couchdb-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Anthony Waddell" <>
Subject Re: Solution on a Memory Stick?
Date Thu, 03 Jul 2008 07:25:03 GMT
Thanks for the response

I guess I can handle security of the database by putting some kind of
lock on on the memory stick, or the entire app, but what about
replication.  I understand it now replicates on port 80:
- is it robust?
- Any ideas how could I encrypt the transmission over public access networks?


2008/7/2 Jan Lehnardt <>:
> Hi Anthony,
> On Jul 2, 2008, at 10:53, Anthony Waddell wrote:
>> Greetings
>> I was about to start developing an app that is to be used in various
>> remote rural areas with little/no internet connectivity. The data
>> generated in these sites needs to be replicated back to a central
>> database.  I was going to do it in Notes/Domino, running the app on a
>> memory stick that could then be taken by the remote worker to their
>> nearest town, zapped into a workstation with internet access, and so
>> replicated to the central server.  Now I'm wondering if I should do it
>> in Couchdb instead?
> The scenario you describe is the exact kind of which CouchDB was
> designed to handle.
>> Some questions:
>> - can it run an instance (programme and data) on a memory stick?
> Yeah. A simple example is the just-released binary package for
> MacOS X I threw together:
> This is obviously Mac-only, but equally possible to do for other
> platforms.
>> - does this sound feasible in this environment?
>> - Is Couchdb ready for showtime in an environment like this.  I need
>> to feel confident that these isolated offices that are a long way from
>> any help are going to get a robust solution
> Maybe not yet fo your purposes. CouchDB does not have a sense
> of security (it will, though). So you'd need to write a small
> proxy application that would do authentication and whatever else
> you need. I guess you don't want to open each local instance to
> everybody for replication :)
> With large views, we are still missing database compaction, but
> that won't be much of a problem with small data sets or small
> views.
>> (P.S. apologies in advance if this is the wrong forum)
> This is the exact right place to discuss this :) Feel free to send in
> any follow-up questions you might have. And have a look at
> our documentation:
> Cheers
> Jan
> --

View raw message