Aiko,

I believe the standard approach is to assign an allowed range of keys to the laptop. This adds a bookkeeping job, as you can no longer causally increment a single integer, but it gets the job done.

>I have a question for anyone with experience might be able to help. I am trying to design this system that allow user to receive a schedule table that the user need perform for next 2 weeks This table is loaded to laptop and used as schedule table. This record will be marked as done each time the user finish tasks. Then this table would be uploaded to office system and be processed accodingly. Updating dates and information is stored in the respective tables So this schedule table be going back and force.
>
>A new record would be added in the laptop as an ad hoc task too.
>This could be system generated surrogate key but. I could receive
>a schedule file with surrogate key last number 100
>The next woker could be receiving 110. Then I could add new record
>in my laptop and reach 110 number. From the 110 up my surrogate
>key has duplicate numbers as the next worker receved.
>
> Should I have surrogate key for this table? But if so
>how I am going to deal with duplicating key situation.
>I feel I need to take another approach for this case.
>But poor me for the moment I have no good idear.
>
>I am sure people out there must have faced with such
>prediicaments before me.
>
>What is the best solution for this kind of case. Sorry to describe such a situation to you all.
>Any suggestion is welcome.
>
>Aiko in DC