Bill Janssen wrote:

>Joseph Harris writes:
>
>
>>And what share of the download are various formats getting.
>>
>>More hard knowledge of this sort will inform ever better decision making.
>>
>>
>
>I wonder. I think our obsession with formats is sort of dumb. And
>certainly boring. I have to keep track of them for technical reasons
>having to do with my work. But I think in the business of ebooks,
>formats and format wars are just noise.
>
>

Ah, the subconsious spin of value-laden words.

I don't think the discussion of formats, here and elsewhere, can
possibly rise to the level of being deemed an obsession (although it
would be hard to argue that it's not boring). Nor can most of the
discussion of formats be considered a "war". Each format has strengths
and weaknesses, and a frank discussion of those strengths and weaknesses
is hardly a war (although I confess it can sometimes lead to
"skirmishes"). And, as I hope to argue convincingly below, an
understanding of format differences is a central part of the business of
e-books.

>Look at BlackMask.com, for instance. I was looking for a fresh copy
>of JS Mill's Autobiography the other day (my copy seems to have
>bit-rotted somewhat inexplicably), and found it there. In 6 formats:
>MS Reader, Acrobat, Rocket, iSilo, Mobipocket, and zipped HTML!
>
>No one on earth should have any problem reading that ebook. And all
>the books on that site are like that.
>
>

Unless, of course, you own a Palm Pilot, or one of it's derivatives, and
your User Agent is eReader, which can't display _any_ of these six
formats. I suppose you could argue that because MobiPocket and iSilo
both manufacture User Agents for the PalmOS the books are "accessible"
by everyone, just so long as you don't mind
buying/downloading/installing/using the unique User Agent capable of
displaying a particular format.

>Looking at Powell's for Shirley Hazzard's THE GREAT FIRE, I see it in
>Palm Reader (eReader, really), PDF, and MS Reader format, all for
>$2.00 less than the trade paperback price. Again, very readable --
>Palm Reader for my handheld/laptop. Though you really should get all
>three versions when you purchase it -- you probably don't simply
>because of the download size. They should offer it as an option.
>
>

These three formats, among the most limiting of them all, practically
leave me with only three choices for a User Agent: Microsoft Reader,
eReader, and Adobe Reader for the PocketPC. MSReader and eReader and
adequate, if not exceptional, User Agents, but forcing a consumer to
choose between one of these two readers, or to forgo e-books at all, is
not a very market-friendly solution.

>So who cares what the format is? Let the reader decide, based on what
>hardware/software they like to use. Offer each book in the three
>formats Powell's uses, or the six BlackMask uses, and you're golden.
>
>

I'm a reader, and I want to decide. That means if you're a
publisher/e-bookseller you have to give me meaningful choices.

There are a lot of consumers who have not yet made the paradigm shift
from paper presentation to electronic presentation, and they really like
the way PDF mimics paper. And due to a number of reasons, not the least
of which is Adobe marketing, there are a lot of consumers who equate
e-books to PDF. So, if you're an e-publisher/e-bookseller, and you want
the business of those who think electronic images of paper is the best
solution, you have to give them PDF.

I read almost exclusively on my PocketPC. I refuse to purchase (or,
indeed, accept) PDF files for e-reading, and I am aware of a number of
consumers who feel the same way I do. PDF is great for downloading and
printing IRS forms, and is a wonderful desktop publishing format, but
for e-reading, it sucks. So, if you're an e-publisher/e-bookseller, and
you want my business and the business of others like me, you have to
give me an alternative format to PDF.

Microsoft Reader probably has the best typographical richness of all the
non-PDF User Agents (although iSilo is probably a close second), and is
pre-installed on every PocketPC sold. There are a large number of
readers who love Microsoft Reader for these two reasons. So, if you're
an e-publisher/e-bookseller, and you want to tap into that large
pre-installed base of customers, you have to give them Microsoft Reader
format.

Because of their BlurType(tm) technology (which cannot be disabled) I
can't read in Microsoft Reader for more than 15 minutes without getting
headaches (my eyes seem to think that they _know_ they should be able to
focus if they just try hard enough). And, of course, Microsoft Reader
runs only on devices using Microsoft OS's, so owners of PalmOS devices
are unable to read books released in Microsoft Reader format. So, if
you're an e-publisher/e-bookseller, and you want my business and the
business of PalmOS owners, you have to give me an alternative format to
Microsoft Reader.

This is why the format question is _so_ important to the _business_ of
e-books.

Silvia Hartmann has decided, for whatever reason, to only furnish her
e-books in PDF format. The consequence of that decision is that she has
also decided to only pursue that segment of the e-book market which will
accept PDF format. This is not necessarily a bad decision. It is
possible that her books require such typographical complexity that only
PDF will work as an electonic format (IIRC, of all the major e-book
formats, only PDF supports scalable Vector Graphics). It is possible
that her e-books are an adjunct to print books and she uses Acrobat or
some other PDF composing tool to create the print books; in this case
the PDF version is essentially free, but if the market for her
particular genre is so small that the cost of converting and maintaining
e-books in PDF-alternate formats cannot be recovered by tapping into
that market segment, it doesn't make any sense to pursue it. But
whatever the reason, the decision to choose PDF format only _is_ a
business decision, and has _business_ consequences.

Brad Enslen points out that PDF is the preferred format for the RPG
community, "because most gamers will print out manuals for actual play."
(In this context, RPG is _not_ a Rocket Propelled Grenade). In other
words, the RPG market is principally a Print On Demand market, whether
the POD occurs at a fulfillment house or on the player's own printer. An
RPG publisher would choose to publish only in PDF because PDF is an
_excellent_ POD format, and that's what it's consumers want. In this
case, the selection of a format is an important _business_ decision.

Realistically, there are only four formats that offer Technological
Protection Measures in an attempt to prevent copying: PDF,
MicrosoftReader, eReader, and MobiPocket's XDOC (iSilo also offers
encryption, but is really not a player in the market). If you are an
e-publisher/e-bookseller who decides you want TPM (a trade-off which
involves further restricting your market in the hope that it will
generate sales that might otherwise be lost to unauthorized copying --
yet another business decision) you are pretty much limited to these four
formats. Offering PDF format will satisfy a certain market segment, but
will pretty much do nothing to satisfy the small device
(PocketPC/PalmPilot Clone) market. Offering a Microsoft Reader format
will allow you to tap in to the pre-installed base on PocketPC's, but
will do nothing for the Palm market. Offering a MobiPocket format gets
both the PocketPC and PalmOS markets, and also caters to some markets
which are now marginal (such as the eBookMan, and some cell phones) but
the presentation of MobiPocket books is so poor (just a step above
Impoverished Text) that you risk being seen by consumers as providing a
"low-value" product. eReader is clearly the best TPM-enabled User Agent
for the PalmOS market, so if you're serious about the PalmOS segment of
the market, and you are committed to TPM, you pretty much have to
support the eReader format. Selecting among these competing TPM-enabled
User Agents is an important _business_ decision.

None of these format issues can be considered "noise" in the business of
e-books, and Mr. Harris' desire for more "hard knowledge" about format
acceptance is both understandable and useful. Getting from (good) zipped
HTML to non-secure MSReader and Mobipocket formats (and even non-secure
PDF) is a relatively low cost proposition, but getting from RTF to good
XHTML can be fairly costly; if you're starting from RTF knowing the
market acceptance of these end-user formats is important information.
Getting from PDF to any of these other formats can be _very_ expensive,
so if you're starting from PDF it's important to know just how much of
the market you may be missing if you stay PDF-only; can the increased
market penetration justify the cost of conversion? An eReader User Agent
exist for both the Microsoft-based and PalmOS devices; will consumers
accept an eReader-only format if no other format is offered, or will
they reject the e-book completely in favor of some other book which is
more widely formatted? If supporting a wide range of formats is a good
idea from a business perspective, does it make sense to change the
business processes to start from good XHMTL (or more likely, OEBPS) and
derive all the other formats from that (because transformation from
OEBPS to all other formats is relatively inexpensive), or is the expense
of the process change, and the accompanying required paradigm-shift,
simply too costly?

Inquiring business minds want to know.

Now I happen to agree with you about your recommendation about the six
Blackmask formats, although I would add a seventh format: eReader.
Blackmask doesn't offer eReader format because there is currently no
non-commercial way to distribute e-books in eReader format; you could
certainly use html2pml
(http://www.dysfunctionals.org/~networker/html2pml.zip) to convert from
HTML to PML, but the licensing for DropBook does not permit you to share
e-books you have created this way. Even the open-format "Palm" books
offered by FictionWise are PalmDOC format, and not eReader format.

Offering e-books in these seven formats will probably catch every market
niche there is. But in every case _some_ expense is involved in
converting and maintaining a different formats, and it is important to
be able to know whether market demand for a particular format is worth
the expense required.