ant-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Steve Loughran" <>
Subject Re: <available> / <condition> breaking immutability
Date Mon, 26 Nov 2001 20:05:31 GMT

----- Original Message -----
From: "Bevan Arps" <>
To: "Ant Developers List" <>
Sent: Monday, November 26, 2001 11:45 AM
Subject: Re: <available> / <condition> breaking immutability

> As a user with some moderately complex Build files, I agree with Erik's
> above comments about "expected" behaviour.

> However, I'd actually go one step further. To me, if the property is
> already set before <available> (and similar tasks) fire, you have a broken
> build and you should get some sort of fatal error.

I dunno about fatal error, but it reporting it at the MSG_WARN level may
make sense -if the assignment was to have been different. That way if you
have a child build file that checks for junit.available and the parent did
already, then all is well.

> The decision has been made that properties are immutable - I have no
> problem with this (at least they aren't called variables, like they are in
> XSLT!). So, they really should be immutable and *any* task that changes
> this should be considered broken.

I think they probably have been broken, but nobody noticed till now. That
is, unless people have been using these tasks for the purpose of overwriting
properties without letting on...

> This communication  is confidential  to ACT  Financial  Systems  (Asia
> Pacific)  and is intended for  use only by the  addressee.   The  views
> opinions  expressed in  this email  are the senders  own and do not
> represent  the  views  and  opinions of  ACT  Financial  Systems  (Asia
> Pacific).

So what is the official ACT  Financial  Systems  (Asia Pacific) stance on
property immutability in ant then :-)


To unsubscribe, e-mail:   <>
For additional commands, e-mail: <>

View raw message