incubator-couchdb-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dale Harvey <d...@arandomurl.com>
Subject Re: Fauxton + Pouchdb
Date Thu, 07 Nov 2013 19:28:08 GMT
Firefox has unlimited storage, Chrome has heuristics for calculating
storage but it seems to sit somewhere around 20-50MB, Mobile safari moans
around 5MB, the w3c spec I think actually specifies that storage should be
unlimited (pending used agreement) so hopefully a matter of waiting for
browser to fall in line (chrome apps are unlimited, phonegap also unlimited
I believe) (theres a faq @ http://pouchdb.com/faq.html)

Also in the meantime yeh, Pouch is more designed for the initial use cases
of keeping local immediate data stored.

The node implementation is pretty much a full fledged couch replacement, it
doesnt implement configuration in the same way and is missing a few
couchappy type features, although they would be trivial to implement (I
expect there is a plugin lying around somewhere)

The Pouch node implementation is no where near as stable or well tested as
couch, it will likely not perform as better and will most certainly break a
lot more, it was a pretty awesome contribution however I am more focussed
on the browser use case so it hasnt seen much maintenance recently
(although it hasnt needed any, it still passes all tests), since its vastly
easier to install than couch, and now with fauxton experimental support I
expect it can be a good alternative to get started quickly and develop
against, while I would recommend for production environments to be using
CouchDB / etc

Cheers for this Garren + Fauxton team (and thanks to Nick for express-pouch
+ pouchdb-server, they are awesome)



On 7 November 2013 08:00, Jens Alfke <jens@couchbase.com> wrote:

>
> On Nov 7, 2013, at 5:09 AM, Manokaran K <manokaran@gmail.com> wrote:
>
> > isn't the size of local storage limited to a few MB - 5 in Firefox and 10
> > in IE? In which case what'll be the use of a couchdb clone in the
> browser?
>
> Many applications won’t need that much data; the clichéd shopping-list,
> for example.
> The web-app can use a filtered pull replication to only sync the subset of
> the main database that it needs.
>
> —Jens

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message