|Electronic version of Datafile|
Message #1 Posted by Bruce Horrocks on 26 Jan 2005, 7:06 p.m.
Gene's suggestion of a PDF version of Datafile is very timely as I was
about to write an article canvassing members' opinions on that very
subject. Let me set out the pros and cons from my point of view as
There are few or no technical production issues - Datafile is currently
produced in Word, converted to PDF and sent to the printer as a PDF
file. The original use of a PDF was simply to improve production
quality: prior to that we had been laser printing masters which the
printer copied from. This worked well enough but we didn't have an
oversize printer capable of putting crop marks on the page so pages were
often not centred correctly etc. Since moving to PDF, that kind of
physical print quality has improved considerably and is extremely
consistent from one issue to the next.
Incidentally, I only use Word because it is familiar software to many
people. If I were ill, I know that I could easily hand over the
editorship to someone else for one issue with few or no problems. If I
were using Scribus, say, instead, then I don't know of anyone else in
the Club who could pick up the role nearly so easily.
Distributing the final PDF electronically, does provide some challenges
whether it is encrypted or not. Assuming an unencrypted version - i.e.
everyone gets the same edition - then it could be downloaded from the
web or emailed out by me.
Downloading is a non-starter because we could not control access. I do
not want HPCC to get into the game of running a secure download site.
There is too much hassle with managing passwords & accounts, dealing
with queries etc. Also people are very intolerant of delay when dealing
with all things electronic. Just look at the criticism that Joe Horn has
received in this forum simply for having the misfortune to fall ill and
be unable to communicate that fact. I wouldn't want be on the receiving
end of that level of criticism nor create another official "role" within
HPCC of "electronic membership secretary" and ask someone else to
volunteer for it if they are going to become the target instead.
Emailing poses its own problems. The (letter) postal service is
exceptionally reliable. We post 120 or so issues 6 times a year, so
roughly 720 issues and over that year we might get one complaint from a
member that their issue didn't arrive. Distributing the same number
electronically, I would expect perhaps 5 or 10 *per issue* to fail or be
rejected in some way - either because the user has changed their email
address, has gone on holiday and their mail box has filled up, has an
overactive spam filter or any one of many, many more reasons. The worst
part from my point of view, is that - from talking to people who do run
mailing lists - many bounce messages do not necessarily indicate which
address caused the bounce. So I would be left knowing that someone
didn't receive their issue but with no way of knowing who it was. Of
course, you as the unsuspecting non-recipient have no way of knowing
this so you get mad with me.
Some of the above problems can be avoided by distributing a version of
Datafile that is individually encrypted to each subscriber. This is
useful in that it prevents re-distribution (other than by printing
multiple copies) since each copy is traceable to the original owner, but
would require distribution of keys to each subscriber, which means a
whole new set of headaches dealing with lost keys because of hard-disk
failures, upgrading to a new PC etc. etc. Also I would have to buy or
write some software to automate the encryption process and be sure that
I can make it work for re-running previous issues when, inevitably,
someone says "My hard disk crashed, could you just send me all of my
issues again please?"
A solution to all of these issues would be to use a 3rd party company to
distribute Datafile electronically and let them handle all the problems.
There is one big advantage to this approach and that is a subscription
need not be for a whole year - people could pay per issue. That would
allow those interested in a particular issue to easily buy just that
issue. The disadvantages are that more money goes to the 3rd party
provider and not HPCC and that these types of provider, if they are to
offer the secure, non-copiable documents needed to prevent electronic
redistribution, usually require the use of their own proprietary,
Windows-only, DRM enforcing software. This would prevent users of non-
Windows machines and older, but still perfectly serviceable, machines
from accessing Datafile.
Lastly, one practical problem is that Datafile is generated assuming an
A4 sized page that is scaled down to A5 when printed. While this means
that European subscribers could easily print their own copies, US
subscribers using US letter paper would have a funny looking copy with
excessive margins and very small text. Producing two versions - one A4
and one Letter is not possible - articles are often tweaked to make them
fit the page size neatly and I don't really have time nor the
inclination to do it twice over.
From an editorial point of view there are a number of problems with
going to an electronic issue, but they are not necessarily
I'm sure you're all well aware that printing has a high fixed cost and
very low marginal costs. Because of this it might take only a relatively
small number of current subscribers to switch from the print to an
electronic subscription to make the whole print edition uneconomic.
Therefore, once an electronic version starts there is a very real risk
that the print version would soon finish. (I suppose that if there were
only a few dozen print subscribers left, I could go down to a copy shop
and get them run off each time. It's the half-way house situation that
is the problem - too many to print individually but too few to print by
a commercial printer.)
Currently, there is a practical limit of 48 (52 at a pinch) pages to an
issue. (Any more and it won't fold neatly as a booklet and really needs
to be bound properly.) If I have extra material I can always hold it
over to the next issue which enables me to smooth things out over the
year. If there were a completely electronic issue then there would be
pressure from authors to "print" their article in the very next issue,
regardless of how big that issue might be and how small the following
one might be.
I could always explain that some readers print their copies as booklets
and so it can't get too large, but upsetting authors is not a good way
to ensure a healthy supply of articles.
If Datafile did go all electronic, could it survive? That's very hard to
answer. I like to think that there is a market for considered,
researched articles as opposed to the majority of the internet which is
instant opinion. However, the museum here has a separate section for
"proper" articles but there is nothing like the volume that appears here
in the main forum. It is very likely that articles would dry up: why
bother writing one for Datafile when I could post it on the forum page
instead? The knowledge that it goes down on paper (and copies are lodged
with the British Library) for posterity must help a little: not quite
like being set in stone but paper may be the next best thing in the
Also, if Datafile did go all electronic there would be pressure for it
to become free - after all there are no printing costs. The trouble is
that HPCC has no other sources of income so paying for the (tiny) costs
of web-hosting our site (www.hpcc.org), the mail redirection for our
email, our domain name even, ultimately become problematic. These might
only add up to one penny per issue, but these kinds of micro-payment are
not easily collectable over the web in small volumes. Admittedly this is
less of a problem for the time-being but still something to think about.
So there you have it, a run down of what might or might not happen. For
me the biggest issue is the one of email distribution; if we host on a
web page then we might as well make it free and that leads to all sorts
Bruce Horrocks (Datafile Editor)