FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Gentoo > Gentoo Development

 
 
LinkBack Thread Tools
 
Old 06-08-2008, 03:55 PM
"Santiago M. Mola"
 
Default die/QA notice for do* failure?

On Sun, Jun 8, 2008 at 4:48 PM, Arun Raghavan <arunisgod@gmail.com> wrote:
> Hello All,
> We were just discussing if it makes sense to either die or issue a QA
> notice if one of the do* functions fail. It turns out that there's
> already a bug for this [1]. This potentially applies to all helper
> functions that don't currently die on failure.
>

That's indeed a good idea, and we can do a step further.

We can define a new EAPI which makes all do* functions die on error,
and provide non-fatal versions as trydo* where it makes it sense.
emake could also die since we currently add die calls for all emake
occurences.

Regards,
--
Santiago M. Mola
Jabber ID: cooldwind@gmail.com
--
gentoo-dev@lists.gentoo.org mailing list
 
Old 06-08-2008, 03:58 PM
"Arun Raghavan"
 
Default die/QA notice for do* failure?

On Sun, Jun 8, 2008 at 8:57 PM, Ciaran McCreesh
<ciaran.mccreesh@googlemail.com> wrote:
[...]
>> I didn't understand you. Even if the external binary can't call die,
>> what's to prevent the caller from dying based on the return value of
>> the called binary?
>
> Then we're back to having people do dobin || die, which is precisely
> what we're trying to solve.

Not really. Can't dobin be like so:

fail() {
if hasq strict FEATURES; then
die "$@"
else
ewarn "QA Notice: ${@}. blah foo"
}

dobin() {
dobin.sh "${@}" || fail "dobin failed"
}

>> It should not be necessary to define a new EAPI to make sure packages
>> are not broken.
>
> Yes it should. It's a change in behaviour in functionality upon which
> quite a lot of things depend.

This is not functionality. It is the lack thereof. Making this part of
an EAPI makes it opt-in, which it shouldn't be. It is important for QA
and should be mandatory for all ebuilds.

Regards,
--
Arun Raghavan
(http://nemesis.accosted.net)
v2sw5Chw4+5ln4pr6$OFck2ma4+9u8w3+1!m?l7+9GSCKi056
e6+9i4b8/9HTAen4+5g4/8APa2Xs8r1/2p5-8 hackerkey.com
--
gentoo-dev@lists.gentoo.org mailing list
 
Old 06-08-2008, 04:51 PM
Ciaran McCreesh
 
Default die/QA notice for do* failure?

On Sun, 8 Jun 2008 21:28:00 +0530
"Arun Raghavan" <arunisgod@gmail.com> wrote:
> > Then we're back to having people do dobin || die, which is precisely
> > what we're trying to solve.
>
> Not really. Can't dobin be like so:
>
> fail() {
> if hasq strict FEATURES; then
> die "$@"
> else
> ewarn "QA Notice: ${@}. blah foo"
> }
>
> dobin() {
> dobin.sh "${@}" || fail "dobin failed"
> }

Like I said... A lot of these utilities have to work with xargs.

> >> It should not be necessary to define a new EAPI to make sure
> >> packages are not broken.
> >
> > Yes it should. It's a change in behaviour in functionality upon
> > which quite a lot of things depend.
>
> This is not functionality. It is the lack thereof. Making this part of
> an EAPI makes it opt-in, which it shouldn't be. It is important for QA
> and should be mandatory for all ebuilds.

That's not how it works. We've seen plenty of times in the past
that forcing QA by making users' systems break (which is how far these
things get before they're fixed) just leads to lots of annoyed users.
EAPI, plus slowly moving things towards new EAPIs on version bumps once
newer EAPIs are widely supported, is the clean way of doing this.

--
Ciaran McCreesh
 
Old 06-08-2008, 05:19 PM
"Arun Raghavan"
 
Default die/QA notice for do* failure?

On Sun, Jun 8, 2008 at 10:21 PM, Ciaran McCreesh
<ciaran.mccreesh@googlemail.com> wrote:
[...]
> That's not how it works. We've seen plenty of times in the past
> that forcing QA by making users' systems break (which is how far these
> things get before they're fixed) just leads to lots of annoyed users.
> EAPI, plus slowly moving things towards new EAPIs on version bumps once
> newer EAPIs are widely supported, is the clean way of doing this.

This might be the clean way to do it, but the unfortunate truth is
that new EAPIs seem to be becoming "standard" pretty darn slowly, and
counting on one to implement a feature that is definitely very useful
for QA seems to be miring ourselves in unnecessary bureaucracy.

Also, this does not have to cause breakage if done incrementally, so
the net loss is nil, and the net gain is getting a useful feature in a
relatively short, deterministic period of time rather than otherwise.

--
Arun Raghavan
(http://nemesis.accosted.net)
v2sw5Chw4+5ln4pr6$OFck2ma4+9u8w3+1!m?l7+9GSCKi056
e6+9i4b8/9HTAen4+5g4/8APa2Xs8r1/2p5-8 hackerkey.com
--
gentoo-dev@lists.gentoo.org mailing list
 
Old 06-08-2008, 05:34 PM
"Santiago M. Mola"
 
Default die/QA notice for do* failure?

On Sun, Jun 8, 2008 at 7:19 PM, Arun Raghavan <arunisgod@gmail.com> wrote:
> On Sun, Jun 8, 2008 at 10:21 PM, Ciaran McCreesh
> <ciaran.mccreesh@googlemail.com> wrote:
> [...]
>> That's not how it works. We've seen plenty of times in the past
>> that forcing QA by making users' systems break (which is how far these
>> things get before they're fixed) just leads to lots of annoyed users.
>> EAPI, plus slowly moving things towards new EAPIs on version bumps once
>> newer EAPIs are widely supported, is the clean way of doing this.
>
> This might be the clean way to do it, but the unfortunate truth is
> that new EAPIs seem to be becoming "standard" pretty darn slowly, and
> counting on one to implement a feature that is definitely very useful
> for QA seems to be miring ourselves in unnecessary bureaucracy.

(Replying to a random snippet)

There has been previous discussion on
https://bugs.gentoo.org/show_bug.cgi?id=138792

Regards,
--
Santiago M. Mola
Jabber ID: cooldwind@gmail.com
--
gentoo-dev@lists.gentoo.org mailing list
 
Old 06-08-2008, 05:45 PM
"Arun Raghavan"
 
Default die/QA notice for do* failure?

On Sun, Jun 8, 2008 at 11:04 PM, Santiago M. Mola <coldwind@gentoo.org> wrote:
[...]
> There has been previous discussion on
> https://bugs.gentoo.org/show_bug.cgi?id=138792

Right, I'd mentioned that in the original post. I posted here in order
to get some general consensus on the matter and take it forward.

Cheers,
--
Arun Raghavan
(http://nemesis.accosted.net)
v2sw5Chw4+5ln4pr6$OFck2ma4+9u8w3+1!m?l7+9GSCKi056
e6+9i4b8/9HTAen4+5g4/8APa2Xs8r1/2p5-8 hackerkey.com
--
gentoo-dev@lists.gentoo.org mailing list
 
Old 06-09-2008, 04:26 AM
Ciaran McCreesh
 
Default die/QA notice for do* failure?

On Sun, 8 Jun 2008 22:49:18 +0530
"Arun Raghavan" <arunisgod@gmail.com> wrote:
> This might be the clean way to do it, but the unfortunate truth is
> that new EAPIs seem to be becoming "standard" pretty darn slowly, and
> counting on one to implement a feature that is definitely very useful
> for QA seems to be miring ourselves in unnecessary bureaucracy.
>
> Also, this does not have to cause breakage if done incrementally, so
> the net loss is nil, and the net gain is getting a useful feature in a
> relatively short, deterministic period of time rather than otherwise.

Why don't you ask the Portage people to implement what you're after as
EAPI 2 next week?

EAPI in no way slows down progress -- quite the opposite, since it
allows changes to go through, safely, straight away. What slows down
progress is that the Portage people don't care enough to implement most
things, no matter how trivial and useful they are.

--
Ciaran McCreesh
 
Old 06-09-2008, 08:49 PM
Enrico Weigelt
 
Default die/QA notice for do* failure?

* Arun Raghavan <arunisgod@gmail.com> schrieb:

Hi,

> We were just discussing if it makes sense to either die or issue a QA
> notice if one of the do* functions fail. It turns out that there's
> already a bug for this [1]. This potentially applies to all helper
> functions that don't currently die on failure.

I'd like to see it this way: introduce a new FEATURE (eg. "hard-qa"),
which causes QA warnings to fail. Then evryone's free to enable it.
Maintainers always should run their tests with hard-qa, unless there's
are *really* good reason.

IMHO, this wouldn't harm any users who don't insist in it.


cu
--
---------------------------------------------------------------------
Enrico Weigelt == metux IT service - http://www.metux.de/
---------------------------------------------------------------------
Please visit the OpenSource QM Taskforce:
http://wiki.metux.de/public/OpenSource_QM_Taskforce
Patches / Fixes for a lot dozens of packages in dozens of versions:
http://patches.metux.de/
---------------------------------------------------------------------
--
gentoo-dev@lists.gentoo.org mailing list
 

Thread Tools




All times are GMT. The time now is 12:13 PM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org