Talk:Dynamic DNS

From Wikipedia, the free encyclopedia

[edit] I intend to delete the list of external providers

Any objections? Josh Parris ✉ 23:19, 18 May 2005 (UTC)

Yes, unless you explain why. – Smyth\^talk 14:17, 19 May 2005 (UTC)

The list of providers is transient and non encyclopedic. Wikipedia:What Wikipedia is not#Wikipedia is not a mirror or a repository of links, images, or media files. Surely there is a list of Dynamic DNS providers already kicking around on the web somewhere. Josh Parris ✉ 00:33, 20 May 2005 (UTC)

Alright, I have replaced it with a link to Google Directory. – Smyth\^talk 17:29, 20 May 2005 (UTC)

[edit] http update

Scottprive seems to be describing the CGI code used to call nsupdate by providers like dyndns.org. I know of no DNS server that is listening on port 80. I think the claim of update via http should be reverted unless it can be backed up. Smallpond

Yes, I think that the original author attempted to say that many proprietary services that call theirselves "dynamic DNS" do not actually implement the RFC2136/2845 protocol, but use a proprietary protocol instead, that's often HTTP-based. The paragraph is really unclear and confusing about this. The DDNS RFCs don't even mention HTTP, so it's probably safe to rewrite it. -- intgr 15:54, 11 October 2006 (UTC)

Actually, I'm willing to bet that the providers use RFC-standard methods internally, they just put the webserver between the customer and the DNS updates for ease of use and security. Anyway, I changed the section to describe the HTTP protocol updates, see if it's clear. Smallpond 21:45, 12 October 2006 (UTC)

Comment: The assumption they just put the webserver between the customer and the DNS updates for ease of use and security is not correct. Security has nothing to do with the reasons for the proprietary (HTTP) dynamic DNS formats. There was not any standard in RFC-based dynamic DNS updates to allow for "user" and "account" centric responses ('update abuse', 'no host', 'expired account', 'call support' etc). Commercial and free dynamic DNS providers are all 'proprietary' in this respect. Scottprive 16:52, 20 November 2006 (UTC)

I removed the "(DDNS)" acronym since it's not used for referring to "dynamic DNS" in the general case, but rather the RFC-specified DNS protocol extension [as far as I know]. I think the next section about the standard DDNS (currently saying "'Dynamic DNS' is documented by RFC 2136 ...") should also be rewritten to be more clear that it's just one of all the protocols capable of updating DNS; also, it should probably be mentioned earlier in the article that there is a standard as well as proprietary protocols, just to make it clear since that's likely to confuse people otherwise. -- intgr 22:26, 12 October 2006 (UTC)

'dynamic DNS' is not a single method for updating a DNS server: it's also a service industry, a proprietary protocol, a method of working on limited or degraded Internet connections, and it has peripheral relationships to 'port blocking'. It's not just what it says in the current article. I'll post another talk on this. -- Scottprive 16:52, 20 November 2006 (UTC)

[edit] Artifically narrow definition for dynamic DNS

There are multiple definitions for 'dynamic dns'. For whatever reason, this article periodically expands to cover both definitions, and then contracts when someone deletes an edit that is 'proprietary'.

RFC based 'dynamic dns' implementations do not get deleted from this article, but 'other' definitions of dynamic DNS ("proprietary") do get deleted. The policing on this article actually harms the average person unfamiliar with the different technologies, and who has been told by someone to 'get dynamic DNS, so you can host a security camera online". People are getting confused or misdirected.

Excluding the concept of vendor supplied dynamic DNS will change usage. There is another context here, one which is used by almost every router on the planet and how people use it in networking forums. The reason folks are injecting things here (which eventually get deleted or edited) is because the article lacks full scope.

I am suggesting we cover the FULL scope of 'dynamic DNS' in a way that is not specific to any vendor or implementation. Where appropriate we can move expanded details off to new stub articles: 'Dynamic_DNS_proprietary' and 'Dynamic_DNS_RFCs'. Agreed?

Note: I am from a DDNS vendor, and if I lead on this edit spree I am susceptible to accusations of bias. Would someone please consider this argument and make a pass at this suggestion? -- Scottprive 18:27, 20 November 2006 (UTC)

I can't see why anyone would oppose splitting the article into three sections (introduction+generic meaning, DDNS defined by the RFC, and proprietary dynamic DNS) – it's screaming for sections anyway. I promise to back you up in any edit wars you might encounter. :)

An alternative would be splitting the article in two with a disambiguation page, but I don't think the concepts are different enough to warrant that -- intgr 19:05, 20 November 2006 (UTC)