5.0.10 Windows versions include the use of encrypted distribution files.
In order to install, you will need an Authentication Key file which you can
get from our download site or can request from customer service. You will
need the End-user Company Name in order to request/download this keys file
(All) #577
HTML :IN :ML generated "MAXLEN=n" rather than "MAXLENGTH=n" attribute.
(5.0) #582
The TVM_*() functions did not handle comma-type edits for input.
(ie: "12,345.67" was treated as "12".)
(4.9) #581
Add PFMEMOINSERTMODE=ON to set insert mode on by default in the
memo editor.
(All) #585
PFCHECKLOCK does not report errors on "-p" lookups if you
modify fields after a WRITE.
(All) #???
Enhancement to PFTMP for specifying temp directory:
If PFTMP is not set, then use TMP. If both are unset, use TEMP.
(All) #???
Tweak to the above: some versions of Windows allow for a list of
directories in the TMP/TEMP variables, separated by semicolons.
In such instances, filePro will use the first directory listed.
(And, change the Unix version to do the same, should there be a
version of Unix that allows a colon-separated list of directories.)
(All) #615
runmenu was using PFTEMP rather than PFTMP to store menu batch files.
(Quickstart) #616
SHOW statements were truncated to 255 characters.
(All) #617
Fuzzy browse lookups with processing caused non-global dummy fields
to be cleared. ("Normal" browse lookups were okay.)
(All) #619
Fix case where "node_addr(x!=y)" error would be generated.
(All) #621
If you have an unprotected lookup to the current file, and find
the current record, do a GETNEXT, modify that record, and CLOSE
the lookup, the modified record will remain locked.
(All) #628
If you have a lookup dash in @ENTSEL processing on a file with no
automatic processing, it is possible to get DKNF errors displayed.
(Note that the indexes aren't actually corrupted in this case.)
(All *nix) #???
Pressing "N" or "I" in the monochrome attribute editor of dscreen
caused the display to fill with text.
(fileProGI) #???
Shrinking a memo in the GI memo editor caused the tail end of
the original memo to remain.
(Some *nix) #???
Running menu items with the "no return" flag ("#") could cause
the stty settings to get confused.
(All) #275
Using F8/Options to set the processing password in *cabe would
cause subsequent tables loaded in the same session to have the
same password assigned.
(All) #588
A "lookup - r=nn" which fails due to an EOF condition would cause
the current record to be unlocked.
(All) #601
Using block functions to copy processing lines from beyond the
last one could cause garbage to be copied.
(All) #602
fPTransfer now copies "map.new" files as a map.
(All) #604
Pressing F6 in ddir does not bring up the config editor if the
highlighted name is not a valid filePro file.
(All) #610
PFLOOKWIZPROT=ON was not recognized properly.
(All) #612
ddir did not empty the blob file on kill data / retain formats.
(All) #613
Lookup aliases in the format "letter-number-others" (ie: "d4a")
would fail syntax on assignments.
(All) #614
You can now do COntains comparisons on memo fields.
(fPSQL) #618
fPSQL did not recognize 4-digit-year date fields, nor PFSYSYR4=ON
(UnixWare only) #640
Some configurations of UnixWare do not allow a setuid program
when running as real-uid root to execute child processes.
(UnixWare only) #641
An incompatibility in UnixWare running OpenServer binaries caused
SLEEP to be ignored.
(All) #635
dxmaint -M "" did not override PFQUAL environment setting.
(Linux only) #642
Fuzzy searches on Linux did not return results consistent with
other filePro systems.
(All) #637
Extended selection sets didn't work if comparing to a value with
a colon in it.
(All) #591
RAND("-1") didn't properly seed the random number generator on
some systems.