info needed: what happens when ints have different sizes ?

[DexterMagnific]

Dear developer,

If I serialize an int on a 64 bit machine, an deserialize the tpl on a machine whose ints are 32 bits, what would happen ?

Another related question: if I serialize an int whose value is 5 on the 64 bit machine, does the tpl image allocates 8 bytes for it (plus any metadata it needs) or does it optimize it (e.g. by allocating only 1 byte). I'm asking this because we have big size restrictions on our embedded system.

Thanks

[troydhanson]

On Jun 19, 2019, at 2:18 AM, DexterMagnific ***@***.***> wrote: Dear developer, If I serialize an int on a 64 bit machine, an deserialize the tpl on a machine whose ints are 32 bits, what would happen ?

This is why tpl has different format characters for each size of int (i and I, u and U, etc). When you unpack, the same character has to be used in the format string, that was used to pack it. So a U can only get unpacked to a U (64 bit) and so on. I recommend using int32_t and other explicit sized types in programs that serialize or unserialize just to keep things very clear.

Another related question: if I serialize an int whose value is 5 on the 64 bit machine, does the tpl image allocates 8 bytes for it (plus any metadata it needs) or does it optimize it (e.g. by allocating only 1 byte). I'm asking this because we have big size restrictions on our embedded system.

Unfortunately it has no compression.

…

Thanks — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.