQL than I do PostgreSQL and that plays a huge part in it. Like this person I am qualified to tell you how totally awesome MSSQL is. Also, like this person I AM NOT qualified to compare it to PostgreSQL even though I do use PG too. This isn't a problem of bias, this is a lack of knowledge on the part of the author. I don't know how to do X in MSSQL so PostgreSQL is clearly better. What needs to happen is experts on both sides sit down and do a true comparative analysis and then the reader can decide which product suits their needs best.
2 comments

Thanks, saved me the bother of saying this as well.

That section on "Reliability" is pure comedy, the only time I had SQL Server "crash" was due to faulty hardware. And despite the best efforts of our previous data centre company (who we've since ditched) dropping the ball and losing power across the site multiple times over two years our MS SQL servers never lost any data, despite the rug being pulled whilst under some fairly heavy workloads. I should add that neither did any of the MySQL fleet, even the ones running replication.

One day when I get time I'll get around disembollocking this flawed article. I don't say that as a MS SQL "fanboy", but as a DBA with coming on for 20 years experience managing and programming SQL Server in banks, blue chips and ISP's (yeah I know, "appeal to authority"-fail, but who the hell is this anonymous author, and what are his credentials?).

Edit: sorry I should add I quite like PostgreSQL, and I'm hoping to roll it out as a service offering to our client base in the next few months, so no axe to grind from me with regards to its features and capabilities.

My experience: SQL Server kicks its NAS off the network due to the NAS's own flakiness, but SQL Server has always recovered fine. I've had to boot it in single-user mode many times to cut the thundering herd down to size on startup/recovery, but then it's back to business.
>>> What needs to happen is experts on both sides sit down and do a true comparative analysis and then the reader can decide which product suits their needs best.

Postgre wins for people who have no money.

SQL Servers wins for people who have money AND are somewhat invested in the Microsoft ecosystem.

By the way the article is from 2014 and states that he's been doing that for a decade.

An article on the capabilities of postgre (or lack thereof) in 2004 and 2009, that'd be fun.