On 8/17/01 11:04 PM, "Geir Magnusson Jr." <[EMAIL PROTECTED]> wrote:
> Jason van Zyl wrote:
>>
>> On 8/16/01 12:04 PM, "Sam Ruby" <[EMAIL PROTECTED]> wrote:
>>
>>> Jason van Zyl wrote:
>>>>
>>>> The dependencies are not detected for you. Geir's little dep engine
>>> handles
>>>> this nicely. I started with what I thought I needed for the TDK and
>>> assumed
>>>> the dependencies would be detected but they're not. Not a big deal as I
>>> only
>>>> have to do this once, but as projects changed it would be nice for these
>>> to
>>>> be detected so that you don't have to edit your profile when all the
>>>> projects change their definitions.
>>>
>>> Where can I find Geir's little dep engine? Is it something that could be
>>> called by gen.java?
>>
>> I think he put it in the commons ... hmmm it's not there so but it is in the
>> commons sandbox along inside jjar. But the dep engine will go into the
>> commons utils, that's the plan anyway.
>
> Sure, if it has a general use.
>
>>> In prior discussions with Geir, I got the impression that this was
>>> intertwined with the downloading and recursion, but if I could have a
>>> shallow set of dependencies provided, that would be great!
>>
>> I used the little dep engine for the java version and it works great.
>
> This is starting to sound like duplication of what already exists in the
> jjar 'stuff', and is leading to a duplication - would it be possible to
> simply use jjar to get the dependencies? That way, there would only
> one set of dependency descriptors that could be used both for gumping as
> well as for general project build and deployment use.
Gump has been around a lot longer than jjar and has comprehensive dependency
info that can easily be used for a lot more projects. I don't think using
jjar to get the dependencies is the way to go. Alexandria and Gump, I feel,
are more appropriate places for this information and that's what I'm going
to continue with this weekend. I honestly feel that jjar should have been
based on the descriptors long present in Gump.
I think jjar can easily use what's in Gump. They are already cleaned up and
there will be code for parsing these XML descriptors into Java objects and
this code can be reused in many building related tools. Alexandria/Gump has
always been slated to be the storehouse for this information.
> geir
--
jvz.
Jason van Zyl
http://tambora.zenplex.org
http://jakarta.apache.org/turbine
http://jakarta.apache.org/velocity
http://jakarta.apache.org/alexandria
http://jakarta.apache.org/commons
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
