On 7/3/01 8:10 PM, "Jon Stevens" <[EMAIL PROTECTED]> wrote:
> on 7/3/01 5:04 PM, "Jason van Zyl" <[EMAIL PROTECTED]> wrote:
>
>> I try to oblige, but I think that is wholly an unreasonable
>> request. You as a project have decided to bind yourselves
>> to HEAD and Turbine's build system. I say that's your problem
>> entirely. I myself am trying to accomdate you but I think you
>> have a very unorthodox build system because you require
>> bleeding edge code and I don't think people who make
>> changes to improve the build system should be chastised
>> because it breaks scarab. You are tying yourself to
>> a train and all you're wearing is sneakers.
>>
>> You are developing against HEAD so you have to help us out.
>> It is really hard to experiment and try new things out
>> because scarab updates turbine everyday. I am as frustrated
>> as you are, believe me. But lets try to help each other. I think
>> what results will end up being better for everyone.
>>
>> I have held off making a big commit because I want to have
>> a patch ready for scarab. We are trying, be patient.
>
> The thing is that it is easy to test against Scarab's build system. It takes
> all of 20 seconds and it also gives you a really good idea of what you are
> breaking when you make changes.
It takes more than 20 seconds when HEAD changes.
> This goes back to the whole dependency tracking thing...we need to be really
> conscience of this...
I know exactly what has changed between releases, exactly. On a daily
basis it is much harder and in most cases meaningless because the
experiments done in HEAD may or may not persist.
> because Scarab is going against head gives us an early
> warning system just like Gump provides us...without it, we would be stuck
> with not knowing what we are breaking for other people.
This is not true actually. HEAD would probably break your code on a daily
basis, when 2.2 is done and all the adapter code is in and we release a beta
that will be the real test. Scarab keeping up with HEAD doesn't do
much for compatibility testing really because you are no longer a 2.1
app. I could radically alter HEAD and keep scarab up to date, at this
point it doesn't mean a whole lot in terms of using Scarab as a test
for the majority of our users: users of 2.1.
Testing with Scarab on a daily basis as a test of 2.2 I think is a great
idea. It is a very good test to see that altered code is in fact performing:
scarab is a much better 'real life' test than the TDK, which I hope will
one day be the ultimate test. Scarab is trying to stay in tune with HEAD,
so I think as a common courtesy testing with Scarab is the right thing
to do. All I was saying is that people making changes to help improve
our 'lot of turbine land' often don't have as much time to kill as I do :-)
Some mistakes will definitely be made. We'll figure it out.
> -jon
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
--
jvz.
Jason van Zyl
http://tambora.zenplex.org
http://jakarta.apache.org/turbine
http://jakarta.apache.org/velocity
http://jakarta.apache.org/alexandria
http://jakarta.apache.org/commons
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
