New Robot from NPG
Cameron Neylon
at:
http://blogs.nature.com/wp/nascent/2009/07/igor_a_google_wave_robot_to_ma.html
Cheers
Cameron
Bruce D'Arcus
This is exactly the sort of fragmentation of effort I would have liked
to avoid by establishing this group. My comment on the blog post:
"I think your going down the wrong path here in tying citations to
particular record ids, in particular accounts. I think the better path
is to have global identifiers (URIs) embedded in the Waves, and for
the robot to be configured with one-or-more trusted sources, which
could include those accounts."
Bruce
Nick Stenning
>
> "I think your going down the wrong path here in tying citations to
> particular record ids, in particular accounts. I think the better path
> is to have global identifiers (URIs) embedded in the Waves, and for
> the robot to be configured with one-or-more trusted sources, which
> could include those accounts."
I'm not sure I'd agree with this sentiment. While globally-unique
identifiers for publications are important, I don't think this is the
place for them. Personally, if I'm mid-sentence and want to cite
something, I'm quite likely to know the author's name or the paper's
title, but the chance of me remembering a DOI or some publishing group
URI for the paper is close to zero.
The beauty of this approach is that each writer manages their own
reference library, with their own tags and folders and whatnot to keep
them organised. The use of a URI/GUID/DOI is in helping communication
*between* people who might otherwise not know if they're talking about
the same paper. There's nothing stopping Igor including such an
identifier in the footnotes/bibliography.
So, I agree that globally-unique-identifiers are important, but I
disagree with the suggestion that Igor is "doing it wrong."
-N
Andrew Lang
Bruce D'Arcus
>
> On Mon, Jul 27, 2009 at 13:51, Bruce D'Arcus<bda...@gmail.com> wrote:
>>
>> "I think your going down the wrong path here in tying citations to
>> particular record ids, in particular accounts. I think the better path
>> is to have global identifiers (URIs) embedded in the Waves, and for
>> the robot to be configured with one-or-more trusted sources, which
>> could include those accounts."
>
> I'm not sure I'd agree with this sentiment. While globally-unique
> identifiers for publications are important, I don't think this is the
> place for them. Personally, if I'm mid-sentence and want to cite
> something, I'm quite likely to know the author's name or the paper's
> title, but the chance of me remembering a DOI or some publishing group
> URI for the paper is close to zero.
As a user-interface issue, I accept that. I'm more talking about
what's going on behind-the-scenes.
> The beauty of this approach is that each writer manages their own
> reference library, with their own tags and folders and whatnot to keep
> them organised.
And what happens if you have ten contributors, each of which use a
different reference management solution?
I don't use Connotea or CUL, nor do I ever intend to*.
> The use of a URI/GUID/DOI is in helping communication
> *between* people who might otherwise not know if they're talking about
> the same paper. There's nothing stopping Igor including such an
> identifier in the footnotes/bibliography.
>
> So, I agree that globally-unique-identifiers are important, but I
> disagree with the suggestion that Igor is "doing it wrong."
If you think it's fine that everyone is forced to use the same
services, then I guess we'll have to agree to disagree. But are you
really saying that? Or are you just talking about the UI issues?
Bruce
* I use a combination of Zotero, delicious, google reader, and my own
home-grown RDF-based solution.
Bruce D'Arcus
> Anybody get it to work? Not working for me.
Alas, I still have no account.
Bruce
Fenner...@mh-hannover.de
If at all possible, I want the references in my Waves to not be tied to a particular reference manager, so that I can use Connotea (or Papers) and Bruce can use Zotero when we work on a document together.
Martin
-----Ursprüngliche Nachricht-----
Von: knowled...@googlegroups.com [mailto:knowled...@googlegroups.com] Im Auftrag von Bruce D'Arcus
Gesendet: Montag, 27. Juli 2009 16:35
An: knowled...@googlegroups.com
Betreff: Re: New Robot from NPG
Andrew Lang
> Subject: AW: New Robot from NPG
> Date: Mon, 27 Jul 2009 16:41:20 +0200
> From: Fenner...@MH-Hannover.de
> To: knowled...@googlegroups.com
Nick Stenning
>
> And what happens if you have ten contributors, each of which use a
> different reference management solution?
From the video, it appears the answer to that's quite clear, you have
to specify the account name as well as the service: (cite citeulike
nickstenning)(cite foo). If you're unhappy with that as a UI solution,
then I'd be interested to know what you're proposing as an
alternative. If you want to get round the fact that neither you nor I
want to go dig out a DOI just to make a citation, when we have
plain-English keywords we can use to identify the paper, you obviously
have to have some method of translating human-speak into
machine-speak. You haven't suggested how to address this problem, and
as a user of citeulike, this problem has already been solved for me.
> I don't use Connotea or CUL, nor do I ever intend to*.
Right, and that's fine, and I'd be disappointed if Igor only ever
supported those two. If this were to be really useful, I'd want it to
pull citations from JSTOR, Pubmed, arXiv, the works...
> If you think it's fine that everyone is forced to use the same
> services, then I guess we'll have to agree to disagree. But are you
> really saying that? Or are you just talking about the UI issues?
First of all, no-one is *forced* to use anything. It's only a matter
of whether or not Igor works for you.
I don't have an account, so I can't tell exactly how Igor's working,
but it doesn't look to me like it's tying references to a particular
service, rather it's pulling them in from a particular service as and
when you ask it to. That's not lock-in, and if it works how I think it
works (and it is of course entirely possible that it doesn't) then
users of different services can collaborate on the same document,
without any problems.
The only "problem" I can see with this model is that people will have
to write adapters for their favourite services. As for you, with parts
of your system stored locally on your machine, you'll presumably have
to find a way to push those entries out to the web. I can't see that
there's anything in principle stopping you adding an adapter to Igor
so you can type (cite brucerdf)(cite Joe Bloggs) -- except of course
that it's not currently obvious how you can contribute to Igor's
source code.
-N
Bruce D'Arcus
>
> On Mon, Jul 27, 2009 at 15:29, Bruce D'Arcus<bda...@gmail.com> wrote:
>>
>> And what happens if you have ten contributors, each of which use a
>> different reference management solution?
>
> From the video, it appears the answer to that's quite clear, you have
> to specify the account name as well as the service: (cite citeulike
> nickstenning)(cite foo). If you're unhappy with that as a UI solution,
> then I'd be interested to know what you're proposing as an
> alternative. If you want to get round the fact that neither you nor I
> want to go dig out a DOI just to make a citation, when we have
> plain-English keywords we can use to identify the paper, you obviously
> have to have some method of translating human-speak into
> machine-speak. You haven't suggested how to address this problem, and
> as a user of citeulike, this problem has already been solved for me.
I think in general it's important, before settling on solutions to
hard problems, to first identify the problem.
But I think my notion of being able to configure a trusted list of
source might work.
It would help if these services could move towards a common, simple, API.
>> I don't use Connotea or CUL, nor do I ever intend to*.
>
> Right, and that's fine, and I'd be disappointed if Igor only ever
> supported those two. If this were to be really useful, I'd want it to
> pull citations from JSTOR, Pubmed, arXiv, the works...
>
>> If you think it's fine that everyone is forced to use the same
>> services, then I guess we'll have to agree to disagree. But are you
>> really saying that? Or are you just talking about the UI issues?
>
> First of all, no-one is *forced* to use anything. It's only a matter
> of whether or not Igor works for you.
>
> I don't have an account, so I can't tell exactly how Igor's working,
> but it doesn't look to me like it's tying references to a particular
> service, rather it's pulling them in from a particular service as and
> when you ask it to. That's not lock-in, and if it works how I think it
> works (and it is of course entirely possible that it doesn't) then
> users of different services can collaborate on the same document,
> without any problems.
>
> The only "problem" I can see with this model is that people will have
> to write adapters for their favourite services. As for you, with parts
> of your system stored locally on your machine, you'll presumably have
> to find a way to push those entries out to the web. I can't see that
> there's anything in principle stopping you adding an adapter to Igor
> so you can type (cite brucerdf)(cite Joe Bloggs)
Re: my above comments, what's wrong with simply {cite: bloggs, some
title} and for the robot to be smart enough to grab it from the
appropriate source?
> -- except of course
> that it's not currently obvious how you can contribute to Igor's
> source code.
An off-list comment from one of the NPG guys suggests it will be open sourced.
Bruce
Frank Bennett
> I don't have an account, so I can't tell exactly how Igor's working,
> but it doesn't look to me like it's tying references to a particular
> service, rather it's pulling them in from a particular service as and
> when you ask it to. That's not lock-in, and if it works how I think it
> works (and it is of course entirely possible that it doesn't) then
> users of different services can collaborate on the same document,
> without any problems.
Exciting to see working code so early. A thought, though. From the
description and the video, it looks like Igor pulls cites
pre-formatted from the services. Blundering forward with that
assumption (which might be wrong) ... I wonder what would then happen
if different users have their service accounts configured for
different citation styles ... or if the group decides to switch styles
in mid-draft ... or if a slice of the document is embedded in another
Wave containing references in a different style.
Bruce D'Arcus
In general, I'd say you want the style to be set for the Wave; not per
user. If you're using CSL, you could simply add style options as you
would add a feed URI to a feed reader.
Bruce
Ian Mulvany
> assumption (which might be wrong) ... I wonder what would then happen
> if different users have their service accounts configured for
> different citation styles ... or if the group decides to switch styles
> in mid-draft ... or if a slice of the document is embedded in another
> Wave containing references in a different style.
independent.
Martin Fenner
providing feedback (bugs and feature requests) to Euan Adie. Drop me a note
(mfe...@wavesandbox.com) if you want to be added to the Wave.
Martin
Am 27.07.09 17:25 schrieb "Bruce D'Arcus" unter <bda...@gmail.com>:
Euan Adie
Igor is just a proof of concept to see what we can and can't do with
Wave from a technical pov, as opposed to any sort of new initiative.
I'll open source the code as soon as I can write better tests remove
the swear words from the comments.
> This is exactly the sort of fragmentation of effort I would have liked
> to avoid by establishing this group
guidelines for stuff like references in Waves coming out of this group
and to share source code for anything we implement. I don't think
participating here can / should preclude any of us trying out our own
ideas, though... in fact I think we need as many demonstrations and
prototypes as possible at this point (er, though admittedly lack of
developer accounts makes this harder).
e
On Jul 27, 1:51 pm, "Bruce D'Arcus" <bdar...@gmail.com> wrote:
> On Mon, Jul 27, 2009 at 3:42 AM, Cameron
>
>
> > Euan describes a reference management Robot that has been put together
> > at:
>