Haptics-e The Electronic Journal of Haptics Research

Correspondence with NIH re E-Biomed Project

Date: Wed May 12 08:49:45 1999

From: Roberta L Klatzky 
To: Blake Hannaford 
Subject: NIH input on e media, FYI
Cc: lederman@psyc.queensu.ca
In-Reply-To: <4.1.19990311094225.00a3e930@brl.ee.washington.edu>
References: <4.1.19990311094225.00a3e930@brl.ee.washington.edu>
Content-Type: text
X-UIDL: bc0e7828747f895ad41c1f520f70161e


---------- Forwarded message begins here ----------
>>NIH SEEKING INPUT ON ELECTRONIC PUBLISHING
>>
>>NIH Director Harold Varmus, M.D., has created a Web page to discuss
>>the role of electronic media in disseminating the results of
>>biomedical research.  The site is a follow-up to a proposal
>>distributed by Dr. Varmus, in which he sought input on using the
>>Internet, e-mail, and other electronic media to enhance the
>>publication of research results.   NIH officials hope the proposal and
>>the subsequent discussion will lead to publication techniques that,
>>among other things, accelerate dissemination of information, reduce
>>frustrations with traditional mechanisms for publication, and save
>>substantial sums of public and private money.  The Web site can be
>>found at http://www.nih.gov/welcome/director/ebiomed/ebiomed.htm.
>

Date: Wed May 12 10:19:44 1999

To: Roberta L Klatzky 
From: Blake Hannaford 
Subject: Re: NIH input on e media, FYI

Thanks Bobby!

I'll send a reply based on our Haptics-e experience.

Blake


Wed May 12 10:19:44 1999

Date: Wed, 12 May 1999 09:41:21 -0700
To: execsec1@od.nih.gov
From: Blake Hannaford 
Subject: E-Biomed Comments

Dr. Varmus,

I applaud your proposal for E-Biomed.  As founding editor of 
an electronic journal I have been thinking about the issues 
highlighted in your proposal for several years along with a
distinguished international group of colleagues.  Our new journal
is called Haptics-e, The Electronic
Journal of Haptics Research.  Haptics is the science and technology
of interfacing computers to humans via the sense of touch.  The
Haptics-e web site contains our policy documents and governance
structure.  Currently we have 5 papers under review and none published
yet. 

In forming our new journal, we have come across some complex issues
which we are still working out.  I believe your proposal would be
strengthened by consideration of these:

1) Archiving.

This is perhaps the biggest concern of more conservative scholars. 
Nevertheless it is a valid one.  How can we guarantee that an electronic
document will be around and accessible for 10, 20, 30 and more years?

The risk of data loss due to hardware or software failure can be mediated
by conventional backup and site mirroring techniques.  The bigger and much
thornier issue is that of formats.  Our journal has standardized on .pdf
documents.  This format has many advantages over all others at the present
time.  However, it is highly unlikely that pdf will be the best format in 
10 or 20 years.  At that point when something better comes along, for how
long will pdf readers stay around?  Who will perform conversion of old pdf
files into the new format? As time goes on the conversion issue becomes
bigger and bigger as more and more documents are added.

Digital archives are an active area of computer science research, but right 
now no guarantees can be made.

Our new journal is taking two steps to address this problem now:  

  1) We have a small network of research librarians around the world who
     have agreed to print out and archive hard copy of all articles 
     published in Haptics-e.  Should archiving problems arise in the future,
     we can revert to paper publication. 

  2) We are working with a digital library archiving project at Carnegie
      Melon University to explore electronic means. 

This problem is especially acute for enhancements such as multi-media and
software publishing.  It seems particularly impossible to guarantee the 
future readability of these forms of communication.  Therefore we have a 
policy that in the review process, the publication decision must be made 
entirely on the text and figures portion of the submittal and these portions
must stand alone as a whole.  Multimedia material must be supporting or 
enhancing only. 

2) Resume Building

There is also apprehension among junior investigators that an e-journal will
not be recognized at promotion time. 

We have no solution to this problem and at this point are advising young
academics not to submit work to our journal.  As we get more established and
reputable, we expect this stigma to fade. 

We have also created an innovative relationship with a technical society, 
IEEE Robotics and Automation Society, of "technical co-sponsorship".
This provides something of a seal of approval for the journal but does not
go all the way towards full respectability.

3) Costs

Clearly electronic publication radically changes the economics.  This is why
traditional publishers have been so slow to embrace the www.  Most commercial 
electronic adjuncts charge EXTRA over and above the paper subscription for 
electronic access. 

In fact, since reproduction, distribution are now free, and page
composition can
be easily done by the authors, the true remaining cost is the time spent by 
the editor and reviewers.  This cost has never been accounted for in the old
model.  In fact, the reviewers and editors and their employers have subsidized 
the costs of journals for years. 

The point is that NIH and funders should NOT fund operating costs of journals
UNLESS they fund stipends for editors and reviewers.  Everyone has a computer 
now and the costs of serving an electronic journal on the www are near zero. 
Government funds should not be spent on expensive support staffs etc. they are
not necessary.  

What then should NIH do to encourage e-publication?  Some ideas:

   * Encourage investigators to include www links in grant applications so that
     review panels could get timely updates via the www when making funding 
     decisions.

   * Involve librarians - they are highly tuned in to these issues and 
     have some great ideas. 

   * Coordinate internationally - E-journals will have huge benefits to 
     investigators in other countries (as you point out).  In fact, even
     advanced countries such as western European ones typically have 
     weaker libraries than we do (with respect to scientific journals), and
     so would benefit tremendously.   They should share in the promotion 
     effort. 

   * Support high profile blue ribbon panels which certify the quality of 
      e-journals.  Perhaps there could be an NIH official listing of
      high quality e-journals. 

   * Encourage networks, workshops and conferences of e-journal editors. 

   * Fund research, in partnership with other agencies, on electronic
archiving.



Regards,

Blake Hannaford
Editor, Haptics-e
Professor of Electrical Engineering
Adjunct Prof. of Bioengineering and Surgery
University of Washington
Seattle WA, 98195-2500
206 543 2197


----------------------------------------------------------------------
        Haptics-e, the electronic journal of haptics research.
Blake Hannaford, Editor.                          editor@haptics-e.org
                     http://www.haptics-e.org
----------------------------------------------------------------------

Date: Fri, 14 May 1999 12:29:55 -0400

From: "Exec Sec1 (OD)" 
To: "'Blake Hannaford'" 
Subject: RE: E-Biomed Comments

Thank you for your e-mail regarding the E-biomed proposal.  Your message is
being reviewed personally by Dr. Varmus.  


Date: Fri, 10 Sep 1999 09:11:12 -0400

To: blake2@isdl.ee.washington.edu
From: Judith Turner 
Subject: Re: E-Biomed Comments
Cc: execsec1@od.nih.gov

Dear Dr.Hannaford:

Dr. Varmus has asked me to reply to your e-mail message of May 12 in 
support of the E-biomed proposal.

The hundreds of letters and e-mail messages we have received have 
been very helpful. Through that process the proposal has changed, and 
last week we published the outline of what we are calling PubMed 
Central, the NIH's Web-based repository for barrier-free access to 
primary reports in the life sciences. (You may find it at 
http://www.nih.gov/welcome/director/pubmedcentral/pubmedcentral.htm)

Under this new format the NIH will have no role in determining 
content in PubMed Central; all articles, whether peer reviewed or 
screened, will be deposited by outside organizations. The screening 
will be handled by groups that include at least three grant holders, 
giving authors an opportunity to get articles published that might 
otherwise not be made public, such as those that report negative 
results.

Thank you for sharing your experience with Haptics-e. Clearly you 
have given the issues a great deal of thought. Let me tell you where 
we are on them:

Archiving: Our plan is to accept articles in SGML only, in one of the 
standard published DTDs. We already have software that will convert 
those DTDs to a utility DTD for presentation. The material will be 
stored in our DTD, and we will continually upgrade the DTD to meet 
new capabilities in the technology. SGML was created for just such an 
eventuality, and the fact that this is a living archive in constant 
use will help guarantee that it will be adapted.

Resume building: In part to meet that need, we are depending on 
existing journals, new journals and recognized groups of leaders in 
the field to vet all submissions. Authors will still be able to put 
on their resumes the fact that they were accepted by prestigious 
journals.

Costs: The NIH will allow researchers to use grant money to pay for 
the costs of review, redaction, and formatting. We expect that 
journals and publishers will set prices that will allow them to 
furnish those services and make an appropriate profit, yet be 
competitive for the best articles. We expect that other funding 
agencies in the United States and elsewhere will follow our example. 
We expect that the costs of printing and distributing journals will 
continue to be picked up by library subscriptions for the next few 
years, at least.

Thank you for adding to the discussion. It is through this exchange 
of ideas that the PubMed Central concept is taking shape.

       --Judith Turner



         Judith Axler Turner
         Special Assistant for PubMed Central
         National Institutes of Health
         jaturner@mail.nih.gov
         (202) 986-3463

Date: Fri Sep 10 10:17:38 1999

To: Judith Turner 
From: Blake Hannaford 
Subject: Re: E-Biomed Comments

Ms. Turner,

Thank you for your reply to our inputs on electronic publishing.  I am 
pleased to receive a personalized reply.  I have two questions about 
your new plan:

1) Does the new plan really add much value?  Aren't you now just proposing 
that NIH will run a big server farm?  Aren't the Societies and groups 
that you will work with able to do it anyway?  We can run our entire journal 
on a Linux PC which cost $800.

2) How do authors prepare SGML documents?  So far SGML is not widely 
supported (i.e. by Microsoft Word).

Good luck with this ambitious effort.

Blake Hannaford

Date: Thu Sep 23 09:19:44 1999

To: Blake Hannaford 
From: Judith Turner 
Subject: Re: E-Biomed Comments

Dear Dr. Hannaford:

Good questions! Let me see if I can answer them to your satisfaction.

1) The strength of PubMed Central is twofold: It allows researchers 
to search across journals and across publishers to get information 
they are seeking. Findings from experiments such as PEAK show that 
those who use large databases of material from many journals end up 
accessing articles in journals they otherwise would not normally 
read. Researchers who go to your Linux PC can search only your 
journal; they cannot search a rich database of articles in the field. 
In addition, PubMed Central is a commitment by the NIH to keep a 
perpetual archive of life-sciences research reports. We will adapt 
the entire archive to new technologies as they are developed, 
guaranteeing that no material gets lost in the eight-track tape 
graveyard. That second feature is actually a benefit for publishers 
like yourself who no longer will have to worry about keeping material 
current with the latest technologies.

2) We don't expect authors to prepare SGML documents, we expect that 
journals will do so because it gives them the greatest long-term 
flexibility for work that is enduring and valuable. SGML's strength 
is that it can be transformed into many different outputs: print, 
online, CD-ROM, and technologies yet to be invented. It does not 
require re-keying (as print does), or re-tagging and formatting (as 
typesetting files do). SGML tagging is typically done by the 
publisher because each publisher has its own markup style, called a 
DTD. The cost can be high at first, but most printing companies are 
prepared to work with publishers to make the process easy. And 
several organizations have indicated to us that they want to help 
publishers move their material into a format that will work for 
PubMed Central. We expect that those organizations will start 
publicizing their services soon.

Thank you for giving me an opportunity to answer your questions.

       --Judith Turner

Date: Tue 11:28:32 Sep 28, 1999

To: Judith Turner 
From: Blake Hannaford 
Subject: Re: E-Biomed Comments

Judith,

Thanks again for this response.   Additional comments:

Re: point 1)    Aren't MEDLINE and other indices already providing
this function?  It seems that the real value NIH could contribute
is the archiving - a guarantee that documents will be available
forever.  However don't underestimate this task. Be sure that you 
are advised by top computer scientists.  SGML is not universally 
recognized as a solution to this problem. For example, what about
images, equations,  and multi-media content??

Point 2)  SGML works for existing paper-based journals which have 
investments in professional publishing software which can generate
SGML.  However Haptics-e and other ALL-ELECTRONIC journals are 
really just a dialog among scholars.  There is no staff or money 
for special software so all production is done by the authors. In 
our field at least this is not bothering authors because they are
used to producing camera-ready papers for conference proceedings. 
So, be careful that PubMed Central is open also to this new emerging
type of E-journal that is produced entirely by authors.  The electronic
data formats of these journals are constrained by what authors actually
have in their PCs.

Regards,

Blake Hannaford



Blake Hannaford, Sept. 28, 1999