[sage-devel] Should we ship vanilla upstream tarballs or stripped-down ones?
Dear all, Should we go on stripping down upstream tarballs from stuff we don't use when there is some substantial gain? See http://trac.sagemath.org/ticket/17169 (GCC) and http://trac.sagemath.org/ticket/15015 (MPIR). We can debate forever on when the gain becomes substantial in comparison with running a script automating the task when it exists or crafting a new one surely based on other ones hen it does not already exist... It can also be a good thing to just ship upstream tarball without any modification (stripping down, recompression, renaming) for better tracability and security. Whatsoever I think this should be discussed here. Best, JP -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Should we ship vanilla upstream tarballs or stripped-down ones?
The following is just a statement, not an opinion: The new GCC spkg is 50MB larger than the old stripped-down one (86MB vs. 36MB) The largest packages currently in Sage are: 86MB gcc-4.9.2.tar.bz2 35MB sagenb-0.11.0.tar 30MB jmol-14.2.4_2014.08.03.tar.bz2 28MB r-3.1.1.tar.gz 16MB ppl-1.1.tar.bz2 15MB python-2.7.8.tar.gz 13MB maxima-5.34.1.tar.bz2 12MB ipython-2.3.0.tar.gz 11MB matplotlib-1.3.1.tar.gz 10MB scipy-0.14.0.tar.gz -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Should we ship vanilla upstream tarballs or stripped-down ones?
I personally think we should stick with vanilla upstream as much as possible. There are currently two cases where sliming down the tarball is probably useful: * the latest gcc * gap - which is not in Jeroen’s list because it is already completely stripped down. You should check upstream size if you think gcc is bad. I would rather keep the sliming down as an exception rather than a rule. Francois > On 31/10/2014, at 23:06, Jeroen Demeyer wrote: > > The following is just a statement, not an opinion: > > The new GCC spkg is 50MB larger than the old stripped-down one (86MB vs. 36MB) > > The largest packages currently in Sage are: > > 86MB gcc-4.9.2.tar.bz2 > 35MB sagenb-0.11.0.tar > 30MB jmol-14.2.4_2014.08.03.tar.bz2 > 28MB r-3.1.1.tar.gz > 16MB ppl-1.1.tar.bz2 > 15MB python-2.7.8.tar.gz > 13MB maxima-5.34.1.tar.bz2 > 12MB ipython-2.3.0.tar.gz > 11MB matplotlib-1.3.1.tar.gz > 10MB scipy-0.14.0.tar.gz > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Should we ship vanilla upstream tarballs or stripped-down ones?
IMHO we should only modify upstream tarballs if we have to (e.g. strip out non-free parts). The upstream tarballs are cached, so its just a one-time download anyways. On Friday, October 31, 2014 9:54:56 AM UTC, Jean-Pierre Flori wrote: > > Dear all, > > Should we go on stripping down upstream tarballs from stuff we don't use > when there is some substantial gain? > See http://trac.sagemath.org/ticket/17169 (GCC) and > http://trac.sagemath.org/ticket/15015 (MPIR). > We can debate forever on when the gain becomes substantial in comparison > with running a script automating the task when it exists or crafting a new > one surely based on other ones hen it does not already exist... > It can also be a good thing to just ship upstream tarball without any > modification (stripping down, recompression, renaming) for better > tracability and security. > > Whatsoever I think this should be discussed here. > > Best, > JP > > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Sage grant
Yo ! Does Math directorate pay for programmers to write open-source versions > of commercial software? > Of course not. It is just that when commercial softwares fail to do the job, we have to do it in their stead. And we cannot seriously expect them to implement what we need for our research, very often that code is only interesting/useful to researchers. Or are these topics designating novel algorithms and data structures? > Well. For instance you will find this feature quite useless to non-researchers: sage: print designs.orthogonal_arrays.explain_construction(10,814) Construction 3.4 with n=17,m=47,r=6,s=9 from: Julian R. Abel, Nicholas Cavenagh Concerning eight mutually orthogonal latin squares, Vol. 15, n.3, pp. 255-261, Journal of Combinatorial Designs, 2007 sage: print designs.orthogonal_arrays.explain_construction(22,792) Lemma 4.1 with n=25,m=28 from: Charles J.Colbourn, Jeffrey H. Dinitz, Mieczyslaw Wojtas, Thwarts in transversal designs, Designs, Codes and Cryptography 5, no. 3 (1995): 189-197. It tells you in which paper was proved the existence of an orthogonal array OA(10,814) and OA(22,792). If not for Sage, it is just impossible to find out this kind of information (*). That's not really computer science, that's more archeology than mathematics, but it can be useful to (some) mathematicians. Nathann (*) It is not just a database. We implement different recursive constructions from different papers, Sage computes all possible combinations of them and find out which leads to the result. I dare you to do it with a paper and pen :-P -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Should we ship vanilla upstream tarballs or stripped-down ones?
Hi Volker, On 2014-10-31, Volker Braun wrote: > --=_Part_1546_392576408.1414752441684 > Content-Type: text/plain; charset=UTF-8 > > IMHO we should only modify upstream tarballs if we have to (e.g. strip out > non-free parts). The upstream tarballs are cached, so its just a one-time > download anyways. There are people who have a very bad band-with. In my case, it's fine when I am at university. But other than that, I only have a mobile internet stick, for which 50MB more or less really matters. Best regards, Simon -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Should we ship vanilla upstream tarballs or stripped-down ones?
> > > > > > IMHO we should only modify upstream tarballs if we have to (e.g. strip > out > > non-free parts). The upstream tarballs are cached, so its just a > one-time > > download anyways. > > There are people who have a very bad band-with. In my case, it's fine > when I am at university. But other than that, I only have a mobile > internet stick, for which 50MB more or less really matters. > And outside the developed world this can be even more of a problem. Especially for upgrades I notice it can take a while even on my relatively fast US connection - gcc took several minutes at least to download on my last `git pull`. -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Should we ship vanilla upstream tarballs or stripped-down ones?
Oh come on, cry me a river. The average top-1000 web page is well over a megabyte nowadays. A one-time download of 50mb is a minor convenience at best. We are talking about the equivalent of maybe 30 page visits. On Friday, October 31, 2014 11:48:47 AM UTC, Simon King wrote: > > Hi Volker, > > On 2014-10-31, Volker Braun > wrote: > > --=_Part_1546_392576408.1414752441684 > > Content-Type: text/plain; charset=UTF-8 > > > > IMHO we should only modify upstream tarballs if we have to (e.g. strip > out > > non-free parts). The upstream tarballs are cached, so its just a > one-time > > download anyways. > > There are people who have a very bad band-with. In my case, it's fine > when I am at university. But other than that, I only have a mobile > internet stick, for which 50MB more or less really matters. > > Best regards, > Simon > > > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Should we ship vanilla upstream tarballs or stripped-down ones?
However (local) caching is used in those cases as well, so it's only a pain the first time you visit a website. The router to where I work in my home setup is not shabby, but it's still took some 10 minutes to download GCC. For a while I thought something had stalled. 86 MB to 36 MB is a huge reduction and just makes things better with bad or laggy internet connections. Even worse is when the connection drops and you'd have to restart the download over again. Best, Travis On Friday, October 31, 2014 6:44:07 AM UTC-7, Volker Braun wrote: > > Oh come on, cry me a river. The average top-1000 web page is well over a > megabyte nowadays. A one-time download of 50mb is a minor convenience at > best. We are talking about the equivalent of maybe 30 page visits. > > > > On Friday, October 31, 2014 11:48:47 AM UTC, Simon King wrote: >> >> Hi Volker, >> >> On 2014-10-31, Volker Braun wrote: >> > --=_Part_1546_392576408.1414752441684 >> > Content-Type: text/plain; charset=UTF-8 >> > >> > IMHO we should only modify upstream tarballs if we have to (e.g. strip >> out >> > non-free parts). The upstream tarballs are cached, so its just a >> one-time >> > download anyways. >> >> There are people who have a very bad band-with. In my case, it's fine >> when I am at university. But other than that, I only have a mobile >> internet stick, for which 50MB more or less really matters. >> >> Best regards, >> Simon >> >> >> -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] "Can We Trust Computer Algebra Systems?"
> > How hard would it be to switch from the singular interface to the > MAcualay2 one for the polynomial stuff? > I believe Macaulay 2 also uses Singular for its basic polynomial stuff: http://www.math.uiuc.edu/Macaulay2/doc/Macaulay2-1.6/share/doc/Macaulay2/Macaulay2Doc/html/___Acknowledgements.html Best, Nathan -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Sage grant
I looked at the NSF-OCI solicitation. Maybe I'm not seeing what you think is a target. I found this - Enable academic departments, disciplinary and cross-disciplinary units, or multi-organization consortia to renovate research facilities through the addition or augmentation of cyberinfrastructure, other than general-purpose computing systems or data storage systems, to create environments that enhance research and integrate research with education. - If so, then the argument should be made as to how you are augmenting cyberinfrastructure. I suppose the example below suggests that you can write programs to do something that would be more time-consuming without computers. I would not n There are lots of thing that fall into this category. I don't know what NSF is aiming for as "cyberinfrastructure". For example, I could propose a project that did the following: (a) made a list of all humans identifiable on the internet. (b) as far as possible, determined their date of birth day/month or day/month/year (c) allowed research that would otherwise be quite difficult like 1. What percentage of such people were born in November? 2. What is the distribution of ages of people ... 3. What percentage of people were born on Friday etc. Now is this interesting fundable cyberinfrastructure? (I don't know. Depends on the reviewers I guess). And whether, say http://www.emba.uvm.edu/~jdinitz/contents.1.html or just google can do the job you propose needs new programs. And whether the new programs are self sustaining or would be stale immediately after funding stopped. Just some thoughts. I have not been on an NSF review panel for at least 10 years. RJF On Friday, October 31, 2014 3:51:13 AM UTC-7, Nathann Cohen wrote: > > Yo ! > > Does Math directorate pay for programmers to write open-source versions >> of commercial software? >> > > > Of course not. It is just that when commercial softwares fail to do the > job, we have to do it in their stead. And we cannot seriously expect them > to implement what we need for our research, very often that code is only > interesting/useful to researchers. > To the extent that code you write is only interested in researchers in a narrow area (say in some corner of pure mathematics), the proposal is less interesting than one with broad impact. Just saying. > > > Or are these topics designating novel algorithms and data structures? >> > > Well. For instance you will find this feature quite useless to > non-researchers: > > sage: print designs.orthogonal_arrays.explain_construction(10,814) > Construction 3.4 with n=17,m=47,r=6,s=9 from: > Julian R. Abel, Nicholas Cavenagh > Concerning eight mutually orthogonal latin squares, > Vol. 15, n.3, pp. 255-261, > Journal of Combinatorial Designs, 2007 > sage: print designs.orthogonal_arrays.explain_construction(22,792) > Lemma 4.1 with n=25,m=28 from: >Charles J.Colbourn, Jeffrey H. Dinitz, Mieczyslaw Wojtas, >Thwarts in transversal designs, >Designs, Codes and Cryptography 5, no. 3 (1995): 189-197. > > It tells you in which paper was proved the existence of an orthogonal > array OA(10,814) and OA(22,792). > > If not for Sage, it is just impossible to find out this kind of > information (*). That's not really computer science, that's more archeology > than mathematics, but it can be useful to (some) mathematicians. > Yes, some. Is it the best use of NSF's limited budget? Will it provide leverage in solving the problems of society? (I have no idea how broad your search stuff is. Is it just orthogonal arrays, experimental design ?? All of mathematics??) Good luck > > Nathann > > (*) It is not just a database. We implement different recursive > constructions from different papers, Sage computes all possible > combinations of them and find out which leads to the result. I dare you to > do it with a paper and pen :-P > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] Re: Sage grant
Yo ! Hey, just to make it clear: I am not supporting their grant request industry. I could not care less, and if anything I just hate to see that people use Sage (which includes some of my own work) to request solid money for themselves. I just did not want to let you have the last word when all you do here is say that "mathematica is better". Nothing smart involved. About the actual request: I believe that it contains the usuall bull***, and includes as many fancy words as possible while not claiming that anything too specific will happen, so that you can do whatever you want with the money. Nathann -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
[sage-devel] Re: Should we ship vanilla upstream tarballs or stripped-down ones?
OMG you had to wait 10 minutes instead of 4 minutes On Friday, October 31, 2014 2:59:52 PM UTC, Travis Scrimshaw wrote: > > However (local) caching is used in those cases as well, so it's only a > pain the first time you visit a website. The router to where I work in my > home setup is not shabby, but it's still took some 10 minutes to download > GCC. For a while I thought something had stalled. 86 MB to 36 MB is a huge > reduction and just makes things better with bad or laggy internet > connections. Even worse is when the connection drops and you'd have to > restart the download over again. > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
Re: [sage-devel] "Can We Trust Computer Algebra Systems?"
Thanks for the info. I didn't know that. Anyways, it seems that Macaulay2 only uses Singular-Factory and Singular-Libfac for a few functions (factor, gcd minimalprimes and irreduciblecharacteristicseries). So for all the other stuff, it would stil be useful to compare the results given by Maculay2 and Singular. El viernes, 31 de octubre de 2014 16:04:04 UTC+1, Nathan Dunfield escribió: > > How hard would it be to switch from the singular interface to the >> MAcualay2 one for the polynomial stuff? >> > > I believe Macaulay 2 also uses Singular for its basic polynomial stuff: > > > http://www.math.uiuc.edu/Macaulay2/doc/Macaulay2-1.6/share/doc/Macaulay2/Macaulay2Doc/html/___Acknowledgements.html > > Best, > > Nathan > > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
