Yahoo Tuning Groups Ultimate Backup tuning Re: Ascii or Html FAQ

Robert Walker wrote:

> Seems there are strong views both ways.
>
> Could it be both?
>
> The program could go through the html and strip all the tags except
> a few recognised ones, such as <SUP> -> ^

That'd work if the original file were marked up with the ASCII
conversion in mind. Note that JavaScript code would end up in the
text file.

The biggest problem would be getting internal links to work. With
text FAQs, you usually quote the section number, and the reader can
search for it. With HTML, you'd obviously make it a link. Presumably

<h2><a name="1.1.7" />Tuning a bouzouki</h2>
...
See <a href="#1.1.7">tuning a bouzouki</a>.

Would be converted to

+1.1.7 TUNING A BOUZOUKI
...
See tuning a bouzouki (1.1.7)

replacing

<h\d><a\s+name="#(.+?)"\s*/>(.+)</h\d>

with

+\1 \2

and

<a\s+href="#(.+?)"\s*>(.+)</a>

with

\2 (\1)

would do the trick (for those fluent in regular expressions). Except
for the conversion to all caps.

> Not sure about ascii diagrams though.
>
> One could devise some kind of mechanism for the original author to
> include graphics and ascii diagrams as alternatives, ditto for
> musical
> examples.

Straight graphics would be a problem anyway, as you wouldn't be able
to download the FAQ and move it to a different machine to read. HTML
is a good, simple markup language and can be readable straight up in a
text editor. The other ideas people are having seem more appopriate
to a website. Which is great, I've had a website for a while now and
it'd be nice to see more of them. A FAQ should be a single file (or a
file for each section).

Or the ASCII art could be in the ALT tag. Sounds a bit hairy.

Incidentally, here
<http://help.yahoo.com/help/uk/groups/files/files-10.html> is where it
says that we can automatically e-mail a FAQ to all new subscribers.

The ultimate would be to use our own XML language, and convert to
either text or HTML. But that's probably more complex than we need.

Graham

🔗J Scott <xjscott@earthlink.net>

Hey all, thought I'd register my vote on this issue:

John ff:

>Ho hum. Just when I though that the FAQ might be useful to
>me it seems that it is to be in hypertext, which makes it
>hard to take home and read. OK so I am in a minority, but
>I am not usually attached to the network when I read
>e-mail, and very rarely do I use the web for anything,
>being visual, gaudy, required clicking and all kinds of
>other things from which I prefer to protect myself.

Dave:

>I'm with you John,

> Hey guys, how about making the FAQ plain ASCII with explicit
> URLs? They get turned into links by most email readers anyway.

I am in **100% agreement** with John and Dave and am not
certain they are an actual minority. I myself prefer to
read my FAQs in the bathtub and would like to do so
without electricuting myself. I also require the ability
to mark the bejammers out of everything with my colored
pens, underlining things, drawing arrows and stars, and
writing 'crackpot!' and 'huh?' all over the margins!! :-)
(As well as scribbling out the math for myself to see if
it has any validity.)

Seriously, I love the FAQs that are out there that come in
6 parts each 1 MB long and just ramble on about the topic
from various perspectives at length, interspersing actual
FAQs with actual Best-Ofs with actual FQAs. Some sections
might have multiple definitions like a dictionary does;
other sections might read like a US Supreme Court decision
-- with majority and minority opinions containing
concurrences and dissenting viewpoints. Also, they are
usually structured something like:

1.1.0
1.1.1
1.1.2
1.1.2.1
1.1.2.2
1.2
1.2.1
2.1

etc. -- so that you can say: "for mare about this, see
section 9.5.2.3 -- 'Throat Warbling Temperaments'"

On the otherside, I have no use for HTML FAQS at all and
avoid them like the devil dogs of satan they are :-)
unless they at least have a 'click here to download this
entire site as a .zip file' link like the Monz's.

The problem with hypertext is that it is nonlinear and
cannot therefore be read -- it can only be browsed, and
usually in a trance-like state of confusion. Hypertext
links might be ok to get to section 9.5.2.3 from 7.1.4 but
that leaves you with the problem that either the files are
too big and cannot be easily jumped to without waiting
forever, or are too small and cannot easily be read
straight through.

This is a real problem for those of us who do not have
always-on wireless T3 connections like you fat cats at the
universities! :)

Even so I think a HTML version would be OK for those
who want to maintain it -- but I would definately want to
see a standard USENET type version as well. I think
it would be easier to keep things in ASCII as
the basic raw format. Then those who want to can make up
hypertext sites using whatever format and organization
they fancy.

It seems obvious to me that the job of the editor
is difficult enough to produce a text version.
If the editor has to impose some sort of multidimensional
graph on the whole thing, then the task becomes onerous
and unpleasant.

Also, I do not want to be learning new revision control
systems using untested cutting-edge ASP technology just in
order to be allowed to write faq entries. And I certainly
am not going to be trying to mess around on Yahoos ultra
slow site rooting around in some weird system for draft
faq entries to read and offer comments on. How do I know
it is weird? Cos the simple directions and explanations
offered on how easy it is to use are totally weird! It's
fine for technology addicts (no disrespect intended) but
the FAQ should be open and accesible to everyone, not just
the technology fans.

I hate having to be online! And I have 21 years of
programming experience! So it's not because I am too dumb
to figure it out! (I also refuse to learn to program most
VCRs, to type in vi ("at the center of evil you find vi" -
js) or to mess around with any more badly designed
technology -- I'm more productive ignoring it all and
sticking with what works rather than thrashing around like
an overloaded network server.)

Personally, my plan is that if there is no text version, I
will cut and paste one together myself and put it on my
own website when I get it up and running. And I think
everyone who wants to should also do the same -- grab the
faq and put it on their own sites with a note at the start
saying 'the latest version of this document may be found
at http://wherever.com/tuning-faq.txt'.

I think the interactive stuff and the diagrams and music
examples are really great ideas and that links to them
should be included in the FAQ, but with most of the
discussion in those links being replicated in the FAQ
(minus the examples). That way people can at least know
what they're missing without having to waste time screwing
around on the internet endlessly clicking and not being
able to absorb any ideas because of the distractions of
following links in a nonlinear haphazard arrangement,
which describes almost everything on the web. In the old
days people organized and structured material *a little
bit* so that it could at least be read. With hypertext,
people just randomly connect stuff and there is no
cohesiveness to any of it. I do not fool myself into
imagining that we need a cohesive voice in our FAQ (in
fact I am opposed to it), but I think that we should at
least be able to read the FAQ straight through and bring
it together that way since I know it will not be brought
together by any sort of unifying agreement on the theory,
practice and meaning of musical art. And if it does then
it will be of limited usefulness. (!)

(I should point out here that time and time again I hear
crackpot tuning theories and then I hear the music
realized in the crackpot tuning theory and am surprised to
find that it is great stuff and so maybe the crackpot knew
more than I did! The point of this digression is that
totally incompatible and contradictory theoretical systems
can produce great art.)

As far as automatically mailing the faq to new subscribers
goes, you want to be careful. The faq could reasonably be
several megabytes in length, which even if broken up might
overflow the mailbox size limit (and even automatically
incur additional charges) as well as severely irritate the
hapless recipient. I believe proper net etiquette is that
large files should only be sent by explicit request. The
instructions could be put in the "Welcome to tuning!"
letter that goes out:

The faq for this group may be accessed as a 15 MB file
at:

http://wherever.com/tuning-faq.txt

Or as a hypertext document at:

http://wherever.com/tuning-faq,html

It can also be downloaded in 15 chunks:

http://wherever.com/tuning-faq-1.txt
http://wherever.com/tuning-faq-2.txt
http://wherever.com/tuning-faq-3.txt
...
http://wherever.com/tuning-faq-15.txt

You may also request it be emailed to you by sending
an email to listproc@wherever.com requesting the faq
sections you want by placing any of the following in the
body of your message:

SEND tuning-faq.txt

SEND tuning-faq-1.txt

SEND tuning-faq-1.txt
SEND tuning-faq-2.txt
SEND tuning-faq-3.txt
SEND tuning-faq-4.txt

etc.

(The subject line of your email message is ignored.)

-----

Joseph:

> Anyway, it seems that various people with a significant computer
> prowess on this list are really disagreeing on the format... so
> probably it will have to come down to some kind of vote.

> I, personally, have no problem with the simple ascii (there is a
> new "ascii is sexy" movement just started), but would feel
> strongly that it should have significant web links... to
> lattices, VRML pages, sound examples, etc., that take present
> graphic and web sound technology into account...

Yes, I agree totally with you!! ASCII is hot & sexy
because it downloads fast and is forced to contain content
rather than fluff and eyecandy and clicks. Practical
_is_ sexy! (Joseph, where did you hear about this sexy
ascii movement anyway? I want to join the vanguard!)

HTML is a high maintenence trophy princess. ASCII is a
sturdy woman with large hands and a strong back who knows
how to bake bread, plant a garden and rope a bull. You
will have a much happier life with the latter.

- Jeff

🔗JSZANTO@ADNC.COM

🔗John A. deLaubenfels <jdl@adaptune.com>

🔗JSZANTO@ADNC.COM

🔗Robert Walker <robert_walker@rcwalker.freeserve.co.uk>

Hi Graham,

How about the idea that anyone can add anything, and then the FAQ
gets selected from within the site.

Then at that point, if an entry isn't suitable for use in ascii
as it is, the author can go through and mark out which sections
to leave out, name the bookmarks, and add alternative ascii graphics
if necessary and feasible.

One could then mark ones file as ready for the FAQ by putting
FAQ
as the first three letters in the file, or the first three letters
of text in html (i.e. the first non tag immediately after the <BODY>
statement)

Then the program could make a sub-tree of all the entries marked in
that way, and use that for the ascii faq.

The keeper of the FAQ could then edit it at that point, if
necessary, moving folders around etc, before it is finally output
as html or ascii.

I suggest that both types of entry be encouraged - ones suitable
for an ascii faq, and ones that might be especially welcoming
to a newbie who uses (and perhaps likes) a web browser.

> Or the ASCII art could be in the ALT tag. Sounds a bit hairy.

Probably best idea is if the markings are visible to the author
editing the html.

One very simple technique would just be to mark off the
sections in the text - don't even need to edit the html tags:

Just need some reserved sequence of symbols for it.

Anything unlikely to be used.

How about
/*-------------
(any number of dashes more than 10 and nothing else on the same
line)

-----------*/
for ascii section,
and
/-****************

*****************-/
for the graphics section.

Then to process it for the text faq, remove all graphics sections,

To process it for the graphics, remove the ascii sections.

Could do same for bookmarks.

In html there's something of a tradition of using words for
bookmarks, as one can then re-arrange sections
without having to re-number all the bookmarks.

So if it was an html document, then at the end, maybe only once it
has been volunteered, and selected for the ASCII FAQ, one would just
go through it and add the section headings by hand as

/*1.1.7*/ <h2><a name="tuning_a_bouzouki" />Tuning a bouzouki</h2>
...
See <a href="#tuning_a_bouzouki">tuning a bouzouki</a>.

Here one wouldn't need to look at the raw html - for those using
a html authoring program, all they need to know is to put the
ascii section heading
/*1.1.7*/
before the bookmark flag.
and do that for all the bookmarks in the file.

Rest will then happen automatically.

In ascii it would become

1.1.7 TUNING A BOUZOUKI
(supposing one decided to make h2 correspond to caps)
...
See tuning a bouzouki (1.1.7)

Program knows to substitute (1.1.7) for the #tuning_a_bouzouki
bookmark in ascii because /*1.1.7*/ is the ascii "bookmark" that
immediately preceded it.

The other way, if one was editing ascii, one could mark sections of
it as bookmarks to convert into links in the html version using

/*1.1.7*/ TUNING A BOUZOUKI

....
See tuning a bouzouki (/*to 1.1.7*/)

The /*, */ and /*to will be removed in the ascii faq, and they will
all
be turned into bookmarks and links in the html version

Program doesn't even need to know if the text is ascii or html (e.g.
if you want to paste from an ascii document into an html one)

When a line has a bookmark and an ascii heading on the same line,
one can use either to refer to it later, and if it has just the
ascii, one can use that.

Similarly, one can either use /*to ...*/ or the html ref.

One possible problem to anticipate here.

The /* may occur in an ascii diagram.

I have something of a fondness for /* */ because of c, but of course
one could use any convention.

Maybe some other sequence would be better. What is unlikely to be
used I wonder? How about %%
%%1.1.7%%

Basic assumption behind all this is that while the drafting
is in progress, one keeps the originals of all the
documents. E.g. in a folder that one excludes from browsers in one's
robots.txt file.

However, I'd suggest the program also maintains a backup of
the Ascii/ html combined document, e.g. in the same location
as the html, and same file name but with another file extension,
say, .ahtml (ascii / html)

Once finished, one can delete the file or move it if one wants
- the program will just ignore any urls in the database
that it can't locate.

Presumably it would do a regular backup of the entire site too, e.g.
on a local disk or another site, from which it can be recovered if
necessary.

If one wants to delete an entry from the tree, one could do a
mechanism for that too.

just prefix the first field with
[delete]

(do a second copy of the entire entry if you can't find the original
of it in the database once it gets large)

That's just for a single entry.

Again, that would only delete the reference to the entry, in case of
a mistake the file itself will still be there.

At least those are a few thoughts on how it could work.

Robert

🔗John A. deLaubenfels <jdl@adaptune.com>

[I wrote:]
>>My, my. Such eloquence in favor of Ascii FAQ's! As I have already
>>said, I prefer Ascii myself, when it does the job. My advocacy of
>>html is not, as one list member would have it,
>>to "impress...somebody";

[Jonathan Szanto wrote:]
>Hey John,
>
>Yo! I can take any finger-pointing, so feel free to include my name
>on the above "impress..." quote!

OK, it was _you_ who wrote that, Jon.

[Jon:]
>But just to be clear, that was just a little 'aside'. My main point
>(and only point, really) is that the content of the FAQ is going to
>have to get written up *no matter what*, and that is all going to
>come in a very plain, text format. Once there, it can be poured into
>whatever type containers you like: .html, .pdf, anything.

I would, however, like to avoid the complication of supporting two or
more formats, unless others would agree to volunteer as well.

[Jon:]
>A very plain format has benefits outweighing it's first-impression,
>well, "plain-ness": a good place to start; the content to place in
>more elegant containers; a basic format very easy to deliver (on a
>web page, in a .zip or .sit. or [name your archive] file, for
>a 'printer-friendly' page).

See below.

[JdL:]
>>I must confess that I do not understand the fear and/or loathing
>>that hyper-links seem to inspire in some people.

>Unless I am mistaken, didn't you fess up to very little web design
>experience?

Quite so. And, as part of my very minimal knowledge, I know that it
is possible, using the <pre> tag, to produce nice basic text (with a
fixed-pitch font, for those lattice diagrams) which, however, can
contain links (by name) directly to any other place. In this format,
ordered insertions are easy, since sequential numbers are unnecessary.

[Jon:]
>Having maintained and designed websites for over 5 years,
>I neither fear nor loathe them, but *respect* the extra work
>involved. Links can be added (and constantly watched for changes);
>graphics can be developed, etc. etc. Just seems to make sense to
>start from first principles and then go from there.

But, you don't seem to appreciate my concern that maintaining an Ascii
FAQ (with inevitable insertions) will be _more_ work than maintaining
a very simple html FAQ. Believe me, I'm not looking to make my life
more challenging over this; quite the opposite, along with my goal to
provide the most useful FAQ possible.

JdL

🔗Jay Williams <jaywill@tscnet.com>

🔗Robert Walker <robert_walker@rcwalker.freeserve.co.uk>

> Hi Robert!

> How would one put in a simple "question and answer" kind of FAQ in
> this database??

Sorry, forgot to explain that.

I made a folder called faq_tree in the TL files section.

So all one does is to make a file in that folder with one's
ascii question and answer.

Then reference it in the database.

Because of an eccentricity of the java tree control, best to make
files with no spaces in the names - replace by underlines if nec.

If one has several entries to add, or might do, it would be best
to make a subfolder in the faq_tree, to make them easier to find.
- again make sure the subfolder has no spaces in its name.
E.g. robert_walker rather than "robert walker"

E.g. suppose you want to start a new section called notation.

To make it, add an entry in the database

folder; short descr. (of sub folder this time)

main: 3/mar/2000 notation;

3/mar/2000 here would be the date one added it.

Rest of the fields are left blank

Now suppose you want to add a FAQ entry for 19-tet notation.

The fields would be

3/mar/2000 notation
19-tet notations
How sharps, flats, half sharps and half flats are notated in 19-tet
groups.yahoo.com/group/tuning/files/faq_tree/josephpehrson/19tet_notations.txt
Joseph Pehrson

Thing to be careful about is to get the name and date right for the
folder as that is the only way the program will know where to put it

Now you need to upload the file 19tet_notations.txt
to faq_tree/josephpehrson in the files area
(where I'm assuming you've already made that folder)

And you are done, that's it.

Any question? Why not try a test entry, - we can always delete
it if it doesn't work.

>
> From your discussion, I am under the impression that you could,
> pretty much automatically, turn any entries in the database into
an
> HTML tree. Is that correct??

Yes, with not much in the way of programming. Once I've written
the program, won't be hardly anything one needs do except run
it now and again to refresh the tree.

(except if one needs to re-arrange the folder structure, which I
don't imagine will happen often)

The nice thing is that the draft version of the tree directly links
to the original files, so is always up to date, and you can edit
your solutions at any time.

The tree also needs to be remade properly on-site, so that one
can see the urls, in order to cross-link between sections,

The links between different pages of the tree will only work in the
on-site tree, which will be a little while behind the date of the
up to date one, depending on how often the tree gets refreshed.

> Anyway, it seems that various people with a significant computer
> prowess on this list are really disagreeing on the format... so
> probably it will have to come down to some kind of vote.
>
> I, personally, have no problem with the simple ascii (there is a
new
> "ascii is sexy" movement just started), but would feel strongly
that
> it should have significant web links... to lattices, VRML pages,
> sound examples, etc., that take present graphic and web sound
> technology into account...

I like the look of ascii graphics too. Reminds me of the days when
people did ascii art, and you got your computer print outs with
each letter of your name large, and made out of the same letter
to a smaller size.

Robert

🔗JSZANTO@ADNC.COM

John,

--- In tuning@y..., "John A. deLaubenfels" <jdl@a...> wrote:
> But, you don't seem to appreciate my concern that maintaining an
> Ascii FAQ (with inevitable insertions) will be _more_ work than
> maintaining a very simple html FAQ. Believe me, I'm not looking
> to make my life more challenging over this; quite the opposite,
> along with my goal to provide the most useful FAQ possible.

I'm sorry that I gave that impression -- believe me, I want whatever
solution to be the least amount of work as possible, for the person
in charge of it. I haven't had too much time to follow this, but in
just reading a couple of Robert Walker's posts, I can't believe the
complexity that is being built up -- all in the name of being simple!

(Not a slight on Robert's skills, but maybe it's a programmer's
solution [speaking as a programmer myself!]).

The only other thought I had was this: the whole thing had started
with people like Daniel Wolf submitting a couple of drafts in an
email message. Would it be possible to take each FAQ topic as a
separate text file, and then have an index of all the questions? Each
topic, as a page, could contain links and would always have a link
back to the main index page. If there were additions/changes to the
text files it would be easy to incorporate them, and there are a lot
of tools available to take a very plain text file and convert it to a
simple .html file. Then the entire directory could be zipped for
download, and, once opened, you'd only need to open the
index.html/FAQ.html page in the browser and then browse to all the
topics.

(If you'd like, I could put up a dummy version of this idea on a temp
page on my website, as an example...)

Of course, I would always end with: the format chosen should match
what the editor of the FAQ would like to work with -- anything else
is a burdon on the volunteer effort. So, John, if you would want to
work with .html and there isn't some landslide in another direction,
that's what you should do. I'd be happy to point to .html resources
to make the job easier if you like (for instance, quite a few
freeware HTML-2-ASCII convertors...)

All for now,
Jon

🔗jpehrson@rcn.com

--- In tuning@y..., "Robert Walker" <robert_walker@r...> wrote:

/tuning/topicId_19651.html#19677

> > Hi Robert!
>
> > How would one put in a simple "question and answer" kind of FAQ in
> > this database??
>
> Sorry, forgot to explain that.
>
> I made a folder called faq_tree in the TL files section.
>

Hi Robert!

I just see a folder called "faq" NOT "faq_tree" in the files section.
Am I missing something??

> So all one does is to make a file in that folder with one's
> ascii question and answer.
>
> Then reference it in the database.
>
> Because of an eccentricity of the java tree control, best to make
> files with no spaces in the names - replace by underlines if nec.
>

OK, I just tried that...

>
> 3/mar/2000 notation
> 19-tet notations
> How sharps, flats, half sharps and half flats are notated in 19-tet
>
groups.yahoo.com/group/tuning/files/faq_tree/josephpehrson/19tet_notat
ions.txt
> Joseph Pehrson
>
> Thing to be careful about is to get the name and date right for the
> folder as that is the only way the program will know where to put it
>
> Now you need to upload the file 19tet_notations.txt
> to faq_tree/josephpehrson in the files area
> (where I'm assuming you've already made that folder)
>
> And you are done, that's it.
>
> Any question? Why not try a test entry, - we can always delete
> it if it doesn't work.
>
>

OK... Now when I made my text entry in the Files area I was allowed
to create text right on the screen in the Website... I did not have
a .txt file, per se. Is this ALSO a possible way to do it??

I notice, though, that using that method, I get a kind of "weird"
filename. In this case my path is:

/tuning/files/faq/19-tET%20notation

A bit strange, but will it work??

I then referenced the path in the database...

Did I do it correctly, or must I upload a "true" .txt file??

In any case, SOMETHING is showing in the Database.

Will it work??

Actually, this seems like a great system, since people can just add
whatever they want in the files section and then we can decide what
to reference or NOT to reference in the database.

It looks like a great system for this project...

Apparently, then, you can just take this database tree and turn it
into EITHER a tree-based website or ascii...

Is that correct??

Thanks!

_______ _____ ____ _
Joseph Pehrson

🔗Robert Walker <robert_walker@rcwalker.freeserve.co.uk>

Hi Joseph!

> I just see a folder called "faq" NOT "faq_tree" in the files section.
> Am I missing something??

You are perfectly right. Forgot that I'd renamed it to "faq". Sorry.

> OK... Now when I made my text entry in the Files area I was allowed
> to create text right on the screen in the Website... I did not have
> a .txt file, per se. Is this ALSO a possible way to do it??

Yes, that's fine.

> I notice, though, that using that method, I get a kind of "weird"
> filename. In this case my path is:

> /tuning/files/faq/19-tET%20notation

> A bit strange, but will it work??

NO, it won't.

Since a space character isn't url-safe, one uses %20 if one wants
to use one in a url (and % followed by other numbers for other non
url safe characters).

For instance, Windows is happy with spaces in file names,
but I think Linux / Unix doesn't allow it yet, and Windows 3.1
certainly didn't.

Often it's fine to do this, but unfortunately, the tree control doesn't
understand this notation.

So what it means is that you had a space in the file name - would
have given it as

"19-tet notation"

naturally enough.

What you have to do is to rename you file so that it has no spaces
in the name.

A very common convention here is to use an underline instead of a space:

"19-tet_notation"

This is worth doing anyway - it is going to confuse people if they
want to link to an entry in the faq and see a %20 in the url , and they'll
probably wonder like you what it is all about.

This is clearly going to be a faq about the faq tree, and I'm not
sure how to make sure users of the tree know about it.

Perhaps best to suggest is to make it the very first thing you
see when you load the tree (it is already, but a bit of
a way down the page)

> I then referenced the path in the database...

> Did I do it correctly, or must I upload a "true" .txt file??

> In any case, SOMETHING is showing in the Database.

Yes that's correct, you just need to give the path as you did
(of course you need to edit it to the new version of the filename
with the underline)

It has also brought up a couple of bugs in the system already - but
this is how we will be able to fix them, hopefully!

You gave "Test file" as the short description ,which is what one
will see in the tree itself, the text you click on to show it,
and "19-tET notation" as the longer description, which will show
in the status bar, and be used when converting hyperlinks
to ascii.

It's possible you won't have realised that's what wil happen??

So perhaps we should just have the short description, and use that
everywhere, including when one converts hyperlinks to your file to
ascii as well - i.e. delete the long description file.

It is just a refinement; not essential if it is going to cause
confusion.

What do you think?

The other bug is in the system I'd worked out for saying where you want your
entry to go.

The first column is supposed to have two things in it, the folder you want
your file to go into plus the date of creation of that folder.

Better just to have the folder name. But then, what do we do about
duplicated names??

E.g. there may be several folders on the tree called notation.

The simple solution here is just to say that one must give each new
folder a unique name.

So one has
notations
notations (Gothic)
notations (Erv Wilson)

etc.

The program could prepare an alphabetical list of all the folders whenever
the tree is refreshed (linked to as one of the first entries in the tree).

So when one wants to add a new folder, one first checks to see if the name
you want to use is in the list already. If it is, you have to use a slightly
different name (at last resort, use notations (a), notations (b), etc. if one
can't think of anything else).

Since the folder "notations" hasn't been created yet, you also need to tell the
program where to put it.

I imagine that is a suitable one to go at the top level.

So as a final step, to make it, add a new field
folder ; descr (of the sub folder in this case)
=
main ; notations

If you wanted to put it in another folder, you'd do it as
calculations ; notations

to put it in the calculations folder

> Will it work??

Yes,

> Actually, this seems like a great system, since people can just add
> whatever they want in the files section and then we can decide what
> to reference or NOT to reference in the database.

> It looks like a great system for this project...

> Apparently, then, you can just take this database tree and turn it
> into EITHER a tree-based website or ascii...

> Is that correct??

Yes, that's true.

Either a tree based web site, or place all the messages one after another
in the same order as they are in the tree.

Also when you edit your original file, the changes will immediately show
up in the "Work in progress" tree, even if John hasn't refreshed it
recently, since the url still points to your original draft.

The "On-site" tree will show your change next time he refreshes it.

Except that I have to write the program to read the database and make the trees
- haven't done so yet because of the debate, but it wouldn't take long
to get at least the basics up and running - i.e. to make the
"work in progress" tree up and running for any type of file,
and to get the "on-site" tree working for ascii.

Then a bit more work to get the on-site tree working for graphics
sound and VRML etc.

Final stage would be to make it easy to re-organise the tree and
to help with the final FAQ preparation - that can wait for a bit later.

I'll wait for your reply to see what you think, then perhaps redo the
table with no date, and leave out the long description, to see how
that works.

You can then have a go at adding the notations folder and see
how that works perhaps, and anyone else who wants to try adding
test folders and entries.

I'll probably also add a final field,
"Text or simple html suitable for conversion to Ascii (Y/N)"

Yes for a text file, and for simple html files.

No for a html file that depends heavily on hyperlinks, or graphics,
VRML etc,

So you will need to say something for that, obviously, "yes"

I can then do the program, which won't take long, and take it from
there.

Only way really to see if it will work is to try it out!

Thanks for "taking the plunge"!

Robert

🔗Robert Walker <robert_walker@rcwalker.freeserve.co.uk>

🔗JSZANTO@ADNC.COM

🔗J Scott <xjscott@earthlink.net>

Hello JdL and Robert Walker,

John, OK so I understand you are the editor?
That is great!

I sympathize with your aversion to renumbering
sections and I also defer to you for final decision
of how you want to structure it.

I understand Robert's system a little better now.

And I am coming out of my Luddite cave.

We can put FAQ entries in a directory tree on yahoo.
These FAQ entries can be plain ASCII text if we
want or they can be HTML files - whatever the
author of the FAQ entry feels comfortable with.

Robert is writing a program that will take a ftp'd
snapshot of the directory tree and automatically
convert it into two versions:

1. A fancy HTML thing.
2. A straight out ASCII USENET style FAQ.

Also Robert acknowledges that some of the files
may contain graphics and such any their authors
may want it that way so those files can just be
referenced in the ASCII version by their section
title and a URL, and perhaps a short description.

I want to commend Robert for being willing to do
all this work that will make _maintaining_ and
_generating_ the FAQ so much easier for JdL.

I have a suggestion that I think possibly might
work for everyone:

If you want to, you can post your FAQ revisions
to the list as you always have. Then we can all
debate the FAQ and revise it as usual. Once the
FAQ entry is settled down, the author can
post it as "FAQ FINAL: What is Meantone?" or
such - including FAQ FINAL. Then JdL, as editor
can collect those and put them where he thinks
best in Walker's tree system and, as editor,
add links or special $$ tags or whatever,
or nothing at all -- whatever editing for form
he likes. Since JdL puts the file in the
tree, the Luddite FAQ entry author is protected
from the system which for many of us can function
as a black box which Robert and JdL are the
keepers of. The black box will eventually produce
an HTML website AND an ASCII USENET file with
numbered section headers so bathtub dwellers
can still navigate the document. FAQ authors who
want to get more involved can put the files in
the yahoo tree themselves and perhaps learn
Robert's system for organizing GIF's and MPEGs
in folders and so forth. They can then post
a note saying that a new revision up with
the yahoo address. Debate of these pages can
take place on the list or privately with the
author - whatever the debater wants.

It sounds like an OK system as long as:

1. FAQ authors don't have to use the system if
they don't have web access, or don't want to use
the system.

2. A text version with navigatable (numbered) cross
references is one of the products of the black box.
(all the numbers should just be automatically generated
from the links JdL and the authors put in the docs).

So, like I said I do appreciate Robert's good work
on this and patience in trying to explain it to us,
as well as JdL's boldness to accept this position.

Does this sound workable to any one else?

- Jeff

🔗jpehrson@rcn.com

--- In tuning@y..., "Robert Walker" <robert_walker@r...> wrote:

/tuning/topicId_19651.html#19685

> Hi Joseph!
>
> > I just see a folder called "faq" NOT "faq_tree" in the files
section. Am I missing something??
>
> You are perfectly right. Forgot that I'd renamed it to "faq". Sorry.
>
> > OK... Now when I made my text entry in the Files area I was
allowed to create text right on the screen in the Website... I did
not have a .txt file, per se. Is this ALSO a possible way to do it??
>
> Yes, that's fine.
>
> > I notice, though, that using that method, I get a kind of "weird"
> > filename. In this case my path is:
>
> > /tuning/files/faq/19-tET%20notation
>
> > A bit strange, but will it work??
>
> NO, it won't.
>
> Since a space character isn't url-safe, one uses %20 if one wants
> to use one in a url (and % followed by other numbers for other non
> url safe characters).
>
> For instance, Windows is happy with spaces in file names,
> but I think Linux / Unix doesn't allow it yet, and Windows 3.1
> certainly didn't.
>
> Often it's fine to do this, but unfortunately, the tree control
doesn't understand this notation.
>
> So what it means is that you had a space in the file name - would
> have given it as
>
> "19-tet notation"
>
> naturally enough.
>
> What you have to do is to rename you file so that it has no spaces
> in the name.
>
> A very common convention here is to use an underline instead of a
space:
>
> "19-tet_notation"
>

HI ROBERT!

Of course... I was wondering what the %20 was all about! I have
renamed the file and, yes, I did know about that "convention..."

> > Did I do it correctly, or must I upload a "true" .txt file??
>
> > In any case, SOMETHING is showing in the Database.
>
> Yes that's correct, you just need to give the path as you did
> (of course you need to edit it to the new version of the filename
> with the underline)
>
> It has also brought up a couple of bugs in the system already - but
> this is how we will be able to fix them, hopefully!
>
> You gave "Test file" as the short description ,which is what one
> will see in the tree itself, the text you click on to show it,
> and "19-tET notation" as the longer description, which will show
> in the status bar, and be used when converting hyperlinks
> to ascii.
>

I noticed this, and it has now been corrected!

> It's possible you won't have realised that's what wil happen??
>
> So perhaps we should just have the short description, and use that
> everywhere, including when one converts hyperlinks to your file to
> ascii as well - i.e. delete the long description file.
>
> It is just a refinement; not essential if it is going to cause
> confusion.
>
> What do you think?
>

I used the same "description" for both... so I guess the answer would
be "yes..."

> The other bug is in the system I'd worked out for saying where you
want your entry to go.
>
> The first column is supposed to have two things in it, the folder
you want your file to go into plus the date of creation of that
folder.
>
> Better just to have the folder name. But then, what do we do about
> duplicated names??
>
> E.g. there may be several folders on the tree called notation.
>
> The simple solution here is just to say that one must give each new
> folder a unique name.
>
> So one has
> notations
> notations (Gothic)
> notations (Erv Wilson)
>
> etc.
>
> The program could prepare an alphabetical list of all the folders
whenever the tree is refreshed (linked to as one of the first entries
in the tree).
>
> So when one wants to add a new folder, one first checks to see if
the name you want to use is in the list already. If it is, you have
to
use a slightly different name (at last resort, use notations (a),
notations (b), etc. if one
> can't think of anything else).
>
> Since the folder "notations" hasn't been created yet, you also need
to tell the program where to put it.
>
> I imagine that is a suitable one to go at the top level.
>
> So as a final step, to make it, add a new field
> folder ; descr (of the sub folder in this case)
> =
> main ; notations
>
> If you wanted to put it in another folder, you'd do it as
> calculations ; notations

Actually, Robert, for now I have created a folder "notations" and put
my "notation" FAQ in it... but I agree we probably will end up with
MORE "notations" folders before we're through!

>>
> > Will it work??
>
> Yes,
>
> > Actually, this seems like a great system, since people can just
add whatever they want in the files section and then we can decide
what to reference or NOT to reference in the database.
>
> > It looks like a great system for this project...
>
> > Apparently, then, you can just take this database tree and turn
it
> > into EITHER a tree-based website or ascii...
>
> > Is that correct??
>
> Yes, that's true.
>

WELL THEN, LET'S GET TOO IT! (Sorry Jon Szanto for the caps!)

THIS IS GREAT!

> Either a tree based web site, or place all the messages one after
another in the same order as they are in the tree.
>
> Also when you edit your original file, the changes will immediately
show up in the "Work in progress" tree, even if John hasn't refreshed
it recently, since the url still points to your original draft.
>
> The "On-site" tree will show your change next time he refreshes it.
>
> Except that I have to write the program to read the database and
make the trees - haven't done so yet because of the debate, but it
wouldn't take long to get at least the basics up and running - i.e.
to
make the "work in progress" tree up and running for any type of file,
> and to get the "on-site" tree working for ascii.
>
> Then a bit more work to get the on-site tree working for graphics
> sound and VRML etc.
>
> Final stage would be to make it easy to re-organise the tree and
> to help with the final FAQ preparation - that can wait for a bit
later.
>
> I'll wait for your reply to see what you think, then perhaps redo
the table with no date, and leave out the long description, to see how
> that works.
>
> You can then have a go at adding the notations folder and see
> how that works perhaps, and anyone else who wants to try adding
> test folders and entries.
>

Yes, I did this... as you will see...

> I'll probably also add a final field,
> "Text or simple html suitable for conversion to Ascii (Y/N)"
>
> Yes for a text file, and for simple html files.
>
> No for a html file that depends heavily on hyperlinks, or graphics,
> VRML etc,
>
> So you will need to say something for that, obviously, "yes"
>

GOOD IDEA.

> I can then do the program, which won't take long, and take it from
> there.
>
> Only way really to see if it will work is to try it out!
>
> Thanks for "taking the plunge"!
>

You're welcome, Robert. This system makes a lot of sense, and I
really hope the other members of the list agree to it.

In essence, it is really SIMPLE, and, perhaps some people have been
"put off" by all the details...

The best part is the fact that all people have to do is put files in
the files database! That's about it... and organize folders.

THEN, the FAQ editor could be the person to decide how to organize
the
DATABASE.

Don't you think that might be the best "role" for the FAQ editor??

Then, actually, John deLaubenfels, if I am understanding this
correctly, won't have to have anything AT ALL sent to him!

The editor can just decide which completed FAQ entries in the FILES
section are ready to be incorporated into the tree. And he will add
them.

There will be debate on the Tuning List of the various entries which
we can view in the "files" area until they are complete... or copy
some of the text to the list for commentary and revision.

Then YOU, Robert, will do your "magic" and turn the whole thing into
a
beautiful HTML web tree OR, for the "asciiphiles" a nice, simple text
that one can read while sitting on the... well, divan..

WONDERFUL!

________ _____ ______ _
Joseph Pehrson

🔗jpehrson@rcn.com

--- In tuning@y..., "J Scott" <xjscott@e...> wrote:

/tuning/topicId_19651.html#19690

>>
> I understand Robert's system a little better now.
>

It's actually quite easy, Jeff. Robert has just been giving a fair
amount of "detail" :) that's all!

>
> We can put FAQ entries in a directory tree on yahoo.
> These FAQ entries can be plain ASCII text if we
> want or they can be HTML files - whatever the
> author of the FAQ entry feels comfortable with.
>
> Robert is writing a program that will take a ftp'd
> snapshot of the directory tree and automatically
> convert it into two versions:
>
> 1. A fancy HTML thing.
> 2. A straight out ASCII USENET style FAQ.
>

YA GOTS IT

> Also Robert acknowledges that some of the files
> may contain graphics and such any their authors
> may want it that way so those files can just be
> referenced in the ASCII version by their section
> title and a URL, and perhaps a short description.
>
> I want to commend Robert for being willing to do
> all this work that will make _maintaining_ and
> _generating_ the FAQ so much easier for JdL.
>
> I have a suggestion that I think possibly might
> work for everyone:
>
> If you want to, you can post your FAQ revisions
> to the list as you always have. Then we can all
> debate the FAQ and revise it as usual. Once the
> FAQ entry is settled down, the author can
> post it as "FAQ FINAL: What is Meantone?" or
> such - including FAQ FINAL. Then JdL, as editor
> can collect those and put them where he thinks
> best in Walker's tree system and, as editor,
> add links or special $$ tags or whatever,
> or nothing at all -- whatever editing for form
> he likes. Since JdL puts the file in the
> tree, the Luddite FAQ entry author is protected
> from the system which for many of us can function
> as a black box which Robert and JdL are the
> keepers of.

Actually, this is a good idea. HOW ABOUT IT ROBERT... maybe ONLY
John
deLaubenfels should have access to the "database" tree.

But, maybe YAHOO doesn't work like that... In fact, it rather looks
like it doesn't... EVERYBODY has access to the database that's on the
group.

The black box will eventually produce
> an HTML website AND an ASCII USENET file with
> numbered section headers so bathtub dwellers
> can still navigate the document.

And great "good clean fun" at that!

FAQ authors who want to get more involved can put the files in
> the yahoo tree themselves and perhaps learn
> Robert's system for organizing GIF's and MPEGs
> in folders and so forth.

This is really not all that hard... It's just that we're seeing lots
of "details" right now. The basic system is almost "moronically"
simple..

The best part is that there is no "sending back and forth" of files,
and the sending to John of revision after revision.

John will like that. Of course, John is out all next week, so we may
get a slow start...

They can then post
> a note saying that a new revision up with
> the yahoo address. Debate of these pages can
> take place on the list or privately with the
> author - whatever the debater wants.
>
> It sounds like an OK system as long as:
>
> 1. FAQ authors don't have to use the system if
> they don't have web access, or don't want to use
> the system.
>

As long as the FAQ is posted to, and debated on the list, it can
easily be COPIED as text to a file in the files area. Anybody could
do this for the "web adverse..."

> 2. A text version with navigatable (numbered) cross
> references is one of the products of the black box.
> (all the numbers should just be automatically generated
> from the links JdL and the authors put in the docs).
>

Well, this aspect Robert is still working on, so we must "bear with
him..."

> So, like I said I do appreciate Robert's good work
> on this and patience in trying to explain it to us,
> as well as JdL's boldness to accept this position.
>
> Does this sound workable to any one else?
>
> - Jeff

THIS IS GREAT. BETTER THAN SLICED BREAD! (Sorry, Jon Szanto for the
caps!)

________ ______ _____ _
Joseph Pehrson

🔗graham@microtonal.co.uk

Robert Walker wrote:

> How about the idea that anyone can add anything, and then the FAQ
> gets selected from within the site.

First part's great, it's the idea behind the World Wide Web, and I see no
need to reinvent it.

How about people submit FAQ entries to the list, others suggest
alterations, and when everybody agrees with it the FAQ editor adds it to
the FAQ?

> The keeper of the FAQ could then edit it at that point, if
> necessary, moving folders around etc, before it is finally output
> as html or ascii.

The keeper of the FAQ can maintain the FAQ however they like. I'm
certainly not taking on the work.

> I suggest that both types of entry be encouraged - ones suitable
> for an ascii faq, and ones that might be especially welcoming
> to a newbie who uses (and perhaps likes) a web browser.

I fully encourage web pages, and will host them at www.microtonal.co.uk if
the author can't get their own site together. Bigger, multimedia content
I could put at meantone.com but no long-term promises on that one.

> One very simple technique would just be to mark off the
> sections in the text - don't even need to edit the html tags:
>
> Just need some reserved sequence of symbols for it.
>
> Anything unlikely to be used.
>
> How about
> /*-------------
> (any number of dashes more than 10 and nothing else on the same
> line)
>
> -----------*/
> for ascii section,
> and
> /-****************
>
>
> *****************-/
> for the graphics section.

Or how about

<ascii-section>
...
</ascii-section>

and

<graphics-section>
...
</graphics-section>

An HTML parser, or maybe even an XSL stylesheet should be able to handle
it.

> Then to process it for the text faq, remove all graphics sections,
>
> To process it for the graphics, remove the ascii sections.

The problem is that you then have three versions: the original, the text
version and the HTML version. That sounds like overkill, although it
might be possible with an XML language. I'm not sure if XSLT on its own
can render a good enough text version, but a few regular expressions
should do the trick.

> Could do same for bookmarks.
>
> In html there's something of a tradition of using words for
> bookmarks, as one can then re-arrange sections
> without having to re-number all the bookmarks.
>
> So if it was an html document, then at the end, maybe only once it
> has been volunteered, and selected for the ASCII FAQ, one would just
> go through it and add the section headings by hand as
>
> /*1.1.7*/ <h2><a name="tuning_a_bouzouki" />Tuning a bouzouki</h2>
> ...
> See <a href="#tuning_a_bouzouki">tuning a bouzouki</a>.

It means the information is duplicated in different ways. If the numbers
were the same for HTML and ASCII, you could go through checking all the
links in a web browser and know they must be right in text as well.

You could try using an XML format so that you could work out the numbers
automatically. Like

<section name="practical instructions">
...
<section name="tuning_a_bouzouki">
<title>Tuning a Bouzouki</title>
<content>
...
</content>
</section>
...
</section>
...
see <link target="tuning_a_bouzouki">tuning a bouzouki</link>

And you could run through assigning section numbers and references to
them. It'd need to be done as two passes, so I can't see XSLT working.
I'm sure I could write a parser in Python to do it, but it would take a
while. If I even had time to edit the FAQ, I'd prefer to add entries to
the end. I can't see why you'd ever need to insert an entry: the order
shouldn't matter within a section. If you do feel a need, you can use an
"a". If you take out an entry, don't change the numbering so you don't
break links.

> I have something of a fondness for /* */ because of c, but of course
> one could use any convention.
>
> Maybe some other sequence would be better. What is unlikely to be
> used I wonder? How about %%
> %%1.1.7%%

How about  ? Then you wouldn't even need to take them out of
the HTML version.

Another thought: let's avoid links to within the files section, because
last I checked it wasn't publicly accessible.

Graham

🔗Seth Austen <klezmusic@earthlink.net>

🔗jpehrson@rcn.com

🔗John A. deLaubenfels <jdl@adaptune.com>

[I wrote:]
>>But, you don't seem to appreciate my concern that maintaining an
>>Ascii FAQ (with inevitable insertions) will be _more_ work than
>>maintaining a very simple html FAQ. Believe me, I'm not looking
>>to make my life more challenging over this; quite the opposite,
>>along with my goal to provide the most useful FAQ possible.

[Jon Szanto:]
>I'm sorry that I gave that impression -- believe me, I want whatever
>solution to be the least amount of work as possible, for the person
>in charge of it. I haven't had too much time to follow this, but in
>just reading a couple of Robert Walker's posts, I can't believe the
>complexity that is being built up -- all in the name of being simple!

>(Not a slight on Robert's skills, but maybe it's a programmer's
>solution [speaking as a programmer myself!]).

Well, I'm glad I'm not the only one experiencing that sensation! I
feel like the grandfather in "Moonstruck", when he says, "I'm CONFUSED!"
in that big scene at the end. I _did_ succeed in struggling partway
through Robert's scheme this morning, and, if we go that way, I would
need to be the only person who understands it, since everyone else can
simply e-mail their proposed FAQ's to the list for comment and eventual
incorporation.

>The only other thought I had was this: the whole thing had started
>with people like Daniel Wolf submitting a couple of drafts in an
>email message. Would it be possible to take each FAQ topic as a
>separate text file, and then have an index of all the questions? Each
>topic, as a page, could contain links and would always have a link
>back to the main index page. If there were additions/changes to the
>text files it would be easy to incorporate them, and there are a lot
>of tools available to take a very plain text file and convert it to a
>simple .html file. Then the entire directory could be zipped for
>download, and, once opened, you'd only need to open the
>index.html/FAQ.html page in the browser and then browse to all the
>topics.

>(If you'd like, I could put up a dummy version of this idea on a temp
>page on my website, as an example...)

Thanks for offering! If I end up taking a DIY approach, that'd be
very useful. Let's wait and see how things shake out.

>Of course, I would always end with: the format chosen should match
>what the editor of the FAQ would like to work with -- anything else
>is a burdon on the volunteer effort. So, John, if you would want to
>work with .html and there isn't some landslide in another direction,
>that's what you should do. I'd be happy to point to .html resources
>to make the job easier if you like (for instance, quite a few
>freeware HTML-2-ASCII convertors...)

I'd be glad to know of sources for such tools, but as far as the FAQ
format goes, I don't want to make that decision. So far, list members
seem to be passionately Ascii. Robert's scheme would seem to offer both
formats, and, if I'm understanding correctly, to accomplish renumbering
when something new is inserted. With such a tool, I can accept Ascii.

JdL

Re: Ascii or Html FAQ

🔗Robert Walker <robert_walker@rcwalker.freeserve.co.uk>

🔗jpehrson@rcn.com

🔗Graham Breed <graham@microtonal.co.uk>

🔗J Scott <xjscott@earthlink.net>

🔗JSZANTO@ADNC.COM

🔗John A. deLaubenfels <jdl@adaptune.com>

🔗JSZANTO@ADNC.COM

🔗Robert Walker <robert_walker@rcwalker.freeserve.co.uk>

🔗John A. deLaubenfels <jdl@adaptune.com>

🔗Jay Williams <jaywill@tscnet.com>

🔗Robert Walker <robert_walker@rcwalker.freeserve.co.uk>

🔗JSZANTO@ADNC.COM

🔗jpehrson@rcn.com

🔗Robert Walker <robert_walker@rcwalker.freeserve.co.uk>

🔗Robert Walker <robert_walker@rcwalker.freeserve.co.uk>

🔗JSZANTO@ADNC.COM

🔗J Scott <xjscott@earthlink.net>

🔗jpehrson@rcn.com

🔗jpehrson@rcn.com

🔗graham@microtonal.co.uk

🔗Seth Austen <klezmusic@earthlink.net>

🔗jpehrson@rcn.com

🔗jpehrson@rcn.com

🔗John A. deLaubenfels <jdl@adaptune.com>

🔗J Scott <xjscott@earthlink.net>

🔗Dave Keenan <D.KEENAN@UQ.NET.AU>