Hi All,
This is a very profitable issue, I couldn't miss it....
First of all, the issue in question was assigned to me in
the end of
march, shame on me, I apologize!!!
By that time I manage to study the ANT <exec> task and
we had discussed
some of it in the list to, but since may we haven't talked
about it
since then. The account.properties file was the seconda task
assigned to
me, and this one I failed completely, althoug by your
discussion, I
think I wouldn't have contributed much anyway.... this time
I realy
apologize!!!! Sorry....
I have read some MSN documentation and also downloaded some
MSN
libraries, those Emil sugested. I also have looked the ICQ
implementation, in order to plan the MSN protocol provider
service
implementation.
At the moment I thinking how would I test it, once I my
opinion it is
very important know how to test something before start
developing it.
That is way I became so glad when I read this topic!!!!
Still, I have
some doubts, not in the testing process, which I might have
understood.
Here come the questions (ICQ protocol provide implementation
questions
will come in a different email.):
1) The package net.java.sip.communicator.slick.protocol.icq
on the test
folder became one test bundle, right?
2) What is the test agent??
3) What is the Slick Fixture???
4) Why use the System.getProperty() method and when was il
set???
5) Is it possible to use the debbuger with the service
installed in
OSCAR??? How???
I better whait for those answers, they certainly will clear
my thoughts
and my understanding!!
Thanks and Best regards!!!!
Pedro
Brian Burch wrote:
>On Mon, 19 Jun 2006 19:10:31 +0200, Emil Ivov wrote:
>
>
>>There are two ways to go here:
>>
>>1) During the first phase, when the icq slick is
being loaded, we detect
>>that the accounts.properties file is not there. We
log an error and we
>>bravely continue so that the icq slick and all its
tests get registered.
>>Then we make sure that when the tests are being run
they fail with the
>>appropriate error and wiki reference (or at least
those that require
>>online access).
>>
>>or
>>
>>2) During the first phase, when the icq slick is
being loaded, we detect
>> that the accounts.properties file is not there. At
this point we
>>completely replace the set of tests that we register
for later execution
>>with a single test that always fails with the
appropriate message and
>>wiki reference.
>>
>>I personally believe that 1 is somewhat more logical
but I don't mind 2)
>>either.
>>
>>What do you think?
>>
>>
>
>Thanks for taking the time to explain, Emil. You have
saved me from taking a small innocent step and then being
surprised by not getting the result I expected. Your
description of the structure of the tests is very
>clear and I'm sure will be helpful to others (as well
as me).
>
>At the moment, I am as unsure as you about which is the
"best" of your two approaches. They both seem a
bit sort-of-right and a bit sort-of-wrong, but in different
ways. Given your warning, I prefer to spend more time
>tracing and understanding the flow before I answer you.
In the end, I'll favour an approach that is "most
harmonious" with the existing code, but until I've
understood the alternatives, I won't jump to conclusions.
>
>Since posting my optimistic suggestion, a couple of
urgent problems have turned up with my (paid) work. However,
It looks like I'm not having to fight people off to make
this change, so I'll try to do some research
>later this week. I still won't make a change until
I've discussed it further with you - unless one of the
options turns out to be obviously better.
>Regards,
>
>Brian
>
>
>--------------------------------------------------------
-------------
>To unsubscribe, e-mail: dev-unsubscribe sip-communicator.dev.java.net
>For additional commands, e-mail: dev-help sip-communicator.dev.java.net
>
>
>
>
------------------------------------------------------------
---------
To unsubscribe, e-mail: dev-unsubscribe sip-communicator.dev.java.net
For additional commands, e-mail: dev-help sip-communicator.dev.java.net
|