Ok, great - good to know I wasn't trying to do something that is
outright forbidden by the spec. I'll clean this patch up and submit it.
Phil.
On 17/11/09 15:12, Richard S. Hall wrote:
On 11/17/09 7:29, Phil Messenger wrote:
Ok, so I've encountered another issue that I can't see covered in the
spec.
Regarding the prior issue, the Felix framework is rather simplistic
about determining conflicts, so currently it requires them to be an
exact match. This still abides by the spec, since the spec doesn't
specify what it means to conflict.
Another fragment defines an import or org.w3c, with a version
attribute of "0", which AFAIK is equivalent to "empty version". The
bundle also defines that import, but with no version specified at
all. AFAIK, this /should/ be ok, but
// If different number of attributes, then they conflict.
if (exAttrs.length != addAttrs.length)
{
return true;
}
Causes the requirements to conflict. Is this a malformed bundle and
fragment, or is this a bug?
Well, specifying no version on an import is equivalent to specifying
"0.0.0" and "0" is also equivalent to "0.0.0", so I think it is safe
to say they are equivalent. This is just a result of the exact match
comparison.
If you want to submit a patch for this, I will take a look at it.
-> richard
Phil.
On 17/11/09 11:39, Phil Messenger wrote:
Hi,
I've got a problem using Hibernate as packaged by the folks at
Springsource. I need to use the Hibernate Annotations fragment,
which has the following Import-Package dependency:
org.slf4j;version="[1.3.1, 2.0.0)"
The Hibernate bundle specifies the following Import-Package dependency:
org.slf4j;version="[1.5.6, 2.0.0)"
Deploying the fragment fails because:
DEBUG: Excluding fragment com.springsource.org.hibernate.annotations
from com.springsource.org.hibernate due to conflict with imported
package org.slf4j from com.springsource.org.hibernate
Looking at the source code, I'm not convinced the range check on
line 437 of FelixResolver is correct - surely it should be checking
that the import range specified by the addition is satisfied by the
import range specified by the main bundle, rather than checking the
ranges are identical?
If that's the case, I have a patch here which I'll add to a bugzilla
issue...
Phil.
This e-mail (and each attachment) is confidential. If you have
received this email in error please delete it and notify us
immediately. Views or opinions in this email do not necessarily
represent those of Miniweb Technologies Limited (registered in
England with number 5897305 ) or its subsidiaries. No representation
is given nor liability accepted for the accuracy or completeness of
any information included in this email.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
This e-mail (and each attachment) is confidential. If you have
received this email in error please delete it and notify us
immediately. Views or opinions in this email do not necessarily
represent those of Miniweb Technologies Limited (registered in
England with number 5897305 ) or its subsidiaries. No representation
is given nor liability accepted for the accuracy or completeness of
any information included in this email.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
This e-mail (and each attachment) is confidential. If you have received this
email in error please delete it and notify us immediately. Views or opinions
in this email do not necessarily represent those of Miniweb Technologies
Limited (registered in England with number 5897305 ) or its subsidiaries. No
representation is given nor liability accepted for the accuracy or completeness
of any information included in this email.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
