Vassun,

thanks for the detailed answer.

>>This of course mean that all tools which use patchad and relied on 
>>undocumented exit codes will need to be now modified. So transition period is 
>>nessesary for this tools and this is why -t option, which means "transition" 
>>was introduced. In general it skips all C-dependancy check and just pass 
>>parameters to patchadd.ksh .
>>
>>We should remove -t option in time.

Not a good solution at all. I'm pretty sure that we are not the only company 
using  patchadd and relying on the return code of patchadd (due to the various 
awful and erroneous patch tools provided by Sun until now for Solaris 9 and 
older)

Anyway, how long do you think (!) will this transition period be?
(Because we are forced to find another solution for our patch process in this 
time frame)

We need a solution so that we can install patches via patchadd in a script and 
so that we can decide in the script if an error is harmless (like the 
returncodes 2 and 8 of the  "old" patchadd) or not.

regards

Bernd
 
 
This message posted from opensolaris.org

Reply via email to