THE WELCH COMPANY
440 Davis Court #1602
San Francisco, CA 94111-2496
415 781 5700
rodwelch@pacbell.net


S U M M A R Y


DIARY: April 26, 2003 10:36 AM Saturday; Rod Welch

Gary reports diary and summary getting out of sync.

1...Summary/Objective

ACTION ITEMS.................. Click here to comment!

1...I think I am going to start on a script to validate at least

CONTACTS 
0201 - Boeing                                                                                                                                                             O-00000816 0505
020101 - Mr. Garold L. Johnson
020103 - Modeling and Simulation                                                                                                                                          O-00000816 0505

SUBJECTS
04 Pointer Data Base for 01 Lines in the Diary are Getting Out of Syn
Johnson, Gary Diary Out of Sync with Diary Summary.
25.bat Global pointers, 910304
Archiving Diary Records, and SAA
Maintenance Pointers Verify Align with Actual Records Gary Planning P

1107 -
1107 -    ..
1108 - Summary/Objective
1109 -
110901 - Follow up ref SDS 1 0000.
110902 -
110903 - Received a letter from Gary this morning saying that....
110904 -
110905 -    1.  I accidentally discovered a set of missing Diary entries:
110906 -
110907 -        April 13 shows:
110909 -         ..
110910 -        0833  SDS - Problems -- Error Getting Diary For Rod
110911 -        0836  SDS - Using -- Record Threads
110912 -        0918  SDS++ -- Flag Records Converted to the Web
110913 -        0928  SDS - Learning -- Formatting Email with Multiple Messages
110914 -        1035  SDS - KM -- Projects, Contexts, Links
110916 -         ..
110917 -    2.  The directory for that date shows:
110918 -
110919 -        071058 -- missing - fixed
110920 -        082447 -- missing - fixed
110921 -        083346
110922 -        083613
110923 -        091832
110924 -        092823
110925 -        103500
110927 -         ..
110928 -    3.  The issue isn't that these were difficult to fix once I knew
110929 -        they were missing, but how it is that they went missing in the
110930 -        first place.
110932 -  ..
110933 - I am not sure, but it seems like Gary is pointing out that
110934 - occassionally, the 01 line for the record description is lost in the
110935 - 04 pointer data base file of an archived diary record in the 02 data
110936 - base directory.
110938 -  ..
110939 - I have noticed this as well in the past few months.  Seems to occur
110940 - maybe once a week, once every two two weeks.  This can be considered
110941 - infrequent, but raises the question of why this occurs, and what to
110942 - do about it.
110943 -
110944 -       [On 031101 developed better method to recover from errors in the
110945 -       data base that show up in the Diary Summary and in Subject
110946 -       reports.
110947 -
110949 -  ..
110950 - My sense is that fixing this problem is somewhat difficult, involiving
110951 - opening the record at issue using F1 F3 file management, may not a
110952 - favorite tool, assigning the SDS profile of 47 2, and the scrolling
110953 - through the profile levels 1 2 3 1, so that the SDS profile is
110954 - assigned, then creating a link to the SDS record of origin, and then
110955 - opening the target record and pressing F2 to update the 04 pointers.
110957 -  ..
110958 - Another way to fix it is to open the approprite 04 data base and
110959 - manually enter the 01 record description in the correct place in the
110960 - file using cut and paste from the diary record in the 02 directory, or
110961 - just keying it it.
110963 -  ..
110964 - Since this seems like a complex set of steps, and Gary says that
110965 - fixing these problems is not difficult, not sure if we are talking
110966 - about the same thing.
110968 -  ..
110969 - Gary continues...
110970 -
110971 -    4.  Once a record disappears from the system, finding that it is
110972 -        missing becomes a matter of accident.
110974 -  ..
110975 - In my experience, missing diary records are somewhat obvious because
110976 - the time invested to create them.  This builds an expectation that a
110977 - record exists, and so provides a mental alert, when the record is
110978 - missing from the Diary Summary, assuming this is the problem Gary is
110979 - describing.
110981 -  ..
110982 - Gary continues.
110983 -
110984 -    5.  I think I am going to start on a script to validate at least
110985 -        some of the pointer files against the actual record data.  I
110986 -        don't think I know enough to get this right all at once, but I
110987 -        ma getting tired of encountering this particular flavor of
110988 -        problem. It looks to me to be worth spending some energy on.
110990 -  ..
110991 - I like this idea of at least verifying the 04 pointers for the
110992 - current date.  I am inclined to the idea this is a memory problem
110993 - related to 640K issue.
110995 -  ..
110996 - Not sure if we want to be validating the entire data base of pointers
110997 - every day.  Would like to hear Gary's ideas.
110998 -
110999 -      [On 030602 Gary encounters problem again, and threatens a
111000 -      solution. ref SDS 2 K56J
111001 -
111002 -
111003 -
111004 -
111005 -
111006 -
111007 -
111008 -
111009 -
111010 -
111011 -
1111 -
Distribution. . . . See "CONTACTS"