country attribute: "ZZ" for unspecified?
Dear DB-WG, might there be a consent for introducing "ZZ" as an allowed value for the "country"-attribute for "unspecified"? Reasoning: - ISO 3166-1 [1] specifies "ZZ" as "User-assigned code elements are codes at the disposal of users who need to add further names of countries, territories, or other geographical entities to their in-house application of ISO 3166-1, and the ISO 3166/MA will never use these codes in the updating process of the standard.". - Many current entries do not specify the correct/meaningful value for country. Examples includes: [2a] RIPE-NCC-K-ROOT: NL [2b] TELTEK-1: GB, but: country is really Europe and Middle East [2c] ORG-AFNC1-RIPE: EU, but: country is really somewhere in African Region [2d] IANA-BBLK-RESERVED1: EU, but RFC1918 network [2e] EU-ZZ-163: EU, but: Country is really world wide [2f] LUXLINK-003: DE, but: These IP's are really World-Wide - For automated statistics it would be nice to have a special designation for IPs which do not correspond to a specific location, so that they are easy to filter. - It could be used for (global) anycast services, too. - ARIN & AfriNIC legacy entries could get a "correct" country. - Documentation says: "There are no rules defined for this attribute.", but the attribute is mandatory? Shouldn't we change that? Either make the attribute optional or specify, what it is supposed to contain. - I am dreaming: Wouldn't it be <http://dict.leo.org/ende?lp=ende&p=5tY9AA&search=desirable>desirable to have a designation for "world", too? - We already have a least on strange countries "EU", which does not seem to be part of ISO 3166-1... I'd like to hear your opinion... Best regards, Christian [1] <http://en.wikipedia.org/wiki/ISO_3166-1> [2a] <http://www.db.ripe.net/whois?searchtext=193.0.14.129&object_type=inetnum> [2b] <http://www.db.ripe.net/whois?searchtext=193.9.152.0&object_type=inetnum> [2c] <http://www.db.ripe.net/whois?searchtext=193.189.128.0&object_type=inetnum> [2d] <http://www.db.ripe.net/whois?searchtext=172.16.0.0&object_type=inetnum> [2e] <http://www.db.ripe.net/whois?searchtext=163.0.0.0&object_type=inetnum> [2f] <http://www.db.ripe.net/whois?searchtext=91.142.59.0&object_type=inetnum>
Christian 'wiwi' Wittenhorst schrieb:
might there be a consent for introducing "ZZ" as an allowed value for the "country"-attribute for "unspecified"?
It might be easier and more usefull to declare the country attribute optional and leave it out if you cannot and do not want to assign a proper country location for an object. Cheers Hendrik -- Hendrik Voelker Server & Services Management EMEA Server Operations Verizon Deutschland GmbH Sebrathweg 20 D-44149 Dortmund GERMANY http://www.verizonbusiness.com/de/ tel +49-231-972-1565 fax +49-231-972-2587 PubKey 1024D/92479A5D EC1B 76F2 D69D 11C6 2611 5CD1 5269 9351 9247 9A5D
Well, my personal preference would be to: a) make the country attribute optional and discourage its use at the same time, and b) allow something like ZZ (in case it isn't already) if an organisation wants to keep this attriute but doesn't want to, or cannot (reasonably), indicate the country "relevance". For any reasonable definition of country relevance, that is ;-) Wilfried. Christian 'wiwi' Wittenhorst wrote:
Dear DB-WG,
might there be a consent for introducing "ZZ" as an allowed value for the "country"-attribute for "unspecified"?
Reasoning:
- ISO 3166-1 [1] specifies "ZZ" as "User-assigned code elements are codes at the disposal of users who need to add further names of countries, territories, or other geographical entities to their in-house application of ISO 3166-1, and the ISO 3166/MA will never use these codes in the updating process of the standard.".
- Many current entries do not specify the correct/meaningful value for country. Examples includes: [2a] RIPE-NCC-K-ROOT: NL [2b] TELTEK-1: GB, but: country is really Europe and Middle East [2c] ORG-AFNC1-RIPE: EU, but: country is really somewhere in African Region [2d] IANA-BBLK-RESERVED1: EU, but RFC1918 network [2e] EU-ZZ-163: EU, but: Country is really world wide [2f] LUXLINK-003: DE, but: These IP's are really World-Wide
- For automated statistics it would be nice to have a special designation for IPs which do not correspond to a specific location, so that they are easy to filter.
- It could be used for (global) anycast services, too.
- ARIN & AfriNIC legacy entries could get a "correct" country.
- Documentation says: "There are no rules defined for this attribute.", but the attribute is mandatory? Shouldn't we change that? Either make the attribute optional or specify, what it is supposed to contain.
- I am dreaming: Wouldn't it be <http://dict.leo.org/ende?lp=ende&p=5tY9AA&search=desirable>desirable to have a designation for "world", too?
- We already have a least on strange countries "EU", which does not seem to be part of ISO 3166-1...
I'd like to hear your opinion...
Best regards,
Christian
[1] <http://en.wikipedia.org/wiki/ISO_3166-1> [2a] <http://www.db.ripe.net/whois?searchtext=193.0.14.129&object_type=inetnum> [2b] <http://www.db.ripe.net/whois?searchtext=193.9.152.0&object_type=inetnum> [2c] <http://www.db.ripe.net/whois?searchtext=193.189.128.0&object_type=inetnum> [2d] <http://www.db.ripe.net/whois?searchtext=172.16.0.0&object_type=inetnum> [2e] <http://www.db.ripe.net/whois?searchtext=163.0.0.0&object_type=inetnum> [2f] <http://www.db.ripe.net/whois?searchtext=91.142.59.0&object_type=inetnum>
Christian, On 2010-02-07 21:31, Christian 'wiwi' Wittenhorst wrote:
might there be a consent for introducing "ZZ" as an allowed value for the "country"-attribute for "unspecified"?
History ------- A proposal to do just this was made at RIPE 48 (Value "ZZ" for country: attribute): http://tinyurl.com/yauje3s You can see the minutes here: https://ripe59.ripe.net/ripe/wg/db/minutes/ripe-48.html There was consensus for this in the database working group shortly after this. It became an action for the address policy working group at RIPE 49: https://ripe59.ripe.net/ripe/wg/db/minutes/ripe-49.html You can look at the minutes from RIPE 50: http://www.ripe.net/ripe/wg/db/minutes/ripe-50.html Basically, the address policy working group said "non!" and that was it. This was 5 years ago, so maybe the opposition has changed its mind (or retired). An attempt was made later, which you can see in the minutes from RIPE 54: https://ripe59.ripe.net/ripe/wg/db/minutes/ripe-54.html It got dropped between RIPE 55 and RIPE 56, I'm not sure why. Opinion ------- My theory is that people *want* the "country:" attribute to be useful so badly they don't care that it does not actually work. And of course it is useful in most cases. But in some cases it is not. And in some cases it is misleading or otherwise harmful. I think making it optional makes complete sense. Or allowing ZZ. Or both. :) -- Shane
Shane Kerr wrote: [...]
An attempt was made later, which you can see in the minutes from RIPE 54:
https://ripe59.ripe.net/ripe/wg/db/minutes/ripe-54.html
It got dropped between RIPE 55 and RIPE 56, I'm not sure why.
One of the reasons why we didn't follow up and keep pushing (direction?), was input received from dependable sources in the hallways: that *keeping* the country: attribute - even if it is vague, misleading or plain wrong (in some cases) - overall it were still good enough for generating coarse-grained statistics.
Opinion -------
My theory is that people *want* the "country:" attribute to be useful so badly they don't care that it does not actually work.
And of course it is useful in most cases. But in some cases it is not. And in some cases it is misleading or otherwise harmful.
I think making it optional makes complete sense. Or allowing ZZ. Or both. :)
-- Shane
Wilfried.
History -------
A proposal to do just this was made at RIPE 48 (Value "ZZ" for country: attribute):
for the security conscious among us, could we please have a direct url. thanks randy
Randy, On 2010-02-08 13:12, Randy Bush wrote:
History -------
A proposal to do just this was made at RIPE 48 (Value "ZZ" for country: attribute):
for the security conscious among us, could we please have a direct url.
It wraps on my mail client, but you can edit it to use on your end: http://www.ripe.net/ripe/meetings/ripe-48/presentations/ripe48-db-country.pd... Alternatively you can go here: http://www.ripe.net/ripe/meetings/ripe-48/presentations/#db Then click on the "country" attribute presentation by Engin. -- Shane
hi shane,
A proposal to do just this was made at RIPE 48 (Value "ZZ" for country: attribute):
for the security conscious among us, could we please have a direct url.
It wraps on my mail client, but you can edit it to use on your end:
http://www.ripe.net/ripe/meetings/ripe-48/presentations/ripe48-db-country.pd...
looks great here, no need to edit to me, preso appears quite germane today randy
Note that this quote - ISO 3166-1 [1] specifies "ZZ" as "User-assigned code elements are codes at the disposal of users who need to add further names of countries, territories, or other geographical entities to their in-house application of ISO 3166-1, and the ISO 3166/MA will never use these codes in the updating process of the standard.". [1] <http://en.wikipedia.org/wiki/ISO_3166-1> is a bit short on the details. The standard ISO 3166-1:2006(E/F) actually says: 8.1.3 User-assigned code elements If users need code elements to represent country names not included in this part of ISO 3166, the series of letters AA, QM to QZ, XA to XZ, and ZZ, and the series AAA to AAZ, QMA to QZZ, XAA to XZZ, and ZZA to ZZZ respectively, and the series of numbers 900 to 999 are available. These users should inform the ISO 3166/MA of such use. Short of buying the standard documents, some more details about 3166 and user-assigned code elements can be found at: <http://www.iso.org/iso/customizing_iso_3166-1.htm>. jaap
participants (6)
-
Christian 'wiwi' Wittenhorst -
Hendrik T. Voelker -
Jaap Akkerhuis -
Randy Bush -
Shane Kerr -
Wilfried Woeber, UniVie/ACOnet