On Thu, Aug 27, 2009 at 03:53:36PM -0700, Philip Brown wrote: > My reason in bringing this up again, in this forum, is to ask the IPS team > how they plan to make IPS magically fix this sort of problem, when it is > caused by the sun packaging/patch teams' PROCESSES, not lack of technology.
One supposes that patching technology could be layered above IPS in bad ways, yes. I certainly hope that won't happen. I imagine that the IPS team must have a good handle on this. > patching 121975-01 FAILS. Because it depends on 121975-01. > patch 121975-01 fails, because SUNWdtdte is not installed. Patch clusters result in an aggregate set of dependencies that doesn't necessarily follow from the pkg dependencies. This is a problem with patching, not packaging. > But we dont WANT SUNWdtdte installed! That includes stuff like dtlogin, and > all kinds of other cruft that dont belong on our servers! Package dependencies sometimes are incorrect. Other times they are merely unfortunate (why did X need a dependency on Y? why couldn't the developer use Z instead? ...). Now, if IPS is used correctly for patching, then this particular problem should disappear because there should be no such thing as a patch cluster where artificial dependencies get created. Unfortunate and incorrect pkg dependencies may still arise; the latter are easy to fix, once one has determined that the dependency was incorrect, while the former is harder. > That fairly tightly fits my own personal definition of > "broken patch creation policies/procedures". Or, rather, technology. Patch clusters happen because one patch at some point grows a dependency on another patch which also patches things that are unrelated to the first and which might be optional. There's nothing wrong with that. The problem is that patch clustering can make optional contents required for no good reason. That's a technology problem. Of course, it's also possible that new dependencies really do happen on patch. > So, how is switching to the "better technology" of IPS, going to solve this > problem of broken patch creation process inside of Sun? Mostly by resetting patching technology. But also by providing a base packaging technology that makes some things easier/default. Technically though, Solaris patching could have used features of SVR4 packaging that are rarely used, and the result could have been better. (I'm thinking of package updates, which require setting MAXINST=1 and admin(4) conflict=overwrite -- update should have been a first class operation, rather than a round-about feature.) Nico -- _______________________________________________ pkg-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
