recommendations of email account providers or have a go at my C question.

[Mark Blackman]

On Thu, Jun 21, 2001 at 04:03:06PM +0100, Tim Borgeaud wrote:
> 
> 
> My old University account is about to expire. So I'm looking for 
> a new email account somewhere. I'd love shell/web/ftp services but
> a simple email account would do. 
> 
> The account would be for personal and FreeBSD list related email.
> I'd like to be able to enable some kind of filtering (on the host
> machine) and to be able to get my email easily from home and work 
> so I'd prefer not to use POP.
<plug class="blatant">
could try netscalibur.co.uk :) and use IMAP.
sign up online (with CC handy) at ..

http://www.netscalibur.co.uk/services/1_1_1_dial.html
</plug>

> 
> 
> Here's the C question - it's not really a problem but I'd like to
> know if something can be done or not (I suspect not).
> 
> Imagine I have a data array or some sort and a void pointer to this
> data:
> 
> void *My_data;
> 
> Now suppose that the data is a series of 2 byte integers (short). 
> I decide that I want one of these integers found at an offset.
> 
>  ((short *)My_data + offset) is then the pointer to this integer. 
> 
> Now I have a function foo which is defined as
> 
> double foo(int *bar);       the int is 32bit
> 
> I'd like to use the My_data array and offset directly (no matter 
> how ugly the code looks) :
> 
> foo(  &((int)*((short *)My_data + offset))
> 
> But I can't do this because of the cast from short to int.
> I can't use the & operator on the casted data - 
>  (int)*((short *)My_data + offset) is not an l-value.
> 
> I am assuming this is equivalent of something like
> 
> foo( &(atoi("10")) );   
> 
> But is there something about casts that I don't know that would allow
> another way of doing this?
> 

This is de-referencing a int pointer and it will actually give you
the 32bit value residing at that offset (i.e. two shorts slammed together).

You should either shift the short value into a temporary 32 bit container
or rewrite the entire data array as ints. Does foo attempt to read more
than the first 32bit int at *bar?

- Mark