Update fails, if run in tweo instances

[lscheffler]

📝 Provide detailed reproduction steps (if any)

1. Run RunThor.prg WITH n, where n>0 on startup routine
2. Let more then n days pass by
3. Run one IDE of VFP with the startup
4. Run immediately a second IDE of VFP with the startup

✔️ Expected result

Update should proccedd in the first instance, ignoring the second

❌ Actual result

Second instance stops with an error. (no debug info since the first instance altered the prg, so source is out of date)

[Jimrnelson]

@lscheffler

The parameter to RunThor is used to compare today's date with the last date that Thor's CFU completed successfully (not the last date is was started).

Since the last completion date had not changed when you immediately started your second VFP session, that session would try to run CFU as well.

[rschummer]

One solution is to lock the record in the table that holds the previous update date and keep the lock until the update run is complete. Any other instances can attempt to lock the record and if they fail, don't run the update process.