ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Peter Donald <dona...@apache.org>
Subject Re: [DISC] XML issues
Date Tue, 03 Apr 2001 01:55:05 GMT
At 10:27  2/4/01 +0200, Stefan Bodewig wrote:
>> It was just an idea I had ages ago (or maybe it was suggested on the
>> list?). Essentially associated with every task there are certain
>> "aspects" that have little to do with what the task does but do
>> relate to how we "host" the task. Some aspects that have been
>> suggested so far include
>
>I agree that we should move things common to all tasks out of the Task
>interface itself, what I'm not convinced of is the use of namespaces
>here. Why should a build file writer be bothered to know whether
>failonerror is handled by the task itself or something else?

It really depends on what class of user we are aiming at I guess. The user
who wants absolute simplicity would get zero advantage - except maybe for
naming where tasks come from - such as

<tasklib name="jdk" location="..." />

<jdk:javac ... />

Looks easier to me especially when you have a range of possible tasklibs
available.

However it is at higher complexity build files (medium->high complexity)
where the advantage would be shown. Mainly as it would allow core to be
clean and handling of tasks predictable. For big build processes it would
allow customisation without magic variables (ala GUMPs sysclasspath) and
added value (ie assign fee: namespace to something specific to buisness).

For medium size build processes it could be an advantage because other
people have already created handlers for particular facilities for you to
rely upon. You can plug in the antdoc facility and generate nice
documentation outlining your build-process, plug in blame: namespace
handler and all your failures will be directed to particular individual
people. Plug in "interactive:" handler and you can get questions asked at
build time etc.

Whether these things are to be considered good or bad is largely irrelevent
in my consideration of the architecture. It does allow others to do
whatever they want with the tool (in theory) - it may turn out they do bad
things and then again a few may turn out exceptionally good. 


Cheers,

Pete

*-----------------------------------------------------*
| "Faced with the choice between changing one's mind, |
| and proving that there is no need to do so - almost |
| everyone gets busy on the proof."                   |
|              - John Kenneth Galbraith               |
*-----------------------------------------------------*


Mime
View raw message