Adriano Crestani wrote:
Hi Haleh,
This way we would be using the Cxxtest api, and according to Simon it's
considered derived work, so we couldn't distribute it, right Simon?
This comment came from Pete, not from me. I'm not familiar with the
stdcxx license so I'll defer to Pete to explain why it's a concern.
Simon
Adriano Crestani
On 10/23/07, haleh mahbod <[EMAIL PROTECTED]> wrote:
But if you go with what Simon suggested, you leave the tests and test tool
outside of distribution. Wouldn't that work?
On 10/23/07, Adriano Crestani <[EMAIL PROTECTED]> wrote:
I think this is one for the legal discuss list. This has been
discussed before and I think the conclusion was that because you code
to the cxxtest apis to write your test code it could be considered a
derivative work.
So, does it mean we cannot distribute a code using a api that we cannot
distribute? Then we should start looking for another unit test. I was
looking on the web site I commented before, most of them are GPL : (,
but
I
found this 2:
http://unittest-cpp.sourceforge.net/
http://tut-framework.sourceforge.net/
Their license seems to have almost no restriction, but I cannot tell for
sure if they are compatible with ASF license.
Regards,
Adriano Crestani
On 10/23/07, Pete Robbins <[EMAIL PROTECTED]> wrote:
I think this is one for the legal discuss list. This has been
discussed before and I think the conclusion was that because you code
to the cxxtest apis to write your test code it could be considered a
derivative work.
Cheers,
On 23/10/2007, Simon Nash <[EMAIL PROTECTED]> wrote:
I think it's fine to distribute unit test source and tell people
what
tool they need to build and run the tests. And I agree that having
a
list of suitable unit test tools on the Web site is helpful.
Simon
Adriano Crestani wrote:
Hi Simon,
Yes, you are right, I forgot this option, there is no problem to
distribute
the unit test source code :P. But anyway, the list contained on
the
web site
I could be helpful :)
Regards,
Adriano Crestani
On 10/22/07, Simon Nash <[EMAIL PROTECTED]> wrote:
Why does the test tool need to be distributed with a Tuscany
release?
If the build depends on having the tool available, then I can see
some
justification for this, but even then it would be possible for
people
who build the source to download the tool separately.
Simon
Adriano Crestani wrote:
Hi,
Brady suggested to use CxxTest only on development process and
don't
distribute it with the released source. However, whoever wants to
modify
the
code from a release would want to test it, to check if the
modifications
does not compromise the software. So, I suggest to look for
another
text
unit tool that could be distributed with the released source. I
really
dont
know any other, but searching on web I found a list of open
source
C/C++
unit test tools on [1].
[1] http://www.opensourcetesting.org/unit_c.php
Regards,
Adriano Crestani
On 8/10/07, Brady Johnson <[EMAIL PROTECTED]> wrote:
Good idea, I always prefer to see plenty of documentation. I
updated
the
wiki with a documentation feature.
http://cwiki.apache.org/confluence/display/TUSCANYWIKI/SCA+Native+Next+R
elease+Contents
What sort of help do you think I'll have with these features?
--------------------
Brady Johnson
Lead Software Developer - HydraSCA
Rogue Wave Software - [EMAIL PROTECTED]
-----Original Message-----
From: haleh mahbod [mailto:[EMAIL PROTECTED]
Sent: Friday, August 10, 2007 3:36 PM
To: [email protected]
Subject: Re: [SCA Native] next release content [was: Tuscany
roadmap]
How about enhancing the documentation (architecture, get started
and
user
doc) to help new people come on board faster?
Another thought might be to have an integration story between
Native
and
Java. Some of this work started for OSCon, for example a sample
of
a
composite which include C++ and Java components.
On 7/26/07, Pete Robbins <[EMAIL PROTECTED]> wrote:
That looks good. I think there is more than enough in that list
to
justify a release. My priorities would be:
1) upgrade to the sca 1.0 spec levels (assembly and cpp).
2) build system move to ant
(enough there for a release)
We should discuss your ideas for the rearchitecture of the data
model.
It sounds like a good idea so maybe we can flesh out a proposal
for
that.
Cheers,
On 26/07/07, Brady Johnson <[EMAIL PROTECTED]> wrote:
Hello all,
I created a wiki page detailing the TuscanySCA Native Next
Release
Contents, which will probably be called M4.
http://cwiki.apache.org/confluence/display/TUSCANYWIKI/SCA+Native+Ne
xt+R
elease+Contents
Can I get some feedback on the items listed there. Also,
what's
the
Apache procedure to start planning and implementing the
changes?
--------------------
Brady Johnson
Lead Software Developer - HydraSCA
Rogue Wave Software - [EMAIL PROTECTED]
-----Original Message-----
From: Pete Robbins [mailto:[EMAIL PROTECTED]
Sent: Thursday, July 12, 2007 11:00 AM
To: [email protected]
Subject: Re: [SCA Native] next release content [was: Tuscany
roadmap]
On 12/07/07, Brady Johnson <[EMAIL PROTECTED]> wrote:
I forgot to mention another one in my previous post:
- get the test suite up to date and working. I don't like
making
changes to code without running a good unit/basic test suite.
We do not have ANY test suite. I run through the samples to
test
changes. The code under tuscany/cpp/sca/test is not maintained
and
should probably be discarded. I think we need to build up a
unit
test suite and would welcome suggestions on how to start this
(use
cppunit?)
I can start a separate thread for the ant vs make discussion.
Basically, I think it would be easier to simplify the build
process using make. I've looked through some of the makefiles
and
they're horrendous. :)
Let's discuss it here then. We need to be able to build from
source
on windows, linux and Mac. On Windows we settled on MSVC 8 so
it
can
build with the free studio express. For linux we settled on
automake
as it seemed to be fairly standard for C/C++ open source
projects.
In doing this I had to learn automake and learnt to hate it
;-) ...
and as you say some of the makefiles are ugly. If you believe
an
ant
based build would be better then I'll happily go along with
that.
Perhaps you could start this off by showing us what the build
would
look like for, say, cpp/sca/runtime/core ??
--------------------
Brady Johnson
Lead Software Developer - HydraSCA Rogue Wave Software -
[EMAIL PROTECTED]
-----Original Message-----
From: Pete Robbins [mailto:[EMAIL PROTECTED]
Sent: Thursday, July 12, 2007 9:53 AM
To: [email protected]
Subject: [SCA Native] next release content [was: Tuscany
roadmap]
We should definitely start planning some content for the next
SCA
Native release.
On 12/07/07, Brady Johnson <[EMAIL PROTECTED]> wrote:
Is there some sort of TuscanySCA roadmap? I've looked around
a
bit and
haven't found one. I was curious what the future plans for
TuscanySCA CPP were in particular. I have a few ideas and I
was
curious if they had been contemplated yet.
- Move from Assembly Model 0.96 to 1.0
Definitely. We also need to move the CPP extension to the
1.0C++
C&I spec version
- Move to ant instead of make
I need to understand this proposal a little better. Can you
elaborate?
Probably worth starting a separate thread to discuss this.
I'm
all
for
simplifying the build though!
- Remove runtime dependancy on model data structure (slight
changes to
data/model shouldnt affect runtime usage)
ok
- Support additional WSDL bindings: RPC, DOC encoded...
sounds good.
--------------------
Brady Johnson
Lead Software Developer - HydraSCA Rogue Wave Software -
[EMAIL PROTECTED]
Cheers,
--
Pete
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
--
Pete
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
