[ncc-services-wg] Discrepancy between WHOIS and delegated-ripencc

Hello!

This might not be the appropriate place for my question; if that is so, I
apologize in advance.

I've noticed that the data provided by the RIPE WHOIS database is 
sometimes inconsistent with the data shown in the delegated-ripencc-latest 
file, available from the ftp://ftp.ripe.net/ripe/stats/ server. For 
example, consider these two entries:

-------------------------------------------------------
$ whois 91.212.135.0/24

inetnum:        91.212.135.0 - 91.212.135.255
netname:        YABA-NET
descr:          YabaMedia Ltd
country:        RU

$ grep 91.212.135.0 delegated-ripencc-latest
ripencc|CZ|ipv4|91.212.135.0|256|20090409|assigned
-------------------------------------------------------
$ whois 193.105.207.0/24

inetnum:        193.105.207.0 - 193.105.207.255
netname:        ALFAHOSTNET
descr:          Alfa-host LLP.
country:        KZ

$ grep 193.105.207.0 delegated-ripencc-latest
ripencc|CZ|ipv4|193.105.207.0|256|20100315|assigned
-------------------------------------------------------

As you can see, the data differs in the "country" field -- in both cases, 
the range is attributed to Czech Republic, while the block is actually 
assigned to anothers.  Since the delegated-ripencc-latest file is used by 
some reputable and often-used sources (like Team Cymru; 
http://www.team-cymru.org/Services/ip-to-asn.html), this can sometime 
result in misleading and completely useless reports (as you can probably 
guess, it was Czech Republic getting unfair load of hostile sites 
attributed to it).

There are two questions now:
1) What is the source of this discrepancy?

2) If it was caused by some kind of an error, can the databases be put in
    sync? If it is intentional, which source is more "reliable"?

Thanks,

Peter

-- 
[Name] Peter Kosinar   [Quote] 2B | ~2B = exp(i*PI)   [ICQ] 134813278