pkgsrc-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Why are depends rebuilt on "bmake update" ?



Greg Troxel <gdt%ir.bbn.com@localhost> writes:

> Aleksej Saushev <asau%inbox.ru@localhost> writes:
>
>> Greg Troxel <gdt%ir.bbn.com@localhost> writes:
>>
>>> Joerg Sonnenberger <joerg%britannica.bec.de@localhost> writes:
>>>
>>>> On Wed, Oct 24, 2012 at 01:14:12PM +0200, Dario Niedermann wrote:
>>>>> When I run 'bmake update' all the dependencies of the package I'm
>>>>> updating are rebuilt, even if there aren't newer versions available.
>>>>> Is there a way to avoid this?
>>>>
>>>> You are thinking of "bmake replace". It's essentially doing what your
>>>> recipe is doing.
>>>
>>> Also, see sysutils/pkg_rolling-replace, which is a way to recover from
>>> the inconsistencies that make replace can cause.
>>
>> I had experience of exactly the opposite in past.
>
> I can't parse your sentence.    In all the reports of problems with
> pkg_rr, they have boiled down to one of:
>
>   some package didn't build when doing make replace (this has nothing to
>   do with make replace or pkg_rr, but is commonly reported as such).
>
>   a package won't build/install because of something else installs, and
>   if one had the data about what to do and could remove 1 package and
>   install 2, or something like that, it would have worked.
>
>   the wrong package to replace was guessed at (typically about python
>   versions, or something else about packages with complicated names)
>
> Are you having some other problem?

I tried to recover from inconsistent "make replace" with pkg_rolling-replace
several times and always ended in a state with even more inconsistencies.
This led to rebuilding of all packages at least two times, and at least one
time I could get out with pkg_chk. In other cases I performed manual updates
to recover. This makes me believe that pkg_rolling-replace is not safe 
mechanism.

I still don't understand why it is recommended to recover from inconsistencies.
The way it functions (utilizing unsafe operations) doesn't make impression of
doing it correctly either.


-- 
HE CE3OH...



Home | Main Index | Thread Index | Old Index