Speller Bug

[Robert Holmgren]

OK, here's how to force PCLEXAM to crash. Make sure that:
 D DT=0 (thus, eXPanded mode is your default)
 D UD=name of your current PCLEXAM .DIC file, e.g. "D UD=US"
 D LX="d:\path" of the current (& all other) .DIC file(s)
 PCLEXAM.DL0|1|2|3 (4 files) are located in the dir
  referenced in ®VANL¯ (usually the same dir as EDITOR.EXE)
 Now, CAll a file of any size. Locate cursor near TOF. Toggle "D OM=0"
(switches to PCLEXAM) and "D OM=31" (restores default Microlytics) a few times,
and use the Thesaurus in each dictionary. End up in Microlytics. Now CAll a
large-ish file, say 100K+, with a fair (typical) amount of formatting. It will
open in eXPanded mode. You're at TOF. Go directly to BOF. PageUp one page.
Put cursor on a word with synonyms. Toggle D OM=0. Command "FUNC WG".
Command "FUNC SY". My system hangs. For example, putting my cursor on the
word "elaborated", which I know has PCLEXAM synonyms, the system beeps and the
PRompt line says "No alternates for :\TEMP\BLOEMEN2.TMP". C:\TEMP\BLOEMEN2.TMP
happens to be the current filename.

Hint: VN$ZZ should contain "elaborated"; instead it contains
":\TEMP\BLOEMEN2.TMP"

Note that my procedure is "magical" (in the religious or anthropological
sense), i.e. if you follow *exactly* this procedure, you get -- theoretically
-- exactly this result every time. In fact, half of the elements in the
procedure may be unnecessary, like toggling D OM=n a few times. Probably only
one element is key.

The prophylactic fix is contained in routine {{5goThes}} in DICTION1.HLP, which
I nclosed with msg 9211 (actually, I see that I nclosed DICTION1.>D