# Workaround for align without using amsmath

I'm preparing a manuscript for a journal that demands a style file (copernicus2.cls) that's not compatible with amsmath. I want to display some aligned equations, but the lack of amsmath means I can't use the align environment. eqnarray's inconsistent spacing offends me, so I'm wondering if there's some kind of workaround that will let me achieve the effect of align without including the amsmath package?

-
@Dilaton I'd like to continue this in chat but I don't know how. Can you create a room? –  Nathaniel Nov 15 '13 at 10:58
@dilaton I agree with you, I think. I'm on a mobile device so can't easily check all the questions you posted, but generally I think those kinds of high level questions can be detagged, perhaps with some modification in some cases. –  Nathaniel Nov 15 '13 at 13:32
@dilaton I think the definition in emilio pisanty's proposal is going in the direction you suggest. I'll do what I can to make sure it gets there. I'll be busy this weekend too though unfortunately. (Many apologies to the mods here for the irrelevant thread, I just don't want to try and figure chat out while I'm browsing on a phone) –  Nathaniel Nov 15 '13 at 13:42

I guess you mean the copernicus2.cls provided by the Biogeosciences. The files can be found on there website.

The class file loads the package txfonts which must be loaded after amsmath. The package txfonts has a mechanism to redefine some internal commands of amsmath which amsmath doesn't have.

You can load the package amsmath as follows to use align:

\RequirePackage[fleqn]{amsmath}
\mathindent0pt
\documentclass[journal abbreviation]{copernicus2}
\begin{document}
\begin{align}
1+1&=2 \\
2&=2
\end{align}
Test
$1+1=2$
\end{document}

EDIT: Set option fleqn and mathindent=0pt

-
Great, this seems to work, thanks! –  Nathaniel Mar 9 '12 at 12:16
Hm, actually there's still one issue, which is that the copernicus2.cls file aligns equations to the left instead of centering them, but the ones I enter with align are still centered. What's the best way to make the align ones match the others? –  Nathaniel Mar 9 '12 at 12:29
@Nathaniel: Please see my edit –  Marco Daniel Mar 9 '12 at 12:38
sorry for being dense - fleqn still indents the align equation a bit, and putting \setlength{\mathindent}{0mm} (or any other value) in my header doesn't seem to affect it. –  Nathaniel Mar 9 '12 at 14:49
Also, mathindent=0pt as an option for amsmath doesn't work either, it just says it's an unknown option. –  Nathaniel Mar 9 '12 at 14:50

@Marco, amsmath does not load amssymb. The problem is two fold.

copernicus2.cls load txfonts.sty which define various \XintY macros which amsmath also define. Since amsmath define these using \newcommand the OP get an error because they are already defined.

This works

\documentclass[journal abbreviation]{copernicus2}
\let\iint\relax
\let\iiint\relax
\let\iiiint\relax
\let\idotsint\relax
\usepackage{amssymb,amsmath}
\begin{document}
\begin{align}
1+1&=2 \\
2&=2
\end{align}
\end{document}

Though Marcos solution may be better, because the solution above may use a different font for those integral symbols than the Journal intended.

-
Please see my edit (17min ago) ;-) –  Marco Daniel Mar 9 '12 at 10:10
better, but the journal should provide a better style (is there still kerning issues with txfonts?) –  daleif Mar 9 '12 at 12:06
@Deleif: A user of copernicus should send a request to Biogeosciences. The kerning issue is the same as usual ;-) –  Marco Daniel Mar 9 '12 at 12:17