From: "Chris Antos (Exchange)" 
i see from subsequent email the problem was found.
for future reference, the way you'd debug it is this:
1.  compile the macro and the menu (FixScanMenu) as a .MAC file instead
of burning it in as your .UI.
2.  Macro->Debug, select your .S file that contains the macros.
3.  in the debugger, go to the first line of the TwoSpaces function, and
press F2 (Set Breakpoint).
4.  press F5 (Run) and then hit your F11 key to bring up the
FixScanMenu, then select the 2-Space command.
5.  the debugger will kick back in (because of the breakpoint you set on
the 2-Space function).
6.  you can press F7 to execute each line of the macro, one at a time.
7.  after executing a line, press Ctrl-B to look at the buffer to see if
that line did what you expected.  (hit ESC to go back)
8.  go back to step 6, etc, until you find the problem.
//chris
> -----Original Message-----
> From:	Eric Stork [SMTP:estork@erols.com]
> Sent:	Saturday, July 26, 1997 5:07 AM
> To:	Chris Antos (Exchange)
> Subject:	RE: Possible Bug in TSE2.5e, or maybe my error
> 
> At 01:26 PM 7/24/97 -0700, you wrote:
> >i cannot reproduce your problem with TSE 2.5a or TSE 2.8.
> >
> >i'll bet you have more than just those two replace commands in your
> >2_Space macro.  one of the other replace lines is causing this
> >side-effect.  you might even have a replace line that says
lReplace(".
> >(", ".  (", "gnx")...
> >
> Chris:
> Below is what I sent to Kyle Watkins.  It's really a puzzle, in case
you're
> interested.
> Eric
> -----------------------------------
> Kyle:
> Here's my report.  I experimented with Macro/Debug, but did not get
very far
> -- I did not figure out how to test a particular macro on a particular
file.
> But as far as I got, Debug indicated that there is no problem with my
> 2_SPACE.S file.
> 
> What I did find is really weird and makes no sense to me, but after
checking
> and re-checking, here is what happens:
> 
> Normally, I run 2_SPACE from a menu of various macros that manipulate
> scanned newspaper articles, e.g., New York Times and Wall Street
Journal
> articles.  The menu, invoked with F11, is:
> 
> //
> menu FixScanMenu()    // invoked with F11
>      "&Insert CR/LF in ColBlock",       mInsCrLf()
>      "&CleanUp Hy @ EOL",               mCleanUp_Hy()
>      "Set R Margin to &45 cpi",         m45cpiRightMargin()
>      "&Wrap Paragraphs to EOF",         mWrapAllPara()
>      "Wrap &Blocked Paragraphs",        mWrapBlockedPara()
>      "&Dele Spaces @ BegLin",           mDelInitSpace()
>      "Dele I&nitial Spaces - Block",    mDelInitSpaceBlock()
>      "&2-Space after Periods",          TwoSpaces() //<------------
>      "Delete Bo&x Lines",               mDelBoxLines()
>     ""                          ,                   ,   Divide
>      "Find &Tabs",                      tRangeFind(8,10)
>      "Find Contro&l Characters",        tRangeFindMenu()
>      "T&oggle between LPT1/2"      ,    mTogglePrintDevice()
>      "&Save File Status",               mSaveState()
>      "&Restore File Status",            mRestoreState()
>     ""                          ,                   ,   Divide
>     "&Align ColBlk on a Chr"     ,      mAlign2Char()
>     "Su&m Column Block"          ,      sumMenu()
> end FixScanMenu
> //
> 
> The 2_SPACE macro is #INCLUDEd in my eos.ui file, i.e.:
>        #include ["2_space.s"]          // mapped to F11 menu
> 
> When I run my test by invoking 2_SPACE from the above menu on this
file:
> 
>             OK for "xxx.) xxxx"
> 
>             But for: "xxx. (xxxx" I get "xxx.  (xxxx"
> 
> the second line is NOT correct.
> 
> But when I run another test from a separately compiled macro (i.e.,
running
> SC 2_SPACE.S to create 2_SPACE.MAC) that I load from MACRO/EXECUTE
MACRO,
> the macro performs as it should and the 2nd line is correct.
> 
> I have looked at everything several times and cannot figure out why
this
> happens.  To let you try to recreate the problem, I  attach a ZIPped
file
> (S_FILES.ZIP) consisting of EOS.UI (my base file) and all included 
> xxx.S files.
> 
> I do not claim that this conundrum is a serious problem, but it sure
is a
> serious puzzle.
> 
> Eric
> July 25, 1997
> 
> 
> ----------
> Eric Stork
> e-mail address: estork@erols.com > 
---
---------------