ok, i've written some posts that've been "on hold",
with others still in the process of being composed,
but i'll let this thought jump the queue for #1492.
***
roger said:
> I've come to the conclusion that
> working at the page-level is sufficient but inefficient, and
> that working at the book-level is efficient but insufficient.
despite the fact that roger hasn't collected nearly enough data
to be coming to conclusions, his statement is _fairly_ accurate,
albeit highly trite and quite misguided.
first, the page-level _is_ "inefficient"; however, it's necessary.
further, the page-level is _not_ "sufficient; there are _plenty_
of decisions which simply cannot be made at the page-level...
roger's done enough postprocessing that he can name a few.
second, the book-level _is_ "efficient"; that's been my _point_
for literally _years_ now. i'm glad roger finally came around.
but nobody has _ever_ claimed the book-level is "sufficient";
i've always said that "smoothreaders" are a vital component.
these "beta readers" can catch stuff that even proofers miss,
because they are reading the book for the _content_ within.
and again, roger's well aware of their valuable contribution.
so much for the "highly trite" constructive criticism.
now let us go on to the "quite misguided" counterpart.
roger said:
> I concatenated all the text files first
> and did the global corrections.
> A lot of improvements were made quickly.
> When I was convinced I had done all I could,
> I burst that back into individual pages
> and loaded it into PPE, the page-at-a-time editor
> which presents the image, and edit window,
> and an analysis window simultaneously.
i find it quite interesting that roger feels the need to
explain the exact characteristics of his "ppe" machine.
perhaps roger didn't monitor this list 18 months ago,
when i presented this screenshot of such an interface:
> http://z-m-l.com/go/triple2010.png
> http://z-m-l.com/go/triple2010.html
(ironically, that screenshot was done as i was working
on the "sitka" book, which had been scanned by roger.)
that's dated march 29th of 2010. but the reason that
it has a "2010" there in the filename is because i had
uploaded an identical version back on july 7th, 2007:
> http://z-m-l.com/go/triple.png
> http://z-m-l.com/go/triple.html
the 2010 version has more buttons, labeled as "buttons",
because i have plenty of ideas for routines one could run,
but note the 2007 version even has a dedicated button for
every word which you might want to add to the dictionary,
which is a nice touch, doncha think? i've been here a while.
not that i claim any originality on such on interface, since
the earliest o.c.r. program i ever used had a similar one...
there are also lots of "obviousness" considerations here...
i am only pointing out that roger is now fishing in a hole
where i have already caught some nice trout for breakfast.
but what's more important is that roger seems to believe
that this interface belongs to the "page-at-a-time" mode.
nothing could be further from the truth!
i use this mode when i'm doing my "book-level" editing...
roger seems to believe that book-level editing can only
be done in a text-editor, using a concatenated text-file,
and that it disallows any reference to the page-images...
wrong!
it _can_ be done that way, but not to its full effect, and
you're always taking a risk if you make an edit _blindly_.
(at times the risk is so minuscule that it would be stupid
to waste any time "verifying" the change against the scan
-- e.g., when you delete the spaces in front of commas --
but there's still a "risk" that any specific edit is incorrect.)
so, like i said, for almost all the book-level editing i do,
i use the very same tool that i use for page-level editing,
an app that shows the text on one side, scan on the other.
(it didn't take me long to abandon the 3-pane interface
for my tools, since all machines don't have huge screens,
though i still use it occasionally on my cinema-monitor.)
so what does differentiate "book-level" from "page-level"?
the way that you navigate the text-file is one factor.
another is the amount of information you can access.
starting with the second one, my book-level tools can
access the entire book. so if you wanna know whether
"car-fare" appears as "car fare" elsewhere in the book,
a book-level tool will be able to inform you about that.
how many times did "chet" mis-scan as "chef"? three.
is "kamon" really "ramon", or another name? yes, no.
"page-level" doesn't even let you ask such questions...
"book-level" encourages them, and then answers them.
but just as important, if not more, is the _navigation_,
or how you approach the text, and handle the tasks...
when i do something at "book-level", i do _that_thing_,
and not any other thing. my attention is laser-focused.
if i am correcting scannos, that's _all_ that i am doing...
if i'm verifying paragraphing, that's the only thing i do.
if i'm checking for italics, checking for italics is all i do.
if i notice something else, out of the corner of my eye,
i might fix it right away, _but_only_ if i won't lose focus!
otherwise, i will just make a note, and continue on-task.
because it is your _focus_ which makes book-level rock.
i might be jumping all 'round the book, from page 16 to
page 242, back to page 88, to page 7, to page 181, 211,
125, 39, and 123. and then i'm done with one task, so
i go on to the next task, which is yet another whirlwind.
but my attention is always solid, because i'm _focused_.
contrast this with the "page-level" approach.
the essence of a "page-by-page" method is that you do
all the tasks for one page, and then go to the next page,
where you'll do all the tasks for _that_ page, and so on...
there's _no_focus_ on a task, none!, with page-by-page.
that's why it's inefficient. it's also why it's error-prone.
without an explicit checklist of "do this, that, and thus",
a person is bound to forget to execute one task or other.
heck, even when you have a laser-focus, you miss stuff!
i did a focused check of the paragraphing in "betty lee",
where that was the _only_ thing i looked at in the book,
and i _still_ managed to make two mistakes nonetheless.
that's why we need more than one set of eyes on a book.
but if you fracture the focus of your proofers, you can have
a half-dozen of 'em check a book, and still have bugs in it.
anyway...
one of my "on-hold" posts is an extensive treatment of
book-level versus page-level, so i'll just finish that off
and send it, rather than say anything more about it now.
i just wanted to show that roger was very misguided here
about what constitutes "book-level" versus "page-level".
-bowerbird