Bug 389184

Summary:	isSatisfied() (or isRelevant()) reports differently with and without setOnlyRequires(true)
Product:	[openSUSE] openSUSE 11.0	Reporter:	Jan Kupec <jkupec>
Component:	libzypp	Assignee:	Stefan Schubert <schubi>
Status:	RESOLVED DUPLICATE	QA Contact:	Duncan Mac-Vicar <dmacvicar>
Severity:	Normal
Priority:	P5 - None	CC:	mls
Version:	Factory
Target Milestone:	---
Hardware:	Other
OS:	Other
Whiteboard:
Found By:	---	Services Priority:
Business Priority:		Blocker:	---
Marketing QA Status:	---	IT Deployment:	---
Attachments:	zypper.log zypper up --no-recommends --debug-solver

Description Jan Kupec 2008-05-11 22:10:21 UTC

Created attachment 214251 [details]
zypper.log

$ zypper up --no-recommends
Reading installed packages...
Nothing to do.
$ zypper up
Reading installed packages...

The following NEW patches are going to be installed:
  kernel-12 kmps-11


No additional space will be used or freed after the operation.
Continue? [YES/no]:


The same environment as in bug #389182. Zypper marks the patches for installation if (patch->isRelevant() && !patch->isSatisfied()). So one of the two methods must return something else if setOnlyRequires(true) is used, which should AFAICS not happen (whether a solvable's (hard) dependencies are satisfied or not should not depend on whether user wishes to dragg the weak depenencies in or not).

Comment 1 Jan Kupec 2008-05-11 22:28:28 UTC

Created attachment 214253 [details]
zypper up --no-recommends --debug-solver

zypper --debug-solver needed a fix to set 'only-requires' before generating the test case (r10041)

Comment 2 Stefan Schubert 2008-05-15 17:07:23 UTC

"zypper --debug-solver needed a fix to set 'only-requires' before generating the
test case (r10041)"
And the testcase should document it too (like force-resolve)

Comment 3 Jan Kupec 2008-05-20 17:34:27 UTC

(In reply to comment #2 from Stefan Schubert)
> And the testcase should document it too (like force-resolve)
you mean to log it or save it in the xmls? it is being logged now

Comment 4 Stefan Schubert 2008-05-21 09:40:19 UTC

Yes, I have added it to the testcases. The rest is duplicate.

*** This bug has been marked as a duplicate of bug 389182 ***

Comment 5 Jan Kupec 2008-05-21 12:49:55 UTC

Is it really a duplicate? bug 389182 is about plain 'zypper up'. This one is about the difference if --no-recommends is used and not used. Or is it dup. of bug 389694 ??

Comment 6 Jan Kupec 2008-05-21 12:50:45 UTC

(In reply to comment #5 from Ján Kupec)
> Is it really a duplicate? bug 389182 is about plain 'zypper up'. This one is
> about the difference if --no-recommends is used and not used. Or is it dup. of
> bug 389694 ??

I mean, does bug 389694 apply also to patches?

Comment 7 Stefan Schubert 2008-05-21 15:08:40 UTC

comment #5
I think yes cause it is the same behaviour here:
"The following NEW patches are going to be installed:
  kernel-12 kmps-11"

comment #6
I think no, this bug has to do with recommends/suggestions.
Our bug has to to with broken patches