couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jan Lehnardt <>
Subject Re: "Personal" Couch DB.
Date Thu, 25 Jun 2009 07:27:00 GMT
Hi Dale,

I like your points :) A user-friendly database is something that CouchDB
aims to be. But I think we're mixing issues here. Wether CouchDB is
SQLite-style embedded or an HTTP server has zero impact on an
Excel/Futon hybrid that Aunt Velma could plan birthday parties with.

Packaging is key. CouchDBX* on the Mac is already behaves like an
integrated solution. The fact that CouchDB runs as a networked server
is nicely hidden from the user. I don't see why this wouldn't work for
other operating systems, too. In fact, the core** of CouchDBX is a
cross platform standalone CouchDB installation. I'd love to see
Linux & Windows versions of the UI runner that can host any


libcouchdb (a la libsqlite) is gonna be tricky since there is no  
that could use (except on Windows) and I hear it is not a priority at
Ericsson to get that going on Unix systems, too. I'd love to see it,
but it is not happening anytime soon. Plus, the "Personal Couch"
doesn't really care if it runs embedded or as a local web server.

Let's get our families on the couch!


On 25 Jun 2009, at 05:25, Dale Wiles wrote:

> "]" is Wojciech Kaczmarek <>
>    on Wed, 24 Jun 2009 18:03:12 +0200
> ] On Wed, Jun 24, 2009 at 04:17, Dale Wiles ">" wrote:
> ] Replication and distribution is where CouchDB really shines and  
> it's not
> ] scary at all :). It's good especially for business because you can  
> save
> ] lots of money this way, and many people recognize this value, I'd
> ] guess much more
> The advantages of replication and distribution are obvious from a  
> business
> sense.  That's not the target of the mythical "Personal" CouchDB.
> ] than 10% of users "out there" (btw, define "out there" -- who's  
> the target
> I'm glad you asked :0).
> My target is my Dad.  Maybe your parents too.
> He's a "clever" user.  Not a l33t haxor, but if he gets interested  
> he'll
> poke around in a program and make it do interesting things.  I think
> clever users are under served in the computer world.
> Most databases that people use are "casual" databases.  They tend to  
> be
> small, cobbled together and edited by one person.  "Who's turn is it  
> to
> buy bagles?  Let me check my list."  Cobble Cobble Cobble.
> The most popular database out there is either the text file or a  
> *shudder*
> the Microsoft Excel spreadsheet.  Most businesses are up to their  
> eyeballs
> in them, and they all email them to me.  Even tho they're rotten  
> fits for
> the problems, often they're much better fits than the hard core  
> databases
> out there.
> Anyone can open a blank spreadsheet, force a problem into a bunch of
> overly long rows, get no insight into the data, and then save the  
> file.
> The data is sitting right there in front of them.  If they want to
> replicate the data, they drag it onto a USB fob or email it to Google.
> They can then take the spreadsheet to any other computer (because
> *everyone* uses their version of Excel (Oy!)).  If they find another
> Windows box then they have access to the data.
> ] While the http interface is a Good Thing to have these days and it
> ] attracts lots of people, it's not what makes this database  
> different.
> What makes it interesting from my perspective is that the data isn't  
> in
> scary SQL tables, but in a more normal format that my Dad, or even,  
> dare I
> say it, my manager can understand.  However, getting a manager to  
> move from
> a safe, pain in the butt spreadsheet to a server based database is a  
> much
> bigger step than going from a file based spreadsheet to file based DB.
> ] My point is that if you have a good hammer then everything looks  
> like a
> ] nail.
> Misapplied that's also called "Dancing in the Turing tar pit."   
> Everything
> is possible, but nothing is interesting.  (Stolen from Alan Perlis.)
> I'm not really interested in solving an address list for myself, I've
> pretty much got that one beat.  I, personally, play with everything  
> from
> Postgress to a 2 line Dump/Eval Perl database.  It's the upside of  
> being a
> geek.
> What I'm looking for, in this context, is a database that someone  
> who's
> clever could use to do things like plan a birthday party or wedding.
> SQLite is cool for me, but my Dad isn't going to even try to learn  
> SQL and
> database normalizing no matter how useful it is in the long run. This
> facet of CouchDB is a major plus.  It's one of the reasons that I'm
> interested in it.
> ] But maybe for address list and other simple structured data the
> ] easiest way to go is using yaml files or sth?
> If I want a hacky database I just write nested Perl hashes and eval  
> the
> file in.  Works for me, but not really useful in the general sense.
> ] Which subset of features would you like to see in such serverless
> ] product?
> Everything that it already has, with the exception of anything dies  
> in the
> night or moves the data off of that file on my Dad's desk top.   
> Curse you
> zombie robot uprising!
> ] Maybe it would narrow down to something which exists already?
> If you have any favorites that fit the bill, let me know.  I'm a  
> blogger
> ( as you may have gathered from this
> pontificating reply and I'm currently gathering data points for a blog
> post tentatively titled "My Dad Married a Data Base from Open  
> Source."  I
> live for titles like that!
> ] > I think a *lot* of potential casual database users would be  
> interested in
> ] > no hassle/no mystery version of CouchDB they could play with.   
> It's
> ] > something to think about.
> ] IMO CouchDB is no mystery already. Exceptionally easy to configure  
> and
> ] very well documented.
> Your lovable Aunt Gurtrude begs to differ.
> Pop writes a DB.  How does get it off his computer and over Gurtie's?
> Remember, Gurtie is currently chucking rocks in Iran, so no  
> replicating
> across the Internet tubes.  It's going to have to be air dropped in.
> ] Excuse my skeptic tone,
> No problem.
> ] but I'm raising this because you just can't please everyone.
> I agree.  But sometimes you can kill two birds with one stone, and  
> as I
> said before, I think the single author database is a huge, mostly  
> untapped
> sector.
> As for CouchDb, I have no idea if a serverless, file on the desk top  
> based
> version is practical, trivial or absurd.  Why guess when you can ask?
> - Dale

View raw message