Peter,

>On Fri, 29 Mar 2002 21:55, Paul Hammant wrote:
>
>>We will have gone from a position of an easy build of a big jar (that
>>itself being an acceptance problem) to a difficult build of many small
>>jars.  I think the new MEME to be associated with Avalon would be "I
>>can't build it, I have no idea what it is or how useful it is".
>>
>
>Funny - I would have said the opposite. That used to be the old problem. 
>Given that I just tried to build ftpserver and it FAILED, then I tried 
>to build db and it FAILED, a few weeks ago I tried to build phoenix and it 
>passed but with CORRUPT jars, a week before that the same occured with 
>phoenix, and a week before that pheonix was also broken. 
>
>So exactly how is it easier to have these corruptions and failures?
>
With respect dude, I am not defending the previous position....

>I am TIRED of having to fix build problems that arise because people are 
>using jakarta-avalon/tools and those stupid build.* scripts. Look at all the 
>ugly hacks in the build files that are present to get around the failure of 
>that.
>
Agree.

>I have no problem putting some jars in CVS - especially ones of highly 
>"dynamic" projects. However things like xalan, xerces, ant, junit and so 
>forth should not be put in CVS because inevitably they will cause headaces 
>and incompatabiliteis - especially now that JDK1.4 includes xalan and crimson.
>
Agree.  

- Paul



--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to