# CSL 1.0 in the wild

**URL:** <https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895>\
**Category:** CSL Development\
**Created:** [October 11, 2010, 4:54pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895 "2010-10-11T16:54:51Z")\
**Posts on this page:** 13\
**Page:** 1

<div class="post-metadata">

**Author:** ![Rintze\_Zelle](https://yyz1.discourse-cdn.com/flex029/user_avatar/discourse.citationstyles.org/rintze_zelle/32/175_2.png) [@Rintze\_Zelle](https://discourse.citationstyles.org/u/Rintze_Zelle)\
**Post date:** [October 11, 2010, 4:54pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/1 "2010-10-11T16:54:51Z")

</div>

FYI: I haven’t run the software yet, but apparently the latest stable  
release of Mendeley now also supports CSL 1.0:

[http://www.mendeley.com/release-notes/v0\_9\_8/](http://www.mendeley.com/release-notes/v0_9_8/)

and the link below suggests that a fork of the CSL repository has been made  
(see the comment by Robert Knight):

[http://feedback.mendeley.com/forums/4941-mendeley-feedback/suggestions/1121321-american-chemical-society-fetching-is-incorrect](http://feedback.mendeley.com/forums/4941-mendeley-feedback/suggestions/1121321-american-chemical-society-fetching-is-incorrect)

Rintze

---

<div class="post-metadata">

**Author:** ![Carles\_Pina](https://avatars.discourse-cdn.com/v4/letter/c/dbc845/32.png) [@Carles\_Pina](https://discourse.citationstyles.org/u/Carles_Pina)\
**Post date:** [October 11, 2010, 5:10pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/2 "2010-10-11T17:10:12Z")

</div>

Hello,

> FYI: I haven’t run the software yet, but apparently the latest stable  
> release of Mendeley now also supports CSL 1.0:
> 
> [http://www.mendeley.com/release-notes/v0\_9\_8/](http://www.mendeley.com/release-notes/v0_9_8/)

I thought that someone would say before than us… I have a mail in  
the drafts waiting to release Mendeley Desktop 0.9.8.1 to announce it  
here and in citeproc-js.

Also, CSL 1.0 and citeproc-js was in the wild but only web-wise. See  
the Mendeley catalog, e.g.:

> **[Citation Statistics](https://www.mendeley.com/catalogue/daba9dd7-b967-35fb-841a-e06e40519788/)**
>
> (2009) Adler et al. Statistical Science. Rejoinder to "Citation Statistics" \[arXiv:0910.3529\]

(it was properly announced here)

> and the link below suggests that a fork of the CSL repository has been made  
> (see the comment by Robert Knight):  
> [http://feedback.mendeley.com/forums/4941-mendeley-feedback/suggestions/1121321-american-chemical-society-fetching-is-incorrect](http://feedback.mendeley.com/forums/4941-mendeley-feedback/suggestions/1121321-american-chemical-society-fetching-is-incorrect)

At the moment Mendeley Desktop gets the styles from a repository in  
Mendeley Web, the idea is that this repository should be synced with  
the “main repository” that every now and then we discuss here.  
Personally I like that Mendeley Desktop communicates only, as far as  
it’s possible, with Mendeley servers (easier to handle things like SSL  
certificates, uptimes, firewalls in universities/companies, etc.)

I’ll use this email to say thank you once more to everyone,

---

<div class="post-metadata">

**Author:** ![Rintze\_Zelle](https://yyz1.discourse-cdn.com/flex029/user_avatar/discourse.citationstyles.org/rintze_zelle/32/175_2.png) [@Rintze\_Zelle](https://discourse.citationstyles.org/u/Rintze_Zelle)\
**Post date:** [October 11, 2010, 5:51pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/3 "2010-10-11T17:51:14Z")

</div>

Well, there is also Zotero 2.1b1, released September 17th, which uses  
citeproc-js and CSL 1.0 :). But it’s great to see CSL 1.0 and citeproc-js  
are being used in production.

I’ll use this email to say thank you once more to everyone

I’m also very happy to see CSL and citeproc-js being attributed at

[http://www.mendeley.com/release-notes/v0\_9\_8/](http://www.mendeley.com/release-notes/v0_9_8/)  
and  
[http://www.mendeley.com/citationstyles/](http://www.mendeley.com/citationstyles/)

RintzeOn Mon, Oct 11, 2010 at 1:10 PM, Carles Pina \<@Carles\_Pina\>wrote:

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [October 11, 2010, 9:12pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/4 "2010-10-11T21:12:48Z")

</div>

> Hello,
> 
> > FYI: I haven’t run the software yet, but apparently the latest stable  
> > release of Mendeley now also supports CSL 1.0:
> > 
> > [http://www.mendeley.com/release-notes/v0\_9\_8/](http://www.mendeley.com/release-notes/v0_9_8/)
> 
> I thought that someone would say before than us… I have a mail in  
> the drafts waiting to release Mendeley Desktop 0.9.8.1 to announce it  
> here and in citeproc-js.
> 
> Also, CSL 1.0 and citeproc-js was in the wild but only web-wise. See  
> the Mendeley catalog, e.g.:  
> [Citation Statistics - preview & related info | Mendeley](http://www.mendeley.com/research/citation-statistics/)  
> (it was properly announced here)

Out of curiosity, are the different styles generated when they page loads,  
or when one cliks the tabs? Seems pretty snappy, so assume the former (not  
that citeproc-js isn’t fast enough or anything)?

> > and the link below suggests that a fork of the CSL repository has been  
> > made  
> > (see the comment by Robert Knight):
> 
> [http://feedback.mendeley.com/forums/4941-mendeley-feedback/suggestions/1121321-american-chemical-society-fetching-is-incorrect](http://feedback.mendeley.com/forums/4941-mendeley-feedback/suggestions/1121321-american-chemical-society-fetching-is-incorrect)
> 
> At the moment Mendeley Desktop gets the styles from a repository in  
> Mendeley Web, the idea is that this repository should be synced with  
> the “main repository” that every now and then we discuss here.  
> Personally I like that Mendeley Desktop communicates only, as far as  
> it’s possible, with Mendeley servers (easier to handle things like SSL  
> certificates, uptimes, firewalls in universities/companies, etc.)

Still wondering how we move forward on that idea we were kicking around of  
the github-based repo (and get in some of the remaining features on the  
editor).

BruceOn Mon, Oct 11, 2010 at 1:10 PM, Carles Pina \<@Carles\_Pina\>wrote:

---

<div class="post-metadata">

**Author:** ![Dan\_Stillman](https://avatars.discourse-cdn.com/v4/letter/d/ecb155/32.png) [@Dan\_Stillman](https://discourse.citationstyles.org/u/Dan_Stillman)\
**Post date:** [October 12, 2010, 6:36am UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/5 "2010-10-12T06:36:51Z")

</div>

Yeah, I think I’m going to have to echo Carles’s sentiments here. While  
a central repository would be a great resource, I’d be inclined to have  
Zotero still get most of its styles directly from our servers (even if  
those are somehow synced with a central repo), since, among other  
things, that’s where it gets translators from as well, and we can manage  
and optimize style downloading as part of the same process, including  
making decisions about when to push particular versions. This doesn’t  
mean that there couldn’t be a separate updating mechanism for styles  
downloaded from other URLs (e.g., central repo or publisher site), but  
at this point, at least, I don’t think that would be the default behavior.

I’m not sure how best to reconcile support for separate projects,  
though. For example, CSL developers in the Zotero community have made  
various implementation decisions based on discussions in the Zotero  
forums, and they can explain those decisions to users seeking support  
there and make adjustments in response to feedback. I don’t know how  
that will work when multiple projects are using the same master styles.  
(Of course, technically this is already the case, but I don’t know  
whether Mendeley developers make independent alterations to the Zotero  
styles or simply pull new versions from the Zotero repository as they’re  
made available—I don’t think any changes have been committed back more  
than once or twice, so this hasn’t come up on our end.)

So I’d defer to Rintze and other style developers for guidance here.  
There might be some good parallels in other software projects, but none  
spring immediately to mind.

- Dan

---

<div class="post-metadata">

**Author:** ![Carles\_Pina](https://avatars.discourse-cdn.com/v4/letter/c/dbc845/32.png) [@Carles\_Pina](https://discourse.citationstyles.org/u/Carles_Pina)\
**Post date:** [October 12, 2010, 9:27am UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/6 "2010-10-12T09:27:04Z")

</div>

Hello,On 11 October 2010 22:12, Bruce D’Arcus \<@Bruce\_D\_Arcus1\> wrote:

> On Mon, Oct 11, 2010 at 1:10 PM, Carles Pina \<@Carles\_Pina\> \> wrote:
> 
> > Hello,
> > 
> > On 11 October 2010 17:54, Rintze Zelle \<@Rintze\_Zelle\> wrote:
> > 
> > > FYI: I haven’t run the software yet, but apparently the latest stable  
> > > release of Mendeley now also supports CSL 1.0:
> > > 
> > > [http://www.mendeley.com/release-notes/v0\_9\_8/](http://www.mendeley.com/release-notes/v0_9_8/)
> > 
> > I thought that someone would say before than us… I have a mail in  
> > the drafts waiting to release Mendeley Desktop 0.9.8.1 to announce it  
> > here and in citeproc-js.
> > 
> > Also, CSL 1.0 and citeproc-js was in the wild but only web-wise. See  
> > the Mendeley catalog, e.g.:  
> > [Citation Statistics - preview & related info | Mendeley](http://www.mendeley.com/research/citation-statistics/)  
> > (it was properly announced here)
> 
> Out of curiosity, are the different styles generated when they page loads,  
> or when one cliks the tabs? Seems pretty snappy, so assume the former (not  
> that citeproc-js isn’t fast enough or anything)?

I haven’t done the web part, but I’ve just asked and they are all done  
when the page loads.

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [October 12, 2010, 12:04pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/7 "2010-10-12T12:04:35Z")

</div>

> Yeah, I think I’m going to have to echo Carles’s sentiments here. While a  
> central repository would be a great resource, I’d be inclined to have Zotero  
> still get most of its styles directly from our servers (even if those are  
> somehow synced with a central repo), since, among other things, that’s where  
> it gets translators from as well, and we can manage and optimize style  
> downloading as part of the same process, including making decisions about  
> when to push particular versions. This doesn’t mean that there couldn’t be a  
> separate updating mechanism for styles downloaded from other URLs (e.g.,  
> central repo or publisher site), but at this point, at least, I don’t think  
> that would be the default behavior.

I don’t have an opinion on this since I don’t understand the technical  
point that Carles was making. The idea of basing this on a DVCS could  
make my concerns about this impulse ease a fair bit.

> I’m not sure how best to reconcile support for separate projects, though.  
> For example, CSL developers in the Zotero community have made various  
> implementation decisions based on discussions in the Zotero forums, and they  
> can explain those decisions to users seeking support there and make  
> adjustments in response to feedback. I don’t know how that will work when  
> multiple projects are using the same master styles. (Of course, technically  
> this is already the case, but I don’t know whether Mendeley developers make  
> independent alterations to the Zotero styles or simply pull new versions  
> from the Zotero repository as they’re made available—I don’t think any  
> changes have been committed back more than once or twice, so this hasn’t  
> come up on our end.)
> 
> So I’d defer to Rintze and other style developers for guidance here. There  
> might be some good parallels in other software projects, but none spring  
> immediately to mind.

In a year, if someone posts a “I have a problem with X style” question  
on the Zotero or Mendeley forums, I’d like to point them to  
[http://citationstyles.org/styles/x](http://citationstyles.org/styles/x), where there would be a comment  
area. Discussions which also deal with client issues would remain  
where they were.

Bruce

---

<div class="post-metadata">

**Author:** ![Ian\_Mulvany](https://avatars.discourse-cdn.com/v4/letter/i/a9adbd/32.png) [@Ian\_Mulvany](https://discourse.citationstyles.org/u/Ian_Mulvany)\
**Post date:** [October 12, 2010, 2:03pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/8 "2010-10-12T14:03:08Z")

</div>

Speaking from the Mendeley side, we will also want to host a local  
copy of any repository styles, so that we can ensure uptime for our  
users, but we are happy to have these as a mirror of a community  
repository.

We don’t tend to make adjustments to the CSL styles once we are using them.

- Ian

---

<div class="post-metadata">

**Author:** ![Robert\_Knight](https://avatars.discourse-cdn.com/v4/letter/r/ba9def/32.png) [@Robert\_Knight](https://discourse.citationstyles.org/u/Robert_Knight)\
**Post date:** [October 12, 2010, 2:12pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/9 "2010-10-12T14:12:37Z")

</div>

> but we are happy to have these as a mirror of a community repository.

Or even just a cache. CSL styles are identified by a URL so if Zotero  
/ Mendeley servers are asked to fetch  
a style they don’t have they always know where to get it from.

Regards,  
Rob.

---

<div class="post-metadata">

**Author:** ![Rintze\_Zelle](https://yyz1.discourse-cdn.com/flex029/user_avatar/discourse.citationstyles.org/rintze_zelle/32/175_2.png) [@Rintze\_Zelle](https://discourse.citationstyles.org/u/Rintze_Zelle)\
**Post date:** [October 13, 2010, 2:55am UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/10 "2010-10-13T02:55:19Z")

</div>

I think it might be quite a challenge to keep the repository centralized. In  
the Zotero project, CSL styles have been mostly created and edited by a  
relatively small group of people. The learning curve (knowing how to  
manually edit XML, learning CSL, using SVN), combined with either the need  
for a registered SVN account, or peer-review if a style developer commits a  
style for someone else, has ensured most commits to the Zotero Style  
Repository are of a high quality.

If Mendeley’s CSL editor comes into use, I think two things might complicate  
things. First, the editor might output styles that are relatively difficult  
to edit further by hand (I haven’t studied the styles generated by the tool,  
but I expect this to be the case as the editor is mostly item type-based).  
Secondly, there might be too many edits (which will most probably include  
quite a few wrong ones) to easily review. As a result, I’m not sure it’s  
desirable to automatically push styles edited with Mendeley’s CSL editor  
upstream to the mainstream repository (this would be less of an issue for  
newly created styles).

Rintze

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [October 13, 2010, 2:01pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/11 "2010-10-13T14:01:58Z")

</div>

I think you’re absolutely right about potential problems, but that  
your conclusion is wrong. We should consider it a requirement of the  
editor that it can serve as the front-end for a central repository  
such that nobody ever has to hand-edit the styles. E.g. it should  
create “good” styles, and it should be easy to make corrections to  
them in the interface.

Of necessity, this will mean further work on the editor, but I really  
don’t see the point of deploying it at all if it can’t do what I’m  
suggesting.

Bruce

---

<div class="post-metadata">

**Author:** ![Rintze\_Zelle](https://yyz1.discourse-cdn.com/flex029/user_avatar/discourse.citationstyles.org/rintze_zelle/32/175_2.png) [@Rintze\_Zelle](https://discourse.citationstyles.org/u/Rintze_Zelle)\
**Post date:** [October 13, 2010, 2:47pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/12 "2010-10-13T14:47:57Z")

</div>

Right. I guess I’ve become slightly biased towards manual editing of CSL  
styles over the years. I’ll try to test/review the CSL editor a bit more  
thoroughly one of these days to see how it’s coming along.

Rintze

---

<div class="post-metadata">

**Author:** ![Bruce\_D\_Arcus1](https://avatars.discourse-cdn.com/v4/letter/b/4bbf92/32.png) [@Bruce\_D\_Arcus1](https://discourse.citationstyles.org/u/Bruce_D_Arcus1)\
**Post date:** [October 23, 2010, 3:38pm UTC](https://discourse.citationstyles.org/t/csl-1-0-in-the-wild/895/13 "2010-10-23T15:38:03Z")

</div>

So I just want to underline this point …

> > Yeah, I think I’m going to have to echo Carles’s sentiments here. While a  
> > central repository would be a great resource, I’d be inclined to have Zotero  
> > still get most of its styles directly from our servers (even if those are  
> > somehow synced with a central repo), since, among other things, that’s where  
> > it gets translators from as well, and we can manage and optimize style  
> > downloading as part of the same process, including making decisions about  
> > when to push particular versions. This doesn’t mean that there couldn’t be a  
> > separate updating mechanism for styles downloaded from other URLs (e.g.,  
> > central repo or publisher site), but at this point, at least, I don’t think  
> > that would be the default behavior.
> 
> I don’t have an opinion on this since I don’t understand the technical  
> point that Carles was making. The idea of basing this on a DVCS could  
> make my concerns about this impulse ease a fair bit.
> 
> > I’m not sure how best to reconcile support for separate projects, though.  
> > For example, CSL developers in the Zotero community have made various  
> > implementation decisions based on discussions in the Zotero forums, and they  
> > can explain those decisions to users seeking support there and make  
> > adjustments in response to feedback. I don’t know how that will work when  
> > multiple projects are using the same master styles. (Of course, technically  
> > this is already the case, but I don’t know whether Mendeley developers make  
> > independent alterations to the Zotero styles or simply pull new versions  
> > from the Zotero repository as they’re made available—I don’t think any  
> > changes have been committed back more than once or twice, so this hasn’t  
> > come up on our end.)
> > 
> > So I’d defer to Rintze and other style developers for guidance here. There  
> > might be some good parallels in other software projects, but none spring  
> > immediately to mind.
> 
> In a year, if someone posts a “I have a problem with X style” question  
> on the Zotero or Mendeley forums, I’d like to point them to  
> [http://citationstyles.org/styles/x](http://citationstyles.org/styles/x), where there would be a comment  
> area. Discussions which also deal with client issues would remain  
> where they were.

If you want to understand the insanity of the current situation, read  
through all the Zotero forum posts on citation style bugs.

Then go here:

[http://feedback.mendeley.com/pages/4941-mendeley-feedback/suggestions/80941-manually-edit-citation-styles-in-mendeley-desktop](http://feedback.mendeley.com/pages/4941-mendeley-feedback/suggestions/80941-manually-edit-citation-styles-in-mendeley-desktop)

Almost total disconnect, and potential for opening up a massive  
duplication of effort for different communities of CSL users,  
developers, and support people. Let’s stop the insanity please.

Bruce