Which database system is the most in the spirit of FreeBSD?

[dbdemon]

Is there a database system which is created, developed and/or designed more "in the spirit" of FreeBSD?

Obviously, Berkley DB was originally created and developed as part of BSD, so is probably a strong contender.

If the question was about Unix rather than FreeBSD specifically, would the answer be different?

 

[rootbert]

PostgreSQL: similar licence and a strong focus on clean design

 

[msplsh]

Postgres

 

[covacat]

here it is the most bsd-ish db first distribution, including documentation and source code

whereis -q awk sh grep comm paste sort join|tr " " "\n"|tar --files-from - -cvf freebsd-db-0.0.0.0.1.tar

 

[eternal_noob]

PostgreSQL is nice once you get used to it. But at first, you have to memorize the weird replacements for MySQL commands, e.g.

SHOW TABLES;

becomes

\dt

and so on.

 

[covacat]

PostgreSQL is nice once you get used to it. But at first, you have to memorize the weird replacements for MySQL commands, e.g.

SHOW TABLES;

becomes

\dt

and so on.

that's why you use flat files and grep
you can't forget ls ...

 

[eternal_noob]

Yeah, you definitely need a cheat sheet for the first few weeks.

 

[mer]

OP, please define your "needs" in a database.
Do you need SQL ability?
Or do you need other capabilities?

flat text files and grep/sed/awk work pretty well for a large chunk of things.
But SQL works for different things.

 

[dbdemon]

OP, please define your "needs" in a database.
Do you need SQL ability?
Or do you need other capabilities?

flat text files and grep/sed/awk work pretty well for a large chunk of things.
But SQL works for different things.

grep/sed/awk are cute, but in my view not sufficient for a serious database system.

Personally, I like SQL, but if there is a serious NoSQL solution, then it should be considered. Berkeley DB was originally "NoSQL", but Oracle has since added an SQL interface, if I understand correctly.

 

[mer]

grep/sed/awk are cute, but in my view not sufficient for a serious database system.

But you need to define your specific needs. flat files with grep/sed/awk is more than sufficient for some use cases, but not all. For some use cases, full SQL capabilities are overkill. So for some use cases MySQL is fine, others Berkley DB is fine, others you need more.

 

[hruodr]

I have some plain text dbs that look up with grep and scripts,
that works faster than many db engines.
If I want SQL, I try first sqlite3. It is FreeBSD spirit, because pkg uses it.
For client server I used MySQL and tried PostgreSQL, I like the second more.

 

[dbdemon]

here it is the most bsd-ish db first distribution, including documentation and source code

whereis -q awk sh grep comm paste sort join|tr " " "\n"|tar --files-from - -cvf freebsd-db-0.0.0.0.1.tar

flat text files and grep/sed/awk work pretty well for a large chunk of things.

On this topic, I came across this article the other day:

There's a Relational Database in Your Unix CLI

With the unix join command, you can process relational data at the command line. Here's a brief introduction on how to use it.

spin.atomicobject.com

I thought that was novel. But not nearly sufficient for what I want in a database system.

 

[mer]

I thought that was novel. But not nearly sufficient for what I want in a database system.

the second sentence here drives what people will say. I've gone back through this whole thread a couple times and I can't find where you say what you want in a database system.

lay out your requirements and then folks can have a better answer to your question.

 

[dbdemon]

the second sentence here drives what people will say. I've gone back through this whole thread a couple times and I can't find where you say what you want in a database system.

lay out your requirements and then folks can have a better answer to your question.

I simply wish to learn more about which database systems knowledgable FreeBSD users/admins/developers consider to be "in the spirit" of FreeBSD and Unix. I don't have requirements or needs as such.

I have noted that some of you think sed/grep/awk/sort/join combined with flat text files are one possible option. Thank you.

 

[astyle]

To answer the question in the Opening Post, I'd say Berkeley DB - it's what sendmail uses, and it's the original. Yeah, it's not a proper 'relational' database, but it is part of the FreeBSD base.
--
The article OP references in post #12 is most likely about this utility: join(1). A quick read through the manpage reveals that it's really meant to work on a text file. I would not treat it as an alternative to SQL proper. And /usr/bin/join is not going to work on BDB or any RDBMS - just totally separate and different things.

 

[Alain De Vos]

that's why you use flat files and grep
you can't forget ls ...

csv
PS: kyotocabinet & tokyocabinet are interesting.

 

[covacat]

PS: kyotocabinet & tokyocabinet are interesting.

are those namco/sega/konami/capcom arcade machines ?

 

[covacat]

for quick & dirty hacking, embedding sqlite3 > *
client / server postgres is more bsd-ish than mysql

 

[ralphbsz]

more "in the spirit" of FreeBSD?

What do you think is the spirit of FreeBSD?

There are lots of coffee places on Shattuck Avenue. I bet lots of database code was fueled by espressos served there. As an example, the origin of Berkeley RISC project (later commercialized as the SPARC) was that John Cocke (the inventor of the first RISC machine) used to go drink coffee with Dave Patterson every few weeks.

Or maybe the spirit of BSD can be found in Kirk McKusick's wine cellar (which has been famously computer monitored for decades, same with his hot tub).

OK, now serious: Instead of jokes about what people in Berkeley drink, please define what you think is that elusive "spirit", and then maybe we can answer it.

 

[eternal_noob]

Kirk McKusick ... computer monitored ... hot tub

I never thought bad about Kirk but this is weird. (If it's true).

 

[blind0ne]

here it is the most bsd-ish db first distribution, including documentation and source code

whereis -q awk sh grep comm paste sort join|tr " " "\n"|tar --files-from - -cvf freebsd-db-0.0.0.0.1.tar

tar --files-from - -cvf freebsd-db-0.0.0.0.1.tar
what is this --files-from - ?
The set of programs is quite good advice, but it would be much more pleasant to find out about common practices and set of rules to follow in order to achieve the similar functionality and throw off the shackles of databases at least for the thoughts time.

 

[hardworkingnewbie]

INGRES of course, since it was developed back in the old days at Berkeley.

 

[covacat]

tar --files-from - -cvf freebsd-db-0.0.0.0.1.tar
what is this --files-from - ?
The set of programs is quite good advice, but it would be much more pleasant to find out about common practices and set of rules to follow in order to achieve the similar functionality and throw off the shackles of databases at least for the thoughts time.

read file names to archive from stdin

 

[dbdemon]

OK, now serious: Instead of jokes about what people in Berkeley drink, please define what you think is that elusive "spirit", and then maybe we can answer it.

I suppose the design philosophy, licencing, development model, and to a lesser degree also origins and history on BSD systems or other Unix systems. And so on ...

But I'm happy to hear other interpretations of "spirit of FreeBSD / Unix".

 

[blind0ne]

the second sentence here drives what people will say. I've gone back through this whole thread a couple times and I can't find where you say what you want in a database system.

lay out your requirements and then folks can have a better answer to your question.

Are there any working in "production" websites, services on the web that is using such approach as here is mentioned in?

On this topic, I came across this article the other day:

There's a Relational Database in Your Unix CLI

With the unix join command, you can process relational data at the command line. Here's a brief introduction on how to use it.

spin.atomicobject.com

I thought that was novel. But not nearly sufficient for what I want in a database system.