Conference vaxuum::document_ft

    Can't imagine why that is happening, can you capture the
    output and send it to me?  VAXUUM::KOHLBRENNER
                                              
    if you say 
    
    $ SET HOST/LOG 26315
    
    and then login again and run DOCUMENT, then LOGOUT, 
    you'll have a SETHOST.LOG file that has all the terminal
    dialogue in it.                      

    I've seen some inconsistencies and misbehavior, too.  My guess is
    that DOCUMENT isn't getting the status right in all cases.  I'll
    try to get a better handle on this the next time it happens.
    
    I have noticed that if you leave out any of the command line
    parameters on the LSE COMPILE, then the error message you get back
    is not particularly clear, partially due to LSE's limitation to
    a two/three line message window.  I think the ideal solution is
    to not require more than one parameter on the DOCUMENT command line,
    but it may be too late for this.  If DOCUMENT can't do that, then
    I suggest working with the LSE people to improve the interface;
    the LSE project is going to have to study the compile command anyway,
    for some other reasons.
    
       Gary

    This may have nothing to do with the problem, but we are running
    the T1.0-008 version on CAADC.  The system was rebooted and on
    startup DOC$STARTUP did NOT get executed.  From inside LSE the
    compile failed with the -- unable to open .DIA file error.
    
    Dave

    
    no, I don't think that has anything to do with it. COMPILE is not
    working, as is noted in the release notes.
    

    I seem to have the same problem with the COMPILE/REVIEW command
    and would be interested in the answer.
    
    To compound matters, my LSE seems to be generating BL6 syntax.
    We're running Document BL7 and LSE V2. The LSE environment file
    in sys$library is dated February 87 which should be ok. Can you
    suggest what we need to do to fix this.
    
    cheers Mike Thomas
    

Did you previously include the BL6 LSE environment into your system-wide
LSE environment file?  If so, the installation procedure probably didn't
properly update the system-wide environment file.  The simplest fix is to
modify the KITINSTAL.COM file that is in saveset A (I'm assuming that the
kitinstal used in BL7 is similar to the initial BL8 kitinstal which also
had this bug.)  Extract all the files from saveset A into a subdirectory and
in KITINSTAL.COM there probably is a command such as:

$ lsedit/nodisp/nosyst/envi=(SDMLLSE.ENV)/nosect -
  /command=(vmi$kwd:lseupdate_env.tpu)

change it to:

$ lsedit/nodisp/nosyst/envi=(SDML.ENV)/nosect -
  /command=(vmi$kwd:lseupdate_env.tpu)


And rename the file SDMLLSE.ENV to SDML.ENV.  Then BACKUP the latest version
of all the files in that subdirectory to DOC010.A and use this new saveset
A for your installation.

If that command does NOT exist in KITINSTAL.COM, then forget all this
nonsense and try something else.

-tl

    First, under Bl7:
    
    If your buffer is marked as a "Read-only" buffer (not "Write"), it
    seems the .DIA file doesn't get written with a COMPILE. This makes
    sense, since a COMPILE creates a new file (if the buffer has been
    modified).
    
    It also seems that if you have an existing file, it will compile 
    the existing file and not the version in your LSEDIT buffer. This
    sounds like an LSEDIT (not DOCUMENT) behaviour.

    To set the current buffer to Write, enter the LSE command:
    
    	LSE>SET WRITE
    
    Then use:
    
    	LSE>COMPILE/REVIEW $ doctype dest


    Under Bl8:

    You can process the file and create a .DIA file before invoking LSEDIT 
    using the DOCUMENT qualifier /DIAGNOSTIC. You may have to process
    outside LSEDIT if you have to use /BATCH anyhow. An example:
    
    	$ DOCU/DIAG/NOTEXT/BATCH  file.SDML S.R LN03
    
    	$ LSEDIT file.sdml
    	 .
    	 .
     	 .
    	LSE> REVIEW
    
    You can use CTRL/G (GOTO SOURCE), and also CTRL/F and CTRL/B
    to go to the next and previous errors.

    Thanks for the info gentlemen. How soon can I expect Bl8, do you
    think? If it's not long, I can live with the syntax problem until
    we reinstall.
    

    [beginning of compilation]
    
    at text on line 0 in file
    
    doc$bl07_root:[tex.formats]tag$software.stt
    can't open output file:
    
        sampchap1.int_tex
    
    Is this error caused by the compile bug?  I can't compile while
    in LSEDIT.  I can do:
    $ doc/dia file design device 
    and get my .dia file.  I can also do a review while in LSEDIT and
    it will bring up my .dia file into a second window...
    
    thanks,
    Bette Jean

    Yes, I think it is the same error.  it should happen on
    the second attempt to do a COMPILE command, because the
    tag translator left open the output file from the first
    attempt.  When it tries to open the same file again, the
    I/O system says "you must be kidding!" and we exit with
    the error message.  
    
    Fixed in V1.0.
    
    If you exit from LSEDIT and go back into LSEDIT, you should
    then be able to execute the COMPILE/REVIEW command (once).

    This is another "getting started with" problem/question.  I'm also
    a relatively new user of LSE/DOCUMENT.  In glancing through "Step
    by Step" and poring over the online help files, I haven't seen
    anything like init files discussed.  How do you maintain settings
    across sessions, like SET WRAP?
    
    Brenda         

    Questions specific to LSE are best asked in the TLE::LSEDIT notes
    file.  (I know; it can be real tough to determine which notes file
    is most appropriate, and for some questions, it isn't even obvious
    to the developers.)  In general, if your question doesn't refer
    to an SDML tag or to a placeholder generated by SDML, the question
    is probably about LSE.
    
    In this case, any LSE setting that applies to a language can be set by
    including a MODIFY LANGUAGE command in your LSE init file. You can
    determine the available language settings by using the LSE command
    "HELP MODIFY LANGUAGE"; most of these are uninteresting, but some
    are very interesting.   Since WRAP is such a setting, you set it with
    either 
    
    	MODIFY LANGUAGE SDML /WRAP
    
    to just set it for SDML, or
    
    	MODIFY LANGUAGE * /WRAP
    
    to set it for all languages.  If you also want WRAP turned on for
    buffers that do not have a language (e. g. if you use LSE as your
    MAIL editor), you should also include the command
    
    	SET WRAP
    
    which turns on WRAP for the $DEFAULT buffer, and hence for any new
    buffer that has no associated language.
    
    The LSE initializatin file can be specified at the LSE command line,
    by /INITIALIZATION=, or by the LSE$INITIALIZATION logical.
    
       Gary