On Tue, Oct 23, 2012 at 6:21 AM, Amirouche Boubekki
<
amirouche...@gmail.com> wrote:
> I'm all new to unhosted project and maybe also a bit about the problems it
> tackles.
welcome to the mailing list!
> My understanding is that remotestorage is an API for the storage of files,
> how does it relate to S3 or other file hosting services that targets
> application developpers ? Is remotestorage a cloud API ?
It's an alternative to S3. Its advantage is that it support access
from client-side apps, and S3 doesn't (yet). There is no one
'flagship' provider of remotestorage, but there is free software
available with which you can run your own server, and there are a few
commercial providers that are compatible with it.
> What is the difference between synchronization like firefox sync does for
> favorites and remotestorage technically and philosophically if it applies.
firefox sync is built into the browser. remotestorage works entirely
on top of the web without any need to change your browser. Also,
firefox sync uses end-to-end encryption, meaning you can only get to
your data from your own devices, and not from an internet cafe.
> Regarding the applications that runs on top remotestorage, my understanding
> is that they are desktop apps using the cloud with remotestorage API as
> storage and the harddisk as offline and cache facility.
no, they are (generally) html5 apps, built from html, css and
javascript, and running inside the browser. They tend to use html5
localStorage for caching and appcache for application code caching.
> Is remotestorage related to P2P ? If so how ?
no, remotestorage is just per-user storage that is addressable
cross-origin from within the browser. You could distribute unhosted
web apps (that use remotestorage) via p2p distribution though, and
then you have no app servers anymore. An app can then be passed around
and not be shut down by any kill switch.
> What I understand is that unhosted is still centralized given the
> remotestorage rely on a DNS endpoint.
unhosted web apps are just client-side web apps that have no relation
to any specific server. So they in themselves have no reliance on DNS.
When using an unhosted web app, you could store the data only on the
device, in which case you need no DNS, no SSL, you don't even need
TCP/IP. :) But usually when we use an app, we want to save the data,
so that's where remotestorage comes in.
To get to your remotestorage server, most people would use DNS and
SSL, so there you rely on those two centralized systems, yes. We don't
solve that part of the puzzle, but there are other projects doing
that.
> So in the the end, even if the backend
> remotestorage is implemented using P2P or any other protocol,
it's not, remotestorage uses simple http. what makes it
'decentralized' is that everybody can easily run their own
remotestorage server wherever they want.
> it's still centralized, so hosted.
Centralized and hosted are two different things. Remotestorage relies
on DNS and SSL, yes, so it's not as decentralized as it could be if we
would build it outside the browser. And it is still hosted, though, in
the sense of relying on a server. Your data is hosted on your
remotestorage server, server-side.
It is the apps that become 100% client-side. and so that's why we call
them 'unhosted' web apps.
hope this helps! :)
Ciao,
Michiel