From: "Michał Górny" <mgorny@gentoo.org>
To: gentoo-dev@lists.gentoo.org
Subject: Re: [gentoo-dev] [RFC] Forced/automatic USE flag constraints (codename: ENFORCED_USE)
Date: Fri, 02 Jun 2017 15:49:51 +0200 [thread overview]
Message-ID: <1496411391.29233.3.camel@gentoo.org> (raw)
In-Reply-To: <20170602131806.6efc7b0e@gentoo.org>
[-- Attachment #1: Type: text/plain, Size: 4666 bytes --]
On pią, 2017-06-02 at 13:18 +0200, Alexis Ballier wrote:
> On Fri, 02 Jun 2017 08:37:30 +0200
> Michał Górny <mgorny@gentoo.org> wrote:
>
> > On czw, 2017-06-01 at 23:31 +0200, Michał Górny wrote:
> > > On czw, 2017-06-01 at 10:55 +0200, Alexis Ballier wrote:
> > > > On Wed, 31 May 2017 21:02:24 +0200
> > > > Michał Górny <mgorny@gentoo.org> wrote:
> > > >
> > > > > On śro, 2017-05-31 at 19:39 +0200, Alexis Ballier wrote:
> > > > > > > > Again, you *need* to process the constraints in order.
> > > > > > > > '!a? ( b ) !b? ( a )' is not deterministic when none of a
> > > > > > > > and b are enabled otherwise.
> > > > > > >
> > > > > > > You can't rely on any particular order of constraints,
> > > > > > > especially when eclass stacking comes into play. You could
> > > > > > > try defining some constraint- sorting algorithm but it's
> > > > > > > going to be complex and it's usefulness will be limited
> > > > > > > anyway due to various kinds of grouping.
> > > > > >
> > > > > >
> > > > > > Better have some order: If half of the above constraint comes
> > > > > > from ebuild and the other half comes from eclass, then PM
> > > > > > might toggle 'a' or 'b' depending on the phase of the moon
> > > > > > which is specifically what we're trying to avoid.
> > > > >
> > > > > No, it can't. That's the whole point. The algorithm must be
> > > > > defined so that it is always predictable independently of order
> > > > > (maybe except the ordering inside ^^, ||, ??) and independently
> > > > > of how it's nested (i.e. 'a? ( b? ( c ) )' must give the same
> > > > > result as 'b? ( a? ( c ) )').
> > > >
> > > > This is a lot of handwaving without real description of how it
> > > > would work. This starts to look a lot like confluence in
> > > > rewriting systems and you want developers to write only confluent
> > > > rules and repoman to decide that. Good luck with that.
> > >
> > > I'm sorry, what I said was quite stupid. I did some thinking and
> > > testing today, and the algorithm can be actually quite trivial. The
> > > real issue is getting a correct input, that is REQUIRED_USE
> > > constraints that have only one solution.
>
> Don't be sorry, it's not stupid actually. There are various ways to
> determinize how the constraints should be solved. Once that is defined,
> one just needs to apply the rules to the constraints to get a
> temptative solution, making the algorithm trivial, indeed. Depending on
> how you do actually determinize the whole thing can make ensuring
> uniqueness (or even the existence) of a solution quite a pain. That's
> where the trade-offs between ebuild writers wishes and solving the issue
> come into play. I see this long thread as a quite interesting
> discussion where we're trying to converge towards something that leaves
> the problem easily solvable without asking too much to ebuild writers.
>
> [...]
> > > The biggest issue for me right now is to find a way to determine
> > > whether a particular REQUIRED_USE constraint can have only one
> > > solution, independently of the order of non-n-ary constraints.
> > > However, some of it can be easily eyeball-checked using a graph.
> > >
> >
> > Well, I've looked at it more and you were right. While keeping them
> > order-independent is a noble goal, it's not really feasible,
> > especially if we assume that n-ary constraints can be reordered. So I
> > think we can assume left-to-right ordering now (with multiple passes,
> > if necessary).
>
>
> There are ways to guarantee order-independence, your gtk+ answer in the
> previous email suggests one: At most one constraint solving rule can be
> applied in any situation; that's what you did by putting the ||
> under !xinerama. This is a quite extreme way to ensure confluence but
> that works. I've not thought too much about it but I think we'd lose
> too much in expressivity though.
For the xinerama rule, you may prefer the opposite rule, i.e.:
!X? ( !xinerama )
If you did so, I can't think of a way to express it correctly while
preserving full allowance of reshifting ||. So yes, we'd lose too much.
One thing we lose is the ability to clearly check whether the constraint
'makes sense'. Previously, we could (at least theoretically) check
whether it has a single solution and complain otherwise. With left-to-
right, I don't see any way to tell the developer 'something stupid is
going to happen now, you probably meant something else'. But I guess
that's still an improvement over what we have now.
--
Best regards,
Michał Górny
[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 988 bytes --]
next prev parent reply other threads:[~2017-06-02 13:50 UTC|newest]
Thread overview: 111+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-05-29 15:33 [gentoo-dev] [RFC] Forced/automatic USE flag constraints (codename: ENFORCED_USE) Michał Górny
2017-05-29 16:30 ` Kent Fredric
2017-05-29 16:44 ` Michał Górny
2017-05-29 18:00 ` Alexis Ballier
2017-05-29 21:23 ` Michał Górny
2017-05-29 21:31 ` Ciaran McCreesh
2017-05-29 22:01 ` Ulrich Mueller
2017-05-29 22:05 ` Ciaran McCreesh
2017-05-30 7:47 ` Alexis Ballier
2017-05-30 8:05 ` Ulrich Mueller
2017-05-30 8:10 ` Alexis Ballier
2017-05-30 7:42 ` Alexis Ballier
2017-05-30 8:22 ` Ciaran McCreesh
2017-05-30 8:46 ` Alexis Ballier
2017-05-30 8:56 ` Ciaran McCreesh
2017-05-30 9:25 ` Alexis Ballier
2017-05-30 12:00 ` Ulrich Mueller
2017-05-30 14:33 ` Michał Górny
2017-05-30 15:33 ` Alexis Ballier
2017-05-30 18:11 ` Michał Górny
2017-05-30 18:46 ` Alexis Ballier
2017-05-31 6:55 ` Michał Górny
2017-05-31 7:24 ` Ciaran McCreesh
2017-05-31 7:34 ` Alexis Ballier
2017-05-31 7:35 ` Michał Górny
2017-05-31 7:51 ` Ciaran McCreesh
2017-05-31 7:54 ` Alexis Ballier
2017-05-31 7:56 ` Alexis Ballier
2017-05-31 7:32 ` Alexis Ballier
2017-05-31 8:03 ` Michał Górny
2017-05-31 8:38 ` Alexis Ballier
2017-05-31 13:04 ` Michał Górny
2017-05-31 17:39 ` Alexis Ballier
2017-05-31 19:02 ` Michał Górny
2017-05-31 22:52 ` Ciaran McCreesh
2017-06-01 8:55 ` Alexis Ballier
2017-06-01 21:31 ` Michał Górny
2017-06-02 6:37 ` Michał Górny
2017-06-02 11:18 ` Alexis Ballier
2017-06-02 13:49 ` Michał Górny [this message]
2017-06-02 11:27 ` Alexis Ballier
2017-06-02 13:55 ` Michał Górny
2017-06-02 15:09 ` [gentoo-dev] " Martin Vaeth
2017-06-03 11:00 ` [gentoo-dev] " Alexis Ballier
2017-06-03 15:33 ` Michał Górny
2017-06-03 16:58 ` Alexis Ballier
2017-06-04 8:59 ` Alexis Ballier
2017-06-05 7:55 ` Alexis Ballier
2017-06-05 14:10 ` Michał Górny
2017-06-05 17:24 ` Alexis Ballier
2017-06-05 18:10 ` Michał Górny
2017-06-05 18:15 ` Ciaran McCreesh
2017-06-06 12:08 ` Alexis Ballier
2017-06-06 17:39 ` Michał Górny
2017-06-07 6:49 ` Michał Górny
2017-06-07 8:17 ` Alexis Ballier
2017-06-07 9:27 ` Michał Górny
2017-06-07 9:56 ` Alexis Ballier
2017-06-09 9:19 ` Michał Górny
2017-06-09 11:41 ` Alexis Ballier
2017-06-09 12:54 ` Michał Górny
2017-06-09 14:16 ` Alexis Ballier
2017-06-09 16:21 ` Michał Górny
2017-06-11 16:05 ` Alexis Ballier
2017-06-12 9:08 ` Alexis Ballier
2017-06-12 19:17 ` Michał Górny
2017-06-13 10:27 ` Alexis Ballier
2017-06-13 22:13 ` Michał Górny
2017-06-14 9:06 ` Alexis Ballier
2017-06-14 12:24 ` Michał Górny
2017-06-14 13:16 ` Alexis Ballier
2017-06-14 13:57 ` Michał Górny
2017-06-14 14:09 ` Alexis Ballier
2017-06-15 15:59 ` Michał Górny
2017-06-15 16:07 ` Alexis Ballier
2017-06-15 16:13 ` Ciaran McCreesh
2017-06-15 16:19 ` Alexis Ballier
2017-06-15 16:22 ` Ciaran McCreesh
2017-06-15 16:30 ` Alexis Ballier
2017-06-15 16:32 ` Ciaran McCreesh
2017-06-15 16:37 ` Alexis Ballier
2017-06-15 16:45 ` Ciaran McCreesh
2017-06-15 16:55 ` Alexis Ballier
2017-06-15 17:04 ` Ciaran McCreesh
2017-06-15 17:30 ` Alexis Ballier
2017-06-15 17:48 ` Ciaran McCreesh
2017-06-15 18:09 ` Alexis Ballier
2017-06-15 17:38 ` Michał Górny
2017-06-15 18:05 ` Alexis Ballier
2017-06-14 14:28 ` Alexis Ballier
2017-06-02 12:16 ` Alexis Ballier
2017-06-02 13:57 ` Michał Górny
2017-06-02 14:56 ` [gentoo-dev] " Martin Vaeth
2017-06-02 15:19 ` Ciaran McCreesh
2017-06-02 16:26 ` Martin Vaeth
2017-06-02 18:31 ` Martin Vaeth
2017-06-02 1:17 ` [gentoo-dev] " A. Wilcox
2017-06-02 1:28 ` Rich Freeman
2017-06-02 1:33 ` A. Wilcox
2017-06-02 5:08 ` Michał Górny
2017-05-31 12:38 ` [gentoo-dev] " Duncan
2017-05-30 21:13 ` [gentoo-dev] " Kent Fredric
2017-05-30 8:29 ` Michał Górny
2017-05-30 9:34 ` Alexis Ballier
2017-05-30 14:12 ` Michał Górny
2017-05-29 19:24 ` Ciaran McCreesh
2017-05-29 19:42 ` Michał Górny
2017-05-29 19:44 ` Ciaran McCreesh
2017-06-05 8:26 ` Alexis Ballier
2017-06-09 12:35 ` Jason A. Donenfeld
2017-06-09 12:42 ` Michał Górny
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1496411391.29233.3.camel@gentoo.org \
--to=mgorny@gentoo.org \
--cc=gentoo-dev@lists.gentoo.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox