Hacker News new | ask | show | jobs
by beefield 931 days ago
Legacy? Unpopular? Can you elaborate a bit more. To my understanding all the rumors of SQL dying have been a bit premature so far.
3 comments

Oracle DB is really less and less popular. Oracle DBAs are worried. They know they are like COBOL devs.
It may be less popular for new projects, but every single project/product/service that's using Oracle DB is gonna continue to use, and hire, Oracle DBA for the foreseeable future.

Sure, if you're learning SQL or a technology to market yourself, Oracle DB might not be the best time investment. However, if you're already an expert in that, you shouldn't have trouble finding a role somewhere.

Lol my company is paying a fortune to get off oracle. Anyone that does not need the million features that come with a oracle will eventually move off it. It's way too expensive to use in smaller environments or non mainframe environment's. Oracle has some of the worst pricing and sales techniques.
> Oracle DB is really less and less popular.

Doesn't look like it to me.

https://www.macrotrends.net/stocks/charts/ORCL/oracle/revenu...

I think the reference is to Oracle, not SQL in general
I should clarify that she would be able to pick up MySQL or Postgres without too much trouble.
That really depends on how much weight the "oracle" part of her experience has. Oracle has tons upon tons of customized tools, language add-ons/plugins, and quirks. The SQL part of Oracle might translate (more or less) to Postgres easily, but these other things are unlikely to.
Sorry, I was unclear. I mean Oracle specifically.
Oracle may be "uncool" to the young crowd, but it runs a lot of very large, very important systems.

https://www.thomsondata.com/customer-base/oracle.php

That doesn't even include primarily government contract companies.

I work on oracle dbs at a large non tech company. It is a great performant rdbms with poor tooling that I cannot believe they included for the licensing cost. Sqlplus could be great but totally isn’t good as a shell tool. Sqldeveloper is garbage. Python libraries and connection engines are solid for running anything automated. Dbeaver is a cool IDE alternative for clicking through tables.

From my experience across some MSSQL server, oracle, sqlite, and some non-professional postgres, the given database model trumps the underlying database tech when dealing with pains writing analytics reports. There are many options to mess up a model design or for feature extension to turn the model ugly.

Forget about building a new company on an oracle stack, the cost is prohibitive. The open source rdbms are very good and cloud providers have chosen their champions. But for a large legacy corporation with expensive, sensitive data, oracle makes sense. To an extent with a well trained dba the db is “self-documented”

Lol sometimes dbs aren’t exclusively used as a global web app backend for ultra scaling grocery delivery services or whatever. Sometimes db models persist for years and you need assurance that this tech is archival quality.

An aside, I think it’s funny the link you shared is a php site, another dinosaur!

Absolutely it does. The number of new companies launching on Oracle rounds to near zero though.
>Absolutely it does. The number of new companies launching on Oracle rounds to near zero though.

I'm not sure how correct that is either, but it's incorrect to call it a legacy skillset on an unpopular technology. A good number of those new companies, if successful, get acquired and are forced to integrate if not convert to Oracle.

The world is way bigger than SV.