CONTACTS 
0201 - IBM OS/2                           407 443 2000
020101 - Mr. George Huerta; Windows Apps Analyst =407 443 0493
020102 - Technical Support =800 992 4777 fax 407 443 6940

SUBJECTS
Omnipage Professional, V 2.11 386; SN 2804A C00 001771
OS2, Windows Applications, Omnipage, 930730
6x764,psr  Omnipage 930730, APR pj09766

0605 -    ..
0606 - Summary/Objective
0607 -
060701 - Follow up ref SDS 11 0000.
060702 -
060703 - IBM's records had us mixed up with someone else.
060704 -
060705 - The record indicates that the form of the mem_exclude_region spec must
060706 - be 5 place strings, not 4 as used under DOS 5.0.
060707 -
060708 - Omnipage seems to be running correctly now.
060709 -
060710 -
0608 -
0609 -
0610 - Background
0611 -
061101 - Statement of problem is on 930730. ref SDS 7 0850.
061102 - ..
061103 - Last meaningful IBM contact was on 930909. ref SDS 11 9065
061104 -
061105 -
061106 -  ..
0612 -
0613 -
0614 - Discussion
0615 -
061501 - Talked to Hylema.
061502 -
061503 - Her record shows that on 930909 Nick Yocca asked OS2 Fix, Mike
061504 - Baharmast, to send PJ10073 to us on a 3.5 disk.
061505 -
061506 -      Our record shows we never received this, and were never informed
061507 -      it was a fix for our problem, nor that it was being mailed to us.
061508 -      The only disk we have received is for:
061509 -
061510 -                           PJ09192
061511 -
061512 -         ...which has ATI video drivers.
061513 - ..
061514 - Hylema said IBM's record shows Ben Delapena left us a message on
061515 - 930909 1012 asking us to call back so we can advise IBM of information
061516 - for the "FA 870."  They are waiting for us to call back.
061517 -
061518 -      Our records show this call did not occur.  We have no involvement
061519 -      of any kind with anything related to an "FA 870."
061520 -
061521 -
061522 - On 930909 1031 Nick Yocca says he called us and reported our line was
061523 - busy.
061524 -
061525 -      Our phone can never be busy, because we use Pac Bell Message
061526 -      Center.  It can take unlimited messages.
061527 -
061528 - ..
061529 - On 930909 1245, Mike Baharmast says he recreated scenario
061530 - without a page reader: select and setting multiple zone interactive
061531 - mode, deferred recognition, multiple pages.
061532 -
061533 -      Our records show we are waiting for a call back from Mike on how
061534 -      to solve the Omnipage problem, per ref SDS 11 line 94.
061535 -
061536 -
061537 - As a result of what appear to be incorrect records, decided to get
061538 - the full IBM record, to ascertain their view of the problem.
061539 -
061540 -
061541 -  ..
061542 - Complete IBM Chronology
061543 -
061544 - Following took 60 minutes for Hylema to develop from IBM's automated
061545 - record system.
061546 -
061547 -    930810 1458 fupping to pj09766
061548 -
061549 -         We were advised by George Huerta on 930810 at ref SDS 8 line
061550 -         070908, that we would get a "fix" under pj09766, within 7
061551 -         days, but never received it.
061552 -
061553 -    930818 1620 intgns update: PJ09766 closed 930817 can
061554 -
061555 -    930820 1458 fupping for closure of pj10073
061556 -
061557 -    930828 IBM intelligenius update auto fup complete.
061558 -
061559 -    930904 Problem inactive for 7 days.
061560 -
061561 -    930905 IBM intelligenius update auto fup complete.
061562 -
061563 -    930909 -- see explanation above.
061564 -
061565 -    930916 problem inactive for 7 days, call generated.
061566 -
061567 -    930917 IBM intelligenius update auto fup complete, IBM ACE.
061568 -
061569 -    930924 problem inactive for more than 7 days call generated.
061570 -
061571 -    930925 IBM intelligenius update auto fup complete, IBM ACE.
061572 -
061573 -    931002 problem inactive for more than 7 days call generated.
061574 -
061575 -    931004 0651 George Huerta fupping for closure.
061576 -
061577 -    931011 problem inactive for more than 7 days call generated.
061578 -
061579 -    931012 IBM intelligenius update auto fup complete, IBM ACE.
061580 -
061581 -    931019 problem inactive for more than 7 days call generated.
061582 -
061583 -    931020 IBM intelligenius update auto fup complete, IBM ACE.
061584 -
061585 -    931027 problem inactive for more than 7 days call generated.
061586 -
061587 -    931028 IBM intelligenius update auto fup complete, IBM ACE.
061588 -
061589 -    931104 problem inactive for more than 7 days call generated.
061590 -
061591 -    931105 IBM intelligenius update auto fup complete, IBM ACE.
061592 -
061593 -    931113 problem inactive for more than 7 days call generated.
061594 -
061595 -    931115 George Huerta fupping to pj10398 for closure.
061596 -
061597 -
061598 - ..
061599 - Hylema put us on hold to wait for Windows Support Group
061600 -
061601 -
061602 -
061603 -  ..
0617 -
0618 -
0619 - 0813
0620 -
062001 - Talked to Jim Tomaszewski.  He asked if Mike sent us the fix, pj10973,
062002 - and I advised this did not occur, per above record for 930909.
062003 -
062004 - He asked us if we received pj09766, and I advised we have not received
062005 - it.
062006 -
062007 -       He said pj09766 is a generic analysis to evaluate individual
062008 -       problems which will then be assigned individual pj#.
062009 -
062010 -
062011 -
062012 -  ..
062013 - PJ10073 Incorrect
062014 -
062015 - He said pj10073 shows:
062016 -
062017 -        Rod Welch problem is with the automatic form feeder.
062018 -
062019 -   I advised that this representation is incorrect.  We do not use an
062020 -   automatic form feeder; evidently we have been confused with someone
062021 -   else.
062022 -
062023 -   I asked Jim to find in IBM's record the date when this entry
062024 -   occurred, so we can correlate with our records to discover the
062025 -   source of the error.  He could find no such entry.  He found IBM's
062026 -   record for 930730 that explains the scope of the problem as set out
062027 -   in our records at ref SDS 7 line 072904, but noted it does not say
062028 -   anything about an "automatic form feeder."
062029 -   ..
062030 -   He could find no other entry that indicates where the error in
062031 -   describing the problem occurred.
062032 -
062033 -
062034 -
062035 -  ..
062036 - PJ10398 Shows Correct Scope
062037 -
062038 - Jim said further research shows pj10398 appears to have the scope of
062039 - our problem, per ref SDS 7 line 46, but there has been no activity on
062040 - it since 930909 because all subsequent activity for two months, was
062041 - incorrectly presumed to be related to pj10073.
062042 -
062043 -
062044 -
062045 - Plan of Action
062046 -
062047 - Jim said an analyst will call us back today on a solution to this
062048 - problem.
062049 -
062050 -
062051 -
0621 -
0622 -
0623h- 0952 George Huerta called back
0624 -   ..
062401 - He remembered calling us in August, but could not recall any
062402 - details. I asked about the disposition of PJ09766, which George said
062403 - would be issued to us within 7 days of 930810, to solve our problem,
062404 - ref SDS 8 line 42.
062405 -
062406 -     George could not find the results of pj09766.  He said he did not
062407 -     have an HP Scanjet IIc at that time to evaluate the problem.
062408 -
062409 -
062410 - He said APR #10398 is being "worked" now, but that he has been unable
062411 - to re-create the problem.  I said that Hylema found in IBM's record
062412 - that on 930909 1245, Mike Baharmast reported that he "recreated the
062413 - scenario."
062414 -
062415 -    George said Mike is no longer with IBM and that Mike "took the
062416 -    information with him" on how he recreated the problem.
062417 -
062418 - ..
062419 - George suggests trying:  mem_exclude_region:  c800-cbff
062420 -
062421 -     I checked and found we have no assignment for this param now, but
062422 -     in our discussion with John MacDonald of IBM on 930730, ref SDS 7
062423 -     line 128, we reported using:
062424 -
062425 -                         cc00-cfff
062426 -
062427 -          ...and
062428 -
062429 -                         A000-Afff
062430 -
062431 -
062432 -    John said at that time, he would investigate and call back, but did
062433 -    not do so.
062434 -
062435 - ..
062436 - I asked George where he got the assignment he suggests we try?
062437 -
062438 -    He said he just finished talking to a representative from HP, who
062439 -    recommends using this assignment.
062440 -
062441 -
062442 - I asked what setting he was using prior to talking to the rep today,
062443 - for the trials he made in which he reported being unable to recreate
062444 - the problem?
062445 -
062446 -    George said he was using the same setting HP gave him today.  I
062447 -    asked wh<0373
062448 -    asked why he called the HP rep, if he already had a solution?  He
062449 -    indicated he called to verify his selection was supported by HP.  I
062450 -    asked where he got the assignment.  He said there is a hexidecimal
062451 -    table showing this address is open, in the HP Scanjet IIc manual.
062452 -    I asked for the page where this is shown.  He said he no longer has
062453 -    the manual and could not recall the page or chapter.  I checked and
062454 -    found on page 1-16 a table under the heading:
062455 -               ..
062456 -            HP ScanJet IIc Manual If you<0483 If you have ROM
062457 -            conflicts --------------------------------- Resolve
062458 -            address conflicts by changing the switches on one of the
062459 -            conflicting cards to choose a new location within the
062460 -            computer's address space.  Use the following table to
062461 -            choose new switch positions to resolve the address con-
062462 -            flict:
062463 -
062464 -                  Selecting ROM Address Switch Positions
062465 -
062466 -                  Expansion ROM          Configuration
062467 -                     Address            Switch Settings
062468 -                   Hexidecimal             2   3   4
062469 -
062470 -          -------->C8000-CBFFF             1   1   0 CC000-CFFFF
062471 -          0   1   0          <QN68 D0000-D3FFF             1   0   0
062472 -          D4000-D7FFF             0   0   0          D8000-DBFFF
062473 -          1   1   1          DC000-DFFFF             0   1   1
062474 -          EC000-E3FFF             1   0   1          E4000-E7FFF
062475 -          0   0   1
062476 -
062477 -
062478 - I asked George how he determined which was the "conflicting cards" and
062479 - what switch to adjust, under the explanation given in HP's manual. He
062480 - said some HP Scanjet IIc controllers do not have switches, so the
062481 - settings are done entirely by software, in this case within the OS2
062482 - Windows "Settings" folder for the Omnipage program.
062483 -
062484 - I asked George to try our setting to see if he can recreate the
062485 - error, since that is the usual approach to identify a problem.
062486 - ..
062487 - I will try his ideas, and advise on results.
062488 -
062489 -
062490 -
062491 -
062492 -
0625 -
0626 -
0627 - 1151 Investigated Memory Assignments
0628 -
062801 - Looks like we have to open the computer to set the switches.  Let's
062802 - see if we can determine current hardware settings from the record.
062803 -
062804 -    The installation record on 920115, ref SDS 1 line 212, shows a
062805 -    recommendation from HP ScanJet IIc program, "readme.txt," to set
062806 -    param for emm386 in config.sys under DOS 5.0, as follows:
062807 -
062808 -                         cc00-cfff
062809 -
062810 -    ..
062811 -    Looks like we implemented this recommendation and correlated
062812 -    the switch settings per ref SDS 1 line 268, which says this is the
062813 -    default condition.  Had to do this to get the swtchset.exe to run
062814 -    in order to show the settings for the controller card.
062815 -
062816 -
062817 - Based on the prior record, tried running HP Scanjet diagnostic in an
062818 - os2 DOS window, and got:
062819 -
062820 -        The Recommended switch setting will
062821 -           usually provide a successful installation.
062822 -
062823 -        However, if it doesn't work, try the
062824 -           alternate settings listed.
062825 -        ..
062826 -        Switch number                 1  2  3  4    ROM Address
062827 -        CARD FOUND at Switch Setting  1  0  1  0    CC000-CFFFF
062828 -
062829 -
062830 -    This just seems to be reporting the current assignment, rather
062831 -    than provide a series of "alternate settings listed."
062832 -
062833 -
062834 -     ..
062835 -    Win-Os2 Settings
062836 -
062837 -    Since HP diagnostics report our hardware is configured for
062838 -    cc000-cffff, decided to try this before trying George's
062839 -    recommendation.  I entered in Omnipage folder under Win-OS2
062840 -    Settings for
062841 -
062842 -                mem_exclude_region:   cc000-cffff
062843 -
062844 -
062845 -    Note that is is slightly different from the spec originally tried
062846 -    at ref SDS 3 line 87, showing:
062847 -
062848 -                mem_exclude_region:   cc00-cfff
062849 -
062850 -
062851 -
062852 -
062853 -
062854 -  ..
062855 - Scanner Tested Successful
062856 -
062857 - This ultimately proved successful, indicating the problem was an in-
062858 - correct memory address specification.  We did not use IBM's suggested
062859 - memory address, but did change the form of the specification we have
062860 - been using all along.
062861 -
062862 -
062863 - Here is what was done for testing:
062864 -
062865 - Scanned about 5 pages of OS2 Professional article at ref SDS 12 line
062866 - 121301.  This ran successfully.
062867 -
062868 - ..
062869 - Tried scanning 2 pages of PC Magazine which has much denser text
062870 - than is used in OS2 Professional, and found Omnipage would not boot.
062871 - When I clicked on the Omnipage folder, OS2 read the disk, then
062872 - processing stopped.  Tried pressing <Enter> on folder, and got same
062873 - result. Tried selecting "Program" in "Open" for the Omnipage folder,
062874 - and got same result.
062875 -
062876 - Tried calling Omnipage from Win Program Manager.  Got message that
062877 - Omnipage is already running.
062878 -
062879 - Used Ctrl Esc, and found it listed in "Window List", so opened it.
062880 -
062881 -       Ordinarily if a program is already running, then the icon should
062882 -       appear cross-hatched, and when User clicks on icon, the applica-
062883 -       tion should become the current application.
062884 -       ..
062885 -       In this case neither of these conditions occurred.
062886 -
062887 -
062888 - The first scan job was still showing on the screen, so I closed it,
062889 - then tried to scan the second article. Scanning operation would not
062890 - initiate.  Tried several times.  Closed Omnipage and tried booting it
062891 - again from OS2 desktop.  This worked.
062892 -
062893 - Began scan job using 400 dpi because of small print in magazine.
062894 -
062895 - First page seemed to scan okay.  During scan of page 2, got message:
062896 -
062897 -
062898 -                 A scanner error occurred, aborting
062899 -
062900 -                               OK
062901 -
062902 -
062903 -       ...could only press <Enter>.  Then got message to "Recognize
062904 -       Now?" and I said "yes."
062905 -       ..
062906 -       This seems to have successfully scanned 1 page of the two
062907 -       I tried.
062908 -
062909 -
062910 - Will try closing down OS2 and starting over.
062911 -
062912 -
062913 - Successful Tests
062914 -
062915 - This was successful in scanning about 5 pages of a 2 page article, as
062916 - reported at ref SDS 12 line 62.  For testing purposes, I scanned the
062917 - same page 4 times to get 5 pages, and this ran successfully.
062918 -
062919 -
062920 - I then scanned another article of about 23 pages from PC Magazine,
062921 - reviewing 14 different "Contact Management" software programs, per ref
062922 - SDS 13 line 54.  This ran successfully.
062923 - ..
062924 - I conclude Omnipage and the HP ScanJet IIc are running
062925 - comparable to the performance when running directly under Win 3.1 and
062926 - DOS 5.0, which was the objective of this problem at ref SDS 7 line 46.
062927 -
062928 -
062929 -
062930 -
062931 -
0630 -
0631 -
0632 - Analysis
0633 -
063301 - There are three apparent possible factors that result in a successful
063302 - configuration:
063303 -
063304 -         ..
063305 -     1.  Originally we entered mem_exclue_spec in 4 place
063306 -         strings, because that was how the spec was entered in
063307 -         config.sys for emm386.exe under DOS 5.0, see ref OF 1 line 3.
063308 -         This instruc- tion came from HP Scanjet IIc Readme.txt, per
063309 -         ref SDS 1 line 212, and was confirmed by Omnipage manufacturer
063310 -         rep at ref SDS 2 line 208.
063311 -
063312 -         Perhaps OS2 requires the 5 place strings, as shown in the HP
063313 -         Manual, cited above, but Win 3.1 under DOS 5.0 does not.
063314 -
063315 -     2.  The record of installation for OS2 shows that instead of
063316 -         entering:
063317 -
063318 -                mem_exclude_region:   cc00-cfff
063319 -
063320 -         ...we actually entered:
063321 -
063322 -                mem_exclude_region:   cc00-c000
063323 -         ..
063324 -         This appears most likely to have been a data entry
063325 -         error, since this part of the record at ref SDS 3 line 87, is
063326 -         linked to the record of the original installation that shows
063327 -         the correct string.  Moreover the address appears to be
063328 -         incorrect on its face; however, there remains the possibility
063329 -         that the original entry to OS2 Omnipage folder for mem_exclude
063330 -         region was ...c000 instead of ...cfff.
063331 -
063332 -     3.  We had to install OS2 several times to get a successful con-
063333 -         figuration because IBM's installation instructions were incor-
063334 -         rect, see ref SDS 4 line 57.  On the final installation, we
063335 -         may have either set mem_exclude_region incorrectly, or may not
063336 -         have set it at all, since the discussion with George, above,
063337 -         showed there was no setting today.
063338 -         ..
063339 -         Also noticed that after I set it and did a test that
063340 -         failed, I opened the folder and found it was not really set,
063341 -         evidently because I forgot to "save" it.  May have just double
063342 -         clicked on the close button.
063343 -
063344 -
063345 -
063346 -  ..
063347 - Further Tests Indicate Causation
063348 -
063349 - Tried using 4 place mem strings, per analysis option 1.  Scanner would
063350 - not load, which is similar to one of the problems originally reported
063351 - at ref SDS 7 line 82.  After several tries, scanner loaded and was
063352 - able to scan a single page, but when multiple pages were scanned, the
063353 - program crashed during "recognition" on about page 3, with error mes-
063354 - sage reported for the original problem at ref SDS 7 line 46.
063355 -
063356 - After changing back to 5 place spec strings, was able to run scanner
063357 - successfully, as reported at ref SDS 13 line 58.
063358 -
063359 - This seems to indicate os2 requires 5 place memory address strings,
063360 - but Win 3.1 under DOS 5.0 does not.  Our config.sys under DOS 5.0
063361 - still has 4 place strings, and Omnipage runs well.
063362 -
063363 -