I'm trying to install the latest Scientific Linux on a new machine. I created an installation CD and booted from it, choosing to install by FTP. It asks for a manual TCP/IP configuration. I've done this before, for SL4. Now I can't seem to get the right information. No matter what I put for the IPv4 address, it is rejected with the message, "Missing Information: You must enter both a valid IPv4 address and a network mask or CIDR prefix".

The field to be filled in looks like this:
IPv4 address: _______________ / _______________

I have tried these entries:
<IP address> / 255.255.254.0
<IP address> / 255.255.255.0
<IP address> / 255.255.0.0
<IP address> / 23
<IP address> / /23
255.255.254.0 / <IP address>
23 / <IP address>
/23 / <IP address>
where <IP address> is the address assigned to this machine in dotted-quad format. It's a 9-bit subnet.

I've run out of ideas.

A 9-bit subnet would be /9 or 255.128.0.0.

ie:
<ip address>/255.128.0.0

or

<ip address>/9

Erp. I meant the mask has 9 zeros, not 9 ones.
There are 512 possible addresses in the subnet.
9+23 = 32.

Oops, sorry about that. I was thinking in terms of "subnet mask" instead of just subnet.

Is this a Class A or Class B network? For Class A, with a 23-bit subnet mask, you would have 8 network bits, 15 subnet bits, and 9 host bits. For Class B, you'd have 16 network bits, 7 subnet bits, and 9 host bits. In either case, I believe you would have a maximum of 510 hosts.

I've seen one client that used a 23-bit subnet mask with a Class C network address scheme. What is the first octet of the IP address you are using?