On 28/01/11 13:02, Rob Vesse wrote:
Andy
Yes invoking the commands is easy which is what is nice about the protocol.
Just bit myself by using my pre-existing UriLoader to handle the GET requests
and forgetting that it would automatically cache the responses. Though I think
part of the logic behind this design choice originally was that if the server
returns ETags then the loader does conditional GETs.
Made sure the cache gets invalidated in the appropriate places and all my unit
tests for Fuseki integration now run fine :-)
Batch scripts to invoke a jar would work as you'd hope if someone was trying to
run Fuseki that they'd have Java installed already.
curl is a possibility but not sure whether the average user would have it
installed. Even I have never bothered to install curl on a Windows machine.
Might code up a command line utility myself to add to my existing Toolkit [1]
and you could point to that for Windows users in the future? Will keep you
posted if you are interested?
That would be excellent.
Part of the point of standards is that the client and server can be
different systems.
Andy
Rob
[1] http://www.dotnetrdf.org/content.asp?pageID=Tools
-----Original Message-----
From: Andy Seaborne [mailto:[email protected]]
Sent: 28 January 2011 11:41
To: [email protected]
Subject: Re: Fuseki DELETE doesn't seem to work
Rob,
Glad this is sorted out.
The SOH scripts are in ruby, although only tested on Linux. I'd like to
ship something for Windows as well (or point to other tools - there are
some around but look like they are Linux-y).
What can be relied on Windows for command use? I was hoping to not ship
java client-side code but it would not be hard to produce a single jar
of commands that some bat scripts could call. Invocation of the
protocol operations isn't hard (that's the point!)
I'm slowly building up curl commands - that's a possibility. Is it
practical?
Andy
