Re: [gentoo-dev] eutils changes wrt EAPI-3 - ebeep and epause no longer available
Maciej Mrozowski wrote:
A as result of discussion http://www.mail-archive.com/gentoo-
d...@lists.gentoo.org/msg37300.html
ebeep and epause functions defined in eutils are not available in EAPI = 3.
For interactive installs, PROPERTIES=interactive should be used instead.
Maybe ebeep and epause should be defined in EAPI=3 but a qa warning so
things actually get fixed?
Something like this (not tested):
%% cvs di eutils.eclass
Index: eutils.eclass
===
RCS file: /var/cvsroot/gentoo-x86/eclass/eutils.eclass,v
retrieving revision 1.330
diff -u -r1.330 eutils.eclass
--- eutils.eclass 15 Feb 2010 02:10:39 - 1.330
+++ eutils.eclass 17 Feb 2010 14:13:16 -
@@ -50,6 +50,15 @@
done
fi
}
+else
+ ebeep() {
+ eqawarn ebeep is not defined in EAPI=3, please file a
bug at
+ http://bugs.gentoo.org;
+ }
+ epause() {
+ eqawarn epause is not defined in EAPI=3, please file a
bug at
+ http://bugs.gentoo.org;
+ }
fi
-Jeremy
[gentoo-dev] Re: eutils changes wrt EAPI-3 - ebeep and epause no longer available
* Jeremy Olexa darks...@gentoo.org:
Maciej Mrozowski wrote:
A as result of discussion http://www.mail-archive.com/gentoo-
d...@lists.gentoo.org/msg37300.html
ebeep and epause functions defined in eutils are not available in EAPI = 3.
For interactive installs, PROPERTIES=interactive should be used instead.
Maybe ebeep and epause should be defined in EAPI=3 but a qa warning
so things actually get fixed?
Something like this (not tested):
%% cvs di eutils.eclass
Index: eutils.eclass
===
RCS file: /var/cvsroot/gentoo-x86/eclass/eutils.eclass,v
retrieving revision 1.330
diff -u -r1.330 eutils.eclass
If you'd update your tree you can see that something like this was
committed.
http://sources.gentoo.org/viewcvs.py/gentoo-x86/eclass/eutils.eclass?r1=1.330r2=1.332
--- eutils.eclass 15 Feb 2010 02:10:39 - 1.330
+++ eutils.eclass 17 Feb 2010 14:13:16 -
@@ -50,6 +50,15 @@
done
fi
}
+else
+ ebeep() {
+ eqawarn ebeep is not defined in EAPI=3, please file
The problem here is that eqawarn isn't defined in EAPI 3.
[gentoo-dev] Lastrite: dev-util/aegis
# Samuli Suominen ssuomi...@gentoo.org (17 Feb 2010) # Doesn't compile with stable libc (#295980) # Vulnerable, CVE-2008-4938 (#245760) # Fails parallel make (#297334) # No maintainer # Masked for removal in 30 days dev-util/aegis
Re: [gentoo-dev] Re: eutils changes wrt EAPI-3 - ebeep and epause no longer available
On 17.2.2010 16.33, Torsten Veller wrote:
--- eutils.eclass 15 Feb 2010 02:10:39 - 1.330
+++ eutils.eclass 17 Feb 2010 14:13:16 -
@@ -50,6 +50,15 @@
done
fi
}
+else
+ ebeep() {
+ eqawarn ebeep is not defined in EAPI=3, please file
The problem here is that eqawarn isn't defined in EAPI 3.
Just shows that committing things to central eclasses without review is
a bad thing. I improved the code so that it doesn't at least call
eqawarn without first checking if it exists. Instead of code like this
in the eclasses, I think this should be done by Portage grepping logs. I
think it's already running searches over it for gcc things any way.
Regards,
Petteri
betelge...@pena /usr/portage/eclass $ cvs diff eutils.eclass
Index: eutils.eclass
===
RCS file: /var/cvsroot/gentoo-x86/eclass/eutils.eclass,v
retrieving revision 1.332
diff -u -r1.332 eutils.eclass
--- eutils.eclass 17 Feb 2010 02:22:43 - 1.332
+++ eutils.eclass 17 Feb 2010 17:10:00 -
@@ -54,11 +54,13 @@
else
ebeep() {
- eqawarn QA Notice: ebeep is not defined in EAPI=3, please file a bug
at http://bugs.gentoo.org;
+ [[ $(type -t eqawarn) == function ]] \
+ eqawarn QA Notice: ebeep is not defined in EAPI=3, please file
a bug at http://bugs.gentoo.org;
}
epause() {
- eqawarn QA Notice: epause is not defined in EAPI=3, please file a bug
at http://bugs.gentoo.org;
+ [[ $(type -t eqawarn) == function ]] \
+ eqawarn QA Notice: epause is not defined in EAPI=3, please
file a bug at http://bugs.gentoo.org;
}
fi
Re: [gentoo-dev] Re: eutils changes wrt EAPI-3 - ebeep and epause no longer available
On Wed, 17 Feb 2010 19:13:06 +0200, Petteri Räty betelge...@gentoo.org
wrote:
On 17.2.2010 16.33, Torsten Veller wrote:
--- eutils.eclass 15 Feb 2010 02:10:39 - 1.330
+++ eutils.eclass 17 Feb 2010 14:13:16 -
@@ -50,6 +50,15 @@
done
fi
}
+else
+ ebeep() {
+ eqawarn ebeep is not defined in EAPI=3, please file
The problem here is that eqawarn isn't defined in EAPI 3.
Just shows that committing things to central eclasses without review is
a bad thing. I improved the code so that it doesn't at least call
eqawarn without first checking if it exists. Instead of code like this
in the eclasses, I think this should be done by Portage grepping logs. I
think it's already running searches over it for gcc things any way.
What is going on with all these undocumented changes? When I look at the
council logs to see what is in EAPI3, I don't see anything about removing
functions. This is just silly and wastes alot of people's time for no
practical gain. In my EAPI3 portage, bin/isolated-functions.sh still has
eqawarn() defined. So, what am I missing now?
Also, other people think it is OK to change the behavior of functions and
not document it in devmanual?
Regards,
Petteri
Re: [gentoo-dev] Re: eutils changes wrt EAPI-3 - ebeep and epause no longer available
On Wed, 17 Feb 2010 18:03:42 + Jeremy Olexa darks...@gentoo.org wrote: What is going on with all these undocumented changes? When I look at the council logs to see what is in EAPI3, I don't see anything about removing functions. This is just silly and wastes alot of people's time for no practical gain. Being in eutils, none of this has anything to do with PMS or the Council. In my EAPI3 portage, bin/isolated-functions.sh still has eqawarn() defined. So, what am I missing now? eqawarn has never been defined in any EAPI. It's a Portage internal, not something for ebuilds to use. If you'd like it to be in an EAPI, you can propose it for EAPI 5. -- Ciaran McCreesh signature.asc Description: PGP signature
Re: [gentoo-dev] Re: eutils changes wrt EAPI-3 - ebeep and epause no longer available
On Wed, 17 Feb 2010 18:11:42 +, Ciaran McCreesh
ciaran.mccre...@googlemail.com wrote:
On Wed, 17 Feb 2010 18:03:42 +
Jeremy Olexa darks...@gentoo.org wrote:
What is going on with all these undocumented changes? When I look at
the council logs to see what is in EAPI3, I don't see anything about
removing functions. This is just silly and wastes alot of people's
time for no practical gain.
Being in eutils, none of this has anything to do with PMS or the
Council.
So, I was highlighting a problem. No one wants to write docs, which is
quite bad for the health of Gentoo.
In my EAPI3 portage, bin/isolated-functions.sh still has eqawarn()
defined. So, what am I missing now?
eqawarn has never been defined in any EAPI. It's a Portage internal,
not something for ebuilds to use. If you'd like it to be in an EAPI, you
can propose it for EAPI 5.
Yes, correct. For some reason, I thought it was valid to use eqawarn in
ebuilds. Now I guess a valid fix for eutils.eclass would be:
%% cvs di eutils.eclass
Index: eutils.eclass
===
RCS file: /var/cvsroot/gentoo-x86/eclass/eutils.eclass,v
retrieving revision 1.333
diff -u -r1.333 eutils.eclass
--- eutils.eclass 17 Feb 2010 17:10:23 - 1.333
+++ eutils.eclass 17 Feb 2010 18:20:18 -
@@ -54,13 +54,11 @@
else
ebeep() {
- [[ $(type -t eqawarn) == function ]] \
- eqawarn QA Notice: ebeep is not defined in EAPI=3, please
file a bug at http://bugs.gentoo.org;
+ einfo QA Notice: ebeep is not defined in EAPI=3, please
file a bug at http://bugs.gentoo.org;
}
epause() {
- [[ $(type -t eqawarn) == function ]] \
- eqawarn QA Notice: epause is not defined in EAPI=3,
please file a bug at http://bugs.gentoo.org;
+ einfo QA Notice: epause is not defined in EAPI=3, please
file a bug at http://bugs.gentoo.org;
}
fi
(where einfo is used because it doesn't need to be logged)
-Jeremy
Re: [gentoo-dev] Re: eutils changes wrt EAPI-3 - ebeep and epause no longer available
On 17.2.2010 20.03, Jeremy Olexa wrote: What is going on with all these undocumented changes? When I look at the council logs to see what is in EAPI3, I don't see anything about removing functions. This is just silly and wastes alot of people's time for no practical gain. In my EAPI3 portage, bin/isolated-functions.sh still has eqawarn() defined. So, what am I missing now? The canonical documentation for eclasses is eclass-manpages generated man pages and those where changed in my original commit. Also, other people think it is OK to change the behavior of functions and not document it in devmanual? devmanual should either have a pointer to eclass-manpages or autogenerate from eclass-manpages. The relevant bug is here: https://bugs.gentoo.org/show_bug.cgi?id=202656 Regards, Petteri
