[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [MiNT] appl_getinfo(APGI_VERSION(=96)) implementation
tor, 16,.12.2004 kl. 08.19 +0100, skrev Philipp Donzé:
> I'm just adding some concerns to the latest API improvements:
> > so, we get a new appl_getinfo_str() function:
> Please don't forget to specify the minimum length the user application should provide for each string!
> I know, it's ugly to predefine the length of each string, but if I didn't miss a mail, no one ever talked about this issue.
You missed a mail :-)
> How does the AES know how many bytes can be written to these strings? What happens if there appears an AES replacement with a very
> long name? :-)
I think I speciffied this in my original posting, didn't I? The string
buffers should be made to take 256 bytes each, except from the aes_id
string which needs room for 8 bytes only.
> > I think that AES can put in the 3rd string a free string with all the
> > informations that may describe the AES, for example the version number,
> > the alpha/beta/release status, the target architecture, the compiler used,
> > the date of the compilation, etc...
> And how assure that the AES application allocated enough bytes to hold the whole string?
It is defined in the appl_getinfo(APGI_VERSION) documentation we are
about to agree on, see my original posting ;-)