Matthew Johnson wrote:
> On Mon Apr 26 19:58, Torsten Werner wrote:
>> On Mon, Apr 26, 2010 at 7:45 PM, Niels Thykier wrote:
>>> What do you feel about this solution; we rename "default-jdk-builddep"
>>> to "gcj-native-helper" and have it only pull gcj-jdk (if available) or
>>> nothing at all.
>>>
On Mon Apr 26 19:58, Torsten Werner wrote:
> On Mon, Apr 26, 2010 at 7:45 PM, Niels Thykier wrote:
> > What do you feel about this solution; we rename "default-jdk-builddep"
> > to "gcj-native-helper" and have it only pull gcj-jdk (if available) or
> > nothing at all.
> > When updating Build-rdep
On Mon, Apr 26, 2010 at 7:45 PM, Niels Thykier wrote:
> What do you feel about this solution; we rename "default-jdk-builddep"
> to "gcj-native-helper" and have it only pull gcj-jdk (if available) or
> nothing at all.
> When updating Build-rdepends we will have to add default-jdk as well
> to get
Matthias Klose wrote:
> On 12.04.2010 14:40, Torsten Werner wrote:
>> -BEGIN PGP SIGNED MESSAGE-
>> Hash: SHA1
>>
>> Matthew Johnson schrieb:
>>> AIUI you were complaining about the specific use of gcj-jdk. I'm
>>> suggesting
>>> that we have a meta package for jdk and a metapackage for -gc
On 13.04.2010 00:52, Matthew Johnson wrote:
On Tue Apr 13 00:46, Matthias Klose wrote:
if this is available on all archs and doesn't do anything if gcj is not
available, then yes.
Yes, although if you are trying to build a -gcj package on an architecture
which does not have gcj, possibly faili
On Tue Apr 13 00:46, Matthias Klose wrote:
>>> if this is available on all archs and doesn't do anything if gcj is not
>>> available, then yes.
>>
>> Yes, although if you are trying to build a -gcj package on an architecture
>> which does not have gcj, possibly failing the dependency is actually
>
On 13.04.2010 00:36, Matthew Johnson wrote:
On Mon Apr 12 19:57, Matthias Klose wrote:
On 12.04.2010 14:40, Torsten Werner wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matthew Johnson schrieb:
AIUI you were complaining about the specific use of gcj-jdk. I'm suggesting
that we have a m
On Mon Apr 12 19:57, Matthias Klose wrote:
> On 12.04.2010 14:40, Torsten Werner wrote:
>> -BEGIN PGP SIGNED MESSAGE-
>> Hash: SHA1
>>
>> Matthew Johnson schrieb:
>>> AIUI you were complaining about the specific use of gcj-jdk. I'm suggesting
>>> that we have a meta package for jdk and a me
On 12.04.2010 14:40, Torsten Werner wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matthew Johnson schrieb:
AIUI you were complaining about the specific use of gcj-jdk. I'm suggesting
that we have a meta package for jdk and a metapackage for -gcj packages and
depend on jdk, -gcj; rather t
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Matthew Johnson schrieb:
> AIUI you were complaining about the specific use of gcj-jdk. I'm suggesting
> that we have a meta package for jdk and a metapackage for -gcj packages and
> depend on jdk, -gcj; rather than what we have at the moment which is
On Mon Apr 12 13:58, Matthias Klose wrote:
>>
>> -gcj please, it's not needed just for for JNI, that should be clear. I also
>> agree that there's no need to have a default-jdk+gcj builddep, you can just
>> depend on both if you need both. I don't know whether gcj-jdk is suitable for
>> that, if no
On Mon, Apr 12, 2010 at 01:08:24PM +0200, Vincent Fourmond wrote:
> Probably completely dropping this paragraph is the best solution:
>
> "The same technique is for example adopted by the Java maintainers
> without using build-essential but by providing a default-jdk-builddep
> metapackage that
On 12.04.2010 12:42, Matthew Johnson wrote:
On Mon Apr 12 10:56, Vincent Fourmond wrote:
On Mon, Apr 12, 2010 at 8:36 AM, Niels Thykier wrote:
As some of you know, default-jdk-builddep (usually) pulls in two JDKs
(openjdk-6 and gcj/gij) to create -gcj packages.
However, some people are not a
On 12.04.2010 13:08, Vincent Fourmond wrote:
On Mon, Apr 12, 2010 at 12:52 PM, Enrico Zini wrote:
On Mon, Apr 12, 2010 at 12:26:24PM +0200, Matthias Klose wrote:
The change was discussed here on the ML. I don't mind about the
name, but this should be a distinct package.
CC'ing Enrico; please
On Mon, Apr 12, 2010 at 12:52 PM, Enrico Zini wrote:
> On Mon, Apr 12, 2010 at 12:26:24PM +0200, Matthias Klose wrote:
>
>> The change was discussed here on the ML. I don't mind about the
>> name, but this should be a distinct package.
>>
>> CC'ing Enrico; please change that in [1] for now.
>> [1]
On Mon, Apr 12, 2010 at 12:26:24PM +0200, Matthias Klose wrote:
> The change was discussed here on the ML. I don't mind about the
> name, but this should be a distinct package.
>
> CC'ing Enrico; please change that in [1] for now.
> [1]
> http://svn.debian.org/viewsvn/nm/trunk/nm-templates/nm_ts
On Mon Apr 12 10:56, Vincent Fourmond wrote:
> On Mon, Apr 12, 2010 at 8:36 AM, Niels Thykier wrote:
> > As some of you know, default-jdk-builddep (usually) pulls in two JDKs
> > (openjdk-6 and gcj/gij) to create -gcj packages.
> > However, some people are not aware of this and looking at the nam
On 12.04.2010 11:27, Torsten Werner wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi,
Niels Thykier schrieb:
I think the best idea is to rename default-jdk-builddep into something
else that does not trigger the "Ah, this is what I should put in
B-D"-instinct of our fellow maintainers an
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi,
Niels Thykier schrieb:
> I think the best idea is to rename default-jdk-builddep into something
> else that does not trigger the "Ah, this is what I should put in
> B-D"-instinct of our fellow maintainers and developers. If you have a
> suggestion
Hello,
On Mon, Apr 12, 2010 at 8:36 AM, Niels Thykier wrote:
> As some of you know, default-jdk-builddep (usually) pulls in two JDKs
> (openjdk-6 and gcj/gij) to create -gcj packages.
> However, some people are not aware of this and looking at the name of
> the package they assume it is the Ja
Hi
As some of you know, default-jdk-builddep (usually) pulls in two JDKs
(openjdk-6 and gcj/gij) to create -gcj packages.
However, some people are not aware of this and looking at the name of
the package they assume it is the Java Team's "Default Build-Dependency"
or in other words the "Right Th
21 matches
Mail list logo
