NT4 client case issues...

[Matt Chapman]

Luke Kenneth Casson Leighton wrote:
> 
> On Fri, 16 Apr 1999, Christopher R. Hertel wrote:
> 
> > My guess is that the code that does the reverse lookup also enumerates the
> > shares and (for no useful reason) the programmer lower-cased them all.
> 
> you know what?  i bet you that someone added an entire set of "new" code
> to deal with *SMBSERVER rather than dig in to the existing stuff.  and
> that [new] programmer added lower-case-of-share-names.

It's even wierder than that. \\2f6-new.newcollege.unsw.edu.au (with
unrelated NetBIOS name, so connection is via *SMBSERVER) results in
the correct capitalisation, while \\129.94.134.157 lowercases.

I would have thought the logic for connecting via TCP/IP would contain
something like (of course, not using sockets)...

if ((ip = inet_addr(lpszHost)) == -1)
{
	if ((lpHostInfo = gethostbyname(lpszHost)) == NULL)
		return STATUS_BAD_NETWORK_PATH;

	memcpy(&ip, lpHostInfo->h_addr, sizeof(ip));
}

Of course, I know that with the complexity of NT there's all sorts of
issues when dealing with TDI and multiple transports. Clearly there is
some wierd wierd code in there somewhere. For instance...

else
{
	bLowercaseForTheFunOfIt = TRUE;
}


-- 
Matt Chapman
New College UNSW