binary_to_term atom table overflow threat

Wiger Ulf ulf.wiger@REDACTED
Wed Mar 12 08:40:04 CET 2003

For the record, UBF is pretty good at creating atoms as well. (:

(In UBF, they're called "constants", but of course, when you type-specify
your protocol, you can avoid using them...)


----- Original Message -----
From: "Shawn Pearce" <spearce@REDACTED>
To: <erlang-questions@REDACTED>
Sent: den 11 mars 2003 21:48
Subject: binary_to_term atom table overflow threat

> So what is the best way to deal with the binary_to_term/1
> atom table overflow problem (you know, 1 million atom
> limit, binary_to_term/1 can make any atoms it pleases)?
> Any chance we might see a binary_to_term/2 where if the
> second argument is false term's which are atoms in the
> input binary and are not already in the atom table are
> returned as strings (lists)?  Might not make sense to
> do though, given that UBF exists...
> --
> Shawn.
>   Your lucky number has been disconnected.

More information about the erlang-questions mailing list