[sage-devel] Re: unicode and pdf documentation build failure
Hi John,
Thank you, that was precisely what I was looking for. I could not
remember how to do that.
Best,
Travis
On Wednesday, June 3, 2020 at 3:28:49 PM UTC+10, John H Palmieri wrote:
>
> Hi Travis,
>
> Can you add a command to `latex_elements['preamble']` in
> src/sage/docs/conf.py`? There are a lot of unicode characters defined there
> already. \DeclareUnicodeCharacter{26AC}{\circ} or something like that?
>
>
>
> On Tuesday, June 2, 2020 at 8:24:52 PM UTC-7, Travis Scrimshaw wrote:
>>
>> Hi everyone,
>>I have a patch with some unicode that uses a character (U+26AC, ⚬)
>> that is not recognized by the pdf docbuild. Is there some way I can add an
>> option for the pdf docbuilder to translate this unicode character into a
>> latex symbol so the PDF doc will build? Or is the best option just to
>> suppress the output?
>>
>> Thanks,
>> Travis
>>
>>
--
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 view this discussion on the web visit
https://groups.google.com/d/msgid/sage-devel/2917b796-ab03-4c2d-8521-9debad4cfe77%40googlegroups.com.
[sage-devel] Re: unicode and pdf documentation build failure
Hi Travis,
Can you add a command to `latex_elements['preamble']` in
src/sage/docs/conf.py`? There are a lot of unicode characters defined there
already. \DeclareUnicodeCharacter{26AC}{\circ} or something like that?
On Tuesday, June 2, 2020 at 8:24:52 PM UTC-7, Travis Scrimshaw wrote:
>
> Hi everyone,
>I have a patch with some unicode that uses a character (U+26AC, ⚬) that
> is not recognized by the pdf docbuild. Is there some way I can add an
> option for the pdf docbuilder to translate this unicode character into a
> latex symbol so the PDF doc will build? Or is the best option just to
> suppress the output?
>
> Thanks,
> Travis
>
>
--
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 view this discussion on the web visit
https://groups.google.com/d/msgid/sage-devel/fc8c7ea6-4745-4e56-8071-5a049189ef49%40googlegroups.com.
[sage-devel] unicode and pdf documentation build failure
Hi everyone, I have a patch with some unicode that uses a character (U+26AC, ⚬) that is not recognized by the pdf docbuild. Is there some way I can add an option for the pdf docbuilder to translate this unicode character into a latex symbol so the PDF doc will build? Or is the best option just to suppress the output? Thanks, Travis -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/928b1c32-6997-4fe7-9f42-2a7bb124e017%40googlegroups.com.
Re: [sage-devel] Re: line length in docstring and tests
On Tuesday, June 2, 2020 at 10:13:55 AM UTC-7, Michael Orlitzky wrote: > > > You shouldn't force yourself to stick to 80 characters when there's a > technical reason for the line to be longer, but the argument for the > 80-character limit is not some fairy tale about punch cards or > PDP-11/VAX terminals. Find a book. How long are the lines? They're > roughly the length that millennia of "user interface" research has > taught us is most comfortable to read. Depending on the font, the max is > around 80. > True, but indent space does not really eat into the readability there. With 3 to 4 levels of 4-space indent (not uncommon in python), you're at 80+16=96; pretty close to 100. One of the reasons for limited line length (and that's also an argument for column-formatting such as in news papers) is that the right-to-left tracking the human eye has to do from one line to the next is more difficult for lines that are way longer than they are high (so: same reason why tables often have alternating backgrounds across their rows). Computer code has plenty of other cues for line separation too, so I'm not sure the same guidance that applies to body text translates to computer code unchanged. -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/9d7a27e1-f62f-41a5-a3e4-b23779cc63af%40googlegroups.com.
Re: [sage-devel] Re: line length in docstring and tests
On Jun 02, Michael Orlitzky wrote: On 6/2/20 9:52 AM, Volker Braun wrote: I'm with Linus in that the 80 characters limit is an anachronism thats too restrictive in a time and age where large high-resolution displays are ubiqitous. There certainly is a "too long" somewhere but its more in the 100-character region. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=bdc48fa11e46f867ea4d75fa59ee87a7f48be144 You shouldn't force yourself to stick to 80 characters when there's a technical reason for the line to be longer, but the argument for the 80-character limit is not some fairy tale about punch cards or PDP-11/VAX terminals. Find a book. How long are the lines? They're roughly the length that millennia of "user interface" research has taught us is most comfortable to read. Depending on the font, the max is around 80. I didn't mean to start an argument about line length. I am about to submit a very large patch and if I ever hope to get a reviewer, I want to adhere as much as possible to whatever code style is imposed. I wanted to get a feeling if people really expected PEP8 as stated in the Developer's guide or not. Particularly about length 72 on output strings of tests and examples. It seems that I got my answer and it seems to be that it won't be an issue to leave these lines. As of now my code is a mess, but it's slowly improving. Best, R. -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/9fabe31a-0ad8-c0a3-7837-bb8e18a7a63f%40orlitzky.com. -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/20200602192114.GA172000%40vertex. signature.asc Description: PGP signature
Re: [sage-devel] Re: line length in docstring and tests
On 6/2/20 9:52 AM, Volker Braun wrote: > I'm with Linus in that the 80 characters limit is an anachronism thats > too restrictive in a time and age where large high-resolution displays > are ubiqitous. There certainly is a "too long" somewhere but its more in > the 100-character region. > > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=bdc48fa11e46f867ea4d75fa59ee87a7f48be144 > You shouldn't force yourself to stick to 80 characters when there's a technical reason for the line to be longer, but the argument for the 80-character limit is not some fairy tale about punch cards or PDP-11/VAX terminals. Find a book. How long are the lines? They're roughly the length that millennia of "user interface" research has taught us is most comfortable to read. Depending on the font, the max is around 80. -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/9fabe31a-0ad8-c0a3-7837-bb8e18a7a63f%40orlitzky.com.
[sage-devel] Re: Mutliprocessing for Matrix Computations?
Michael Jung wrote: > Ah, interesting. Do you have some literature/references for me? Among the standard references, Chapter 5 of *Modern Computer Algebra* by von zur Gathen & Gerhard discusses the basic evaluation-interpolation algorithm for computing the determinant of polynomial matrices in some detail. For more sophisticated methods and recent research on fast exact linear algebra (including both complexity questions and practical speed issues), look at the papers of the people involved in Linbox. -- Marc -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/rb5p9c%242hd4%241%40ciao.gmane.io.
[sage-devel] Re: line length in docstring and tests
I'm with Linus in that the 80 characters limit is an anachronism thats too restrictive in a time and age where large high-resolution displays are ubiqitous. There certainly is a "too long" somewhere but its more in the 100-character region. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=bdc48fa11e46f867ea4d75fa59ee87a7f48be144 -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/a718bd44-6b38-4da3-9df8-7e8e23121cd8%40googlegroups.com.
[sage-devel] Re: New Python version requirement for Sage?
Thats true, I'll fix it On Monday, June 1, 2020 at 11:08:36 PM UTC+2, Jonathan Kliem wrote: > > At the moment we are guessing that kucalc buildbot runs with ./configure > --with-python=2, because this will lead to exactly this error message (and > it really is the only way that this error message makes sense, because this > module is present in python 3.5.2 and the import error doesn't make any > sense). > > Can someone verify/change that or contact someone who will do so? > > Am Freitag, 29. Mai 2020 01:22:54 UTC+2 schrieb Matthias Koeppe: >> >> On Thursday, May 28, 2020 at 12:15:33 PM UTC-7, Volker Braun wrote: >>> >>> Another datapoint is that the kucalc buildbot runs on Ubuntu 16.04 LTS >>> which has Python 3.5.2 >>> >> >> I see you opened a ticket for a https://trac.sagemath.org/ticket/29753 >> >> It would be good if you could provide substantially more information >> about that build. >> >> >> -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/3e3ec925-cb49-4534-ac34-593efef13deb%40googlegroups.com.
Re: [sage-devel] Problem building R on 10.15.5 with homebrew
Hi Zach, On Tue, Jun 2, 2020 at 2:39 AM Zachary Scherr wrote: > >It looks like both were already linked but I ended up reinstalling them > and everything worked, thank you so much. > > In general, do you recommend installing libraries like flint from > sagemath/homebrew-science before building sage? I see that there are a lot of > formulas there but I'm not sure any of them have Catalina bottles. I don't > know too much about homebrew, but I'd be happy to try to contribute Catalina > bottles if this is something worth doing. > as we are proceeding towards making SageMath a pip-installable library, we are unvendoring everything we can, see https://trac.sagemath.org/ticket/27330 And yes, indeed, maintaining our own patches on many OSs is a lot of effort we can ill afford, so e.g. we very much would like to use Python from the system rather than build our own (we are also planning to allow the use of system Python packages, but this is not yet there) So yes, everything listed in the latter ticket as "already done" or "in progress" ought to get a Brew formula providing the corresponding package. Some existing Brew formulas are not compatible with what Sage needs, e.g. https://formulae.brew.sh/formula/flint does not provide the NTL interface, so we have a few of our own in https://github.com/sagemath/homebrew-science (as you most probably know). Ideally these formulas/updates/fixes should be upstreamed, but without that having bottles for these (as well as everything else needed for Sage!) would be great. E.g. speaking about Flint formula, probably the correct way to deal with it is to use depends_on "ntl" => :optional in its "canonical" formula https://formulae.brew.sh/formula/flint I don't know much about Homebrew, and am not an active macOS user either, so if anything on Sagemath side needs to be done to facilitate this, please speak up. Cheers Dima > Best, > Zach > > > On Monday, June 1, 2020 at 3:51:23 PM UTC-4, Dima Pasechnik wrote: >> >> On Mon, Jun 1, 2020 at 8:30 PM Zachary Scherr wrote: >> > >> > Hi Dima, >> > >> >I already have all of the homebrew packages installed, I'm not sure why >> > it offers that suggestion. >> >> By looking at your config.log, I see that flint and ntl are keg-only, >> which actually results in sqlite ,zlib, and curl tests failing >> (because the tests for libcurl and libsqlite actually try to link and >> run programs linked to an already detected list of libs, >> but locations of some of these libs need extra -L flags, oops) >> >> brew link ntl >> brew link flint >> make zlib-clean sqlite-clean python3-clean >> >> run >> ./configure >> it should tell you that zlib, sqlite, python3, r should come from the >> system... >> >> >> >> >> > I also updated from OS 10.14 to 10.15 right before building sage. I had >> > previously built sage on 10.14 without these problems. >> > >> > Best, >> > Zach >> > >> > On Monday, June 1, 2020 at 3:27:19 PM UTC-4, Dima Pasechnik wrote: >> >> >> >> On Mon, Jun 1, 2020 at 7:37 PM Zachary Scherr wrote: >> >> > >> >> > Hi All, >> >> > >> >> >I recently tried building Sage on Catalina 10.15.5. I have homebrew >> >> > with all recommended packages installed and I tried building sage via >> >> > the commands >> >> > >> >> your config.log says that e.g. sqlite is not installed. >> >> At the end of ./configure run, please have a look at the advice to >> >> install packages it now prints. >> >> >> >> > make distclean >> >> > source .homebrew-build-env >> >> > ./configure >> >> > MAKE='make -j18' make build >> >> > >> >> >and it crashed with the error >> >> > >> >> > Error building Sage. >> >> > >> >> > The following package(s) may have failed to build (not necessarily >> >> > during this run of 'make all-build'): >> >> > >> >> > * package: r-3.6.2.p0 >> >> > last build time: Jun 1 14:21 >> >> > log file:/Users/zscherr/SageMath/logs/pkgs/r-3.6.2.p0.log >> >> > build directory: >> >> > /Users/zscherr/SageMath/local/var/tmp/sage/build/r-3.6.2.p0 >> >> > >> >> >I'm not entirely sure why it's building R in the first place since I >> >> > already have it installed via homebrew. I briefly checked the log file >> >> > and saw ld: library not found for -lssl which I'm thinking might be >> >> > somehow related >> >> > to the fact that openssl and curl-openssl are keg-only. My best guess >> >> > might be that "source .homebrew-build-env" ends up missing some >> >> > environment variables, but I'd really appreciate any help you can offer. >> >> > >> >> > Thanks, >> >> > Zach >> >> > >> >> > -- >> >> > 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-...@googlegroups.com. >> >> > To view this discussion on the web visit >> >> > https://groups.google.com/d/msgid/sage-devel/6a84dd22-2e9e-4139-aab7-6582cb0bf598%40googlegroups.com. >> > >> >
