ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From cost...@covalent.net
Subject Re: Ant 2 et al.
Date Tue, 09 Jul 2002 02:48:14 GMT
On Tue, 9 Jul 2002 dion@multitask.com.au wrote:

> > These are minor issues.
> > We're talking bout the project/task/target/datatype architecture, not of 
> 
> > how it's implemented.
> 
> Well, how it's used is more important than either architecture or 
> implementation to the end users.
> 
> The current 'paradigm' of ant to an end user is not clear. To recap on one 
> of my points to Conor, a 'project' in the context of a build process makes 
> no sense to the first time user, as there is little or no 'project' 
> information in the build file. Call it 'Compiler' and people will think, 
> initially, that it's going to do compilation. A name is very important to 
> adoption.

Not sure I understand what you want. Changing the <project> element name 
in build.xml to use a different name you feel is more apropriate ? Are you kidding ? 

A number of people ( usually those who -1 the adding of scripting 
elements) believe ant should be more 'descriptive', and not 
procedural. That's why it's called <project> - it is intended to
describe the project, including how to build various components.

Most people only 'describe' how to build and test it, and do that
in a procedural way. That's where the need for <if>, <while>, etc comes 
from, and that's why ant files become ugly and hard to understand.  

However many ant asks are pretty high level, and nothing prevent 
adding more 'descriptive' and higher level information ( using data 
types). Whatever is in the gump descriptor could very well be 
in an ant file.

Of course, the biggest focus is on describing how to build various
targets - that's what people need the most. I agree we should add
more 'descriptive'/higher level  data types under <project>, maybe
what gump uses.


> > Antlib will make it more granular, what's the problem?
> The problem is these ideas have been around a long time. It seems a 
> gradual process of adding them in is happening, rather than a concerted 
> look at what the proposals have to offer. I think someone first broached 
> the ant lib concept almost 18 months ago...and an implementation's been in 
> Mutant and Myrmidon for how long....?

And what's wrong with a gradual process ? Especially for important things
I think we should take all the time it is needed. If something is obvious
and all commiters are +1, it'll probably get added fast.

If there are doubts - then we should spend more time finding a better 
solution. 


Costin


--
To unsubscribe, e-mail:   <mailto:ant-dev-unsubscribe@jakarta.apache.org>
For additional commands, e-mail: <mailto:ant-dev-help@jakarta.apache.org>


Mime
View raw message