> I understand that there might me a my.plkr.org in the future, but I'd
> like a solution now.
my.plkr.org will not let you "sync to the server" like AvantGo.
There's several reasons for this, which I'll briefly go over:
1.) Plucker has no "online" mode, there are no hooks to NetLib
through Plucker. It does not talk to a modem or anything but a locally
attached cradle.
There is a way around this, however, which I'm actively exploring
heavily. I just released the latest pilot-link development version for
testing, in preparation for a 0.11 final release. One of the conduits in the
suite of conduits lets you sync to a remote server, and sends you a file.
Currently there is no authentication or validation of clients during this
process, it's just connect, send/receive data, disconnect.
I would like to wrap some authentication around this a bit and allow
people to "log in" over pi-nredir, which I could then tie to the server
itself, and with *NO* local conduit on your machine, you could grab
content.. but it's not as clean as that. You're literally hotsync'ing to the
Plucker server, not doing a conduit connection to it. Many things aren't in
place to handle that yet, such as compression, and client trust mechanisms.
How do you know I'm not downloading your user information for example. I
personally am very anal about security, but I need to ensure that the
clients believe in this trust circle.
2.) The architecture behind Plucker is all client-driven, not server
driven. Based on this, the server would have to be gathering and storing the
content itself. Yes, disk space is cheap, but bandwidth isn't, and nor are
multiple parallel machines with several gigabytes of RAM to scale up to
handle the load. Based on the log statistics I've been doing, if a site went
live with server-side Plucker content, sync-to-server functions in place,
and could deal with logins/scheduling of content, etc. (which is most of
what I'm building into my.plkr.org), it would jump to about 200,000 users in
6 months. There's a *LOT* of Plucker users eager to see it move in this
direction. .
3.) Transcoding of content. The content providers provide
copyrighted content to companies like AvantGo under a license designed to
protect that copyright. If we gather (and store) that content on any server,
in preparation for a user to connect to and download, we're breaching that
copyright. We cannot (and will not) reproduce and re-distribute content that
is protected by copyright.
4.) Patent restrictions from our friends at AvantGo. We're already
legally clear from most of it (I'm talking to a patent lawyer friend to make
sure we're clear of *ALL* of it), but they will still try to come down on
us. We do this in our spare time, and don't have the funds to support legal
attention to defend our innocense.
> 2. Send the pdbs to my Visor by email. How about automated
> installation here?
I already have this worked out, and it's part of the my.plkr.org
solution. I started this in perl several years ago, but will be using some
different methods now. Now the issue is automating the retrieval of that
.zip'd pdb from your email spool, unzipping it, and then installing it in
some automated fashion.
[dd]
