Re: [gentoo-dev] RFC: 0-day bump requests

2008-07-03 Thread Torsten Rehn
On Friday 04 July 2008, Tony Chainsaw Vroon wrote:
 How about a metadata.xml tag that indicates whether early bump requests are
 welcome?

People obviously don't care about what it says on the website, why should they 
start looking into metadata.xml?

I think we should remove the useless restrictions on filing bump requests and 
welcome users to open bugs. Closing (valid) bugs feels good and is also sort 
of a psychological reward for the user who opened the bug, perhaps 
encouraging them to stay in direct contact with Gentoo and contribute other, 
less trivial work.

-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


[gentoo-dev] [RFC] Bugzilla enhancements wrt AT work

2008-03-15 Thread Torsten Rehn
Two weeks ago, dirtyepic suggested making some modifications to how ATs and 
developers interact using Bugzilla [1].

+jakub scel: basically... instead of KEYWORDREQ/STABLEREQ
+jakub create keywording and stabilization components
+jakub and use flags accordingly there
+jakub bugzilla already has the features, why not use them
+jakub also, nuke things like TESTED and STABLE

This is trivial to implement and would greatly improve the ability to search 
for {STABLE,KEYWORD}REQs that have been tested by ATs. All ATs I've spoken to 
(most of the active ones at amd64) would really appreciate this change.

Discuss.

Also notice bug 213514 [2] for actual progress on this.

[1] 
http://archives.gentoo.org/gentoo-dev/msg_a5495659df35010ae44c266dd785ae4e.xml
[2] http://bugs.gentoo.org/213514

-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] [RFC] Bugzilla enhancements wrt AT work

2008-03-15 Thread Torsten Rehn
On Saturday 15 March 2008, Rémi Cardona wrote:
 Just curious, who would be the default assignee for those 2 new components?

I don't think anything should be changed here. Unpriviledged users 
automatically assign to bug-wranglers, everyone else goes for the target 
packages' maintainer.
Do you see any problem with this?

-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] Keyword request interface (SoC candidate?)

2008-02-28 Thread Torsten Rehn
On Thursday 28 February 2008, Santiago M. Mola wrote:
 A lot of users don't feel comfortable using Bugzilla and often are
 lost with our procedures for keyword (both ~ and stable) requests. I
 think we could use an easy web interface for requesting specific
 keywords for packages in a point-and-click fashion.

I have been working on something like this and would like to continue doing so 
for SoC (see [1]). However, it is a fairly large project and I would 
appreciate some input on what specific goals to target for SoC.
Also, I anticipate some resistance, so I'd suggest seeing that project as a 
long term experiment starting with SoC, that may eventually produce something 
that satisfies most of the developer staff.

[1] http://scel.info/blog/posts/google-soc-proposal/

-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] [RFC] Deprecating an eclass

2008-02-18 Thread Torsten Rehn
On Monday 18 February 2008, Doug Klima wrote:
 Potentially doing something like:
 DEPRECIATED=$DEPRECATED $ECLASS
 at the top of each deprecated eclass.

Adding deprecation info directly into the eclass file feels wrong to me. 
(Eclasses are free software after all and can be reused - ok, nobody will 
ever do that, but we're talking theory here - so we shouldn't put Gentoo 
policy in there.)

What about /usr/portage/profiles/deprecated_eclasses looking like

old_eclass_1,old_eclass_2:new_eclass_1,new_eclass_2

indicating that the old_eclasses have been deprecated by the new_eclasses. 
Having multiple eclasses deprecated by multiple eclasses may not be that 
common, but this kind of syntax allows for some grouping of related eclasses 
being replaced together.
-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] public system of ratings

2008-02-16 Thread Torsten Rehn
On Saturday 16 February 2008, Oleg Puchinin wrote:
  is not compiled At all 
 works

All stable ebuilds in the tree should work, if they don't that's an exception 
and will probably get fixed soon after the failure is discovered anyway.
Anyone who is willing to test a new keyword in an ebuild shouldn't have a 
problem filing a KEYWORDREQ in Bugzilla.

 It is excellent

If you want package ratings, take a look at ohloh.net or something, shouldn't 
be hard (for a human) to map the package names there to the ones in Gentoo.

 I wish to see a package in Gentoo 

Again, all the most important software is already in the tree, if you need 
something else, filing a bug is not too much effort.

-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] How to get more involved?

2008-01-30 Thread Torsten Rehn
On Thursday 31 January 2008, Mateusz Mierzwinski wrote:
 What should I do to make ebuilds and upload it to main portage mirror?

Take a look at the Gentoo docs on writing[1] and contributing[2] ebuilds. If 
you've got any further questions, try the gentoo-dev-help mailing list or the 
#gentoo-dev-help irc channel. Please note that this list is intended to be 
used for discussion of Gentoo development, not ebuild development.

[1] http://www.gentoo.org/proj/en/devrel/handbook/handbook.xml?part=2chap=1
[2] http://www.gentoo.org/doc/en/ebuild-submit.xml

-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] New eclass: emul-linux-x86.eclass

2007-11-14 Thread Torsten Rehn
On Wednesday 14 November 2007, Mike Doty wrote:
 S=${WORKDIR}

Shouldn't ${WORKDIR} be quoted here, too?

Kind regards

-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] New eclass: emul-linux-x86.eclass

2007-11-14 Thread Torsten Rehn
On Wednesday 14 November 2007 14:21, you wrote:
 But why is it standard to quote other assignments like in DESCRIPTION and
 HOMEPAGE then?

When assigning literal values as in DESCRIPTION and HOMEPAGE, you have to 
quote. ${WORKDIR} is quoted on its assignment and therefore does not have to 
be quoted when assigning its value to another variable (${S}) by reference.

[EMAIL PROTECTED]  cat test.sh
#!/bin/bash
TEXT=A B
Q=${TEXT}
echo ${TEXT}
FAIL=DONT TRYTHISATHOME
[EMAIL PROTECTED]  ./test.sh
A B
./test.sh: line 5: TRYTHISATHOME: command not found

-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] New eclass: emul-linux-x86.eclass

2007-11-14 Thread Torsten Rehn
On Wednesday 14 November 2007, Ferris McCormick wrote:
 Note, however, that for example in /usr/lib/portage/bin/ebuild.sh, it's
 always quoted (except once, which looks like an oversight).

I was talking about referencing variables when declaring other variables. 
Whenever you actually use a variable in-line (e.g. as a parameter to a 
command), you have to quote as well of course.
i.e.
A=${B}
and not
cd ${B}

The difference is that bash handles variable assignments all interally and is 
smart enough to handle ${B} properly, while vars that are passed to some 
command are just evaluated and passed to the command in question and cannot 
be protected anymore by bash. AFAIK.

-- 
Torsten Rehn [EMAIL PROTECTED]
Gentoo AMD64 Arch Tester
http://scel.info


signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-dev] Nero-3.0.0.0 license needs RESTRICT=fetch ?

2007-07-06 Thread Torsten Rehn
On Friday 06 July 2007 19:56, Harald van Dijk wrote:
 And what if they decide they don't accept the license on the first run?
 They'll already have installed the software, which requires acceptance
 of the license.

No violation here: [...] IF YOU DO NOT AGREE TO THE TERMS OF THIS
AGREEMENT, PROMPTLY UNINSTALL AND DELETE THE SOFTWARE [...]

cheers


pgpGSz0kzU2Mk.pgp
Description: PGP signature