Comments on: NoSQL: Why it’s So Damn Sticky

By: ehcache.net

ehcache.net — Sat, 12 Mar 2011 09:20:37 +0000

NoSQL: Why it’s So Damn Sticky…

We wouldn’t care as much about NoSQL if it wasn’t called NoSQL. Even the word itself is triggering existential crises among some very smart, seasoned people….

By: NoSQL Daily – Fri Oct 22 › PHP App Engine

NoSQL Daily – Fri Oct 22 › PHP App Engine — Fri, 22 Oct 2010 01:15:38 +0000

[...] NoSQL: Why it’s So Damn Sticky | Road to Failure: Scalability, Startups, Computer Science, and… [...]

By: araybould

araybould — Mon, 17 May 2010 13:37:04 +0000

We should distinguish between SQL and the relational model of databases. The latter is of considerable value, regardless of whether one is using SQL (and SQL itself is criticized by relational proponents as being a poor implementation of the model.)

We have been here before. In the beginning, RDBMSs had performance problems, and were not used for intensive data processing. Moore’s law took care of that for a while, but then the growth of the internet, together with new, data-hungry techniques for analysis, have brought the wheel full cycle. Moore’s law has an interesting role in this cycle, having made those new techniques possible (or at least practical).

By: Bradford

Bradford — Tue, 13 Apr 2010 20:15:43 +0000

Sasa — it is a bit of an oversimplification, but there’s a kernel of truth. NoSQL is all about being aware of the sacrifices you need to make for scalability up front. The RDBMS is “everything to everyone”, and it’s difficult to scale.

By: Saša Tomislav Mataić

Saša Tomislav Mataić — Mon, 12 Apr 2010 07:41:58 +0000

Could it be stated that SQL is one approach, and the idea behind NoSQL is the other approach to data organization, management and structure. Maybe SQL is “thesis” (being really thought through, engineered and sucessfully used), NoSQL is “antithesis” (pragmatic answer for data management) and we have yet to see emergence of a “synthesis”, an approach which combines these two in one scalable, simple and quick way of handling data?
It seems to me that SQL is a bit over-engineered, extremely smart solution, and NoSQL is a discovered pattern of use which solves 90% of problems which SQL suffers from. But, I may be oversimplifying :)

By: Merv Adrian

Merv Adrian — Fri, 09 Apr 2010 22:57:33 +0000

Wonderfully written and entertaining. Thanks. At its heart, you’ve nailed the problem as being in the difference from the norm. I’d add that it challenges a belief, propagated for 3 decades, that the RDBMS is the universal solution to anything that requires data persistence. That’s always been nonsense. And it gets farther from reality as more data types, more usages, and more volume make that fact more apparent.

By: Bradford

Bradford — Fri, 09 Apr 2010 18:18:09 +0000

Tim Anglade gave a talk recently with the statistic (I believe the $150B included software and services).

By: eric norlin

eric norlin — Fri, 09 Apr 2010 12:06:15 +0000

Bradford –

file under “great minds think alike” –
http://gluecon.ipower.com/blog/?p=201

great post.

By: Jan-Piet Mens » CouchDB (and a bit of Lotus Notes)

Jan-Piet Mens » CouchDB (and a bit of Lotus Notes) — Fri, 09 Apr 2010 12:04:10 +0000

[...] database, a non-relational datastore, one which belongs to the group of NoSQL databases. (More on NoSQL here.) Contrary to a relational database (RDBMS) in which data is contained in tables, rows, and [...]

By: Chad

Chad — Fri, 09 Apr 2010 06:50:44 +0000

Where are you getting $150M for the database market size? IDC says ~10M for database software.