|
Re: Source plugins via url.zipped [message #549772 is a reply to message #549698] |
Tue, 27 July 2010 22:06 |
|
Hi Thorsten,
Try using the 'url' reader type instead of the 'url.zipped'. The latter assumes that you want to treat the zip as a
folder and will unpack it in order to find meta-data inside of it. That does not seem to be what you want. The component
type should be 'unknown'.
HTH,
- thomas
On 07/27/2010 05:08 PM, Thorsten Meinl wrote:
> Hi all,
>
> I want to include a JRE in our RCP products. We currently support five
> different platforms, so I created five features that specify the
> corresponding JRE as rootfiles in their build.properties. However, I do not
> want to add several hundreds of megabytes of binaries to our source code
> repository. Therefore my idea was to put the five features an a web server
> and fetch them via the url.zipped reader. However, this doesn't work as
> expected.
> If I put them as .zip files, Buckminster tries to find files such as
>
> jre.x.y.z.zip/features/jre.x.y.z.jar
>
> (the URI format in the reader is ".../{0}.zip" with {0} being a reference to
> buckminster.component). If I rename them to .jar (and update the URI format
> accordingly) Buckminster finds them but extracts all of them directly into
> ..buckminster/features (the feature name is lost somehow). What I want is the
> url.zipped reader working similar to the svn reader in that it extracts the
> zipped features and puts them into the workspace as if they were fetched via
> svn. Setting the reader's source-property to true didn't help.
> Is this possible somehow? Or is there a completely different best-practice
> way of managing JREs? I don't want to rely on some fixed filesystem
> location.
>
> Best regards,
>
> Thorsten
|
|
|
|
Powered by
FUDForum. Page generated in 0.05044 seconds