Conference vaxuum::document_ft

	One more:

	    <ENDTITLEPAGE> should be <ENDTITLE_PAGE> !

	While you're looking at the SDML-definitions: Could you
	change the "<LE>" definition to "<LE> {text}" please?
	This makes the entering of lists a "real" pleasure.

	detlef.

re: .1

The same is true of some of the templates like MEMO.  If you bring the
memo template in, you get taken to the middle of the template, since thats
the first place a placeholder such as {text} exists.  Seems that most of
the places a user can enter text in a template ought to have a placeholder
such as {text}.

-tl

Not meaning to pick on the SDML environment for memos/letters, but its all
I've really had time to try:

1) The template for letters includes after the body or closing a tag
   called <NAME> which is unrecognized.

2) The arguments to things like <MEMO_FROM> that have multiple lines
   that must be seperated by \'s should put those \'s in for you.

3) As a rule, my recollection is that with other LSEDIT environments a user
   doesn't really need to know the syntax of the language, that is
   partly what LSEDIT is supposed to do.


-tl

    <LE{text} will be in the update...thanks for pointing out the missing
    underscore in <ENDTITLEPAGE>, also!
    
    Julie

Also, <HYPHENTATE> is an unusual spelling.....

					I

    Julie, also read note 55.2, which describes a problem with the
    <condition> tag leaving a space in arg list.

    The identifier characters for the SDML language definition must
    not include parentheses.  Having them there will break things like
    the LSE HELP/INDICATED command, once full on-line help becomes
    available.  (Feel free to contact me for a detailed explanation.)
    
    If you're willing to require LSE V2.0, then angle brackets should
    also be excluded from the identifier character list.  This will
    improve the behavior of GOTO WORD, ERASE WORD, etc. without preventing
    token definitions containing angle brackets.
    
    Finally, if you can find the time, it would be worth the effort to
    define corresponding tokens without the angle brackets; LSE is supposed
    to make life easy, but hitting the angle bracket on an LK201 is not
    particularly easy.  Unlike the points raised in the two paragraphs
    above, this one requires non-trivial work and is somewhat religious in
    nature (personally, I can live without the <foo> tokens altogether).
    Hence, I'd put this at a low priority. 
    
       Gary

    I have struck a few more LSE bugs as well.

    <LIST>(CALLOUTS) should be CALLOUT     

    Several of the <END~ tags have the _ in the wrong place. (like 
    <END_FRONTMATTER>.
    
    Regards
    
    Kevin M
      

    If you create a file with lsedit, the [text_element_list] tag
    disappears if you close the file and open it again.
    
    if my memory serves me right, the equivalent prompt always showed
    when using SDMLLSE.
    
    Peter

    You can simply retype [text-element-list], move the cursor into
    that field, and press CTRL/E. Or, simply add a key combination
    in your lse init file to insert it.
        
    If [test-element-list] did not disappear, you would find in your
    output file--unless it was commented out. 

    The behaviour did change between bl6 abd bl7.  

    thanks for the help
    
    Peter


                       

I can't agree with .10.  If I don't delete [text_element_list] from my
source file, why should I expect LSE to delete it?  And of course its
going to appear in my output file.  If it didn't, I'd be surprised.

Until the marrying of LSE with the language/text processors is to the
level that the language/text processors can detect LSE tags,
templates, placeholders, tokens, and routines, and ignore them, I
don't think LSE should be performing the current BL7 action. 

-tl

    I disagree with .12
    
    I for one do not want to see any [placeholders] in my output file after
    I leave LSE, BUT I do want to see it at the top of my file when I come
    back in LSE just in case I need it again.
    
    One of the reasons I am saying this, is not just for me, but we
    are trying to convince our secretaries to use this product.  Ever
    tried to do that to a bunch who only know WPS??????
    
    IT OUGHT TO BE AS FOOL-PROOF AS POSSIBLE or the secretaries will
    ignore it.
    
    Regards.

    Jerry,
    
    I suggest you have them generate DOCUMENT files with OUTLINE instead.
    Though Jack hasn't gotten all the DOCUMENT structures built into
    OUTLINE, it does handle a good deal of the easy stuff.  The OUTLINE
    notes file is at:
    
    	SHALOT::OUTLINE
    
    or just press <SELECT> or <KP7>


    An OUTLINE user

    Re: .13
    
    In LSE, the right way to get that behavior is to define a token
    that duplicates the behavior of the placeholder.  There's no guarantee
    that the top of the file, bottom of the file, or anywhere else will
    be the right place for the placeholder.  By having a synonymous
    token, you can start typing anywhere in the file, expand, and get
    the menu of items.
    
       Gary

    Sorry if you've already discovered this, but the {text} placeholder
    is not defined.
    

     Shouldn't the [text-element-list] placeholder go away when you
     call in a profile template (the placeholder gets put after the
     <ENDPROFILE> tag)?  After all, you can't have any of the other
     text elements in a profile...
						-- Nina

    Suggestion: let users have the title and the endtitle_section
    tag if they specify the title_section tag in the menu. The one is
    not much use without the others.  
    
    There should be some way to get a variety of the tags specific to
    the 'article' document style.  Selecting them one at a time from
    the menu is clumsy. [Good thing I have terminal cut and paste on
    my VAXstation.]

    Now that I've had a chance to play a bit more, I think I'm beginning
    to understand the difficulties that people are having.
    
    The starting string for the language ought to correspond to what
    is currently [special-templates], with perhaps an additional entry
    for included files.  Each item under special templates
    ought to expand to an outline of the intended document.  For example,
    the MEMO choice under special-templates does this now, but the ARTICLE
    choice isn't even close.
    
    The current starting placeholder, text-element-list, should probably
    go away.
    
    I'd be glad to help work on the overall LSE template design with
    you.
    
       Gary

    If you expand the <align_char> tag, then 
    
    -	enter # for the character
    -	^K to get rid of the \DELTA
    
    you will end up with a space following the #.

    Normally, DOCUMENT ignores spaces.  But here, when you try to
    use the #, DOCUMENT seems to expect the # to be followed by
    a space.  Solution:  get rid of the space.  It would be nice
    if the LSE could do so.

    I don't know if this has been found yet but...
    
    TABLE expands to <TABLE_UNIT>
    			.
    			.
    			.

This isn't a bug but...

It would be nice if SDML LSE would treat backslash as a word separator.
That would help when deleting by word or moving by word through the document.

    Re: .22:
    
    You should be able to use  MODIFY LANGUAGE SDML /PUNCTUATION="\"
    to define the backslash as a delimiter.

    I don't know if you want bugs in different doctypes but here goes:
    
    Also, how do we put blank lines at the top of the letter?
    We have a <Date> tag as our first tag.  The letter is short enough
    not to fill the page.  We tried <literal> etc. None of them put
    blanks in.
    
    The letter looks like this initially -- we moved the <FROM_ADDRESS>
    to the bottom of the letter.
    
    
<DATE>
<TO_ADDRESS>(f\
	     g\
	     h\
	     i\
	     j )

<SALUTATION>({text})

asdfasdlkfj;lksdfj

<CLOSING>(sincerely,\herb)
<NAME>(john smith)
    
<FROM_ADDRESS>(a\
	       b\
	       c\
	       d\
	       e )

    In the letter doctype the initial expansion brings up a <NAME> tag.
    DOCUMENT fails by not finding the tag.
                <NAME>(john smith)
    %TAG-W-TAGNOTDEF, at text on line 19 in file
	   NCADC1$DUA0:[PEREZ]LETT.GNC;1
	   Tag <NAME> is undefined

    The <CLOSING> with two lines results in the following in the final
    output.
    
    sincerely,
    
    
    
    a
    b
    c
    d
    e
    
    and the second argument is lost.  If, however, we put a blank space
    after the second argument before the "\" the second argument prints.
    
    {text} tag is also undefined.
    
    
    This is not a major doctype for us, but as we used it today, these
    things came up.  Since it was our unit manager that was using it,
    his concerns became my concerns!!  It always flows down hill!!!
    
    Dave

    
     
   

    How do I get LSE to do automatic wordwrap at, say, column 79?
    

RE: < Note 8.25 by DELNI::TURBETT "DSS Product Manager" >

>>    How do I get LSE to do automatic wordwrap at, say, column 79?

There are a few ways to do this.  The simplest way is to issue the LSE command
"set wrap".  This will wrap at column 80.  If you want to use column 79 you also
need to issue "set right_margin 79".

If you don't want to issue this command for each buffer, during each LSE
session, you may want to put it in your LSE initialization file.  If for
example, you want to have all you SDML buffers do automatic word wrap, you
should add the following line to your LSE intialization file: 

modify language sdml /wrap

Or you could add the following line to your tpu$local_init routine, if you
already have (or you want to create) a LSE section file:

lse$do_command("modify language sdml /wrap")

For more information on initialization and section files see the LSE/SCA manual
or the LSE notes conference, TLE::LSEDIT.

             ---Karen

    An init file for LSEDIT and DOCUMENT use that does the wrapping
    (and a few other things) exists in:
    
    	NEWDOC::[writer.docv1]LSE_KEYS.INIT