I cannot reproduce it with the Limbo Beta... Can you retry it with another host?
nmap-2.54.36-1 is worth another try.

I am pretty much sure, that the whole thing is a compiler issue: I've
compiled nmap sources on both, Red Hat 7.3 and "limbo". The whole 
compilation takes a few seconds on RH 7.3 and essentially longer with
a universe of warnings on "limbo" which comes with gcc 3.1.x.
Finally, nmap compiled on RH 7.3 works with all versions of it. On the
other hand, no version of nmap compiled on "limbo" with gcc 3.1.x does
work properly.
In the latter case it finds open (but not filtered) ports on localhost
and a machine on a "private" network (e.g., 192.168.0.x) directly 
connected to the initial box. 
Although "ping" and "traceroute", "dig" etc. and other network services
work fine, nmap does not find any open ports on a machine which is 
somewhere on the internet.
Stracing the whole thing on "limbo" with "-o logbook" gives a lot of 
messages of type:

select(5, [4], NULL, NULL, {0, 20000})  = 0 (Timeout)
gettimeofday({1026139456, 651161}, NULL) = 0

in the logbook where the numbers in the latter line differ from attempt
to attempt.

(The 1539 ports scanned but not shown below are in state: closed)
Port       State       Service
22/tcp     open        ssh                     
25/tcp     open        smtp                    
53/tcp     open        domain                  
80/tcp     open        http                    
111/tcp    filtered    sunrpc                  
113/tcp    open        auth                    
135/tcp    filtered    loc-srv                 
139/tcp    filtered    netbios-ssn             
443/tcp    open        https                   
445/tcp    filtered    microsoft-ds            
465/tcp    open        smtps                   
651/tcp    open        unknown                 
993/tcp    open        imaps                   
995/tcp    open        pop3s                   
3306/tcp   open        mysql                   
10000/tcp  open        snet-sensor-mgmt        
27374/tcp  filtered    subseven                

Nmap run completed -- 1 IP address (1 host up) scanned in 50 seconds

Works fine for me.

Tested the whole thing today again (about five different install sessions on
different machines). It worked fine, what a surprise (!) although for my first 
3-5 install sessions - also on different machines - it didn't work.
Can't figure out what exactly the problem is since all install sessions went
through without errors ... it's like russian roulette ... maybe a dependency
problem of different packages.
'Give up ... sorry for bothering.

BTW, what is bug 67218 about ... I "do not have the permissions necessary to 
see bug 67218" ???