Talk:Document-oriented database

Single quotes and double quotes in one of the examples
Should it be one or the other? (fotoguzzi)69.64.235.42 (talk) 15:53, 8 August 2010 (UTC)

Language in the "XML database implementations" Subsection
The language column in this table needs explanation. Does it refer to the language in which the database was implemented, or does it refer to the query language? Several entries are listed with language 'C++', which is likely an implementation language and very unlikely to be a query language. Others are listed with language 'PL/SQL' which is likely to be a query language and very unlikely to be an implementation language.

Could someone with domain expertise clarify and/or correct this table? — Preceding unsigned comment added by 140.180.26.209 (talk) 15:53, 30 July 2011 (UTC)
 * ✅--Bxj (talk) 02:03, 3 August 2011 (UTC)

Non-notable entries
The table seems to have accumulated a large number of non-notable entries. I propose going through a purging those that don't have a wiki page, either for the database or the company behind it. I'll check, of course, for those which have pages under different names. Stuartyeates (talk) 19:30, 2 December 2011 (UTC)

I agree Eedeebee (talk) 01:30, 27 April 2016 (UTC)

How does indexing work?
Some info on document retrieval, querying and the underlying indexing that supports this would be useful. --Dan Bolser (talk) 07:40, 22 May 2012 (UTC)

Today, Indexing isn't something that is specific to a document database implementation so it's quite challenging to say much more than that they are usually indexed by key. Eedeebee (talk) 01:31, 27 April 2016 (UTC)

Addition of another document database implementation
I have created a Wikipedia page for "AurinkoDB" and added it into the table under "Implementations" section, as AurinkoDB is a major document database implementation in Clojure programming language and also the only document database that uses extensible data notation for storage and network transfer. — Preceding unsigned comment added by Hzguo (talk • contribs) 23:16, 7 December 2012 (UTC)
 * Since both articles you created are currently tagged for deletion on the basis of failing basic notability, I undid your addition to the list for now. If the article(s) are kept then you can add the link back again. § FreeRangeFrog 20:01, 8 December 2012 (UTC)

Seems this section should be removed from Talk Eedeebee (talk) 01:33, 27 April 2016 (UTC)

Add Cloudant to implementations?
I work for Cloudant and have a COI.

What do people think about adding http://en.wikipedia.org/wiki/Cloudant to this table of document-oriented DB implementations? Cloudant is built on our open-source fork of CouchDB, http://en.wikipedia.org/wiki/BigCouch. Many of our developers are contributing code back to the Apache CouchDB project. Cloudant is proprietary, has a RESTful API, and uses Erlang for its core along with a mix of Java, Scalang, and C. So I think it is sufficiently different, but some feedback would be great. Thanks. Mbroberg (talk) 15:37, 11 March 2013 (UTC)


 * Cloudant is in the table now. 87.195.184.143 (talk) 07:52, 22 January 2016 (UTC)

Now that is is, can we just junk this section in talk? Eedeebee (talk) 01:32, 27 April 2016 (UTC)

Remove Apache cassandra from implementation
Apache Cassandra is not document oriented database, Cassandra is Wide Column store or column families. as Eric Evans's said here. Maybe, JSON data provided by Cassandra are made of Thrift. Septianw (talk) 10:44, 12 April 2013 (UTC)

Hosted document-oriented databases
I recently added a section to the NoSQL page called NoSQL databases on the cloud, which describes ways of deploying NoSQL databases on the cloud (virtual machine, database as a service, and native cloud databases), and provides a few examples of notable implementations. As it happens most of the examples I provided there are of document-oriented DBs - namely Mongo, Cassandra and Redis. I propose adapting the section and adding it here as well. Any thoughts? Anne.naimoli (talk) 13:17, 5 January 2014 (UTC)

The Section titled Documents should be rewritten
This section was earlier just a brief summary of structured documents. Now it rambles on about non-document databases in a way that is unhelpful and confusing. Document databases store documents structured in some form (binary, text, json, pdf, xml, whatever). Enough said. — Preceding unsigned comment added by Eedeebee (talk • contribs) 04:40, 4 February 2015 (UTC)

Adding Crate
I would like to add CrateIO to the listings, I do work for them but I feel we fit into the current selection and are an open source project. Let me know if anyone has any objections or I will go ahead. ChrisChinchillaWard (talk) 07:29, 12 June 2015 (UTC)

Perfect article.
Fantastic explanation.

This must be in the foremost example of how technical wiki articles must be written. 164.100.31.82 (talk) 04:25, 8 September 2015 (UTC)

NPOV / expert needed
I have added these tags to the article. Aside from the existing issue of relying too much on primary sources, this article reads like it was written exclusively by fans of document-oriented databases. The content constantly harps on the (unsourced) advantages of such systems while not even touching on criticisms or disadvantages. --Chris (talk) 16:14, 9 November 2015 (UTC)


 * I think your assessment is too harsh. As a longtime user of relational databases who didn't know much about NoSQL databases, this article was very helpful to me.  It is exceptionally well written and the explanations are clear.  I came to it from the NoSQL page, https://en.wikipedia.org/wiki/NoSQL, which does discuss the merits of various forms of NoSQL databases, but which does not get into enough detail so that a reader can understand the architectural differences in detail.  I'm not disagreeing that this article could use more information on criticisms and disadvantages of document databases, but it should not be just labeled as a stub either --it's considerably better than that.Saddledbyclouds (talk) 20:17, 7 December 2015 (UTC)

I don't think it's required to list comparisons to other technologies in any more detail than we have so far. Yes, it can be improved, but I agree with the second comment here as of this date Eedeebee (talk) 03:26, 27 April 2016 (UTC) Also, what criteria would be needed for expertise? I have academic credentials as well as real world experience. Eedeebee (talk) 04:42, 27 April 2016 (UTC)

How is the second example on the page XML
The article text states that "A second document might be encoded in XML as:" and then proceeds with something that is clearly not an XML but another JSON document. Was this intentional? — Preceding unsigned comment added by 78.130.131.170 (talk) 07:14, 27 April 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 3 one external links on Document-oriented database. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20111020074113/http://couchdb.apache.org:80/docs/overview.html to http://couchdb.apache.org/docs/overview.html
 * Added archive https://web.archive.org/web/20150622174526/https://crate.io/docs/stable/sql/rest.html to https://crate.io/docs/stable/sql/rest.html
 * Added archive https://web.archive.org/web/20080327222152/http://blog.labnotes.org:80/2007/09/20/read-consistency-dumb-databases-smart-services/ to http://blog.labnotes.org/2007/09/20/read-consistency-dumb-databases-smart-services/

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at ).

Cheers.— InternetArchiveBot  (Report bug) 14:47, 14 December 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 2 external links on Document-oriented database. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20120820182153/http://www.couchbase.com/docs/ to http://www.couchbase.com/docs/
 * Corrected formatting/usage for http://blog.labnotes.org/2007/09/20/read-consistency-dumb-databases-smart-services/

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 23:25, 11 September 2017 (UTC)

Rocket Software
Rocket's offerings Universe and Unidata are not document-oriented databases. Although they are key-value stores, they do not implement metadata stored within the structure in the way this article describes.Wjhonson (talk) 22:27, 2 March 2018 (UTC)

Compare with native files
Document Databases provide functionality that native named files on disk don't.  But if that extra functionality is not needed, eg the files are images for download into browsers, I imagine it is more efficient (faster) to access named files on disk. I think this option should be added to the comparison section of the article. After all, don't we all have files in folders on our computers rather than us inserting them into Document Databases? Yes, it is a genuine alternative. FreeFlow99 (talk) 10:29, 21 November 2021 (UTC)

Example of query wanted
If I understand the concept correctly, a key point of document databases is that searches done by the database engine can parse the documents to extract some field(s) therein, to e.g. answer the query "give me the addresses and phone numbers of all contacts named Joe"; if the document was just some complex value in a key–value store, then in principle one could achieve such an end by applying some fancy functions to extract components from the document to use in the filter condition, but it probably wouldn't look very nice. To clarify this point, an example of a query could be useful. (Possibly the lack of a standard query language makes this a bit tricky.)

Also, this suggests a different take on the "How does indexing work" question, above. One approach for how indices could work is that they are tables over what you would get if you extracted some specific component from each document in the database, to be used in speeding up searches. The question for the specialists is whether there are other kinds of indices than this. 195.67.63.2 (talk) 11:35, 9 May 2023 (UTC)