Hacker News new | past | comments | ask | show | jobs | submit login

That looks too be exactly what I have been missing. I have yet to see details like that not spread out over 5 pages. Thank you!



Yes, the docs could use indeed a few lighter tutorials for people starting out.

However, once you're over that initial learning hump and start looking for more specific things you'll quickly notice why the postgres manual is often cited as being one of the best documentations ever written (inside and outside the OSS world). It's really that good - once you're familiar with a few basics.


How can I take this seriously.

However, once you're over that initial learning hump...you'll quickly notice why the postgres manual is often cited as being one of the best documentations ever written

"It's really that good - once you're familiar with a few basics."


By realizing that sometimes it takes a little investment for a payoff.

Outside the official docs you can find the same number of newbie tutorials for postgres that you'll find for MySQL. The difference is: Postgres has a stellar documentation once you're leaving the newbie stage. What MySQL calls documentation, let's just not talk about it...


EDIT after a downvote.

I'm sorry if in this comment I spoke out of frustration, but I do stand by the thoughts about what a gold standard of a manual has to at a minimum provide. We're not talking about a "little investment". Read his original comment again.

if someone has enough experience with sql and enough mysql experience that they're really hitting limitations in mysql, then a merely "good" or even "ok" manual should at a minimum be able to enable that person's good-faith, dedicated attempt to start using this alternative to succeed. This is a pure horizontal transition. Postgresql is not only "also a database" but "also sql" and "a direct horizontal competitor to mysql" -- the technologies substitute each other other, often on a drop-in basis (in LAMP-type and modern stacks, especially with a framework that uses an ORM mapper that can use either). If someone is familiar with SQL and familiar with an alternative, the manual should at a minimum help the person transition to using this alternative. If nothing else. I would say this is the absolute minimum that a manual would have to provide to meet its basic mission.

At an absolute minimum, a manual MUST let you move horizontally from a drop-in competitor you have some competency with.

(I feel like, even historically, this has been missed by a lot of people.)

This is why I took umbrage with your attitude "why the postgres manual is often cited as being one of the best documentations ever written" after saying: "Yes, the docs could use indeed a few lighter tutorials for people starting out" in response to the guy who simply could not get past the installation hurdle.

This is like a stereo that someone can't turn on. They've used other stereos, but they just can't figure it out. They read the manual, but still can't figure it out. I would argue that, unless it's a real "duh, I was such an idiot" moment (happens to everyone) and isolated case, and it sounds like this isn't what you're saying, then in such a case the manual failed to do its primary mission. If the guy could have turned it on, he would have figured everything else out by himself.

If this guy can't do what he's trying to do after knowing mysql the way he's talking about and having the experience with attempting postgresql the way he talks about, I'd say that while the manual might be a fine technical reference, it is not yet complete and must not be held up as a great manual until this hole is filled.


It is the best documentation I think I ever have used, but is still has flaws. The greatest one being what moe stated.


that's fair enough. But let's be clear. Just because I always skip the first chapter of my stereo's documentation ("turning on your stereo") since it's so obvious how to connect things and turn it on, all the ports are very clearly labeled, etc, that does not mean that any stereo manual can be considered good if it does not include that chapter :)

I think I'm basically not winning any friends with this line of thinking, but I just see this as a long, recurring problem. It would be the same with a certain open source movement's long-standing alternative to a certain very well entranched operating system: they are horizontal substitutes for each other, and yet people who are expert on the entrenched system (power users, even) have given up on the open source alternative, because they "could not turn it on". (Get to the same place they get to after a fresh install of their entrenched operating system).

I'm not going to say more because these are (whatever those-things-that-are-landmines-floating-in-water are called)-filled waters.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: