Hi,
here's a first version. It createsthe license file as it
should - at least it runs OK here. I attached the new
LicenseTask.cs
-sa
--
sa at programmers-world dot com
http://www.livingit.de
Bookmarks online: http://www.mobile-bookmarks.info
Soon available in english
The
Hello,
After digging a little into the code I
noticed that destination is marked as required and in my build file it is spelt
wrong. Two things from this as a newbie, first the documentation for cvs-checkout
on the website indicates that none of the attributes for cvs-checkout are
Ooooppps, should have sent to NAnt Dev:
My developers just started incorporating NDoc into their projects. My first build has resulted in some errors in the compile. It looks like NAnt is looking for the *.xml file that should result before the compile is done and erroring out because it's not
Double Opps, Jean answered me. I'll bounce back as to whether it worked or not (warning level change). Thanks Jean!Eric Fetzer [EMAIL PROTECTED] wrote:
Ooooppps, should have sent to NAnt Dev:
My developers just started incorporating NDoc into their projects. My first build has resulted in
Woops, just found the build file. I see a couple of problems here:
target name = source depends = clean description = Retrieves the
source from the cvs repository
cvs-checkout cvsroot = :sspi:stealth.rjktech.com:/cvs module =
InteliPlot
desitnation = $(build.dir}/
/target
I don't think that this is the best place for me to have a discussion
about why I am not using the pserver protocol (cough: passwords are sent
as plain text).
I will just have to call the cvs executable directly using the exec
task. Perhaps if I have time I will try to submit a patch to #cvslib
This is true. The other option would be to use ssh (:ext protocol)
which should also be supported by cvs *nix and cvsnt.
Having said that if you have the time to create a patch I think that
would be awesome (and am probably not alone).
Clayton
-Original Message-
From: Scott Ford
- if not defined :
* if it is a project reference then CopyLocal is true
* if anything else (file present in gac or not) then CopyLocal is false
Just to clarify a point. You can *never* reference an assembly in the
GAC. You can reference an assembly somewhere on the filesystem
OK, I'm losing my mind here. I got the latest nightly build, put it on my build machine (to see if my current problems were addressed...) and got an error. Now I can't seem to get back to a place where NAnt works. I've deleted the directory from the 9/16 build and pointed at the latest stable
OK, strangeness. When I got the latest build, I put it on my D drive and pointed to the bin directory in my path. Got that error and went back to the oldpath (kind of)pointing at the previous stable build on my D drive. The problem was in the D drive itself. The D drive onmy build machine is
Hi,
I had a problem where I merge sourcefiles into one assembly,
while I use them normally in more than one assembly to link
only what I need.
I'm using the ndoc task to document, and hence it's missing
the namespace descriptions, if I do not create a new
namespace.xml for the merged assembly.
Could we try and avoid cross posting to both the users and developers
lists. Many of us are subscribed to both. The AssemblyVersion thread was
long enough without having 2 copies of each message.
Ian
---
This sf.net email is sponsored
You can probably rely on a version of perl being installed on almost
every platform you can think of. You'll probably have to go and install
it on windows machines but it will be on all the nix's.
Having said that, compiling a simple tool on the fly using csc is fine
too. Whatever works for
You can use my trick:
The technique consists of:
1. A plain text file that holds nothing but a version number
AA.BB.CC.DD (version.txt)
2. A small C# utility that increments one of the components in
the file (NewBuild.exe)
3.A small C# utility that replaces AssemblyVersion(...)
in all
Title: Message
In ReplaceVersion.cs, I wonder
if DirectoryInfo.GetFiles(string) to see if the AssemblyInfo.cs file
exists.
-
Ants
-Original Message-From:
[EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Jaroslaw
KowalskiSent: Tuesday, 16 September, 2003
Title: Message
Yes, it seems to be legal: a snapshot of my
ILDASM dump:
.assembly extern Sooda{
.publickeytoken = (36 D3 1D A5 0D 00 DE DD
)
// 6... .ver 1234:1234:1234:5}
Jarek
- Original Message -
From:
Anthony LoveFrancisco
To: 'Jaroslaw Kowalski' ;
Title: Message
Because you can either provide a
platform-independend executable and nant willrun it or compile it
on-the-fly using csc task. So you need nothing but NAnt + your
project.The source code for both utilities is trivial as opposed to "sed"
or "awk" sourcecode which requires a
Title: Message
Oops, sorry, my bad. I think
that65535 is the upper limit for build number parts which less than
Int32.Max.
-
Ants
-Original Message-From: Jaroslaw
Kowalski [mailto:[EMAIL PROTECTED] Sent: Tuesday, 16 September, 2003
11:05To: Anthony LoveFrancisco; 'Daniel
Title: Message
Yes, but who would want build numbers to be
that high?
Jarek
- Original Message -
From:
Anthony LoveFrancisco
To: 'Jaroslaw Kowalski' ; 'Daniel Nguyen' ; [EMAIL PROTECTED]
; [EMAIL PROTECTED]
Sent: Tuesday, September 16, 2003 8:15
PM
Subject:
Title: Message
It's just an edge condition
that needs to be checked.
At one point there was
discussion of using the ymmdd as part of the build number for NAnt.
Later I proposed just using number of days since a particular date and just
incrementing that.
-
Ants
-Original
20 matches
Mail list logo
