Re: PL/pgSQL 2

On 02/09/14 17:03, Hannu Krosing wrote:
> On 09/02/2014 11:52 AM, Álvaro Hernández Tortosa wrote:
>>
>> On 02/09/14 11:44, Pavel Stehule wrote:
>>>
>>>
>>>
>>>
>>>
>>> For 9.4, we have the media already saying "Postgres has
>>> NoSQL capabilities" (which is only partially true). For x.y we
>>> could have the media saying "Postgres adds Oracle compatibility"
>>> (which would be only partially true). But that brings a lot of
>>> users to postgres, and that helps us all.
>>>
>>>
>>> Partial true can enforce so lot of people will hate postgres too.
>>> False promises are wrong
>>
>> Then let's stop talking about postgres being NoSQL. NoSQL is
>> basically "schema-less" (really bad name) plus "infinite scalability"
>> (which basically means transparent sharding). We fail to provide the
>> latter very clearly...
> Have you ever tried any of the "real" NoSQL products version of
> "infinite scalability" ?
Yes, and they are absolutely not infinite, and they suck in many
other places. But they scale beyond one node, transparently, something
that postgres doesn't. And regardless, this is what people is buying, we
like it or not.

>
> We are no worse than most if you use just the unstructured part (which
> is what the NoSQL crowd provides) and something like pl/proxy for scaling.

We are definitely worse. This is the problem, we only look to our
own belly bottom (if this expression exists in English). All NoSQL scale
*easily*, *transparently* beyond one node. Postgres doesn't. I'm not
saying they don't suck at many many other things, or that some of them
may be worse solution than the problem. But despite JSON/JSONB in pg is
awesome, it's far far away from what we need to compete agains NoSQL in
these regards.

Ask anyone not in the postgres world to use pl/proxy for scaling
and they will run away to mongo/whatever. Talk about HA... and the
discussion is over :( I know how hard these problems are in the general,
transactional approach that postgres takes, and that NoSQL does this for
very simple, non-ACID cases, but they do. Hence, we cannot claim NoSQL
"compliance", just because we have jsonb. Unfortunately :( (Surely we do
have many other values, but let's not say that we have NoSQL
capabilities, because we don't while others -better or worse- do).

Regards,

Álvaro