Debian package...
Florent Hivert
Disclaimer: I'm not a debian user and my intend is not to launch a flame nor
to disregard the hard work that has been done to have a sage debian package.
However, during sage days 20 as well as during my course at the university of
Rouen, I've got at least a dozen reports of people trying to install sage with
the standard "dpkg -i". Everything, looks fine except that this sage seems to
be broken. Maxima simply does not start (just try x+1). I'm quite concerned
that debian is a quite wide spread distro, and that for all these guys the
image of sage is something huge that simply doesn't work. I was very angry
when I heard this very argument from a colleague and two students. If
confirmed, couldn't we make an official request to debian that this package is
removed from their repositories. This non working sage is a very bad
publicity...
Cheers,
Florent
Burcin Erocal
+10
See also:
http://groups.google.com/group/sage-support/t/1f055a381532b667#3542567a19abd5ea
Cheers,
Burcin
Jason Grout
We might make a PPA for Ubuntu. If someone is interested in an easy way
to install Sage via dpkg, that might be the best option at this point.
I agree that removing sage 3.0.5 (or whatever version it is) from Debian
is probably best, since our first piece of advice to anyone is to
uninstall it and install Sage from scratch.
Thanks,
Jason
William Stein
Yes, +1 to removing sage 3.0.5 from Debian.
But how do we make that happen?
-- William
Dima Pasechnik
thing, how it is done...
Dmitrii
On Mar 5, 11:05 am, William Stein <wst...@gmail.com> wrote:
> On Thu, Mar 4, 2010 at 5:03 PM, Jason Grout <jason-s...@creativetrax.com> wrote:
> > On 03/04/2010 11:14 AM, Burcin Erocal wrote:
>
> >> On Thu, 4 Mar 2010 18:03:47 +0100
> >> Florent Hivert<florent.hiv...@univ-rouen.fr> wrote:
>
> >>> Hi there,
>
> >>> Disclaimer: I'm not a debian user and my intend is not to launch a
> >>> flame nor to disregard the hard work that has been done to have a
> >>> sage debian package.
>
> >>> However, during sage days 20 as well as during my course at the
> >>> university of Rouen, I've got at least a dozen reports of people
> >>> trying to install sage with the standard "dpkg -i". Everything, looks
> >>> fine except that this sage seems to be broken. Maxima simply does not
> >>> start (just try x+1). I'm quite concerned that debian is a quite wide
> >>> spread distro, and that for all these guys the image of sage is
> >>> something huge that simply doesn't work. I was very angry when I
> >>> heard this very argument from a colleague and two students. If
> >>> confirmed, couldn't we make an official request to debian that this
> >>> package is removed from their repositories. This non working sage is
> >>> a very bad publicity...
>
> >> +10
>
> >> See also:
>
> >>http://groups.google.com/group/sage-support/t/1f055a381532b667#354256...
Florent Hivert
> > Disclaimer: I'm not a debian user and my intend is not to launch a
> > flame nor to disregard the hard work that has been done to have a
> > sage debian package.
> >
> > However, during sage days 20 as well as during my course at the
> > university of Rouen, I've got at least a dozen reports of people
> > trying to install sage with the standard "dpkg -i". Everything, looks
> > fine except that this sage seems to be broken. Maxima simply does not
> > start (just try x+1). I'm quite concerned that debian is a quite wide
> > spread distro, and that for all these guys the image of sage is
> > something huge that simply doesn't work. I was very angry when I
> > heard this very argument from a colleague and two students. If
> > confirmed, couldn't we make an official request to debian that this
> > package is removed from their repositories. This non working sage is
> > a very bad publicity...
>
> +10
If you think this is of any use and if I have an official mandate from the
sage community, I'm willing to ping again the maintainer to push those things
forward. This is very irritating and deserve sage a lot.
Cheers,
Florent
Dima Pasechnik
---------- Forwarded message ----------
From: Bill Allombert <Bill.Al...@math.u-bordeaux1.fr>
Date: 5 March 2010 18:49
Subject: Re: Fwd: Debian package...
To: Dima Pasechnik <dim...@gmail.com>
On Thu, Mar 04, 2010 at 08:28:53PM -0800, Dima Pasechnik wrote:
> Dear Bill,
>
> do you know how to handle this?
> We would like remove Sage from debian, as it is horribly
> outdated there, and there is no work being done on fixing
> the debian distribution.
Hello Dima, Sage is not really in Debian, since the "sagemath" Debian package
is only in the 'unstable' distibution an not in 'testing', 'squeeze' or
'stable', 'lenny': This is the list of sagemath in Debian:
sagemath | 3.0.5dfsg-5.1 | unstable | source, amd64
sagemath | 3.0.5dfsg-5.1+b1 | unstable | hppa, i386, ia64,
powerpc, s390, sparc
There are 3 critical bugs reported on the package, so no further action is
required, though you can report a bug with severity grave asking it to
be removed because it is outdated.
However I have some comment on William email below:
> ---------- Forwarded message ----------
> From: William Stein <wst...@gmail.com>
> To: sage-devel
>
> >>> Disclaimer: I'm not a debian user and my intend is not to launch a
> >>> flame nor to disregard the hard work that has been done to have a
> >>> sage debian package.
>
> >>> However, during sage days 20 as well as during my course at the
> >>> university of Rouen, I've got at least a dozen reports of people
> >>> trying to install sage with the standard "dpkg -i". Everything, looks
> >>> fine except that this sage seems to be broken. Maxima simply does not
> >>> start (just try x+1). I'm quite concerned that debian is a quite wide
> >>> spread distro, and that for all these guys the image of sage is
> >>> something huge that simply doesn't work. I was very angry when I
> >>> heard this very argument from a colleague and two students. If
> >>> confirmed, couldn't we make an official request to debian that this
> >>> package is removed from their repositories. This non working sage is
> >>> a very bad publicity...
If the users really used 'dpkg -i' to install the package, then much probably
it was not the official Debian package, because such package are normally
installed by high-level tools like 'apt-get', 'aptitude' or 'synaptics', that
will take care of downloading and installing the numberous dependencies.
'dpkg -i' does not and is likely to fail.
If they actually used 'dpkg -i' they probably downloadied some unofficial
.deb file from some web site (maybe even sagemath.org) and used
dpkg -i *.deb to install it.
> >> See also:
>
> >>http://groups.google.com/group/sage-support/t/1f055a381532b667#354256...
This email mentions problem with Ubuntu. Removing the package from Ubuntu
is a totally different issue from removing it from Debian.
> > We might make a PPA for Ubuntu. If someone is interested in an easy way to
> > install Sage via dpkg, that might be the best option at this point.
>
> > I agree that removing sage 3.0.5 (or whatever version it is) from Debian is
> > probably best, since our first piece of advice to anyone is to uninstall it
> > and install Sage from scratch.
>
> Yes, +1 to removing sage 3.0.5 from Debian.
>
> But how do we make that happen?
Cheers,
Bill
--
Dmitrii Pasechnik
-----
DISCLAIMER: Any text following this sentence does not constitute a
part of this message, and was added automatically during transmission.
William Stein
> here is the response from one of Debian folks
Bill says:
>> >>http://groups.google.com/group/sage-support/t/1f055a381532b667#354256...
>
> This email mentions problem with Ubuntu. Removing the package from Ubuntu
> is a totally different issue from removing it from Debian.
Thanks Bill for explaining everything, which I greatly appreciate.
So the new question (for the list): how do we get Sage out of Ubuntu?
-- William
>
>> > We might make a PPA for Ubuntu. If someone is interested in an easy way to
>> > install Sage via dpkg, that might be the best option at this point.
>>
>> > I agree that removing sage 3.0.5 (or whatever version it is) from Debian is
>> > probably best, since our first piece of advice to anyone is to uninstall it
>> > and install Sage from scratch.
>>
>> Yes, +1 to removing sage 3.0.5 from Debian.
>>
>> But how do we make that happen?
>
> Cheers,
> Bill
>
>
>
> --
> Dmitrii Pasechnik
> -----
> DISCLAIMER: Any text following this sentence does not constitute a
> part of this message, and was added automatically during transmission.
>
> --
> To post to this group, send an email to sage-...@googlegroups.com
> To unsubscribe from this group, send an email to sage-devel+...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/sage-devel
> URL: http://www.sagemath.org
>
--
William Stein
Associate Professor of Mathematics
University of Washington
http://wstein.org
Dr David Kirkby
On 5 Mar, 15:27, William Stein <wst...@gmail.com> wrote:
> On Fri, Mar 5, 2010 at 7:09 AM, Dima Pasechnik <dimp...@gmail.com> wrote:
> > here is the response from one of Debian folks
>
> Bill says:
>
> >> >>http://groups.google.com/group/sage-support/t/1f055a381532b667#354256...
>
> > This email mentions problem with Ubuntu. Removing the package from Ubuntu
> > is a totally different issue from removing it from Debian.
>
> Thanks Bill for explaining everything, which I greatly appreciate.
>
> So the new question (for the list): how do we get Sage out of Ubuntu?
>
> -- William
I don't know, but the obvious way to stop this in future is to issue a
suitable warning when the date on the computer exceeds the Sage
release data by some period of time. I suggest two levels of warnings
- perhaps one after 4 months, and a stronger one after a year. This
is
http://trac.sagemath.org/sage_trac/ticket/8447
It will at least stop this being such an issue in future, as the user
will be reminded each time they start Sage that they are using an old
version.
Dave
Kasper Peeters
Tim Abbott <tab...@mit.edu>
and ask him whether he would be interested in handing over
maintainership to someone with more time to bring the debian package
up to date?
I would be happy to help out with this (including contacting Abbott),
but it will clearly require quite a bit of work because neither Debian
nor Ubuntu like packages which duplicate software already in the
repositories.
Removing sage from Debian/Ubuntu is probably not a good idea, since
those repositories are what those users expect to get their software
from. Usage of my cadabra CAS went up dramatically once it got into
the Ubuntu repositories, even though I had binaries available for
download before that. So it would help a lot to have an up-to-date
sage in those repositories too.
Cheers,
Kasper
Robert Bradshaw
Clearly we want up-to-date packages in the repositories, but the
current situation of a really old, broken Sage is much worse than not
being there at all.
- Robert
Ben Goodrich
think his primary difficulty is keeping up with all the dependencies
of sage, especially when sage releases with a patched version of a
dependency that has not made it upstream yet. Debian package
maintainers are unlikely to quickly apply such patches for the
unstable or testing branch, although they might for the experimental
branch.
Earlier there was some discussion of creating an environmental
variable that would attempt to build sage with system versions of the
libraries and other dependencies, rather than the versions shipped
with sage. Did anything come of that?
Ben
François Bissey
There are things to carefully consider.
Time as always is the biggest obstacle to anything like that.
Francois
Ben Goodrich
> > Earlier there was some discussion of creating an environmental
> > variable that would attempt to build sage with system versions of the
> > libraries and other dependencies, rather than the versions shipped
> > with sage. Did anything come of that?
>
> Nothing came out of that, but it would be useful to us packagers.
> There are things to carefully consider.
> Time as always is the biggest obstacle to anything like that.
>
> Francois
+1 to more consideration, although I don't have much to contribute.
One thing I do think is that if sage is packaged for distributions,
the target audience for those binaries is probably not the people who
are doing cutting-edge mathematical research. So if a dependency has a
bug, certainly the patch should try to go upstream ASAP, but a lot of
people in the target audience wouldn't be affected by that bug, even
though for some people the bug breaks a needed feature or possibly
produces wrong results. If people are doing really important stuff,
then they can get the latest sage from sage's website.
Ben
Harald Schilly
wrote:
> I would be happy to help out with this ...
I just want to add that there is a package for mandriva.
Look here at the list of required packages:
http://fr2.rpmfind.net/linux/RPM/mandriva/devel/2010.1/i586/media/contrib/release/sagemath-4.3.3-2mdv2010.1.i586.html
Since all their modifications are online, it might be useful to look
at their patches how they have done it.
H
Dr. David Kirkby
One way the maintainers might accept the huge Sage bundle is if we could produce
a big list of changes made to standard packages. Even libz has been patched for
OS X. (There is a new beta which will stop that being necessary).
The maintainers logic is clear they don't want to duplicate stuff. I can
appreciate that. I suggest we approach them, saying we understand this, and that
in general it would be silly to include everything. If we then produce a long
list of packages which have needed to be patched, then it is less likely they
will object.
There are several Solaris-specific patches. This could be used to our advantage
by saying that Sage is multi-platform, and some patches are needed for Solaris.
Maintaining two separate versions of the source code for two different platforms
would present us severe difficulties.
I suspect the Debian people are reasonable and could be persuaded to accept
things if there were aware of just how many patches have needed to be made to
'standard' packages.
One package I think we would have a lot of problem justifying is the inclusion
of 'Mercurial'. Whether Mercurial is a perquisite for Sage or not is debatable,
but including its source code seems unnecessary to me. If someone is going to be
submitting patches based on Mercurial, they are probably quite capable of
installing it themselves.
I personally prefer to use a system wide install, as I can then apply patches
without having built Sage.
Ben Goodrich
wrote:
> I suspect the Debian people are reasonable and could be persuaded to accept
> things if there were aware of just how many patches have needed to be made to
> 'standard' packages.
They are reasonable. My guess is they would usually email upstream to
ask about a patch and if upstream agrees but isn't planning to release
for a while, then the maintainers might well apply it, particularly in
experimental. But all that would take time and may have to be
coordinated across several package maintainers. That is why I agree
with you about having an option to build sage with reasonably up-to-
date versions of system libraries and dependencies. I think that would
probably be the biggest thing that could be done to make life easier
on Tim and the other package maintainers.
Also, with Debian in particular, when they freeze the testing branch
to start preparing it for the stable release, there is going to be
extra reluctance to take new patches, often even for the unstable
branch. The maintainers just devote all their energy toward
stabilizing and other things get backlogged. Being Debian, freezes
have been known to last a long time.
Ben
Pierre
"stable" release. If so, it would be brilliant to include this one
with ubuntu, and then not produce another ubuntu package before the
next LTS. The custom with such repositories, even more so with debian,
is certainly not to get the latest version at all cost (which has been
pointed out already a couple of times on this list).
pierre
Kasper Peeters
> appreciate that. I suggest we approach them, saying we understand this, and that
> in general it would be silly to include everything. If we then produce a long
> list of packages which have needed to be patched, then it is less likely they
> will object.
Giving that even the Sage developers probably want those patches to
eventually be incorporated upstream (or you would have to keep
maintaining all those spkg's forever), it is indeed a very good start
to just make a web page with all the patches that Sage requires. That
will also make it much easier to judge how hard it is to package for
debian/ubuntu.
Cheers,
Kasper
Dima Pasechnik
On Mar 6, 7:27 am, "Dr. David Kirkby" <david.kir...@onetel.net> wrote:
> Kasper Peeters wrote:
> > Has anyone considered emailing the official maintainer
>
> > Tim Abbott <tabb...@mit.edu>
well, this is trickier than you think.
E.g. Python 2.6 has not made it into Debian stable yet.
And installing Python 2.6 on Debian stable using the standard Debian
source package
installation mechanisms does not work.
Python is needed for functioning of many other system components on
Debian, meaning
that you just cannot just have a system-wide Python 2.6 on a Debian
stable...
Dmitrii
Ben Goodrich
> well, this is trickier than you think.
> E.g. Python 2.6 has not made it into Debian stable yet.
> And installing Python 2.6 on Debian stable using the standard Debian
> source package
> installation mechanisms does not work.
> Python is needed for functioning of many other system components on
> Debian, meaning
> that you just cannot just have a system-wide Python 2.6 on a Debian
> stable...
>
> Dmitrii
That is true and if I understand correctly Python 2.6 is not even
going to be available in the *next* stable release. But neither would
Debian allow the sagemath package to embed a different version of
Python (or anything else) in a stable release. So having a useful
sagemath in the Debian stable repository seems like an impossible goal
for now. At the moment, we are stuck with sagemath 3.0.5 in unstable
and that is on the verge of being asked to be removed. So, I wish that
something could be done to facilitate getting a more recent version of
sage into unstable or experimental. That would also be the relevant
consideration for Ubuntu releases.
Ben
ma...@mendelu.cz
On 4 bře, 18:03, Florent Hivert <florent.hiv...@univ-rouen.fr> wrote:
> Hi there,
>
> Disclaimer: I'm not a debian user and my intend is not to launch a flame nor
> to disregard the hard work that has been done to have a sage debian package.
>
Hi all,
I have three machines on Debian (notebook, PC, server). Compilation
from sources takes hours, but it is trivial! I compile on one computer
and move everything to the others.
I switched to Debian from ArchLinux, since I appretiate stability. I
upgrade Sage about each four weeks and Debian about once in four
months. I think that this is a typical for Debian users.
I think, it does not have too much sense to have Debian package. It
would be better to put compiled Debian binaries to sagemath.org
download page.
Cheers,
Robert Marik
Ben Goodrich
> I think, it does not have too much sense to have Debian package. It
> would be better to put compiled Debian binaries to sagemath.org
> download page.
>
> Cheers,
>
> Robert Marik
I agree that compiling sage is not too difficult, and I actually use
one of the Ubuntu binaries on a Debian machine, which seems to work
fine. But is your argument specific to Debian / Ubuntu? It seems to me
that either it is worthwhile to try to package sage for popular
distros or it is better to tell everyone to get the appropriate binary
from the sage website only. I can see arguments both ways, but I agree
that it is not good to have a broken sagemath in the Debian repository
without a realistic possibility of being able to package a newer
version.
Ben
memilanuk
Another vote here for getting an old and/or broken version *out* of
Debian unstable/experimental. Seems kind of ridiculous to have a
version that old in what is commonly viewed as the 'cutting edge'
branch.
Just tossing out ideas here... if its too much of a PITA to keep sage
meshed with the politics and policies of Debian, which in turn affects
Ubuntu... would it be possible or feasible to distribute sage for
linux as a virtual box image, similar to what is done for Windows, so
it can run independently of the system libraries and such? Granted it
is even more of a pig of a download, but it might side-step some other
issues in the mean time... and people wanting a more light-weight
version can still compile and install locally.
Just a thought...
Monte
William Stein
> Hello,
>
> Another vote here for getting an old and/or broken version *out* of
> Debian unstable/experimental. Seems kind of ridiculous to have a
> version that old in what is commonly viewed as the 'cutting edge'
> branch.
>
> Just tossing out ideas here... if its too much of a PITA to keep sage
> meshed with the politics and policies of Debian, which in turn affects
It's not clear if it is PITA or not. As far as I can tell, nobody
lifted a finger to work on the Debian/Ubuntu packaging of Sage during
the last 6 months (or more). Nobody is working on it.
> Ubuntu... would it be possible or feasible to distribute sage for
> linux as a virtual box image, similar to what is done for Windows, so
> it can run independently of the system libraries and such? Granted it
> is even more of a pig of a download, but it might side-step some other
> issues in the mean time... and people wanting a more light-weight
> version can still compile and install locally.
One can already use the Sage VirtualBox binary (that we already
distribute "for windows") on Linux, OS X, and Solaris (x86), if one
wants to.
-- William
>
> Just a thought...
>
> Monte
Robert Bradshaw
As William mentioned, the "Windows" VM works fine under other
platforms as well. In fact, the notebook itself is served out of a VM
running on Linux, and we do much of our testing in such VMs as well.
However, the reason Sage can't be just dropped into a a Debian package
is because it already runs independently of the system libraries. (We
ship and build our own Python, zlib, blas, gmp, etc.) About the only
thing the binaries depend on is the System's clib, stdc++, and fortran
(and it's even been suggested to ship those as well, though the
general consensus is that's not a good idea).
- Robert
ma...@mendelu.cz
>
> It's not clear if it is PITA or not. As far as I can tell, nobody
> lifted a finger to work on the Debian/Ubuntu packaging of Sage during
> the last 6 months (or more). Nobody is working on it.
>
Just some ideas. I wonder if the following is possible
Replace sage package in Debian and Ubuntu repositories by a simple
script which does the following:
1. checks if sage is installed, if no, downloads the latest Debian
binary and unpacks (say in /opt/Sage-4.3.3 and makes ln -s /opt/sage /
opt/sage-4.3.3)
2. if yes, checks, that sage binary is not 6 months old
3. runs sage
Robert Marik
Dima Pasechnik
This might actually be the way to go, more or less --- make sage
debian package a downloader.
Debian actually has such packages, e.g., something that downloads free
fonts for OO...
Dima
Harald Schilly
On Mar 7, 9:28 am, Dima Pasechnik <dimp...@gmail.com> wrote:
> +1.
> This might actually be the way to go, more or less --- make sage
> debian package a downloader.
I've coded that some time ago, using aria2c for efficient downloading.
It's quite simple. I just wasn't able to pack this together into
a .deb file because I wasn't able to figure out what to do with the
scripts inside and it always failed. Otherwise we would already have
one ;)
H
Dima Pasechnik
you can have a look at
http://packages.debian.org/lenny/ttf-mscorefonts-installer
for a similar functionality in a debian package.
Dima
Harald Schilly
> ... but I agree
> that it is not good to have a broken sagemath in the Debian repository
> without a realistic possibility of being able to package a newer
> version.
Small follow up to this thread:
https://answers.launchpad.net/ubuntu/+source/sagemath/+question/104036
There are distros that package sagemath, i.e. mandriva. So, it's
possible and it happens right now - but just not with debian!
H
Dima Pasechnik
(i.e. in /usr/local, built from scratch, possibly with more stuff
needed to be built out-of-tree)
So from Debian point of view this means that such a package is only
feasible in the testing distro, where it's possible to have python2.6
installed Debian way.
Dmitrii
On Mar 6, 5:57 pm, Kasper Peeters <kasper.peet...@googlemail.com>
wrote:
Tim Abbott
> > We might make a PPA for Ubuntu. If someone is interested in an easy way to
> > install Sage via dpkg, that might be the best option at this point.
> >
> > I agree that removing sage 3.0.5 (or whatever version it is) from Debian is
> > probably best, since our first piece of advice to anyone is to uninstall it
> > and install Sage from scratch.
> >
>
> Yes, +1 to removing sage 3.0.5 from Debian.
>
> But how do we make that happen?
Sorry for the slow reply, I've been quite busy of late.
I suspect that your problem is actually that the Sage package migrated to
Ubuntu Karmic, which is a release lots of people use. Debian unstable has
essentially no users in comparison.
It's possible that moving the Debian package from unstable to experimental
would help (completely removing it from Debian will mean that if someone
decides to update the Debian Sage package, we'll need to go through the
year-long copyright review process again. That review process is what
killed the Debian sagemath package -- Upgrading during the review process
sends you to the back of the queue, and by a year after my original
submission, I had left MIT graduate school to start a startup, and I no
longer had the time to upgrade past a year of Sage development). But the
Debian unstable package is not important -- we should focus on the 50x
larger problem of Karmic.
As I mentioned on the bug report opened today at
<http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=573538>, I'm talking to
a member of the Ubuntu stable release update team who I know to see what
the procedure would be for getting the package removed from Karmic.
I'll keep this thread updated as that progresses.
(Please note I am not directly subscribed to sage-devel, so you will need
to directly CC me or debian-sage@ if you want me to see your replies in
real time).
-Tim Abbott
Tim Abbott
> As I mentioned on the bug report opened today at
> <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=573538>, I'm talking to
> a member of the Ubuntu stable release update team who I know to see what
> the procedure would be for getting the package removed from Karmic.
The answer I got on this is that there is no procedure for removing a
package from a stable release -- Ubuntu doesn't do that kind of thing.
However, they did say that given the circumstances, it might be possible
to push through a major update to the package to Karmic when once one is
available.
So I think this leaves us at "we need to update the package". Sigh.
-Tim Abbott
Dima Pasechnik
binary installation of sage.
Even Debian has packages of this type.
Dmitrii
Alex Ghitza
Hi Tim,
Thanks for looking into this!
On Sat, 13 Mar 2010 20:02:18 -0500 (EST), Tim Abbott <tab...@MIT.EDU> wrote:
> The answer I got on this is that there is no procedure for removing a
> package from a stable release -- Ubuntu doesn't do that kind of thing.
That's pretty hilarious.
> However, they did say that given the circumstances, it might be possible
> to push through a major update to the package to Karmic when once one is
> available.
>
> So I think this leaves us at "we need to update the package". Sigh.
Maybe we should just take whatever version of Sage is in Ubuntu right
now and patch it so that running sage simply exits with something like
**********************************************************************
THIS VERSION OF SAGE IS RIDICULOUSLY OUT-OF-DATE!
Please download a binary for the latest version of Sage at
or build it yourself from source.
If you insist on running this version, type
./sage -yes_i_know_its_ridiculous_but_i_want_to_anyway
**********************************************************************
What do you think? Too over-the-top?
Best,
Alex
--
Alex Ghitza -- http://aghitza.org/
Lecturer in Mathematics -- The University of Melbourne -- Australia
Tim Abbott
> > However, they did say that given the circumstances, it might be possible
> > to push through a major update to the package to Karmic when once one is
> > available.
> >
> > So I think this leaves us at "we need to update the package". Sigh.
>
> Maybe we should just take whatever version of Sage is in Ubuntu right
> now and patch it so that running sage simply exits with something like
>
> **********************************************************************
> THIS VERSION OF SAGE IS RIDICULOUSLY OUT-OF-DATE!
>
> Please download a binary for the latest version of Sage at
>
> http://www.sagemath.org
>
> or build it yourself from source.
>
> If you insist on running this version, type
>
> ./sage -yes_i_know_its_ridiculous_but_i_want_to_anyway
> **********************************************************************
>
> What do you think? Too over-the-top?
I'd be very surprised if the Ubuntu archive team would accept that sort of
stable release update.
Think about it from their standpoint. They have thousands[1] of people
apparently using this package on Karmic, and so they're not going to be
enthused about intentionally breaking it for them all because the calculus
functionality doesn't work or upstream's newer version is much better.
(They would take e.g. a patch to fix the calculus functionality, though).
I don't like the current situation any more than you do, but the Ubuntu
archive maintainers' policy isn't unreasonable, even if it is frustrating.
[1] Where do I get the number thousands? The Ubuntu popularity contest
shows 1300 installations of sagemath on Ubuntu, and generally only a small
fraction of Ubuntu systems participate in popcon.
-Tim Abbott
Robert Bradshaw
> On Sun, 14 Mar 2010, Alex Ghitza wrote:
>
>>> However, they did say that given the circumstances, it might be
>>> possible
>>> to push through a major update to the package to Karmic when once
>>> one is
>>> available.
>>>
>>> So I think this leaves us at "we need to update the package". Sigh.
>>
>> Maybe we should just take whatever version of Sage is in Ubuntu right
>> now and patch it so that running sage simply exits with something
>> like
>>
>> **********************************************************************
>> THIS VERSION OF SAGE IS RIDICULOUSLY OUT-OF-DATE!
>>
>> Please download a binary for the latest version of Sage at
>>
>> http://www.sagemath.org
>>
>> or build it yourself from source.
>>
>> If you insist on running this version, type
>>
>> ./sage -yes_i_know_its_ridiculous_but_i_want_to_anyway
>> **********************************************************************
>>
>> What do you think? Too over-the-top?
>
> I'd be very surprised if the Ubuntu archive team would accept that
> sort of
> stable release update.
Would they accept an "update" of the download-binary-from-sagemath.org
type?
- Robert
Dr. David Kirkby
> Maybe we should just take whatever version of Sage is in Ubuntu right
> now and patch it so that running sage simply exits with something like
>
> **********************************************************************
> THIS VERSION OF SAGE IS RIDICULOUSLY OUT-OF-DATE!
>
> Please download a binary for the latest version of Sage at
>
> http://www.sagemath.org
>
> or build it yourself from source.
Like Tim says, I doubt they would allow that.
But
http://trac.sagemath.org/sage_trac/ticket/8447
would stop it happening in future.
Perhaps we could create a patched version of Sage, which includes something
along the lines of #8447, but delay the trigger for a few months, so its gets
updated in Debian + Ubuntu, then issues a warning in a few months time.
Dave
Kasper Peeters
> -- Upgrading during the review process
> sends you to the back of the queue, and by a year after my original
> submission, I had left MIT graduate school to start a startup, and I no
> longer had the time to upgrade past a year of Sage development).
Tim, I take it that you no longer have time to make an upgrade of the
sagemath package? Would you be willing to give us/me a brief run-
through
of the main obstacles you expect? I am able to put some time into
making a debian package (and have experience doing that), but it would
be good if I could avoid duplicating the work that you have already
put
into it.
Cheers,
Kasper
William Stein
<not really serious!>
Maybe I could threaten them with a trademark lawsuit? So at least
they would have to call it some funny name (like Basil) instead of
"Sage". Anyway, this situation is given me a newfound appreciation
for why Firefox isn't called "Firefox" in some Linux distros...
</not really serious!>
>
> [1] Where do I get the number thousands? The Ubuntu popularity contest
> shows 1300 installations of sagemath on Ubuntu, and generally only a small
> fraction of Ubuntu systems participate in popcon.
>
> -Tim Abbott
>
> --
> You received this message because you are subscribed to the Google Groups "debian-sage" group.
> To post to this group, send email to debia...@googlegroups.com.
> To unsubscribe from this group, send email to debian-sage...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/debian-sage?hl=en.
Jason Grout
> I don't like the current situation any more than you do, but the Ubuntu
> archive maintainers' policy isn't unreasonable, even if it is frustrating.
>
What about this: we make a PPA of the binary and just direct everyone to
that. That way, there is a one-step, distribution-supported avenue for
getting an up-to-date Sage. At least, this might be a very easy
stop-gap measure, if someone knows how to make a PPA.
Then maybe we could print a big warning for the sagemath 3.0.5
installation procedure that says the user really should use the PPA.
Thanks,
Jason
--
Jason Grout
Dan Drake
> What about this: we make a PPA of the binary and just direct everyone
> to that. That way, there is a one-step, distribution-supported avenue
> for getting an up-to-date Sage. At least, this might be a very easy
> stop-gap measure, if someone knows how to make a PPA.
This has come up before, and it always seems to screech to a halt when
we get to the point where someone needs to make the PPA. :)
In any case, at https://help.launchpad.net/Packaging/PPA it says a PPA
only gets 1 GB of disk space, which is problematic for Sage. (You can
ask for more, though.) Perhaps we should add "create a PPA for Sage" to
our GSoC projects?
Dan
--
--- Dan Drake
----- http://mathsci.kaist.ac.kr/~drake
-------
Robert Bradshaw
> On Mon, 15 Mar 2010 at 09:14PM -0600, Jason Grout wrote:
>> What about this: we make a PPA of the binary and just direct everyone
>> to that. That way, there is a one-step, distribution-supported avenue
>> for getting an up-to-date Sage. At least, this might be a very easy
>> stop-gap measure, if someone knows how to make a PPA.
>
> This has come up before, and it always seems to screech to a halt when
> we get to the point where someone needs to make the PPA. :)
That's another item on my post-thesis todo list, if no one beats me to
it.
> In any case, at https://help.launchpad.net/Packaging/PPA it says a PPA
> only gets 1 GB of disk space, which is problematic for Sage. (You can
> ask for more, though.)
I was browsing the lists and they seem to have given more to anyone
who needs it.
> Perhaps we should add "create a PPA for Sage" to
> our GSoC projects?
+1 Who's willing to mentor?
- Robert
Harald Schilly
> In any case, athttps://help.launchpad.net/Packaging/PPAit says a PPA
> only gets 1 GB of disk space, which is problematic for Sage. ...
No, it's not. Just ad "aria2" as a dependency for the package and grab
the ubuntu binary via
aria2c --seed-time=0 http://.../...metalink
There are already some deb packages downloading their content via
other sources, i.e. flash plugin. I was told that this is a good start
to look at. This way the deb package will only be a few kb.
H
Paulo César Pereira de Andrade
I don't have much experience with Debian build system, other then
building or "hacking" some packages for OEM projects locally. And
have zero experience with the procedure for approval of a package,
etc.
I see there was some work at
http://wiki.sagemath.org/debian/sage-4.0.x-in-experimental
The package in Mandriva is not officially supported, that is, it is not in the
"main" repository, it is in "contrib". But I would really like to see
it packaged
for Debian and Ubuntu as a "system" package. This way, since Sage releases
often, it would have a wider user base to check for problems and besides
possibly problematic at first if Sage lags behind on the version of some
component, at the end it should be better for everyone packaging Sage.
One example, the wiki interface in the Mandriva package had been
broken for quite some time, and only yesterday I noticed, and also
corrected it.... So far, I basically only do a "sage -testall" to check for
packaging problems, and run some of the tutorials.
Just to attempt to make it clear. I am not a Linux or Mandriva evangelist,
if anything, I would be a FreeBSD evangelist as it was my first experience
with OSS, but I no longer do flaming (and, while I was indiferent to license
issues in the past, I would advocate GPL now :-)
> Cheers,
> Kasper
Paulo