Vivek P wrote:
> Hi,
>
> I was using NAnt 0.8.3 for which i created a custom task and am using
> it just fine. However, I am now forced to use NAnt 0.8.2 because of
> the current build process we have here. The problem is that my task is
> not being recognized with 0.8.2.
>
Did you re-build your t
I've also started porting the Ant perforce tasks. I haven't gotten too
far - just the base class and p4 add. We should be able to merge the two
efforts.
Ian
Good stuff.
Any chance of getting those into nantcontrib or whereever they fit best?
Would be nice to have some perforce tasks in a posit
- Original Message -
From: "Jean Rajotte" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Thursday, November 20, 2003 6:30 PM
Subject: RE: [Nant-users] Incorrect dependency evaluation?
> gert, thanks for sticking w/ me on this.
>
> > > your build example could be done using call+force
patrick,
i understand and i thought of that, but it means the caller can't
control whether common-build-init can run again. it's become
common-build-init-run-once-only-ever-amen and that's a loss. now, if
we could disappear existing props... is this in place?
jean
> From: [EMAIL PROTECTED]
FWIW - this is how I changed things to implement an "execute only once"
idiom within a target:
--
Evan,
You need to run vsvar32.bat before starting the build or run the build in
the "Visual Studio .NET Command Prompt" ...
Gert
- Original Message -
From: "Bonnett, Evan A" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Thursday, November 20, 2003 7:33 PM
Subject: [Nant-users] (no su
I'm trying to build nant but this project is
failing. Can anyone lend a hand here?
Thanks,
[nant] tests/NAnt.VisualCpp/NAnt.VisualCpp.build test
Buildfile:
file:///C:/nant/tests/NAnt.VisualCpp/NAnt.VisualCpp.build
build:
william sez:
> I think Jean's point was that since we can't know how our targets are
> going to get used (i.e. depend'd or call'd), we have to
> assume they will
> be called (i.e. their dependencies will be run more than once).
>...
> not being
> able to say that a call task doesn't rerun depe
Gert:
>>
>> might as well say depends= don't count and always use />.
>No, quite the opposite : use depends as much as possible, and only use
the
> task when you want to re-execute the complete stack ...
I think Jean's point was that since we can't know how our targets are
going to get used (i
gert, thanks for sticking w/ me on this.
> > your build example could be done using call+force just as
> readily in
> > the past as you have it now.
>
> feel free to provide a sample to convince me, but I don't
> think we will be change the new behaviour :-)
here's your example:
- Original Message -
From: "Jean Rajotte" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Thursday, November 20, 2003 5:26 PM
Subject: RE: [Nant-users] Incorrect dependency evaluation?
calling a target using the task will now indeed cascade to the
dependencies' dependencies ...
> and
and another thing :)
your build example could be done using call+force just as readily in the
past as you have it now.
my setprops example (30 targets depend on setprops, which should run
only once unless forced otherwise) is broken and i'm not sure how to fix
it. basically, it seems depends= ca
Hi,
I was using NAnt 0.8.3 for which i created a custom task and am using it
just fine. However, I am now forced to use NAnt 0.8.2 because of the
current build process we have here. The problem is that my task is not
being recognized with 0.8.2.
--
i'm getting
Unknown task
--
for 0.8.3 i si
so, you're saying : CALLing a target forces to execute its dependencies.
does this cascade to the dependencies' dependencies (etc) ?
---
This SF.net email is sponsored by: SF.net Giveback Program.
Does SourceForge.net help you be more product
Thanks Gert & Brett.
I also found this out last nite after a few hours of digging and
trial & error. I wish this was a little more documented but hey I
found it in the code. Thanks a bunch guys for making such quick
responses. Looking forward in helping on the 0.8.4 release testing.
My Solutio
- Original Message -
From: "Jean Rajotte" <[EMAIL PROTECTED]>
To: "'Breen, Patrick'" <[EMAIL PROTECTED]>;
<[EMAIL PROTECTED]>
Sent: Thursday, November 20, 2003 3:54 PM
Subject: RE: [Nant-users] Incorrect dependency evaluation?
> please let it be that 0.8.4 behaviour is a bug, and that <0
i guess depends= isn't broken, but call is.
gert,
i can see that it's more "elegant" to get positive feedback when you say
. however, imo, it's just syntactic sugar, not essential -- the
language cost of force= was minor compared to revamping all those build
files invested in the previous behav
why, oh why? what was broken w/ force= ?
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of
> Gert Driesen
> Sent: Thursday, November 20, 2003 10:00
> To: Breen, Patrick; [EMAIL PROTECTED]
> Subject: Re: [Nant-users] Incorrect dependency evaluation?
please let it be that 0.8.4 behaviour is a bug, and that <0.8.4
behaviour is the right one. there's a force= attribute to change that.
i'm heavily invested in only-run-once for depends= (e.g. every target
calls the global init-props target to afford each target granularity).
> -Original Mes
Patrick,
This behaviour has indeed changed in the 0.8.4 release, and is documented as
a breaking change in the 0.8.4 change log / release notes
(http://nant.sourceforge.net/changelog.html).
The task has been changed to always execute the given target, and all
its dependencie, regardless of the v
In version 0.8.4 when a target is defined with a dependency that calls
another target with the same dependency, the target is executed a second
time. In version 0.8.3, the target was only executed once. Which is the
proper / desired behavior? The build scripts we have were counting on
the 0.8.3 b
Hi,
I am still having problems. Having picked up the nightly build
(0.8.4.31114), NANT builds the files but fails to run NUnit. I am running
under .NET 1.0. Has anyone any ideas?
Building...
NAnt version 0.8.4.0 Copyright (C) 2001-2003 Gerry Shaw
http://nant.sourceforge.net
Buildfile: file://
Scott Arbiv wrote:
Could you remove me from this list.
Thanks
Scott Arbiv
[EMAIL PROTECTED]
Use the following instructions that were embedded in each message you
received from the list:
To subscribe or unsubscribe via the World Wide Web, visit
https://lists.sourceforge.net/lists/lis
Good stuff.
Any chance of getting those into nantcontrib or whereever they fit best?
Would be nice to have some perforce tasks in a position that they can be
actively developed.
Brett
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Jean
Rajotte
Sent: Thu
Jaroslaw ,
You're right on the money here. This is somthing I've been thinking
about for a while. I think Gerry wrote a basic EE a while back but
nothing came of it.
It would be awesome of you'd like to build the basic EE stuff. I'll
pitch in as well but it sounds like you have a bit more experi
25 matches
Mail list logo
