Gavin Bell - Likes - FriendFeed

FriendFeed is the easiest way to share online. Learn more »

Nat Torkington

New staff find White House in tech Dark Ages - Washington Post- msnbc.com - http://www.msnbc.msn.com/id...

January 23, 2009 from delicious - Comment - Like - Share

Gavin Bell liked this

poor bastards. it's like moving from startup to enterprise. - Nat Torkington

Cameron Neylon

A specialist OpenID service to provide unique researcher IDs? - http://blog.openwetware.org/science...

January 20, 2009 from Science in the open - Comment - Like - Share

Etienne Joly, Mike Chelen, Daniel Mietchen and 48 other people liked this

OK, lets start a movement. Next paper you write use your openID as your affiliation and contact details - Frank

115 more comments

I think I'd look at it the other way: a service providing unique researcher IDs (and additional things like cv, publication lists and network) that also gives you a URL that can be used as OpenID - Jan Aerts

Also, in the more distant future this can be used for microattribution. For example I could "sign" this comment with my openid so that it would be picked up by the above-mentioned service. This could release the pressure a bit of making sure you publish a lot in high-ranked journals: the little things start to count as well. <signed_by http://jandot.myopenid.com> - Jan Aerts

liking the way this thread is going - Neil Saunders

Frank, that's exactly what a number of us are going to try and do I hope... - Cameron Neylon

The author IDs for a paper should be added when the paper is submitted. So it's really the job for the journal publisher. Therefore publishers need added value from this system, or they won't support it. - Martin Fenner

Jan, why not use the functionality of OpenID as the primary ID - a lot of people bring up the example of doi's and crossref - but do people really need a number/code as well as a URL? Actually many of the existing science social networking sites could do precisely this - but I think it gets the thing backwards. The ID is the key thing, the CV page is the sweetener, the microattribution aggregation grows naturally out of the functionality once it exists - Cameron Neylon

Author ID should come from CrossRef, just like the DOI. I'm still not sure if OpenID is the right, as it tries to solve slightly different problems. But the author ID should be (part of) a URL and should replace the email adress that is usually used now. - Martin Fenner

And now I've put the carriage returns back in it should also be much easier to read! - Cameron Neylon

Something we can do now already (or almost in the case of Nature Network) is using OpenID for blog posts and blog comments. - Martin Fenner

Martin, I know we had this conversation but I'm still unsure as to why you feel Crossref is the right solution. It remains a single service dependent on one organisation (that is funded by publishers - which is a bad start for many people) which could go down for financial reasons. The advantage of OpenID is federation - and if the service was built on an open source platform then even better. - Cameron Neylon

OpenID is coming for comments on my blog. We just need it for friendfeed now! - Cameron Neylon

Cameron, I agree that the OpenID can be used as the primary ID, but it's the services around it that will make the difference. An OpenID is a way to prove that you are who you say you are. It doesn't in itself do anything else. Also, I'm not eager to link all my contact details to my openid profile, however - if the service would provide the LinkedIn functionality - I wouldn't mind sharing those details with people in my network. Again: <signed_by http://jandot.myopenid.com> (need a firefox toolbar button) - Jan Aerts

Cameron, if you want the author ID for papers, you need something the publishers will support. CrossRef is a logical choice because of the DOI. - Martin Fenner

Ah ok - got you - yes making the services work with everything else is crucial. And I take the point about sharing all your contact details. Would specific personas for "general" and "submit grant/paper" satisfy that concern? - Cameron Neylon

I'd like to find something that would look like Benjamin Nowack's knowee ( http://knowee.org/ ) " decentralized online social network, OpenID Signup, Consolidated profile with RDF/FOAF export, Personal SPARQL API ,etc... - Pierre Lindenbaum

Would really really want to start building such a service... If only I had the time (or got enough money out of it so it can compensate for days of) and knew that it'd make an impact. Rails, EC2, ... - Jan Aerts

Above all and openID is just an identifier, for an individual (or document describing and individual) Certain publishers may insist that your openID should also resolve to your academic affiliation (hopfully as Pierre suggests via RDF.FOAF )- this would make sense, but it is not what an openID is. This is where the power lies, in that the openID itself is resolvable a DOI is not (although it can be with a certain pre-fix). - Frank

1)The first issue is that we are beginning to recognise the need for a single, common identifier system for researchers output. 2)Secondly it should conform to existing web standards (e.g http, rdf, foaf) and not re-invent the wheel, 3) The more value we can add to the identifier system the more value we get out of it. This is the important one, openID is just an identifier, by adding affilation/contact details CV such as LinkedIN - more information(value) can be resolved via an identifier - Frank

I share the interest in keeping identities separate - not having a knitting pattern come up in your CV, for example. Seems overly simplistic to say it, but if a few major publishers put a line on their submission form with a little red asterisk next to it... voila - Christina Pikas

Christina, you're right about the publishers. The trouble is they can not do that until there is such a service already running and backed up by a large community or authority. What if they'd request an ID at service A, and there's a new better service B a year later? - Jan Aerts

and this is the value of having crossref do it... if the publishers *jointly* control the thing, they'll probably have more confidence that it will last vs. survival of the fittest that sometimes happens with open distributed thingies - Christina Pikas

Journals exist to facilitate our knowledge, they should not control our identity an open identifier platform is the only way forward, journal can choose to implement it or not and researchers can choose where to publish or not. Remember this identification system is to identify digital output, not just journal articles - Frank

NPG/Nature is not 'only' publishing. See Nature Network and Connotea, - Pierre Lindenbaum

I think there is a virtuous circle here. Journals and funders are very keen on rewarding their referees - and this would provide a mechanisms -so I think there is a real chance of buy in whether it is handled by Crossref or an open standard. And I absolutely agree with Pierre and Frank: existing standards, FOAF/RDF and also with Jan that it should be dealt with on commodity services. Jan, I am seriously thinking of talking to a consortium of people about funding this and I know others are trying as well - Cameron Neylon

We all want an author ID, and efforts are already underway (e.g. ResearcherID). The difficult part is how to get there. For me the first step is consensus among the "authors", as they have different agendas than publishers, database vendors, etc. Just some of the questions we should agree on: 1) Use OpenID? 2) Use CrossRef? 3) Centralized or decentralized? 4) Mandatory or voluntary? 5) Only for future work or also for past work? 6) What should be included in the beginning, e.g. only papers? - Martin Fenner

Cameron, librarians would be interested too; don't hesitate to hit somebody up at CNI, which held a meeting on this topic last year. In my own little corner, IRs have a TERRIBLE time keeping author names straight; an author ID would help us a *lot*. - D0r0th34

OpenID has some persistence issues - as indicated in comments above, you'd need a trusted provider to provide a permanent handle. - Richard Akerman

Martin, I think 1) and 2) can be made into red herrings IF systems interoperate (that is, if it's possible to assert e.g. "this ResearcherID refers to the same person as that OpenID"). This definitely adds another layer of complexity, but unfortunately it may be a necessary layer. - D0r0th34

@Christina - I too would fall into the knitting pattern/professional publications quandry, especially as my openID is connected to my blog, which has a dual identity itself: private posts are about knitting etc, and public posts are about work. Would I need two authorID/openIDs, or would there be some form of tagging on the IDs themselves ;) - Allyson Lister

@Allyson. That's why I would like the possibility to _sign_ certain contributions as a researcher (button-click) instead of just providing my OpenID which is used for non-work stuff as well. Researcher ID != OpenID. - Jan Aerts

CrossReg: contributor ID project by CrossRef (start at slide 25): http://www.slideshare.net/CrossRe... - Martin Fenner

Philip Bourne, author of "I Am Not a Scientist, I Am a Number" (among many other things), sits on CrossRef advisory board. - Martin Fenner

interesting Martin - thanks for them. CrossReg is the kind of thing i had in mind here: http://tinyurl.com/9tv5ex - Chris Leonard

I like the idea of OpenID but I've had problems using it - for example on myExperiment every time I log in I have to get it to ring my cell phone and on my wiki sometimes it just didn't work so I reverted back to username/password- is that a common issue? - Jean-Claude Bradley

My blogger OpenID does not work on myExperiment.org at all... some software bug, of which the authors are informed, or so the web page says... - Egon Willighagen

I've never had a technical problem with an OpenID not working but they do do funny things. JC I would suggest you've got some odd security setting in there somewhere. One of the problems is that the different providers are inconistent in what authentication and security measures they provide to the users. FWIW I log into the ONS challenge wiki and myexperiment using a myopenid.com id and have never had a problem - Cameron Neylon

My concern with this being run by Crossref is primarily that this is much bigger than just publishing - the funders really ought to love this - and it will link into many more things. But I will also admit I am uncomfortable with it being handed to traditional publishers, even at one remove, which is probably just a knee-jerk reaction. But the advantage of an openid/FOAF solution is that you can use multiple services - just assert you're the same person - which can help sort out persistence issues as well - Cameron Neylon

Chris's post at http://tinyurl.com/9tv5ex is well worth a look as well - I should have spotted that earlier - Cameron Neylon

Chris, thanks for the link. It's always a sign when many people thing similar things at the same time. I created the tag "authorid" in Connotea and started to collect resources about this topic: http://www.connotea.org/tag... - Martin Fenner

There are also some previous interesting discussions on FF and some old blog posts about this subject. How can we aggregate this information ? - Pierre Lindenbaum

I know : starting a page on OWW... give me a few minutes... - Pierre Lindenbaum

Give the idea an openid? ;-) - Cameron Neylon

I see mentions of FOAF, RDF, decentralized here. That scares me. Even though those approaches would make a lot of sense, they will be far too advanced for the community as a whole (or the provider) to really take off. I think that it should be kept as simple as possible (!= plain simple): a service that assigns an ID, where you can log in using OpenID and that can list your contributions (papers, grants, FF comments...) - Jan Aerts

Jan, that's essentially my idea. The back end can be sophisticated but the naive user just sees it as similar to setting up any online profile (except better because much of the information is automatically aggregated for you). BioMedExperts gets a fair bit of this right and Pierre's FOAF-o-matic is a good example of what can be done easily. Power users can play behind the scenes if they want of course - Cameron Neylon

Added a page about unique-author-ID in OWW: http://openwetware.org/wiki... - Pierre Lindenbaum

It's a good idea but (as Richard, Martin and others also pointed out) (1) persistence is important - the server needs to be around in ten years. Can you guarantee that? (2) uptime - when your server is down scientists all over the world will be unable to login to any services (3) how do you persuade Scopus, Google Scholar, publishers etc. to accept it? - Euan

(just playing devil's advocate) - Euan

Jan, I agree, you cannot ask someone to build a FOAF file. In a perfect world this file would be built with an interface ( eg. http://www.ldodds.com/foaf... , http://knowee.org/ , scifoaf (I like this one ;-) ) ).this repository would host your foaf file or would check if your remote foaf file has changed. - Pierre Lindenbaum

Euan, You're right, this is why I dreamed the NCBI would host this service (a bad side: it would be just for biologists) - Pierre Lindenbaum

Euan - devils advocate is good. I am thinking we have probably about one chance to get this right so solving the problems is important. Ok. Persistence is a good reason to get CrossRef to do it I admit but is also a good reason for using an open standard. 2. Uptime is an issue, a distributed system on a commodity backend would be my answer. Let AWS deal with the hardware maintenance. 3. I think publishers and funders are actually gagging for something like this - not perhaps as logins but at least as tracki - Cameron Neylon

..tracking tokens. The key is obviously driving rapid uptake. If a well known publisher were to recommend including an openid within an address then I think it could work. But this would require getting buy in early and effectively from as wide a range of groups as possible. - Cameron Neylon

Charging a small feel of a dollar per year would help persistence? - Paulo Nuin

Re: the problem of "difficult systems". No-one expects a linkedin user to directly interact with the backend SQL database for instance. The user should never see FOAF/Dublin Core/RDF or even raw html unless they want to. - Cameron Neylon

Two obvious sources of income are providing services that allow funders and publishers to acknowledge their referees and providing services that make it easier for research departments to manage their researcher's web pages. There are probably other services over the data that could make money but my brain is fried. I'd be worried about charging users - uptake is the big problem already - charging would kill it dead surely? - Cameron Neylon

One dollar a year? - Paulo Nuin

Cameron, the "researcher web page" space is alive and well in libraryland: try Cornell's VIVO http://vivo.cornell.edu/ or BePress's SelectedWorks http://works.bepress.com/ or our BibApp http://bibapp.org/ . One of the things we have in mind for BibApp is federating name-authority information, and a consistent ID scheme could fit into that nicely. - D0r0th34

My $0.02, @Pierre, "Euan, You're right, this is why I dreamed the NCBI would host this service (a bad side: it would be just for biologists) " I can report that I made an informal NCBI enquiry a few mins ago and let's just say that things are already being done to help. I can't say more than that. - Graham Steel

Paulo, yes, I think there is some good evidence that any contribution no matter how small really puts people off this kind of thing. It has to really really really easy - finding a credit card (let alone getting a department to transfer the money) will really put people off - remember most of these people don't have PayPal ;-) - Cameron Neylon

This "contribution" can be added to any subscription the researcher my have (journal, society, etc) that would transfer the money to whoever is taking care of the system. Societies are run and managed by scientists, right? And you don't need to have a PayPal to donate money, a credit and debit card would suffice. - Paulo Nuin

That's true - but then would it not be better to put the cost through to the societies, funders and publishers directly? Then they decide how to pass it on? I take your point - I just worry that, as with author side charges for OA, the response to any new direct cost will be a real downer. But I agree there has to be a coherent and plausible business model in place from the beginning, even if it takes some time to start making money - Cameron Neylon

"Societies are run and managed by scientists"? Maybe some of them, but in my experience large societies are mostly run and managed by Association Professionals...with the members having full confidence that they're making the *really important* decisions. Sort of like how faculty senates run universities? - Walt Crawford

This discussion is a very good example why I like FriendFeed. - Martin Fenner

More on the CrossRef Contributor ID thing: http://www.crossref.org/CrossTe... Definately the party to run a contributor ID service (rather than Elsevier or Thomson). - 'Mummi' Thorisson

I don't think this is so much about the author identifiers per se. There will certainly be some kind of author ID for a given (disambiguated) author used behind the scenes by CrossReg, publishers and related services. What is key here is the role of OpenID in the grander scheme of things for authors to identify themselves and initiate linking services together. Potential scenario would... more... - 'Mummi' Thorisson

Wouldn't it be pretty easy to create a mock up on Google Appspot that did exactly this? Allowed people to login with OpenID and add DOIs. Allowed them to add code to their websites to list their papers, etc. I've some code from a related project that could be used if anyone is interested... - Noel O'Boyle

Noel, that sounds basically all it takes... looking forward to a beta! - Egon Willighagen

@Jean-Claude, @Egon - it's a bit late in the discussion, but... I haven't had any problems with my openID, on myExperiment or elsewhere. Mine is a vox.com openid. - Allyson Lister

Noel, we might be interested in your code. What language is it in, and how is it licensed? - D0r0th34

Given Graham's comment ("... NCBI... things are already being done to help") and Gudmundur's one ("... CrossRef Author ID meeting..."), is it a good idea to start to code something 'de-novo' ? - Pierre Lindenbaum

Python/Django (like Appspot). The code is at http://code.google.com/p.... Apparently it's BSD licensed. It's running at publishious.org but don't bother creating an a/c as I never removed the rough edges. As you can see publishious allows you to add your papers to your own website, and export as JSON, and so on. - Noel O'Boyle

Actually Alf already has something very similar up on Appspot, see http://hublog.hubmed.org/archive.... Works off PubMed rather than DOIs but presumably could be modified. - Noel O'Boyle

Cool, nice to see that the publication-list problem list is basically solved in various ways, with existing PubMed query services etc. So, say if these sites were OpenID enabled, they could assist the user in compiling e.g. publication record for her "CV" page viathe more extensive (i.e. beyond what PubMed covers) CrossRef service and cide DOIs rather than PubMed IDs. An important... more... - 'Mummi' Thorisson

Also, to respond to several earlier comments: I don't really see a need for a dedicated researcher/author OpenID provider, when there's swaths of providers with various levels of service & security for people to choose from. - 'Mummi' Thorisson

this is a great thread, if this was just an open author id and not an openid, that would solve the aggregation point for identity / publication. OpenId does give a lot of benefits, but increases the complexity of what you are trying to do. A 40 year lifespan is also the minumum timeframe to think of (new researcher to retirement) - Gavin Bell

Gavin - The reason I think of an OpenID based approach as a solution is because a number of researcher services are adopting it as a login mechanism. But you're right, it is not necessary unless journals, funders, and databases adopt it as a login mechanism rather than a tracking mechanism. I agree it makes things more complicated but the long term benefits would be greater I think. Is there another option that provides an existing open standard for login with the opportunity for federated services? - Cameron Neylon

Gudmundur - the reason I think a dedicated service (or services!) is worthwhile is because researchers do have specific needs in this area _and_ by providing the service is a way to encourage more researchers to use such an ID. I would guess that 99.9% of scientists have never heard of openID and if anyone suggests they get one will just ask why? That question needs a compelling answer that makes immediate sense to them - Cameron Neylon

Other reasons this would be good as a common standard - to ensure that all authors are informed of a manuscript being submitted in their name (as is required by the copyright waivers). Problem may come where undergrad or graduate students are involved - the database would become quickly clogged with possibly one-off entries - and for those who become researchers, they might forget their id if they don't then publish for a while ... (and have moved institution so can't get to their old institutional email - Anna Croft

to recover their password ...) - Anna Croft

The rationale for a author is to help review boards/journals with diambiguating names. One value-add to a publisher is that they'll get more submissions if they provide it and market its value. - Mr. Gunn

Andy Powell makes some good points about the delegation possibilities of OpenID at http://efoundations.typepad.com/efounda... - Cameron Neylon

Cameron - I guess I should correct my statement to say that an OpenID provider for researcher is not required. Yes, I agree that researchers may have specific needs and a dediced service may be helpful. If somebody wants to run it (!). But, the beaty of OpenID is that if I want to use, say, VeriSign as my provider (because of some security feature or whatever), I certainly can. Or if I don't to use an OpenID at all and just log into CrossReg the normal way to manage my profile, that should be an option. - 'Mummi' Thorisson

+1 Cameron. Completely agree re the dedicated service(s). We have to separate *identification* from *authentication*. A central author ID would act as an identifier, whereas researchers could (or could not) use an OpenID to authenticate with that service. See also http://tinyurl.com/7ct4em on identification vs authentication. - Jan Aerts

To continue: I guess my point is that the author/contributor ID concept is rather separate from the issue of whether the user chooses OpenID to identify himself or not. Using OpenID would be more complicated than the usual way, sure, no way around that. But if the 'carrot' is "...but with OpenID you can do all this other cool stuff!", I'll bet a lot of people will make the effort . - 'Mummi' Thorisson

Sorry mummi, disagree there. 99% of the researchers will not be interested at all in OpenID and probably a lot of them (I can think of one or two) not even in an author ID. So to get the completely uninterested ones on board will be difficult enough, let alone explaining to them the virtues of OpenID. That authentication/verification; not identification. - Jan Aerts

(Note: mummi = Gudmundur) - Jan Aerts

Gudmundur - absolutely agree with you - that was one of the main strengths in my view - that if people don't like any specific "science" provider for whatever reason that they have other options. But I think there has to be one agreed standard in practice - one ID somewhere has to be the primary key/token and I guess at core my suggestion is that that be an OpenID rather than anything else - Cameron Neylon

Jan (beer-man!) - yes and no: the identification as the *researcher* corresponding to an identifier (a particular G. Thorisson) is in my mind a seperate issue from me identifying myself as http://mummi.myopenid.com (which I may or may prefer to be publicly known, and may have multiple personae for as well). But yes, the *authentication* part is seperate from either of those two. - 'Mummi' Thorisson

sorry - things coming in fast now and out of order - not sure my previous comment makes that much sense now - Cameron Neylon

Cameron- I agree that delegation is a cool thing and can be used to remove the dependency on a particular provider which may go bust. But it's not something a non-geek user can be expected to configure, unfortunately. Plus some relying parties (e.g. w/ biomedical data on individuals) will probably have a whitelist of providers they accept (e.g. Microsoft Health Vault) so delegation wouldn't work there. - 'Mummi' Thorisson

Let's see if I get this right (I hope that I'm correct here, because that's what I would suggest): the central service could act as an OpenID provider itself. Every researcher gets a unique ID and can use http://centralservice.org/that_id as an OpenID to prove that he is who he says he is on any website. But it's the that_id that is the unique identifier. Unrelated to this, the researcher could use another OpenID (e.g. myopenid) to authenticate at that central service as well. - Jan Aerts

Jan, if some folks aren't interested (or just don't get) the author identifier concept, then that's basically their loss mostly Not much to do about that. I actually think that saying to people that they can use their Google account to log in (openID under the hood) is a better way to get folks to use OpenID rather then a researcher-oriented specialist OP. - 'Mummi' Thorisson

Cameron - my main point is that researchers should be able to participate in the author/contributor ID scheme *without* having to mess about with an OpenID to register with the service if they don't want to. So I guess I disagree with you in that I do not think the author identifier should be the OpenID itself (which is the original topic of this thread, I believe!). - 'Mummi' Thorisson

Cameron I was raising the idea that the two could be separate, identity and a means of accessing services. Persistence is hard to manage over a long time period, so a family of linked identifiers will become necessary me on pubmed, my researcher_id etc. Delegation is a red herring, only geeks care. I agree with Gudmundur, trying to ride author identity management on the back of OpenID usage is laudable, but suffers from the same uptake issues as OpenID has in the wild. - Gavin Bell

http://openid.net/2009... gives some idea of the challenges ahead for the whole web, I would argue that the science community (other than yourselves) are not early adopters of new web technology. OpenID is different, it is as radical a change as RSS and that took 7-8 years to become mainstream. - Gavin Bell

Very good points, Gavin (and many others here). I would however argue about whether RSS is even yet "mainstream", unbelievably, given how fantastic a tool it is, and that even people like me can use it with aplomb. - Maxine

Maxine RSS as a brand is not mainstream, but the technology is embedded in add to news reader buttons, macosx widgets, and iGoogle widgets which use RSS to transfer information without the explicit mention of RSS. The Twitter embed on Typepad is another good example. OpenID needs to get to the same level of invisibility. - Gavin Bell

Invisibility is definitely a key. The more seamless it works, the lower the threshold. I think OpenID could be well-suited for pre-populating any researcher-ID service. For instance, CrossRef could automatically generate contributor-IDs (which are OpenIDs, but invisibly so to non-specialists) which could be claimed and edited. - Björn Brembs

Specifically, e.g., one could use the ID from CrossRef (and the data from their databases) together with code such as the one for Harzing's PublishorPerish (http://www.harzing.com/pop.htm) to generate a page with papers and citations for each author automatically (I'm currently doing this by hand: http://bjoern.brembs.net/citatio...). Every entry would be clickable and editable. - Björn Brembs

Add the possibility to attach RSS feeds and to aggregate review counts and such from funders and publishers and we're talking serious benefits for everybody, with massive pre-population to boot. BTW, there's one service which comes close in principle but is dead because it's not automated: http://getcited.org (great idea, impractical solution). - Björn Brembs

Early support by key players such as CrossRef, Nature, Science, PLoS, Elsevier and the various funding agencies would get this off to a flying start. - Björn Brembs

You may be interested to know that, in the comments to a blog posting I wrote on this recently, Amanda Hill alerted us to the Names project (funded by JISC) - not seen any mention of it so include the link to her comment here. As I say afterwards, 2 years is too long for a prototype to appear... http://blogs.openaccesscentral.com/blogs... - Chris Leonard

Actually, does anyone have any direct contact to the people behind http://getcited.org ? It may be possible to add some automatization to the service (or, alternatively, pull some code if they would let go of it) - Björn Brembs

Embarrassed that I am so late to this conversation. Been in self-impossed writing retreat. Let me just address a few points: a) Yes, CrossRef is exploring this space. b) For those afraid of CrossRef being in the thrall of "traditional publishers", I will note CrossRef members include PLOS, PubMed Central, The Encyclopedia of Life, Hindawi, Jove, OECD, World Bank, some IRs... In short,... more... - Geoffrey Bilder

e) Cameron is right- the author ID problem is "much bigger than publishers". We are talking to researchers, librarians, funding agencies, etc. about what they would require from a service. We were at the CNI meeting and Cliff Lynch is on our advisory board and is aware of our project. f) We too see OpenID is a critical component of the system, but we don't think OpenID and the... more... - Geoffrey Bilder

g) Gumunder described our approach pretty well. We envision creating a repository of profiles. People could use open-ids (they might have a few) or shibboleth ids to authenticate with the service in order to edit their profiles. OAuth and MicroID might be used for other aspects of the service (e.g. profile exchange, blog signing) - Geoffrey Bilder

(Sorry about length there. I'm sure that's bad form) - Geoffrey Bilder

Further blogs from Paul Walk:http://blog.paulwalk.net/2009... Stephen Downes: http://www.downes.ca/cgi-bin... - Cameron Neylon

Geoffrey - being late is much better than not being here at all! Thanks for the detailed commentary. I should note that I don't necessarily have anything against CrossRef, just that I am concerned about any single organisation holding the only key to my identity - I think the possibility of choosing "another supplier" will be key to getting the geeks involved but whether its important to the majority is another question... - Cameron Neylon

More from Euan at http://www.ghastlyfop.com/blog... - Cameron Neylon

We certainly expect others to work on the problem as well. My personal and unfashionable observation is that "distributed" begets "centralized." For every distributed service created, we've then had to create a centralized service to make it useable again (ICANN, Google, Pirate Bay, CrossRef, DOAJ, ticTocs, WorldCat, etc.). This gets us back to square one and makes me think the real issues to focus on are the data portability policies and "living will" of whatever central organization eventually emerges. - Geoffrey Bilder

Geoffrey - great to see you on this thread! I agree about the centralization of key services and the importance of data portability protocols to link them up with one another in a secure way, and with smaller, more specialized services. BTW are you in a position to offer insights into when CrossRef might have something running (even just a pilot), and what that "something" might include? - 'Mummi' Thorisson

I agree, Gudmundur. What needs to be done next and by whom? If there's anything I can do, I'll do all I can to help get this off the ground. - Björn Brembs

And I still also think asking Eugene Garfield if he would support such a service and serve as figurehead would help. I could do that if others also think it's a good idea. - Björn Brembs

My attempt at summarizing this thread: http://bjoern.brembs.net/news... - Björn Brembs

As a follow on from my last comment on this thread, here is what came back from NCBI after I alerted them to this thread a couple of weeks ago:- "Thanks for the heads up. We have actually started moving on this because we're doing something highly related for NIH-funded scientists". - Graham Steel

Follow-up post here: http://friendfeed.com/e... - 'Mummi' Thorisson

Nat Torkington

Grump. Friendfeed-only replies are evil. Cross-media conversations FTL, so you MUST monitor FF. Evil, I say!

June 23, 2008 from Twitter - Comment - Like - Share

Gavin Bell and David Recordon liked this

Listening to @kveton complain about the exact same thing right now IRL! - David Recordon

Aye, see my feature request http://groups.google.com/group... - Edd Dumbill from twhirl

Dimitri Glazkov

Why doesn't Steve Jobs have a FriendFeed account?

June 9, 2008 - Comment - Like - Share

Scott Koon, Gavin Bell and Atul Arora liked this

Because Steve Jobs doesn't tell you what he likes, Steve Jobs tells you what you like. - DeWitt Clinton

6 more comments

point! DeWitt. :) - felix

Kinda like @bret in that respect. The only person I know that can get 95 people to like a freaking Dilbert cartoon. : ) (See http://friendfeed.com/e...) - DeWitt Clinton

1 'Like' for DeWitt - Steven Hodson

DeWitt++ - EricaJoy

We just need another FriendFeed feature: "Steve" User. Users marked as "steves" will like things for you and add items to your Amazon wish list. - Dimitri Glazkov

A while back a buddy and I had the idea that Apple should let people blind preorder whatever it would be that Jobs announces during his keynotes. No discount, just a guaranteed spot at the head of the fulfillment queue. I bet the number of blind preorders would be an accurate leading indicator of AAPL performance, too. - DeWitt Clinton from fftogo

Blind pre-order FTW! DeWitt, you're a genius. - Dimitri Glazkov

The Life Scientists: Neil Saunders

API design - give us some example applications from What's Next forum on Nature Network - http://network.nature.com/forums...

June 5, 2008 from Bookmarklet - Comment - Like - Share

Gavin Bell, Ian Mulvany and Pawel Szczesny liked this

Neglected thread - please comment. - Neil Saunders

3 more comments

hmmm ... here or there :). That's the problem with NN. Things aren't exactly easy to find and take action on - Deepak Singh

And therein lies the problem. Their RSS is still not comprehensive in terms of content and site coverage, either. I don't see any at that forum post, for example. Anyway - please do comment (there!) if you have a moment. - Neil Saunders

There's something about the tags there too. I tag enthusiastically everywhere - except NN. Why is that? - Neil Saunders

I just don't go there often enough. It's never sucked me in :( - Deepak Singh

Gavin Bell › Likes