Particular concurrent access

[Francisco Carriedo Scher <fc...@gmail.com>]

Hi there,

is it possible in Jackrabbit to start saving a binary file (node with binary
property) and obtain the URL to such file (and start reading it) before it
is completely saved?.

Thanks in advance for your attention!

Re: Particular concurrent access

[Francisco Carriedo Scher <fc...@gmail.com>]

As i do not have the use case accurately defined i can just give the
baseline of the idea: live video sharing. Let's say publishing some media
resource (video) that will be grown by the source and concurrently consumed
by other user(s).

Reading JCR 2.0 specification - 17.7 Effect of a Lock [...] Locked nodes can
always be read and copied by any session [...] - i saw a potential way of
getting it working:

1) a producer store a node with a binary property (the first chunk of the
video) and save it.
2) as the node was saved it will be HTTP-GET accessible to consumer(s)
3) the producer updates the binary property (the video stream itself) and
the consumer(s) gets it as it is being progressively stored (this is the
critical point where i have no idea if Jackrabbit serves a snapshot of the
requested resource or it opens the connection and sends data to the client
until the binary data is finished)

Would the stated above be possible?

Thanks in advance for your attention.




2011/10/29 Jukka Zitting <ju...@gmail.com>

> Hi,
>
> On Fri, Oct 28, 2011 at 11:20 PM, Francisco Carriedo Scher
> <fc...@gmail.com> wrote:
> > is it possible in Jackrabbit to start saving a binary file (node with
> binary
> > property) and obtain the URL to such file (and start reading it) before
> it
> > is completely saved?.
>
> No, that's not possible. What's your use case for doing something like
> that? There's no fundamental reason why such a feature could not be
> implemented, so if there's a general enough need for it we can
> certainly look at how it could be done.
>
> BR,
>
> Jukka Zitting
>

Re: Particular concurrent access

[Jukka Zitting <ju...@gmail.com>]

Hi,

On Fri, Oct 28, 2011 at 11:20 PM, Francisco Carriedo Scher
<fc...@gmail.com> wrote:
> is it possible in Jackrabbit to start saving a binary file (node with binary
> property) and obtain the URL to such file (and start reading it) before it
> is completely saved?.

No, that's not possible. What's your use case for doing something like
that? There's no fundamental reason why such a feature could not be
implemented, so if there's a general enough need for it we can
certainly look at how it could be done.

BR,

Jukka Zitting