stock_market_player wrote on September 15th:
> Thank you very much for your reply Antoine.
You are welcome.
> So RPM only tries to automatically restore the partition
> table when he sees that the partitions in the MBR don't match
> the information in the special RPM partion at the end of the
> disk, correct?
Yes, something along those lines.
To be certain your understanding is correct, you can use the simulation
tools: first using part244sim.exe you create several partitions, some of
them in the VTOC and lacking entries in the MBR (be sure to include a F0
partition); then use part240sim.exe (which only edits the MBR and will
not touch to the VTOC) to modfiy the MBR, to simulate the effect of a
"foreign" tool; last, check what will show part244sim.exe when it mixes
the entries in VTOC _and_ the "new" entries in the MBR: I expect red
entries...
> So, in principle there shouldn't be any problem with removing
> one HDD. From RPM point of view, all the HDDs are coherent,
> in the sense that the information in the MBR matches with the
> info in the special RPM partition at the end of the disk.
>
> > *UNLESS* you are using a strange setup of a customized version of
> > Ranish boot manager to boot directly to the 2nd or 3rd disk _and_
> > hiding some other partitions meanwhile, it is very unlikely to have
> > RPM writing partition tables of the _other_ drives "automagically."
>
> What exactly is the meaning of "to boot directly to the ..."?
Long ago (2.37 ?), RPM allowed to boot directly to the 2nd+ harddisk,
and also allowed to select an entry which would select up to 3
partitions, "hiding" the others; I am not sure official version of RPM
allowed the mix of the two options (i.e. booting the 2nd disk _and_
hiding partitions on say the 1st disk), but it exists boot managers
around which do have such possibility (plus RPM was open sourced.)
I do not believe the 2.4x serie still have such options, but there are
stigmates of them in the UI. And it is certainly possible to build them
in a "customized version"...
Antoine