> What's going on? Let's release sage-5.0.beta15 as sage-5.0!
Come ooooooooooon !! What don't you like about the beta ?? The are so
many of them -- a new one each week -- they are packed with totally
juicy stuff, and each time a new one is released you think "My GOD,
Sage-5.0 is going to be so *GREAT*".
You will be disappointed by Sage 5.0. What is so great about Sage 5.0
is that it "will be released soon, not now, but soon". Sage 5.0 is
great because all developpers are trying to write new patches quickly
hoping that the patches will make it in Sage 5.0. And each time a beta
is released you hope there will be a next one.
I think we should release Sage 4.9.0 and continue with this new
numbering. 4.9.0, 4.9.1, 4.9.2, .... Actually I think Jeroen and you
should make a secret pledge to *NEVER* release Sage 5.0 so that we all
expect the next beta to be Sage 5.0 and be hopefully surprised when we
find out it is just yet another beta.
This way we will always write patches like it is the last one we could
have merged before the release of Sage 5.0.
All hail William ! All hail Jeroen ! All hail Sage 5.0 !
> Come ooooooooooon !! What don't you like about the beta ?? The are so
> many of them -- a new one each week -- they are packed with totally
> juicy stuff, and each time a new one is released you think "My GOD,
> Sage-5.0 is going to be so *GREAT*".
Was this sarkasm? If yes: 8 of 15, yeah! ;-)
My best,
Alexander
-- Dr. rer. nat. Dipl.-Math. Alexander Dreyer
Abteilung "Systemanalyse, Prognose und Regelung"
Fraunhofer Institut f r Techno- und Wirtschaftsmathematik (ITWM)
Fraunhofer-Platz 1
67663 Kaiserslautern
Wow, what a lot of useless responses.
If I don't see a concrete plan to release sage-5.0 very soon, I'm just
going to release it. This is getting ridiculous.
> To post to this group, send email to sage-release@googlegroups.com.
> To unsubscribe from this group, send email to
> sage-release+unsubscribe@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/sage-release?hl=en.
-- William Stein
Professor of Mathematics
University of Washington
http://wstein.org
All I'm saying is, surely there's a reason Jeroen hasn't released beta 14 or beta 15 yet. If you want to immediately release Sage 5.0, which is supposed to be a *stable* version, at least consider picking something which has actually been tested to some extent, like beta 13.
On Thursday, April 26, 2012 1:16:15 PM UTC+8, William Stein wrote:
> Hi,
> Wow, what a lot of useless responses. > If I don't see a concrete plan to release sage-5.0 very soon, I'm just > going to release it. This is getting ridiculous.
> -- William
> On Wed, Apr 25, 2012 at 10:11 PM, Keshav Kini <keshav.k...@gmail.com> > wrote: > > Maybe we should wait for beta14 to be released before jumping to 15.
> > -Keshav
> > ---- > > Join us in #sagemath on irc.freenode.net !
> > On Thursday, April 26, 2012 2:08:19 AM UTC+8, William Stein wrote:
> >> Hi,
> >> What's going on? Let's release sage-5.0.beta15 as sage-5.0!
> >> -- William
> >> -- > >> William Stein > >> Professor of Mathematics > >> University of Washington > >> http://wstein.org
> > To post to this group, send email to sage-release@googlegroups.com. > > To unsubscribe from this group, send email to > > sage-release+unsubscribe@googlegroups.com. > > For more options, visit this group at > > http://groups.google.com/group/sage-release?hl=en.
> -- > William Stein > Professor of Mathematics > University of Washington > http://wstein.org
> Wow, what a lot of useless responses.
> If I don't see a concrete plan to release sage-5.0 very soon, I'm just
> going to release it. This is getting ridiculous.
I consider all these tickets important for sage-5.0, I think we really
should not release sage-5.0 without these. After working on something
carefully for months, you don't want to release a half-finished work.
That would be a real pity.
Here is my personal opinion, which I think is close to the opinion of the
Combinat team (or at least two member of the team: Nicolas and myself).
> > Wow, what a lot of useless responses.
> > If I don't see a concrete plan to release sage-5.0 very soon, I'm just
> > going to release it. This is getting ridiculous.
I fully support the idea of having a release (5.0 or 4.9) soon and in
particular before Sage Days 38 in one week. They are major change in this
release and keeping compatibility for 4.8 and 5.0beta* is a real pain for us.
> I consider all these tickets important for sage-5.0, I think we really
> should not release sage-5.0 without these. After working on something
> carefully for months, you don't want to release a half-finished work.
> That would be a real pity.
Clearly those blockers must be integrated. Having a bad 5.0 release would be
a bad advertising move. If we are not sure we still can have a 4.9. As a side
note they are also two positively reviewed critical ticket (#12718, #12849)
which probably should go (I consider the second one as a blocker).
I'm willing to help reviewing the remaining blockers but some review seems to
be harder than the other. I can do #12857 today (and #12272 if it become
needs-review).
Concerning the 4.7.0 GCC (meta ticket #12751) I don't have it myself and I
even less have a ia64 machine so I'm not sure I can help reviewing seriously
the blockers #12830, #12837. However, I was helping one of my student
yesterday installing its first Sage on a Arch linux which has GCC.4.7.0. He
manage to install 5.0 by replacing the following three package with their
version:
- LinBox #12762 (major)
- gfan 0.4plus #12760 (major)
- Givaro #12761 (major)
So I'm quite surprised to see that #12830, #12837 which concerns itanium
(which is relatively rare for casual users) are blockers whereas the x86 and
x86_64 are only major. I do realize of course that this doesn't help releasing
5.0 soon. but I don't want to support a 5.0 which won't compile on several
linux distros.
> So I'm quite surprised to see that #12830, #12837 which concerns itanium
> (which is relatively rare for casual users) are blockers whereas the x86 and
> x86_64 are only major.
This is because the GCC spkg solves the problems for all packages
*except* those which are prerequisites for GCC itself. That's why MPIR
and MPFR are blockers, but not Linbox for example.
The MPIR spkg has many more fixes than just this one, so I consider it
the single most important ticket for sage-5.0.
On Thu, Apr 26, 2012 at 11:29:29AM +0200, Jeroen Demeyer wrote:
> On 2012-04-26 11:04, Florent Hivert wrote:
> > So I'm quite surprised to see that #12830, #12837 which concerns itanium
> > (which is relatively rare for casual users) are blockers whereas the x86 and
> > x86_64 are only major.
> This is because the GCC spkg solves the problems for all packages
> *except* those which are prerequisites for GCC itself. That's why MPIR
> and MPFR are blockers, but not Linbox for example.
> The MPIR spkg has many more fixes than just this one, so I consider it
> the single most important ticket for sage-5.0.
That makes sense.
I'm done with #12857 (it remains to review a mostly trivial review patch).
Given that I've access to sage, boxen, an I7 x86_64 OpenSuSE 12.1 and a Xeon
amd64 Debian and an older 32bit laptop without Sage. Is there something I can
do to help with the review of #12830, #12837 and #11616 ?
Florent Hivert wrote:
> I fully support the idea of having a release (5.0 or 4.9) soon and in
> particular before Sage Days 38 in one week. They are major change in this
> release and keeping compatibility for 4.8 and 5.0beta* is a real pain for us.
It's not fully clear to me why you [have to] do this, but I take it as you say. (Having a more recent "official" version isn't bad. But still, if upgrading from devel versions would work again, we'd IMHO have less problems.)
>> I consider all these tickets important for sage-5.0, I think we really
>> should not release sage-5.0 without these. After working on something
>> carefully for months, you don't want to release a half-finished work.
>> That would be a real pity.
+1.
Although 5.0 is just a number, if you really want to release a so-called final /quickly/ [for the reasons given above], I'd do a 4.9, and put at least a few things more [which currently aren't marked blockers, partially not even critical] into the big one, 5.0.
Also depending on /when/ a 5.0 gets released, it should IMHO compile with GCC 4.7.x (and, at least partially, also Clang) -- without building the GCC spkg first, which is quite dumb in that case [unless one builds on Itanium, where GCC 4.7.0 does have a serious bug].
> Clearly those blockers must be integrated. Having a bad 5.0 release would be
> a bad advertising move.
> note they are also two positively reviewed critical ticket (#12718, #12849)
> which probably should go (I consider the second one as a blocker).
> I'm willing to help reviewing the remaining blockers but some review seems to
> be harder than the other. I can do #12857 today (and #12272 if it become
> needs-review).
> Concerning the 4.7.0 GCC (meta ticket #12751) I don't have it myself and I
> even less have a ia64 machine so I'm not sure I can help reviewing seriously
> the blockers #12830, #12837. However, I was helping one of my student
> yesterday installing its first Sage on a Arch linux which has GCC.4.7.0. He
> manage to install 5.0 by replacing the following three package with their
> version:
> - LinBox #12762 (major)
> - gfan 0.4plus #12760 (major)
> - Givaro #12761 (major)
> So I'm quite surprised to see that #12830, #12837 which concerns itanium
> (which is relatively rare for casual users) are blockers whereas the x86 and
> x86_64 are only major. I do realize of course that this doesn't help releasing
> 5.0 soon. but I don't want to support a 5.0 which won't compile on several
> linux distros.
Naturally, I cannot review any of the latter... ;-)
(But FWIW, I'll put some more work into the LinBox and GMP-ECM spkgs; actually there's already a new, preliminary ecm-6.3.p8 spkg for #11705 / segfault on Linux PPC64.)
There are a couple of further spkgs [also] fixing build issues, IMHO worth getting into a 5.0, still needing review...
-leif
-- () The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
May I remark the following: On both open blocker tickets, the actual reviewers want to give a positive review. Could it be that both are stale, because the reviews hesitate to contradict an oppinion expressed on the ticket?
With #12857 it is the easiest: One persons has doubts that the patch is necessary. He hasn't (yet) come up with a better idea. Also, I think Jeroen knows what he does. John Palmieri or Benjamin Jones, why don't you go ahead and give it a positive review?
With #11616 it's a bit more intrigued. There is one person, who is a bit uncontent with the way flags are obtained. These concernes have been adressed in parts by Jeroen. Volker Braun said he would give it a positive review, if the changes are commited to the spkg. He really knows what he does. So, please, please, go ahead and commit + give positive review.
I think many people will be thankful to you.
Martin
Am Donnerstag, 26. April 2012 11:04:12 UTC+2 schrieb hivert:
> Here is my personal opinion, which I think is close to the opinion of the > Combinat team (or at least two member of the team: Nicolas and myself).
> > > Wow, what a lot of useless responses. > > > If I don't see a concrete plan to release sage-5.0 very soon, I'm just > > > going to release it. This is getting ridiculous.
> I fully support the idea of having a release (5.0 or 4.9) soon and in > particular before Sage Days 38 in one week. They are major change in this > release and keeping compatibility for 4.8 and 5.0beta* is a real pain for > us.
> > Of course I can't stop you to release sage-5.0 but consider that several > > blocker tickets still need review, see
> > I consider all these tickets important for sage-5.0, I think we really > > should not release sage-5.0 without these. After working on something > > carefully for months, you don't want to release a half-finished work. > > That would be a real pity.
> Clearly those blockers must be integrated. Having a bad 5.0 release would > be > a bad advertising move. If we are not sure we still can have a 4.9. As a > side > note they are also two positively reviewed critical ticket (#12718, > #12849) > which probably should go (I consider the second one as a blocker).
> I'm willing to help reviewing the remaining blockers but some review seems > to > be harder than the other. I can do #12857 today (and #12272 if it become > needs-review).
> Concerning the 4.7.0 GCC (meta ticket #12751) I don't have it myself and I > even less have a ia64 machine so I'm not sure I can help reviewing > seriously > the blockers #12830, #12837. However, I was helping one of my student > yesterday installing its first Sage on a Arch linux which has GCC.4.7.0. > He > manage to install 5.0 by replacing the following three package with their > version: > - LinBox #12762 (major) > - gfan 0.4plus #12760 (major) > - Givaro #12761 (major) > So I'm quite surprised to see that #12830, #12837 which concerns itanium > (which is relatively rare for casual users) are blockers whereas the x86 > and > x86_64 are only major. I do realize of course that this doesn't help > releasing > 5.0 soon. but I don't want to support a 5.0 which won't compile on several > linux distros.
> With #12857 it is the easiest: One persons has doubts that the patch is > necessary. He hasn't (yet) come up with a better idea. Also, I think > Jeroen knows what he does. John Palmieri or Benjamin Jones, why don't you
I just did as I already proposed a review patch which was accepted.
On 26 April 2012 08:23, Jeroen Demeyer <jdeme...@cage.ugent.be> wrote:
> On 2012-04-26 07:16, William Stein wrote:
>> Hi,
>> Wow, what a lot of useless responses.
>> If I don't see a concrete plan to release sage-5.0 very soon, I'm just
>> going to release it. This is getting ridiculous.
> Of course I can't stop you to release sage-5.0 but consider that several
> blocker tickets still need review, see
> http://trac.sagemath.org/sage_trac/query?priority=blocker&status=need...
William should consider the very real possibility that if he makes a
5.0 release, despite the release manager is against the idea, he will
very likely lose a release manager. I can't speak for Jeroen, but I
know if I was in his position, I'd walk away from the job of release
manager if William made the release.
On Sat, Apr 28, 2012 at 8:13 PM, David Kirkby <david.kir...@onetel.net> wrote:
> On 26 April 2012 08:23, Jeroen Demeyer <jdeme...@cage.ugent.be> wrote:
>> On 2012-04-26 07:16, William Stein wrote:
>>> Hi,
>>> Wow, what a lot of useless responses.
>>> If I don't see a concrete plan to release sage-5.0 very soon, I'm just
>>> going to release it. This is getting ridiculous.
>> Of course I can't stop you to release sage-5.0 but consider that several
>> blocker tickets still need review, see
>> http://trac.sagemath.org/sage_trac/query?priority=blocker&status=need...
> William should consider the very real possibility that if he makes a
> 5.0 release, despite the release manager is against the idea, he will
> very likely lose a release manager. I can't speak for Jeroen, but I
> know if I was in his position, I'd walk away from the job of release
> manager if William made the release.
Thanks for encouraging the release manager to quit. That was a very
helpful way to support the Sage project. I really appreciate it.
-- William
> Dave
> Dave
> --
> You received this message because you are subscribed to the Google Groups "sage-release" group.
> To post to this group, send email to sage-release@googlegroups.com.
> To unsubscribe from this group, send email to sage-release+unsubscribe@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/sage-release?hl=en.
-- William Stein
Professor of Mathematics
University of Washington
http://wstein.org
On Sat, Apr 28, 2012 at 8:40 PM, William Stein <wst...@gmail.com> wrote:
> On Sat, Apr 28, 2012 at 8:13 PM, David Kirkby <david.kir...@onetel.net> wrote:
>> On 26 April 2012 08:23, Jeroen Demeyer <jdeme...@cage.ugent.be> wrote:
>>> On 2012-04-26 07:16, William Stein wrote:
>>>> Hi,
>>>> Wow, what a lot of useless responses.
>>>> If I don't see a concrete plan to release sage-5.0 very soon, I'm just
>>>> going to release it. This is getting ridiculous.
>>> Of course I can't stop you to release sage-5.0 but consider that several
>>> blocker tickets still need review, see
>>> http://trac.sagemath.org/sage_trac/query?priority=blocker&status=need...
>> William should consider the very real possibility that if he makes a
>> 5.0 release, despite the release manager is against the idea, he will
>> very likely lose a release manager. I can't speak for Jeroen, but I
>> know if I was in his position, I'd walk away from the job of release
>> manager if William made the release.
> Thanks for encouraging the release manager to quit. That was a very
> helpful way to support the Sage project. I really appreciate it.
I should also take some responsibility here. Jereon has wanted very
much to release sage-5.0 for a long time now (probably 2 months), and
*I* have been the one holding up the release... not the other way
around.
On 29 April 2012 02:40, William Stein <wst...@gmail.com> wrote:
> On Sat, Apr 28, 2012 at 8:13 PM, David Kirkby <david.kir...@onetel.net> wrote:
>> William should consider the very real possibility that if he makes a
>> 5.0 release, despite the release manager is against the idea, he will
>> very likely lose a release manager. I can't speak for Jeroen, but I
>> know if I was in his position, I'd walk away from the job of release
>> manager if William made the release.
> Thanks for encouraging the release manager to quit. That was a very
> helpful way to support the Sage project. I really appreciate it.
> -- William
I'm NOT encouraging Jeroen to quit. I'm pointing out to you that you
risk Jeroen quiting if you made the 5.0 release, without him agreeing to it.
So I'm encouraging you to not to make the release, which is what you
threatened to do.
> On Sat, Apr 28, 2012 at 8:40 PM, William Stein <wst...@gmail.com> wrote:
>> On Sat, Apr 28, 2012 at 8:13 PM, David Kirkby <david.kir...@onetel.net> wrote:
>>>>> Wow, what a lot of useless responses.
>>>>> If I don't see a concrete plan to release sage-5.0 very soon, I'm just
>>>>> going to release it. This is getting ridiculous.
>>> William should consider the very real possibility that if he makes a
>>> 5.0 release, despite the release manager is against the idea, he will
>>> very likely lose a release manager. I can't speak for Jeroen, but I
>>> know if I was in his position, I'd walk away from the job of release
>>> manager if William made the release.
>> Thanks for encouraging the release manager to quit. That was a very
>> helpful way to support the Sage project. I really appreciate it.
> I should also take some responsibility here. Jereon has wanted very
> much to release sage-5.0 for a long time now (probably 2 months), and
> *I* have been the one holding up the release... not the other way
> around.
> -- William
That's a different issue. It may have been acceptable to get it ready
release it 2 months ago, but with more patches gone in, it might not
be now. You and him should discuss it together, but comments like
"Wow, what a lot of useless responses. If I don't see a concrete plan
to release sage-5.0 very soon, I'm just going to release it. This is
getting ridiculous."
are very likely to aggravate a release manager. I can't speak for
Jereon, but I know it would annoy me if I was in his position and you
made the release when he felt it was not ready. I also know it would
annoy a lot of other people put in the same position.
On Sun, Apr 29, 2012 at 16:39, David Kirkby <david.kir...@onetel.net> wrote:
>> I should also take some responsibility here. Jereon has wanted very
>> much to release sage-5.0 for a long time now (probably 2 months), and
>> *I* have been the one holding up the release... not the other way
>> around.
>> -- William
> That's a different issue. It may have been acceptable to get it ready
> release it 2 months ago, but with more patches gone in, it might not
> be now. You and him should discuss it together, but comments like
> "Wow, what a lot of useless responses. If I don't see a concrete plan
> to release sage-5.0 very soon, I'm just going to release it. This is
> getting ridiculous."
> are very likely to aggravate a release manager. I can't speak for
> Jereon, but I know it would annoy me if I was in his position and you
> made the release when he felt it was not ready. I also know it would
> annoy a lot of other people put in the same position.
+1. I have no informed opinion on whether it is better to release 5.0
now or later, but this social dynamics point I can certainly
understand.
William: comparing the two comments of yours that David quoted above,
I fail to understand one point. If you were the person holding up the
release, and now want to release it immediately, then why did you post
this thread in sage-release? Why not just tell Jeroen personally "I no
longer oppose releasing 5.0 immediately"? If Jeroen has wanted very
much to release sage 5.0 [immediately] for a long time then why is
Jeroen now arguing against releasing 5.0 immediately? This doesn't
make sense to me.
(Also, minor point: his name is Jeroen, not "Jereon")
> I should also take some responsibility here. Jereon has wanted very
> much to release sage-5.0 for a long time now (probably 2 months), and
> *I* have been the one holding up the release... not the other way
> around.
I don't think it's anybody fault in particular that sage-5.0 is taking a
long time. I also don't think it's such a big deal. 4 months between
two releases is longer than usual but no disaster.
On Apr 29, 6:45 am, Jeroen Demeyer <jdeme...@cage.ugent.be> wrote:
> On 2012-04-29 03:51, William Stein wrote:
> > I should also take some responsibility here. Jereon has wanted very
> > much to release sage-5.0 for a long time now (probably 2 months), and
> > *I* have been the one holding up the release... not the other way
> > around.
> I don't think it's anybody fault in particular that sage-5.0 is taking a
> long time. I also don't think it's such a big deal. 4 months between
> two releases is longer than usual but no disaster.
+1 esp. if high quality, as jdemeyer puts a LOT of time in ensuring.
Just too bad that sagenb wasn't quite ready :(
