Re: Some Design Principles

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Hi Ivan,

On 01/12/2015 03:59 , Ivan Herman wrote:
> 1. The sad reality is that academic publishing is very
> conservative (because people's livelihood depends on playing by the
> rules), which also means that, at the end of the day, their
> publication must end up by reputable(?) journals. And, at least in
> 2015, most of those journals are still bound by formatting rules
> that are, though antiquated, nevertheless prevalent (the ACM or
> Springer formats are probably the best known examples in computer
> science or mathematics). This means that any environment based on
> SH can be successful only if it is possible to produce, through
> some clever software, HTML *as well as* PDF formats that abide to
> those rules. Similarly, authors still use Microsoft Word, mostly,
> to author their articles and tools must exist to convert those into
> SH. These formats are indeed arbitrary and, as I said, antiquated
> and often motivated by the not-invented-here syndrome (a good
> example is the absolutely crazy proliferation of various formal
> reference formats in the bibliography). But I believe that we have
> to be pragmatic and not forget their existence.
> 
> Whether these considerations will affect the final definition of
> SH: I do not know. Mostly not, but maybe yes in some places (eg,
> the reference format and vocabulary we use). But I think keeping
> this in our back of our mind all the time is important. Silvio's
> RASH format is a good example for such a full(er) environment, but
> I let him comment on the details

I think that there are at least two primary areas in which these
constraints influence SH:

  – Semantics: whatever SH supports needs to be reasonably compatible
with the data model used by publishers. This, of course, varies; but
we should also expect notable similarities at a high level. Having
written a tool to convert between bibliographic formats I'll attest
that we can't perfectly map to everything just in that area; but I am
confident that we can get "close enough".

  — Styling: the format needs to have sufficient markup backbone that
it can be styled for both online and print PDFs (assuming reasonably
recent tooling). I think that's largely achievable with today's tools.

> (A good example from my own experience: I am part of the steering 
> committee for the WWW201X conferences. Our 'proceedings' has also 
> been published by the ACM, in their digital library, for many
> years, although we also maintain the proceedings, free of charge
> for everybody, on the Web. We would like to have a purely HTML
> based proceedings eventually, and we are seriously considering
> doing that for WWW2017. But it is clear for us that having a copy
> of the articles in the ACM DL is important for our constituency.)

Your use case is the use case of a *lot* of people, certainly of a lot
of societies, journals, and conferences. That's why having a pivot
format is useful: either the ACM will start accepting it, or someone
will have written a tool to convert to whatever they accept.

> 2. We should not look at SH in isolation, but should also consider 
> the environment they would be in. Of course, SH is an interchange 
> format but, on long term, we would also like to see academic
> journals that are fully Web based using SH at their core. As an
> example, I am co-author of a paper at PeerJ CS[1]: as an author,
> the advantage is not (only) that it is in HTML, but also the whole
> process of getting there which was offered (superbly, I must add)
> by PeerJ: the reviewing, the commenting, etc. SH should make that
> easy and smooth. Obviously, it is *not* our goal to come up with a
> standard environment like PeerJ. But, again, design consideration
> should keep that in the back of our mind.

Well, our goals are certainly about the full environment :) But given
a sane model, a lot of the specifics more or less Just Work™ (the
"more or less" part being important). To give an example, we made
https://github.com/scienceai/web-verse to support annotation models,
notably for peer review, but it should work perfectly fine with any
reasonable HTML vernacular.

> B.t.w., it would be good to have people from places like PeerJ,
> PLOS, arXiv, etc, in our midst…

If you look closely you'll see that we have a few already ;) (But yes,
we can always have more input.)

> 3. The issue of archiving came up. I think we should also
> seriously consider, from the start, that an SH, more exactly the SH
> plus the surrounding information, should also be storable, for
> offline usage *and* archiving, in EPUB of some version. Doing that
> means we can provide a proper offline usage for the paper (and
> forget about PDF in that respect) as well as ensure a certain level
> of defense against link-rot.
> 
> EPUB 3 definitely has some restrictions on what can or cannot be 
> included and what format can be used; we should know that. Note
> also that the W3C DPUB IG is working on a more general vision, a
> *draft* called Portable Web Publications[2] that may be the right
> environment to consider in the future. Again: we should keep that
> approach for archiving in mind...

Yes, interop with EPUB 3/PWP/E0 is certainly very important.

- -- 
• Robin Berjon - http://berjon.com/ - @robinberjon
• http://science.ai/ — intelligent science publishing
•
-----BEGIN PGP SIGNATURE-----
Version: GnuPG/MacGPG2 v2
Comment: GPGTools - https://gpgtools.org
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=uurV
-----END PGP SIGNATURE-----

Received on Tuesday, 1 December 2015 16:56:43 UTC