borisveytsman / acmart Goto Github PK
View Code? Open in Web Editor NEWACM consolidated LaTeX styles
ACM consolidated LaTeX styles
Floats are implicitly centered in acmart. Either change this or document it.
Thanks to @krono
For a recent paper of mine (http://dl.acm.org/citation.cfm?doid=2746325.2746327), the ACM Reference format given by the ACM DL (see "Export Formats:" in sidebar of above page) is:
Connor Adsit and Matthew Fluet. 2014. An Efficient Type- and Control-Flow Analysis for System F. In Proceedings of the 26nd 2014 International Symposium on Implementation and Application of Functional Languages (IFL '14), Sam Tobin-Hochstadt (Ed.). ACM, New York, NY, USA, , Article 3 , 14 pages. DOI=http://dx.doi.org/10.1145/2746325.2746327
When printing the bib entry for this paper (see "Export Formats:" in sidebar of above page) using ACM-Reference-Format-Journals.bst
, the displayed format is:
Connor Adsit and Matthew Fluet. 2014. An Effcient Type- and Control-Flow Analysis for System F. In Proceedings of the 26nd 2014 International Symposium on Implementation and Application of Functional Languages (IFL ’14). ACM, New York, NY, USA, Article 3, 14 pages. DOI:http://dx.doi.org/10.1145/2746325.2746327
However, when I recreate the paper using acmart.cls
, the displayed ACM Reference format is:
Connor Adsit and Matthew Fluet. 2014. An Efficient Type- and Control-Flow Analysis for System F. IFL’14, Boston, MA, USA, October 01–03, 2014.
DOI: http://dx.doi.org/10.1145/2746325.2746327
Although full article number and page number information may not be available at the time of camera-ready submission, a reference format should at least encourage the use of "In Proceedings of ".
The \grantsponsor and the \grantnum commands require a sponsor id that "is the unique ID used to match grants to sponsors". It is not clear from the documentation where this ID can be found, or whether the authors can come up with their own IDs, as long as they use unique ones for the different sponsors.
For several institutes that are adjunct to an university, it can be necessary to state both the institute and the university. Also, they might have their own departments.
Which of the following is preferred?
Separation with \\
\affiliation{%
\department{Department of Lunar Studies}
\institution{John Doe Institute,\\ University of San Serriffe}
\country{San Serriffe}}
Separation with \par
(Won't work but consistent with how \institution
works right now)
\affiliation{%
\department{Departmen of Lunar Studies}
\institution{John Doe Institute,\par University of San Serriffe}
\country{San Serriffe}}
Separate \institution
\affiliation{%
\department{Departmen of Lunar Studies}
\institution{John Doe Institute}
\institution{University of San Serriffe}
\country{San Serriffe}}
Alternatively, introduce a new \institute
(might be confusing, admittedly)
\affiliation{%
\department{Departmen of Lunar Studies}
\institute{John Doe Institute}
\institution{University of San Serriffe}
\country{San Serriffe}}
This may be an issue with documentation:
On line 2142 of acmart.dtx, it says: "Use Times throughout".
I'm not sure what the intention of this sentence was, but if the new fonts are installed, those are used, and they are not "Times" (which I would associate with \usepackage{times}), but rather, inconsolata and libertine.
Also, I was wondering about the choice of fonts. Can you explain why inconsolata and libertine where chosen? Previous ACM format used computer modern (if I remember correctly) and this seems a "big" change.
Thanks.
If new fonts are not installed (old distribution), use CMR
Footnotes in abstract do not receive a footnote mark, neither at the reference point in the abstract nor in the footnote text.
Related to #24, since the abstract almost certainly falls on first-page.
Commit b2b1091 / v1.13 integrated some formatting changes requested by ACM; among them were additional information in headers. Currently, the headers for proceedings have conference info and authors for both even and odd pages. In particular, the headers never include the paper title. AFAIK, prior proceedings formats have not had headers with this information. If headers are being added, then it seems useful to include the paper title somewhere (on odd pages seems consistent with the journal formats). Certainly when reviewing a stack of anonymized submissions, the "Anon." author information in the header provides no identifying information and when flipping through a printed proceedings, I think that paper title would be more helpful.
Depending on the requirements from ACM, I would think that paper title and authors in the header and conference info in the footer would provide all of the information on each page, though I think that the conference info is extraneous in most contexts.
An article under review has no journal/conference, DOI, ISBN, publication rights, etc. Don't require/display such information for a paper under review.
As an alternative, reserve review=true
for the line numbering and introduce a new preprint
option which would suppress the copyright block (and be required to enable review=true
).
rt
I hope that we don't have anything weird in our preamble that is causing this problem. Anyway, we always write our papers using \shortcite
where appropriate (instead of \cite
) to avoid the "Foo et al. [Foo et al. 2016]" problem with author-year citation styles. With the new release of the style, we dropped it into our upcoming journal submission this morning. I would expect with the new ACM style and its numerical citations, \shortcite
would output the same numerical citations as \cite
, but it instead appears to be outputting a year rather than a number.
\documentclass[format=acmsmall, review=false, screen=true]{acmart}
local operations can communicate through lower-latency, higher-bandwidth mechanisms
like shared memory or warp-wide intrinsics.
Lindholm et al.~\shortcite{Lindholm:2008:NTA} and Nickolls et al.~\shortcite{Nickolls:2008:SPP}
provide more details on GPU hardware and the GPU programming model, respectively.
We use NVIDIA's CUDA as our programming language in this work~\cite{NVIDIA:2016:CUDA}.
CUDA provides several warp-wide voting and shuffling instructions for intra-warp communication
of threads. All threads within a warp can know about a certain predicate as a bitmap variable
returned by \texttt{\_\_ballot(predicate)}~\cite[Ch.~B13]{NVIDIA:2016:CUDA}. Any set bit
\ in section does not work for sig
Given the use of the caption
package, it would seem that the complementary subcaption
package would be the recommended way to obtain subfigures and subtables. Recommend the subcaption
package in the class documentation and demonstrate use in samples.
While it isn't quite possible to generate a fully accessible PDF from LaTeX (or Word) there are some basic steps to making a 'more' accessible one which are documented here:
https://github.com/sigchi/Document-Formats/wiki/Accessibility
and here:
http://www.sigaccess.org/welcome-to-sigaccess/resources/accessible-pdf-author-guide/
it would be good if the templates did their best at making themselves accessible as well as pointed to a guide on how to make it more accessible (like the SIGACCESS one).
When I try to use ccsdesc, I get the following error:
! Package textcomp Error: Symbol \textrightarrow not provided by
(textcomp) font family ptm in TS1 encoding.
(textcomp) Default family used instead.
It can be worked around by changing the fonts with, e.g.:
\usepackage{tgpagella}
But that defeats the entire purpose of the class file wrt fonts.
NB: Using commit acmart d21923a with pdfTeX 3.14159265-2.6-1.40.17 (TeX Live 2016/NixOS.org), using my Racket Scribble acmart package to compile my asdf2017 article.
Some features of natbib
can only be enabled via package options. In particular, the sort
and sort&compress
package options.
Using
\documentclass[sigplan, natbib=false]{acmart}
\usepackage[sort&compress]{natbib}
...
gives
! LaTeX Error: Command \citename already defined.
Or name \end... illegal, see p.192 of the manual.
It also misses out on the other configuration of natbib
in acmart.cls
(e.g., \bibpunct
and \bibfont
).
As described at http://authors.acm.org/main.html, authors choosing ACM's author-pays option with non-exclusive permission to publish (which corresponds to executing \setcopyright{rightsretained}
in an acmart
document) can also choose to display a Creative Commons License on their works. Currently, there is not an easy way to do so with acmart
and certainly not one that easily incorporates into the permission/copyright block. Perhaps a new \setcopyright{rightsretained-cc}
or \setcopyright{creativecommons}
option?
I download it to my macbook air and try to compile it using LaTex typeset in TexShop, but it could not find the acmart.cls file? I tried different typesets but none of them worked. Where is the cls file? Or is it compiled in a different way?
Note: This refers to the version available on ACM website.
In section 2.10, in description for \citestyle
, it reads
the authoryear format acmauthoryear or the numeric format acmnumeric using
the command \sitestyle
It should be
the authoryear format acmauthoryear or the numeric format acmnumeric using
the command \citestyle
I dislike the hyperref default colors for links. In particular I don't like the citation links (green)—take a look at page 2 of sample-acmsmall.pdf—I find that very difficult to read on the screen, and when printed on a black and white printer, the green is light enough to make the text too light to read on paper. In general fully saturated colors are a bad idea, period.
The hyperref package lets you set all colors as extensions, so it's straightforward to include different colors in the class file.
I am happy to suggest colors for the different link types (there's some nice work in the Siggraph community at least, probably CHI as well), but want to hear the opinions of others. In general I would like colors that are a) dark enough to print well on b/w printers and b) not fully saturated.
I'm not sure the degree to which the amsart.pdf
will be the final documentation for the consolidated class, but, since there are a lot of internal changes from sig-alternate.cls
and sigplanconf.cls
(the two predecessors that I'm most familiar with), it would help to have a concise list of the packages that get loaded in the "User's Guide" section; I know that they are there implicitly in the "Implementation" section, but they are spread out and not where a user should need to go for this information.
For example, I needed to discover that the newtxmath
package was being loaded in order to understand what was changing the \forall
and \exists
glyphs (and that alternates were available).
It would be best if the copyright/permission block were a fixed size, independent of the text placed in the block. This would minimize differences in document layouts when switching from preprint (see #8) to final submission to author's version (see #9).
Perhaps also using a fixed size font (rather than \footnotesize
, which is relative to the chosen body text font size).
The \doi
command defined by acmart.cls
conflicts with the standard "\doi" used by BibTeX.
MWE:
doi-bug.tex
:
\documentclass[sigplan, natbib=true, review]{acmart}
\begin{document}
\title{Title}
\author{Author1}
\affiliation{Affiliation1}
\email{[email protected]}
\begin{abstract}
Abstract.
\end{abstract}
\maketitle
Citation to reference with a DOI: \cite{ref-with-doi}.
\bibliographystyle{abbrvnat}
\bibliography{doi-bug}
\end{document}
doi-bug.bib
:
@inproceedings{ref-with-doi,
author = "Author",
title = "Title",
booktitle = "Booktitle",
year = "2000",
doi = "10.1145/XXXXXXX.XXXXXXX",
}
when one uses the natbib=false
option, the command \citationstyle
is left undefined: line 268 of acmart.cls
. (line 250 in the pdf of the manual).
The issue of the \citestyle{acmnumeric}
should be inside the \if@ACM@natbib
block (lines 159-238 in the pdf of the manual)
If \bibliographystyle
is used in the document preamble, then I get
[mtf@uller acmart]$ bibtex sample-sigconf
This is BibTeX, Version 0.99d (TeX Live 2015)
The top-level auxiliary file: sample-sigconf.aux
I found no \bibstyle command---while reading file sample-sigconf.aux
(There was 1 error message)
This seems to be inherited from amsart.cls
, because I get the same error with that class, but not with a variety of other classes.
I know that documentation for BibTex bibliographies generally show using \bibliographystyle
and \bibliography
together (in the document body), but I thought that was just for convenience, both for writing the documentation and for replacing with an explicit \begin{thebibliography}...\end{thebibliography}
.
Some ACM conferences have artifact evaluation. Artifacts successfully evaluated get a badge. The badge currently goes in the top-right corner of the first page but the new style makes that difficult. Please ensure that there's a uniform solution for "what LaTeX code should conference organizers tell AE authors to put into their papers to display the badge". Previous answer is below (from the PPoPP 2016 organizers), but falls off the right side of the page in the new format.
\usepackage[firstpage]{draftwatermark}
\SetWatermarkText{\hspace*{8in}\raisebox{6.5in}{\includegraphics{img/ae-logo.pdf}}}
\SetWatermarkAngle{0}
A number of packages advertised as being used by acmart
have "Not Found" URLs:
kastrup
: http://www.ctan.org/pkg/kastrupmptopdf
: http://www.ctan.org/pkg/mptopdftools
: http://www.ctan.org/pkg/toolsDefault penalties may present problems when space is at premium (Frank Mittelbach). Document the ways to change them.
There can be a page/column break between the "ACM Reference format:" and the following citation, which looks bad.
Frank Mittelbach recommends deletion of the colon at the end of description label
@article
) have months in the bibliography but conference articles (@inproceedings
) do not. Deliberate?numpages
seems to override pages
. I have a bibtex entry with pages = {12:1--12:10}, articleno = 12, numpages = 10,
pages
then certainly how it's done is correct, but if I do have pages
, maybe pages
should take preference? Of course there are tons of bib entries in digital libraries that don't actually have page numbers and so by default, they just start at one, so you have an entire conference with 100 papers all starting at page 1. (IEEE, I'm looking at you.)There are two problems with CCS descriptors, using version 1.25 of acmart
.
the first issue is that the option printccs=false
of \settopmatter
is not recognised.
The CCS terms do not hyphenate, thus in some occasions they run outside the column margin (and potentially in the right column).
The attached zip file contains a tex file and related log file illustrating the two issues above.
Archive.zip
The ACM Copyright Policy allows an author to use their work (e.g., posting on their homepage), but requires a statement "substantially similar" to the following "in connection with any use by the Owner".
© {Owner/Author | ACM} {Year}. This is the author's version of the work. It is posted here for your personal use. Not for redistribution. The definitive Version of Record was published in {Source Publication}, http://dx.doi.org/10.1145/{number}.
It should be easy to generate such a version, say with a \setcopyright{authorversion}
or some such. Note that the choice of "Owner/Author" or "ACM" as the copyright holder depends on the actual eRights chosen for the paper, so a single authorversion
option is not sufficient.
A footnote on the first-page (and first-column of a two-column format) is set below the copyright text (whereas title and author notes are set above).
Possibly related to #10.
The command
\ccsdesc[500]{something~something}
causes warnings in hyperref (even for "valid" values of something~something):
Package hyperref Warning: Token not allowed in a PDF string (PDFDocEncoding):
(hyperref) removing `\to' on input line 156.
One can test this even with the sample-siggconf.tex
example file.
A possible solution may be to use \textrightarrow from the textcomp package (http://tex.stackexchange.com/questions/229289/arrow-in-text-mode#229290)
I haven't tested this.
Add TOPS and TSC to the list of journals
TSC
\def@journalName{ACM Transactions on Social Computing}
\def@journalNameShort{ACM Trans. Soc. Comput.}
\def@permissionCodeOne{2469-7818}
TOPS:
ACM Transactions on Privacy and Security
FREQUENCY: Quarterly
ISSN: 2471-2566
eISSN: 2471-2574
Frank Mittelbach objects agains full column footnote rule. Consider options.
long line on first column and accept the difference
have the real footnotes above the copyright
move the real footnotes into the second column
The \settopmatter
option printbib
is confusingly named, since it does not affect the printing of the paper's bibliography, but rather the printing of the paper's ACM reference citation. Perhaps rename to printacmref
.
When \@acmMonth
or \@acmArticleSeq
are uninitialized by a document using
a journal format (e.g., acmsmall
), typesetting stops with an uninformative:
! Missing number, treated as zero.
This might happen when an author transitions a conference paper to a
journal submission.
It can be confusing to users if the default behavior includes warnings, especially ones that suggest that their document will be rendered with inconsistent layout.
Package Fancyhdr Warning: \headheight is too small (8.0pt):
Make it at least 11.0pt.
We now make it that large for the rest of the document.
This may cause the page layout to be inconsistent, however.
The \@mkabstract
command has \small\noindent
under the \if@ACM@journal
, so the intention seems to be that the abstract is typeset as an unindented paragraph. However, a leading space makes the abstract appear as an indented paragraph (albeit, with the wrong indentation length).
Should the screen mode select the e-issn instead of issn for journals?
Commit aef0057 reduced left margins for lists. The original margins may have been a little large, but the new margins seem too small. In particular, when a list begins after the first line of an indented paragraph, then the list labels hang to the left of the preceding text; this looks bad (to me).
Another aspect inherited from amsart
is that the left margin and label width are adjusted according to the width of the largest label in an enumerate
, but used by both itemize
and enumerate
. The third-level, which uses lower-case roman numerals in enumerate
gets very large spacing. Again it looks bad (to me) that nested itemize
environments do not get equal increases in left margin/label width when the labels at the different levels are (nearly) the same width.
The example given for \thanks is a bit confusing because it contains the address of the authors, while those should be given in \affiliation. Perhaps just the grant information is sufficient in the example, although grant information must also be given in the \ack, using \grantsponsor and \grantnum.
I had
\acmBadgeR[http://ctuning.org/ae/ppopp2016.html]{img/ae-logo}
but that's not obviously in the latest .cls file (or in the documentation). What's the new replacement for adding a badge/logo to the first page of, say, a conference paper?
As noted in #19, it is best if the selected format
option dictates good defaults for natbib
citation style. However, it would be good if the numbers vs. authoryear could be made with the natbib
class option, allowing the selected format
to make the appropriate defaults for that style. At the very least, SIGPLAN has supported both citation styles, but they dictate slightly different specifics in the natbib
\setcitestyle
invocation.
A proposal is a scaled back version of #19, whereby the natbib
option would support four choices:
false
- does not load natbib
authoryear
- uses natbib
with \setcitestyle{authoryear}
, possibly with format
-specific customizationnumbers
- uses natbib
with \setcitestyle{numbers}
, possibly with format
-specific customizationtrue
- behaves like authoryear
or numbers
as chosen by the format
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.