Here is an image of what I mean. I thought our submission was within the 8 page limit because when built locally it is.

It looks like there might be a "leading" (space between lines) issue as well as a definite font size difference for both headers and the body text. But it is not a general typesetting issue because the title and the abstract seem to be rendered with the same spacing properties in both cases.

Also the smallcaps Scipy proceedings header is a serif font in the build server's version and sans-serif in the locally built version.

Not sure what these could be other than default LaTeX parameters that might differ between older and newer distributions. I'm going to guess that the server's is older than my local version, but I could be wrong.

Currently, whatever author is listed first is the corresponding author. However, we would like to use the last author specified as the corresponding author. Will that be possible?

Hello all, this may have come up in conversation in past years (and I talked to @katyhuff a bit about this in Austin), but it would be nice to get the proceedings papers archived permanently (with DOIs) somewhere like Figshare.

Perhaps a Figshare collection could be created for all the papers? The past proceedings websites are all really slick so I don't propose getting rid of those, but this could supplement as a permanent archival for the papers themselves—the PDF links could be replaced with a link to the DOIs.

An issue that got raised as a result of last year's review process is that some more traditional academics would prefer if they could submit papers in LaTeX directly rather than as ReStructuredText

@katyhuff I'm taking the mesa conference proceeding and putting all the code into an ipython notebook to be used as a tutorial. I noticed the from mesa.datacollector import DataCollector line in the Data Collection section should read from mesa.datacollection import DataCollector

(datacollector to datacollection) should this be changed as a PR to be merged?

@dmasad @jackiekazil

Copyright and titling is 2010.. If I get around to it I will submit PR.. Should be done under 2014 branch right?

I don't mean to be harsh, but this long of a delay in publishing the proceedings is simply unacceptable. If the conference wants anyone to bother with submitting a proceedings next year, or for the foreseeable future for that matter, the conference organizers need to step up and completely overhaul this procedure. Some people actually rely on this conference proceedings to get their results out in a published format--it appears that this was a mistake.

Ubuntu 14.04 rendering 00_bibdervalt using sh make_paper.sh gives me a pdf, but with no visible references section and [?] instead of a reference.

I've ran all the apt-get commands for latex packages listed in the README.

In the 2015 proceedings template and build system, are multiple institutions per author accepted/supported?

It seems to work, in terms of generating the correct footnotes for affiliations, but I'm getting incorrect footnote markers next to the names.

Example source:

:author: First author
:email: [email protected]
:institution: Home of Foo
:institution: Home of Bar

:author: Second author
:institution: Home of Bar

:author: Third author
:institution: Yet another place

and the rendered tex/pdf output:

...

Second author should have a single dagger, and only Third author should have the double-dagger.

The readme says 7, but the email said 8.

sb@arcadia:~/projects/scipy_proceedings/publisher$ make all
rm -rf ../output/* _build/*
mkdir -p _build/tex
./build_template.py title.tex
Traceback (most recent call last):
  File "./build_template.py", line 7, in <module>
    import tempita
  File "/home/sb/projects/scipy_proceedings/publisher/tempita/__init__.py", line 295
    except SyntaxError, e:
                      ^
SyntaxError: invalid syntax
make: *** [_build/tex/title.tex] Error 1

wtf

@dalippa suggested (in #135) that you sometimes want the copyright holder to be different than the authors... so... let's make that change to the template.

Hi guys, I am trying to use the scipy proceedings format to self-publish some scientific results (see https://github.com/TheChymera/OPR.OPR-Concepts-Guidelines/blob/master/paper/paper.rst#spi02 for the according guidelines paper I am writing). Previously I tried to do this via LaTeX ( https://github.com/TheChymera/OPR.Concepts-Guidelines-for-OPR ), and even though that approach was inferior - at least I could automatically build and compile my bibliography.

Is there any more automated way to manage the bibliography of a paper written according to the scipy proceedings format? As it is I'm forced to write every bib entry by hand, and being a very cite-happy writer that's a pain.

@stefanv I've been trying to debug this, I closed the PR for objarray and have now reopened it, but on the build server the link for ObjArray always displays the paper for Cesium instead of Objarray. The link for cesium displays the paper for Cesium.

I've cleared the cache and run update_prs, the pr_list file has ObjArray and the right information. I've also been adding print statements throughout the build scripts and all the information looks correct (paper number, url, etc), but somehow the wrong paper shows up at the end. Any help debugging would be appreciated.

See #239 (comment)

Looks like the 2016 branch is currently ahead of master, but those commits are generic.

Can we sync up 2016 with master?

@sarostru @stefanv

Following the link on REAME preview on this page:

https://github.com/scipy-conference/scipy_proceedings

to the review criteria is broken, leading to this url:

https://github.com/scipy-conference/scipy_proceedings/review_criteria.md

a possible alternative URL for this link (which leads to a page on which one can read the review criteria) is:

Quick question: are wide figures appropriate for our publication medium?

If so, how can we convince restructured -> docutils -> latex to use a latex "figure_" (as opposed to "figure"). Would the entire latex figure_ environment and content have to be hard coded in the original rst document?

Best,
Mark

Is there a way to get it to display more than one authors' emails as corresponding authors? Or should that be included in the text manually (e.g., in the conclusion; in a footnote; in acknowledgments; &c.)?

Hello, I have been trying to set up a system for scientific publishing with open accessible revision (as opposed to review!) in mind. For my (rather messy) current LaTeX-based solution please see https://github.com/TheChymera/OPR.Concepts-Guidelines-for-OPR .

The way the scipy proceedings are organized seems a far supperior approach, though. I would like to publish individual papers as separate git repos (so that someone interested in a specific paper can pull and revise it). For this purpose it makes little sense for me to have multiple papers under papers/ .

Any suggestions on how I could best reformat this so that I just have all of my stuff under paper/ ?

Based on my experimentation, it looks like any reference beginning with an author initial must have a second line. Otherwise, there is extra spacing before the author name. And the extra spacing disappears if there isn't a period after the initial, or if there is only a period by itself.

Compare extra spacing before the L. produced by

.. [HEVEA]  L. Maranget. "HEVEA." http://hevea.inria.fr/.

versus the output of

.. [HEVEA] L. Maranget. "HEVEA."
           http://hevea.inria.fr/.

See the discussion in #62 (comment).

Hey everyone,

I'm new to .rst and encountered a few issues in how the .tex document is constructed given the .rst source. If anyone has any expertise in this area, can you help me out and also send your answers to my email address ([email protected])? Thanks.

1. I used numbers as my as variables to my reference my citations. The example document uses words. For example:

[1] vs [ref 1] for reference 1
[2] vs [ref 2] for reference 2

I did this because, if you look at the example document, these variable names do appear in the references section. I think this is in contrast to standard citation practices where the reference list is almost always a list of integers. In doing this, I've screwed up how the references appear in the document. My "Reference" section header does not appear for example.

1. For footnotes, the symbols do not reset on each page. For example, the first footnote on each page ought to be an asterisk; however, the .tex file is going through the full gamut of symbols. This results in some obscure superscripts. Is there a way to correct this?

Thanks.

This can be observed in the demo doc, where table 1 gets booktabs and tables 2 and 3 do not.

This doesn't seem to be a property of the tables themselves, since rearranging them within the source document still ends up with the same behavior.

On a somewhat vanilla Fedora Linux system, the following packages are needed in order to build the example:

texlive-IEEEtran
texlive-everypage
texlive-draftwatermark
texlive-multirow

Looks like that is getting pulled in from

I haven't looked into it yet, but a fresh checkout of the scipy_proceedings followed by

./make_paper.sh papers/00_vanderwalt
*** Warning: /home_mint/bergstra/V/skdata/src/scipy_proceedings/publisher/_static/institutions.json does not exist.
Building: 00_vanderwalt
*** Warning: /home_mint/bergstra/V/skdata/src/scipy_proceedings/publisher/../output/00_vanderwalt/paper_stats.json does not exist.

just hangs there not doing anything. I haven't investigated, but if someone else fixes this before me let me know :)

Bibliographic data (author's names, emails, and institutions) are rendered badly in the article html template.

In the toc.json file created by the build script, this data is represented for each article as (for example):

           "author_institution": [
                "Universidad de Valparaiso, Chile",
                "Advanced Center for Electrical and Electronic Engineering",
                "Universidad de Valparaiso, Chile",
                "Advanced Center for Electrical and Electronic Engineering",
                "Universidad de Valparaiso, Chile",
                "Universidad de Valparaiso, Chile"
            ],
            "bibliography": "",
            "author": [
                "Alejandro Weinstein",
                "Wael El-Deredy",
                "Stéren Chabert",
                "Myriam Fuentes"
            ],
            "author_email": [
                "[email protected]",
                "[email protected]",
                "[email protected]",
                "[email protected]"
            ],

This is then rendered to the template by looping through the lists associated with author and author_instituion, zipped together:

https://github.com/scipy-conference/scipy_proceedings/blob/master/publisher/_templates/article.html.tmpl#L5

If there are multiple institutions listed for any author, these lists will get out of sync and become inaccurate.

Instead, the authors field should be a list of objects with name, institution, and email fields. Then this data should be used in the template to create the expected result.

Long URLs in the reference section may not break appropriately into multiple lines, even when there are hyphens in the URL where it would be appropriate to break. In order to get these URLs to render properly, we need to use the hyphen option in the LaTeX url package:

\usepackage[hyphens]{url}

Thanks to @kyleniemeyer for this point

#205 (comment)

I've been following the instructions in the README and have gotten a PDFLaTeX error when trying to compile my (unedited) copy of the 00_vanderwalt paper. The culprit appears to be a missing file, paper_stats.json

This is a gist of the error and preceding commands, copied from my terminal:

https://gist.github.com/sbenthall/67d0a39d8c7b9efdd961

This seems to be a recurring theme for me. Can we get a tag/specifier for wide tables?

Thanks again for the quick response on the wide figures.

Best,
Mark

https://github.com/scipy-conference/scipy_proceedings/blob/2014/scipy_proc.json#L8

If 2010 was the 9th.. http://conference.scipy.org/proceedings/scipy2010/
would this not be the 13th?

See #239 (comment)

Getting this error on my attempt to build a copy of 00_vanderwalt paper on Ubuntu 14.04

./make_paper.sh papers/sebastian_benthallTraceback (most recent call last):
  File "publisher/build_paper.py", line 4, in <module>
    import docutils.core as dc
ImportError: No module named 'docutils'

In following the instructions after forking the repo, cloning into a local directory, switching to the 2015 branch, making a personal directory (papers/mike_pacer), moving the example paper into my local directory and then trying to run ./make_paper.sh papers/mike_pacer/ I get an error.

Specifically,

  File "publisher/build_paper.py", line 123
    print "PDFLaTeX error output:"
                                 ^
SyntaxError: Missing parentheses in call to 'print'
Error building paper papers/mike_pacer/. Aborting.

Which looks like it both had some kind of problem related to running PDFLaTeX, as well as an error having to do with a python 3 incompatible print statement in its error printing code(though this just happened so I haven't had an opportunity to investigate more thoroughly yet).

This one is quite hard to fix. Because we run the rst converter on processed rst, the line number it throws is different from the original line number.

I don't think it is possible to get the original line number. (Is there a #line instruction in restructured text?)

As a practical alternative, we may want to catch the error, and print some context information from the processed rst to help the authors in finding the error.

I was working on #224 and found this while reconciling changes:

:author: David L
:corresponding: David L
:email: [email protected]
:institution: RealMassive, Inc.

:author: Joe Schmo
:corresponding: Joe Schmo
:email: [email protected]
:institution: Schmo, Inc

:author: A L
:corresponding: A L
:email: [email protected]

:author: Jason V
:corresponding: Jason V
:email: [email protected]
:institution: RealMassive, Inc.

:author: D L
:corresponding: D L
:email: [email protected]

The :author: tag works, but only for the first corresponding author. Adding :corresponding: and :author: tags also cause the error, as does the 00_vanderwalt.rst file on master and 2016 branches.

Stack trace:

Traceback (most recent call last):
  File "publisher/build_paper.py", line 229, in <module>
    build_paper(paper_id)
  File "publisher/build_paper.py", line 214, in build_paper
    rst2tex(in_path, out_path)
  File "publisher/build_paper.py", line 88, in rst2tex
    settings_overrides=settings)
  File "/usr/lib/python2.7/dist-packages/docutils/core.py", line 414, in publish_string
    enable_exit_status=enable_exit_status)
  File "/usr/lib/python2.7/dist-packages/docutils/core.py", line 662, in publish_programmatically
    output = pub.publish(enable_exit_status=enable_exit_status)
  File "/usr/lib/python2.7/dist-packages/docutils/core.py", line 219, in publish
    output = self.writer.write(self.document, self.destination)
  File "/usr/lib/python2.7/dist-packages/docutils/writers/__init__.py", line 80, in write
    self.translate()
  File "/usr/lib/python2.7/dist-packages/docutils/writers/latex2e/__init__.py", line 247, in translate
    self.document.walkabout(visitor)
  File "/usr/lib/python2.7/dist-packages/docutils/nodes.py", line 174, in walkabout
    if child.walkabout(visitor):
  File "/usr/lib/python2.7/dist-packages/docutils/nodes.py", line 174, in walkabout
    if child.walkabout(visitor):
  File "/usr/lib/python2.7/dist-packages/docutils/nodes.py", line 174, in walkabout
    if child.walkabout(visitor):
  File "/usr/lib/python2.7/dist-packages/docutils/nodes.py", line 166, in walkabout
    visitor.dispatch_visit(self)
  File "/usr/lib/python2.7/dist-packages/docutils/nodes.py", line 1882, in dispatch_visit
    return method(node)
  File "/home/dalippa/scipy_proceedings/publisher/writer/__init__.py", line 102, in visit_field_body
    self.corresponding.append(self.author_names[-1])

Filling out paperwork here (needs tracking as Govt funds used). Is there a formal "Publisher" of these proceedings?

I am trying to write the proceedings using the IPython Notebook to write my paper for the SciPy 2014 proceedings, since I am not very familiar with reST (and since the IPython Notebook seems like a good tool for such a document for future reference...)

I have two questions:
(1) What is the correct way to include output from Python code (i.e. notebook output cells) in the proceedings?

(2) The reST conversion using

ipython nbconvert file.ipynb --to rst

produces output which is not recognised (apparently) by the SciPy format, e.g.

.. math:: a = \pm 2^e \times m.

which should be

.. math:: 

    a = \pm 2^e \times m.

Is there a good solution to this (except editing by hand)?

Trying to maintain the conference.scipy.org site, I noticed the following pages are missing images. Is there someplace to find the source?

http://conference.scipy.org/proceedings/scipy2009/
http://conference.scipy.org/proceedings/scipy2008/

In the conference metadata for the proceedigns, scipy_proc.json, there's a place for a list of sponsored students. Entries in this list have three fields: name, org, and location.

Generally speaking, the 'location' field is not used and the result is that the proceedings are published with a trailing comma after each sponsored students.

http://conference.scipy.org/proceedings/scipy2015/students.html

This should be corrected, making the location field optional or non-existent.

Is the use of the year 2010 intentional in these lines? I believe it is causing the wrong year to be rendered at the top of the PDF (assuming it should be 2015).

• list of sponsored students
• committee list
• reviewers list
• intro?

Hello, I am trying to write up some guidelines for revision-oriented publishing ( https://github.com/TheChymera/OPR.Concepts-Guidelines-for-OPR ). As it makes little sense to version figures over git - it would be awesome to have them plotted at compile time (which PythonTeX does). However, though I very much like LaTeX, I am thinking of maybe writing my papers as .rst (so that they can easily be read w/o requiring the user to compile them) - can PytonTeX be used from within .rst documents such as the ones used for the scipy proceedings?

I'd like to put a "note" in my paper. This is part of standard ReST, but in the paper it causes the text to extend across both columns.

This is caused by the following block:

% admonition (specially marked topic)
\providecommand{\DUadmonition}[2][class-arg]{%
  % try \DUadmonition#1{#2}:
  \ifcsname DUadmonition#1\endcsname%
    \csname DUadmonition#1\endcsname{#2}%
  \else
    \begin{center}
      \fbox{\parbox{0.9\textwidth}{#2}}
    \end{center}
  \fi
}

The number before \textwidth should be <0.5, I guess.

Add links to the README pointing to all previous years' published proceedings (P.S., where are they?)