On Thu, 27 Jul 2017 09:52:55 +0200
Roberto <foice.n...@gmail.com> wrote:

> Hi Scott, thanks for your suggestion.
>   The archive does not really do the job because you cannot edit it 
> directly. My goal is to get something like a MS Word document that 
> contains all the pictures you add in it and you can send it as a
> single file.
> Cheers,
> Roberto

Everybody: If you respond to anything in this post, please please
PLEASE interleave post
(https://en.wikipedia.org/wiki/Posting_style#Interleaved_style) so
everyone knows what question you're answering, what assertion you're
commenting on, etc. Software design requires specificity, and is
brought to its knees by ambiguity. If your phone won't let you
interleave post, for the purposes of descendents of this post, please do
it from a regular computer.

Hi Roberto,

The ability to directly edit graphics within a LyX file would be a nice
*option*: perhaps handy feature for some. Personally I prefer having a
graphics file as a graphics file that I can edit in Inkscape or Vim or
a home-grown shellscript. So if it comes to be at all, it should be an
option: Not a forced conversion of workflow.

Beyond that, please understand what you're asking for. You're asking
for a complete restructuring of LyX native format. Such a format change
will break thousands and thousands of scripts LyX users have written
over the years to automate or augment LyX. Occasionally the LyX native
format undergoes major changes and we script-owners have to change our
scripts, but those native format changes are for much more vital
reasons than the ability to click on a picture and pull it up in the
editor of your choice (or navigate a long, convoluted menu to pull it up
in something else). And I'm pretty sure the change you suggest would be
the single biggest change in history to the LyX native format.

You mentioned LyX being more like MSWord or LibreOffice. Copy a
LibreOffice document to an empty,  unused directory, and perform the
following command:

unzip mydoc.odt

Use an editor to really look at all the files you just created. Keep in
mind that all those files are interdependent,  such that putting an
extra sentence in content.xml often requires changes to the other
files. LyX is a fairly DIY friendly native format. LibreOffice (and I'd
presume MSWord) is "no serviceable parts inside." Implementing your
request would eliminate whole categories of current LyX use.

At one point in the thread you mentioned "ordinary users", and how they
won't adopt LyX as-is. That's been my observation, as well. But as
Abraham Lincoln so aptly put it, you can't please all the people all
the time. If you remanufacture software to please the non-technical
drag and drop crowd, you lose the DIY crowd. And vice versa. There are
things you can do to accommodate both, but changing the native format
isn't such a thing.

And there's this question: Imagine you re-manufactured LyX to act like
LibreOffice but with better output. Would people use LyX, or continue
to use LibreOffice? The software world isn't a meritocracy.

Another question: Do you really want "ordinary users" at all? So far,
this mailing list has had the benefit of people who could succinctly
state a symptom description and submit a minimal working example of a
symptom. Do you really want this mailing list cluttered by people
unable to run a command or differentiate between LyX and a browser?

You bring up an interesting point, and your difficulty in
collaboration is both a real problem and one experienced by everyone,
but the cost of your proposed solution is too high.

Roberto, what I'd recommend is that you start a new project that zips
up a directory in an intelligent way, including only files called for
by the master document and its descendents. Your project can include
ways to prevent overriding a newer directory tree.

Is my suggestion as "integrated" as if it were handled by LyX? No. Is
it as pretty? No. Is it as foolproof? No. Is it as easy to program?
Yeah,  by a factor of fifty or more. I bet you can have a prototype
running in a few hours, and a real project ready in 50. When it comes
to GUIizing the program, try Python Kivy: VERY fast development. Or you
could make it a browser app with Python Bottle.

Steve Litt 
July 2017 featured book: Quit Joblessness: Start Your Own Business

Reply via email to