Keeping deleted records

[murray]

I'm hoping someone can help me with a solution to my problem. I would like to stop records being deleted irretrievably from a database by a basic level user, which I can do by password control, but I don't want them to remain indefinitely in the existing database as they will affect calculations.

What I am hoping to do (as simply as possible) is to "delete" the records by first exporting the so-called deleted records, to be saved in another file somewhere before actually being deleted from the file. This way they can be inspected before being actually deleted from the storage file by someone with more access than the basic user.

I would appreciate any advice on the basics of this process, but would also need to do this for about 15 related databases of varying complexity and field count. Ideally, I'd like them all to go to one file rather than having separate "deleted" archives for each database.

Any suggestions are appreciated.

Thanks, Murray

[dbruggmann]

"Ideally, I'd like them all to go to one file rather than having separate 'deleted' archives for each database."

Theoretically this is possible, I think

[murray]

G'day again Detlev,

I thought it would be a pretty hefty challenge but as you know, somethings that are simple for one person are difficult for another.

I have the Audit Tracking working in some of my simple dbs and will attempt to get them working in the more complicated ones, but this is a "per record" method, so once the record is deleted, so is the audit track.

I'm just trying to get the solution as sturdy as possible but I guess I'll have to accept that some user care and common sense will be important.

Thanks for your input.

Cheers, Murray

[dbruggmann]

Hello Murray

"I have the Audit Tracking working in some of my simple dbs and will attempt to get them working in the more complicated ones, but this is a 'per record' method, so once the record is deleted, so is the audit track."

But Ray's

[murray]

Sorry for the delay Detlev,

You make it sound simple!

I would like it to be a seamless process without any dialogs. Is it really that simple?

I can't find any examples anywhere - have you seen any?

Cheers, Murray

PS. I had to get Ray to help me with his audit tracking - I got stuck!

[CobaltSky]

Hi Murray, Hi Detlev,

Murray, if you're looking for a relatively simple way to Transfer records to a holding file rather than deleting them, here is a process you might like to consider:

1. Save a clone of the file you're working on and rename the clone appropriately to identify it as a 'hold' file corresponding to the file you cloned it from.

2. Import a record or two from the original; file into the clone using the 'Matching Names' import option and then create a script in the clone which has only the following steps:

Import Records [Restore, No Dialog]

Show All Records

Close [ ]

- then delete the sample records you imported and go back to the original file.

3 Now, in the original file, create a new script called 'Delete Record' and set it up as follows:

Show All Records

Omit Record

Show Omitted

Perform Script [sub-scripts, External: "YourHoldFile.fp5"]

Delete Record/Request [No Dialog]

Refresh Window [bring to front]

where the external script at step 4 points to the import script you've created in the hold file.

That's basically it. When the Delete Record script runs, the current record will disappear from the main file and be transferred to the hold file.

If you want to add finesse to the process, there are two more things you may wish to consider:

A: add a procedure which 'saves the found set' before the delete procedure and then restores it afterwards (so that, for the sake of convenience, the found set in the main file is not lost when a record is deleted).

B: Set up two more scripts exactly like the above but in reverse (ie the script at Step 2 in the main file and the script at step 3 - named 'Reinstate Record' - in the hold file). That will give you a mechanism by which to efficiently 'undo' an erroneous record deletion.

The chief limitation of the technique is that it requires a separate hold file for each file in your solution. But you need only ever have one hold file open at a time (and then only while you are actually transferring data to or from it). But it is not difficult to set up and will be robust and easy to use.

[murray]

Thanks for that Ray - I'll give it a go.

Cheers,

Murray

[Kundinger]

Hi Murray,

Below is a link to a FMForums post similar to your request...

<<< http://www.fmforums.com/threads/showflat.php?Number=87992 >>>

in it is a reply I posted with some ideas that may work for you.

Good Luck!

Bob Kundinger

[email protected]

[murray]

Thanks Bob,

I'll work through this,

Cheers, Murray