not a valid bookmark - not using a recordset

Collapse
This topic is closed.
X
X
 
  • Time
  • Show
Clear All
new posts
  • ml_sauls

    #1

    not a valid bookmark - not using a recordset

    I've built a system to enter and manage purchase orders. This is in
    use by >10 clients. Some use it in Access 97, most are in A2k. About
    half use it through a Citrix implementation. It is separated into a
    front end and 3 back ends (1 network for permanent storage, 1 local for
    temp storage, and 1 local for storage for particular user).

    One particular form is failing for a single client (Access 2000, using
    Citrix). This form is a form with a subform, used for entering po's.
    The outer form contains summary information, such as vendor. The child
    form is for entering lines. In order, the fields are account,
    quantity, unit price, extended (qty * unit), then two free text fields.
    It is bound to the outer form on a shared id, which uses an autonumber
    key to assure a unique id across all users.

    This single client encounters an error when moving from the unit price
    field to the first free text field. All fields on the current line
    display #error. The error message displayed is "Not a valid bookmark".
    The only code on the unit price field is to set the extended cost
    value equal to quantity * unit price. The error message is not
    produced through my error trap. If I ctrl+break to see what executes
    next, nothing does

    Neither the parent nor child form uses a recordset. I have not
    declared, nor do I use, any bookmarks on either form. I've declared
    recordsets on each, which are used solely within their subs or
    functions, then killed when exiting the sub or function. The parent
    form is unbound, the child form is bound. The error is not consistent.
    Sometimes it fails on the 2nd po to be entered, sometimes the 4th,
    sometimes further out. I have not yet seen it fail on the 1st. If the
    form is closed then reopened, the first entered will not encounter
    errors. Once it has produced the error, even rolling the mouse over
    the subform will produce it. The error message box will sometimes be
    displayed multiple times.

    I cannot find any consistency in the number of times the error is
    displayed, or in the number of entries it will accept prior to
    producing it, though it is typically between 2 and 12.
    I cannot find any consistency in the data used in the record that
    produces the error.
    I have reconfigured the client's application from the master.
    I have relinked another client's application to this client's tables
    and encounter the error.
    I have relinked this client's app to another client's tables and do not
    encounter the error.
    I can use another client's accounts with the client's app & tables and
    do not encounter the error.
    I can use this client's accounts with another client's app & tables and
    do encounter the error.
    I do not see any differences between Citrix vs non-citrix, though
    another tester reports that the error is not encountered in a
    non-Citrix environment.
    I've not yet encountered the error if I enter all detail lines in the
    subform first, then enter data in the outer form.
    I've tried placing break points and error handlers within all code on
    the subform, as well as the subform object on the parent form. None of
    these are encountered when this error is produced.

    I have rebuilt all tables. I noticed when testing, that sorting
    certain fields in the account table would produce this error, when
    viewing them through the table link. These were text fields, using the
    A-Z button to sort asc/desc. I rebuilt that table from scratch. I
    repopulated it by exporting data from the former table to a text file,
    then importing the data from that text file. This did not change the
    testing results.

    My only thought for a fix is to entirely redesign the form to eliminate
    use of a subform, which will take more time than I currently have
    available. Of course, on the other hand, I'd be partly done with it if
    I'd not done so much testing/rebuilding.

    I am at a loss as to what could be causing this error? Any ideas?

  • ml_sauls

    #2
    Re: not a valid bookmark - not using a recordset

    Apologies - both the parent and child forms are bound to tables, each a
    different table. Both are in the local back end.

    Comment

    • deluxeinformation@gmail.com

      #3
      Re: not a valid bookmark - not using a recordset


      ml_sauls wrote:
      Apologies - both the parent and child forms are bound to tables, each a
      different table. Both are in the local back end.
      Doesn't using Citrix mean that you have multiple users using the same
      copy of the front end? Does this also mean that you have multiple
      users using the same copy of the 'local' and 'temporary storage'
      databases? Have you determined whether the problem happens when
      multiple users are accessing these databases via Citrix? You might
      want to revisit the storage architecture of your back end databases.

      Have you compacted and repaired all databases? Does this particular
      Access 2000 installation have the latest service packs installed?

      Bruce

      Comment

      • David W. Fenton

        #4
        Re: not a valid bookmark - not using a recordset

        deluxeinformati on@gmail.com wrote in
        news:1161119003 .760410.146860@ e3g2000cwe.goog legroups.com:
        Doesn't using Citrix mean that you have multiple users using the
        same copy of the front end?
        Only if you've set it up wrong.

        Under a terminal server setup, you'd do exactly the same as running
        on individual workstations -- each user gets an individual copy of
        the front end. Failing to do that leads to the same problems that
        any sharing of a front end causes.

        Someone people seem allergic to doing this with some idea that it's
        wasteful of disk space. Well, unless you have a 1GB front end (which
        would mean you're doing something incredibly wrong), I can't see how
        it would be an issue. The front ends of most of my clients'
        applications are under 10MBs, so even supporting 100 users isn't a
        significant amount of disk space on modern-day servers.

        --
        David W. Fenton http://www.dfenton.com/
        usenet at dfenton dot com http://www.dfenton.com/DFA/

        Comment

        • ml_sauls

          #5
          Re: not a valid bookmark - not using a recordset

          Though it is through Citrix, neither the front end or either of the
          local table files are shared by multiple users. Even if it were, all
          my testing was with only me, which tells me that even if sharing causes
          other issues, it does not seem to be a factor in this instance.

          Yes, I have rebuilt all databases, including compact/repair. The front
          end has been reconfigured from an unused master. I have also
          decompiled/recompiled that. I have been assured that the client is
          running A2k with all service packs installed. Regardless, I know the
          local environment at work has all service packs installed, and again,
          that's where I'm testing and where I'm getting an error. OS is Windows
          xp.

          I've cut my losses on testing and have begun redesigning both the
          parent and subform to be unbound. Though I've been unable to pinpoint
          just *why* this error occurs, I am hoping that making everything
          unbound will eliminate it.

          deluxeinformati on@gmail.com wrote:
          Doesn't using Citrix mean that you have multiple users using the same
          copy of the front end? Does this also mean that you have multiple
          users using the same copy of the 'local' and 'temporary storage'
          databases? Have you determined whether the problem happens when
          multiple users are accessing these databases via Citrix? You might
          want to revisit the storage architecture of your back end databases.
          >
          Have you compacted and repaired all databases? Does this particular
          Access 2000 installation have the latest service packs installed?
          >
          Bruce

          Comment

          • scottkirk@gmail.com

            #6
            Re: not a valid bookmark - not using a recordset


            ml_sauls wrote:
            Though it is through Citrix, neither the front end or either of the
            local table files are shared by multiple users. Even if it were, all
            my testing was with only me, which tells me that even if sharing causes
            other issues, it does not seem to be a factor in this instance.
            >
            Yes, I have rebuilt all databases, including compact/repair. The front
            end has been reconfigured from an unused master. I have also
            decompiled/recompiled that. I have been assured that the client is
            running A2k with all service packs installed. Regardless, I know the
            local environment at work has all service packs installed, and again,
            that's where I'm testing and where I'm getting an error. OS is Windows
            xp.
            >
            I've cut my losses on testing and have begun redesigning both the
            parent and subform to be unbound. Though I've been unable to pinpoint
            just *why* this error occurs, I am hoping that making everything
            unbound will eliminate it.
            >
            deluxeinformati on@gmail.com wrote:
            >
            Doesn't using Citrix mean that you have multiple users using the same
            copy of the front end? Does this also mean that you have multiple
            users using the same copy of the 'local' and 'temporary storage'
            databases? Have you determined whether the problem happens when
            multiple users are accessing these databases via Citrix? You might
            want to revisit the storage architecture of your back end databases.

            Have you compacted and repaired all databases? Does this particular
            Access 2000 installation have the latest service packs installed?

            Bruce

            Please pass along if this resolves the issue. I have had similar
            problem pop up in the last 3 months. My issue is not on a citrix box
            but does occur only on some workstations and randomly during data
            entry. This is ony happening on Access XP or 03 with an access back
            end. If I use Sql Server I have nothing like this.

            I too am pulling my hair out over this issue.

            Thanks

            Scott

            Comment

            • David W. Fenton

              #7
              Re: not a valid bookmark - not using a recordset

              "ml_sauls" <ml_sauls@yahoo .comwrote in
              news:1161139546 .893640.69040@h 48g2000cwc.goog legroups.com:
              I have rebuilt all databases, including compact/repair. The front
              end has been reconfigured from an unused master. I have also
              decompiled/recompiled that. I have been assured that the client
              is running A2k with all service packs installed. Regardless, I
              know the local environment at work has all service packs
              installed, and again, that's where I'm testing and where I'm
              getting an error. OS is Windows xp.
              This is a corruption issue, without doubt, as Peter Miller, who is
              the recognized non-Microsoft expert on Jet, says in this post (all
              on one line, of course):


              4c34872
              I've cut my losses on testing and have begun redesigning both the
              parent and subform to be unbound. Though I've been unable to
              pinpoint just *why* this error occurs, I am hoping that making
              everything unbound will eliminate it.
              Unbound may avoid the error, but it won't eliminate the problem
              that's causing the corruption. You need to fix *that*.

              Keep in mind that the corruption could be in particular records, so
              you may need to import groups of records into your new database,
              rather than just importing the table structures and data all at
              once. Peter advises importing the data structure, then importing
              groups of records until you find the corrupted records.

              --
              David W. Fenton http://www.dfenton.com/
              usenet at dfenton dot com http://www.dfenton.com/DFA/

              Comment

              • scottkirk@gmail.com

                #8
                Re: not a valid bookmark - not using a recordset


                David W. Fenton wrote:
                "ml_sauls" <ml_sauls@yahoo .comwrote in
                news:1161139546 .893640.69040@h 48g2000cwc.goog legroups.com:
                >
                I have rebuilt all databases, including compact/repair. The front
                end has been reconfigured from an unused master. I have also
                decompiled/recompiled that. I have been assured that the client
                is running A2k with all service packs installed. Regardless, I
                know the local environment at work has all service packs
                installed, and again, that's where I'm testing and where I'm
                getting an error. OS is Windows xp.
                >
                This is a corruption issue, without doubt, as Peter Miller, who is
                the recognized non-Microsoft expert on Jet, says in this post (all
                on one line, of course):
                >

                4c34872
                >
                I've cut my losses on testing and have begun redesigning both the
                parent and subform to be unbound. Though I've been unable to
                pinpoint just *why* this error occurs, I am hoping that making
                everything unbound will eliminate it.
                >
                Unbound may avoid the error, but it won't eliminate the problem
                that's causing the corruption. You need to fix *that*.
                >
                Keep in mind that the corruption could be in particular records, so
                you may need to import groups of records into your new database,
                rather than just importing the table structures and data all at
                once. Peter advises importing the data structure, then importing
                groups of records until you find the corrupted records.
                >
                --
                David W. Fenton http://www.dfenton.com/
                usenet at dfenton dot com http://www.dfenton.com/DFA/
                I have created new databases and imported the data in with no errors or
                problems. The error happens during data entry. I always thought
                corruption until I did everything by the book and it still happens.

                Thanks!

                Scott

                Comment

                • David W. Fenton

                  #9
                  Re: not a valid bookmark - not using a recordset

                  scottkirk@gmail .com wrote in
                  news:1161711119 .633439.154370@ h48g2000cwc.goo glegroups.com:
                  >
                  David W. Fenton wrote:
                  >"ml_sauls" <ml_sauls@yahoo .comwrote in
                  >news:116113954 6.893640.69040@ h48g2000cwc.goo glegroups.com:
                  >>
                  I have rebuilt all databases, including compact/repair. The
                  front end has been reconfigured from an unused master. I have
                  also decompiled/recompiled that. I have been assured that the
                  client is running A2k with all service packs installed.
                  Regardless, I know the local environment at work has all
                  service packs installed, and again, that's where I'm testing
                  and where I'm getting an error. OS is Windows xp.
                  >>
                  >This is a corruption issue, without doubt, as Peter Miller, who
                  >is the recognized non-Microsoft expert on Jet, says in this post
                  >(all on one line, of course):
                  >>
                  >http://groups.google.com/group/comp....ess/msg/6a68a5
                  >c1d 4c34872
                  >>
                  I've cut my losses on testing and have begun redesigning both
                  the parent and subform to be unbound. Though I've been unable
                  to pinpoint just *why* this error occurs, I am hoping that
                  making everything unbound will eliminate it.
                  >>
                  >Unbound may avoid the error, but it won't eliminate the problem
                  >that's causing the corruption. You need to fix *that*.
                  >>
                  >Keep in mind that the corruption could be in particular records,
                  >so you may need to import groups of records into your new
                  >database, rather than just importing the table structures and
                  >data all at once. Peter advises importing the data structure,
                  >then importing groups of records until you find the corrupted
                  >records.
                  >
                  I have created new databases and imported the data in with no
                  errors or problems. The error happens during data entry. I
                  always thought corruption until I did everything by the book and
                  it still happens.
                  It *is* corruption.

                  And simply importing the tables into a new database is likely
                  carrying over the corruption from the old data, as it is data-based
                  corruption (i.e., corruption in particular rows), not structural at
                  the table level or higher.

                  You need to follow the instructions in the post cited above, and
                  *not* just do a single import of table and data together, but import
                  structure first and then finite sets of data until the error occurs,
                  at which point you'll have found the corrupted record(s) which you
                  can replace with a new one and then delete the corrupt one.

                  It could also be that the new database you create is fine and as
                  soon as someone edits it, whatever is causing the corruption in the
                  first place kicks in and corrupts it again. You need to solve the
                  real problem, which is not that your database is corrupt, but that
                  something is repeatedly causing your database to become corrupted.
                  Remove the cause and you don't need fix corrupt databases any more.

                  --
                  David W. Fenton http://www.dfenton.com/
                  usenet at dfenton dot com http://www.dfenton.com/DFA/

                  Comment

                  • scottkirk@gmail.com

                    #10
                    Re: not a valid bookmark - not using a recordset



                    On Oct 24, 3:43 pm, "David W. Fenton" <XXXuse...@dfen ton.com.invalid >
                    wrote:
                    scottk...@gmail .com wrote innews:11617111 19.633439.15437 0@h48g2000cwc.g ooglegroups.com :
                    >
                    >
                    >
                    >
                    >
                    >
                    >
                    David W. Fenton wrote:
                    "ml_sauls" <ml_sa...@yahoo .comwrote in
                    >news:116113954 6.893640.69040@ h48g2000cwc.goo glegroups.com:
                    >
                    I have rebuilt all databases, including compact/repair. The
                    front end has been reconfigured from an unused master. I have
                    also decompiled/recompiled that. I have been assured that the
                    client is running A2k with all service packs installed.
                    Regardless, I know the local environment at work has all
                    service packs installed, and again, that's where I'm testing
                    and where I'm getting an error. OS is Windows xp.
                    >
                    This is a corruption issue, without doubt, as Peter Miller, who
                    is the recognized non-Microsoft expert on Jet, says in this post
                    (all on one line, of course):
                    >>
                    I've cut my losses on testing and have begun redesigning both
                    the parent and subform to be unbound. Though I've been unable
                    to pinpoint just *why* this error occurs, I am hoping that
                    making everything unbound will eliminate it.
                    >
                    Unbound may avoid the error, but it won't eliminate the problem
                    that's causing the corruption. You need to fix *that*.
                    >
                    Keep in mind that the corruption could be in particular records,
                    so you may need to import groups of records into your new
                    database, rather than just importing the table structures and
                    data all at once. Peter advises importing the data structure,
                    then importing groups of records until you find the corrupted
                    records.
                    >
                    I have created new databases and imported the data in with no
                    errors or problems. The error happens during data entry. I
                    always thought corruption until I did everything by the book and
                    it still happens.It *is* corruption.
                    >
                    And simply importing the tables into a new database is likely
                    carrying over the corruption from the old data, as it is data-based
                    corruption (i.e., corruption in particular rows),notstruct ural at
                    the table level or higher.
                    >
                    You need to follow the instructions in the post cited above, and
                    *not* just do a single import of table and data together, but import
                    structure first and then finite sets of data until the error occurs,
                    at which point you'll have found the corrupted record(s) which you
                    can replace with a new one and then delete the corrupt one.
                    >
                    It could also be that the new database you create is fine and as
                    soon as someone edits it, whatever is causing the corruption in the
                    first place kicks in and corrupts it again. You need to solve the
                    real problem, which isnotthat your database is corrupt, but that
                    something is repeatedly causing your database to become corrupted.
                    Remove the cause and you don't need fix corrupt databases any more.
                    >
                    --
                    David W. Fenton http://www.dfenton.com/
                    usenet at dfenton dot com http://www.dfenton.com/DFA/- Hide quoted text -- Show quoted text -
                    Do you think its possible that the program or data in the program could
                    cause this?

                    Thanks

                    Scott

                    Comment

                    • David W. Fenton

                      #11
                      Re: not a valid bookmark - not using a recordset

                      scottkirk@gmail .com wrote in
                      news:1161788044 .971589.199960@ k70g2000cwa.goo glegroups.com:
                      On Oct 24, 3:43 pm, "David W. Fenton"
                      <XXXuse...@dfen ton.com.invalid wrote:
                      >You need to follow the instructions in the post cited above, and
                      >*not* just do a single import of table and data together, but
                      >import structure first and then finite sets of data until the
                      >error occurs, at which point you'll have found the corrupted
                      >record(s) which you can replace with a new one and then delete
                      >the corrupt one.
                      >>
                      >It could also be that the new database you create is fine and as
                      >soon as someone edits it, whatever is causing the corruption in
                      >the first place kicks in and corrupts it again. You need to solve
                      >the real problem, which isnotthat your database is corrupt, but
                      >that something is repeatedly causing your database to become
                      >corrupted. Remove the cause and you don't need fix corrupt
                      >databases any more.
                      >
                      Do you think its possible that the program or data in the program
                      could cause this?
                      Corruption? No.

                      Something could be interfering with the file writes, but it won't be
                      internal to the application or to the data. It could be dropped or
                      unreliable network connections. It could be virus scanners. It could
                      be any software running on the server.

                      The only situation where the software makes a difference is when
                      you're using one of the buggy versions of Access or Jet, but you
                      said you were up-to-date on all of those, no? The key one is that
                      you need Jet SP6 or later. The most common versions I run into are 8
                      and 9.

                      --
                      David W. Fenton http://www.dfenton.com/
                      usenet at dfenton dot com http://www.dfenton.com/DFA/

                      Comment

                      • scottkirk@gmail.com

                        #12
                        Re: not a valid bookmark - not using a recordset



                        On Oct 25, 12:58 pm, "David W. Fenton" <XXXuse...@dfen ton.com.invalid >
                        wrote:
                        scottk...@gmail .com wrote innews:11617880 44.971589.19996 0@k70g2000cwa.g ooglegroups.com :
                        >
                        >
                        >
                        >
                        >
                        On Oct 24, 3:43 pm, "David W. Fenton"
                        <XXXuse...@dfen ton.com.invalid wrote:
                        You need to follow the instructions in the post cited above, and
                        *not* just do a single import of table and data together, but
                        import structure first and then finite sets of data until the
                        error occurs, at which point you'll have found the corrupted
                        record(s) which you can replace with a new one and then delete
                        the corrupt one.
                        >
                        It could also be that the new database you create is fine and as
                        soon as someone edits it, whatever is causing the corruption in
                        the first place kicks in and corrupts it again. You need to solve
                        the real problem, which isnotthat your database is corrupt, but
                        that something is repeatedly causing your database to become
                        corrupted. Remove the cause and you don't need fix corrupt
                        databases any more.
                        >
                        Do you think its possible that the program or data in the program
                        could cause this?Corruption ? No.
                        >
                        Something could be interfering with the file writes, but it won't be
                        internal to the application or to the data. It could be dropped or
                        unreliable network connections. It could be virus scanners. It could
                        be any software running on the server.
                        >
                        The only situation where the software makes a difference is when
                        you're using one of the buggy versions of Access or Jet, but you
                        said you were up-to-date on all of those, no? The key one is that
                        you need Jet SP6 or later. The most common versions I run into are 8
                        and 9.
                        >
                        --
                        David W. Fenton http://www.dfenton.com/
                        usenet at dfenton dot com http://www.dfenton.com/DFA/- Hide quoted text -- Show quoted text -
                        All are running latest SP2 and Jet 8. I'll review the data again but
                        would appreciate any other ideas.

                        Scott

                        Comment

                        • David W. Fenton

                          #13
                          Re: not a valid bookmark - not using a recordset

                          scottkirk@gmail .com wrote in
                          news:1161798690 .118435.11960@i 3g2000cwc.googl egroups.com:
                          All are running latest SP2 and Jet 8. I'll review the data again
                          but would appreciate any other ideas.
                          Have you really gone to each PC and checked this? I one time had a
                          case where I was just sure I'd patched all the machines, and then I
                          implemented code in the app to log Access and Jet versions, and
                          discovered ONE MACHINE with the original Access and no Jet service
                          packs. Once it was patched, the corruption stopped for good.

                          Logging that at startup is a good thing, because it's really easy
                          for machines to revert when somebody wipes the drive and rebuilds
                          it, or installs something that overwrites later versions.

                          --
                          David W. Fenton http://www.dfenton.com/
                          usenet at dfenton dot com http://www.dfenton.com/DFA/

                          Comment

                          • scottkirk@gmail.com

                            #14
                            Re: not a valid bookmark - not using a recordset



                            On Oct 25, 4:26 pm, "David W. Fenton" <XXXuse...@dfen ton.com.invalid >
                            wrote:
                            scottk...@gmail .com wrote innews:11617986 90.118435.11960 @i3g2000cwc.goo glegroups.com:
                            >
                            All are running latest SP2 and Jet 8. I'll review the data again
                            but would appreciate any other ideas. Have you really gone to each PC and checked this? I one time had a
                            case where I was just sure I'd patched all the machines, and then I
                            implemented code in the app to log Access and Jet versions, and
                            discovered ONE MACHINE with the original Access and no Jet service
                            packs. Once it was patched, the corruption stopped for good.
                            >
                            Logging that at startup is a good thing, because it's really easy
                            for machines to revert when somebody wipes the drive and rebuilds
                            it, or installs something that overwrites later versions.
                            >
                            --
                            David W. Fenton http://www.dfenton.com/
                            usenet at dfenton dot com http://www.dfenton.com/DFA/
                            I know what you mean. Any recommendations on code to capture the jet
                            version at login?

                            Thanks!

                            Scott

                            Comment

                            • David W. Fenton

                              #15
                              Re: not a valid bookmark - not using a recordset

                              scottkirk@gmail .com wrote in
                              news:1161815185 .032032.170840@ m7g2000cwm.goog legroups.com:
                              On Oct 25, 4:26 pm, "David W. Fenton"
                              <XXXuse...@dfen ton.com.invalid wrote:
                              >scottk...@gmai l.com wrote
                              >innews:1161798 690.118435.1196 0@i3g2000cwc.go oglegroups.com:
                              >>
                              All are running latest SP2 and Jet 8. I'll review the data
                              again but would appreciate any other ideas. Have you really
                              gone to each PC and checked this? I one time had a
                              >case where I was just sure I'd patched all the machines, and then
                              >I implemented code in the app to log Access and Jet versions, and
                              >discovered ONE MACHINE with the original Access and no Jet
                              >service packs. Once it was patched, the corruption stopped for
                              >good.
                              >>
                              >Logging that at startup is a good thing, because it's really easy
                              >for machines to revert when somebody wipes the drive and rebuilds
                              >it, or installs something that overwrites later versions.
                              >
                              I know what you mean. Any recommendations on code to capture the
                              jet version at login?
                              I got the code either from http://mvps.org/access or from Tony Toews
                              website. I can't remember which.

                              --
                              David W. Fenton http://www.dfenton.com/
                              usenet at dfenton dot com http://www.dfenton.com/DFA/

                              Comment

                              Working...