Talk:List of domain name registrars

Remove Webhosting.info References Due to >2.24 Million domain error?

Webhosting.info's figures are over 2.24 Million domains in error when compared to those of the more accurate Registrarstats and Domaintools sites. Should the references to Webhosting.info be removed from the article or flagged as inaccurate? Jmccormac (talk) 14:55, 19 November 2009 (UTC)

The difference between the Registrarstats.com and Webhosting.info counts for is the reason that I have placed a warning on the Webhosting.info figures over their inaccuracy.

Registrarstats.com TLD counts

Webhosting.info TLD counts

Jmccormac (talk) 03:37, 1 March 2010 (UTC)

Good question. I compiled the list of yearly/monthly/weekly/daily stats because there was disagreement over what sources were credible. If you can find positive confirmation that WebHosting.info is inaccurate I'd still recommend that it be left in, with an explanation of why it's inaccurate and a citation of the authoritative source of this info. After all, do we know for certain that it's inaccurate, and not a different metric? (For this reason, I'm changing the word "error" to "discrepancy".) Thirdbeach (talk) 23:08, 1 March 2010 (UTC)

Basically Webhosting.info's statistics are massively out of synch because it is basing its counts on hosters rather than on raw zone counts. The way it is grouping hosters/nameservers has not been updated for years and with new gTLDs and ccTLD nameservers, the numbers have drifted. If the nameservers/hoster does not exist in the Webhosting.info db, then it appears that the domains hosted on these hosters are not counted. The Registrarstats figures are probably raw zone counts and are more accurate. There is a slight difference between the number of domains in a zone and the numbers on valid nameservers as sometimes people will put the IP address down as their nameserver instead of the nameserver's name. Other problems are people using .local or mistyping the TLD of their nameservers. The Domaintools.com DailyChanges figures are also more accurate than the Webhosting.info figures. The same drift problem affected Pingdom's ipwalk.com operation before it finally gave up. The only truly authoritative source for registrar statistics is ICANN. Every thing else is an estimate. Jmccormac (talk) 23:59, 1 March 2010 (UTC)

So given that this article is about the market-share-ranked list of domain name registrars, and there are discrepancies among the published lists, my sense is that the most useful thing we can do for readers is to continue to list all the sites but clearly explain the differences, pros, and cons. You've got new info above (can you provide citations for it?), there's some info in section intro paragraph that should go down into bullets, and I found some about why ccTLDs aren't counted (http://www.icann.org/en/topics/new-gtlds/zfa-concept-paper-18feb10-en.pdf: “In contrast, ccTLDs do not, in general, provide Zone File access. Some of the European ccTLDs that were providing some level of access ended their provision of access circa 2003. The reason generally given was that the data was being abused.”). Let me give it a whirl. Thirdbeach (talk) 01:29, 2 March 2010 (UTC)

ICANN only provides stats on the TLDs for which it is responsible. The ccTLDs have agreements with ICANN but they generally only provide registration totals on their own registry websites rather than complete market breakdowns. The ccTLD markets tend to be far more competitive, especially where the country's market has gone ccTLD positive (more local ccTLD domains registered than gTLD domains). With countries like the UK or Germany, that can cause hosters or registrars to appear small in gTLD terms but when the number of ccTLD domains hosted is combined with the number of gTLDs hosted, the layout of the market changes dramatically. Many of the ccTLDs shut down access in 2003 due to those directory scams (disguised invoices) and domain slamming (where registrants would be bombarded with offers to switch hoster/registrar as their domain expiry date approached). The problem with citations is that I'd probably end up self-citing as there is only a few people doing domain name statistics on a cross TLD level. Jmccormac (talk) 03:04, 2 March 2010 (UTC)

Done. See what you think. I tried to put all info that was specific to one statistics site into that site's bullet, and also tried to make the construction of each bullet similar. It bugs me a little that the bullets are not in a time progression. Did you put them in "credibility rank order"? Heh, a new metric.

Also, I've stuck with Wikipedia's WP:V and WP:NOR policies. Can't cite you, but I did cite the information provided by each site about what information they use. Regards, Thirdbeach (talk) 03:11, 2 March 2010 (UTC)

It looks a lot better now. I put the list in a credibility/confidence order as it made more sense. :) The ZFA paper deals with how zonefile access will evolve as new gTLDs are added. Jmccormac (talk) 03:38, 2 March 2010 (UTC)

Makes sense to me. Thirdbeach (talk) 03:56, 2 March 2010 (UTC)

a.k.a.s and d.b.a.s

Agreeing with a huge percentage of posts of this talk page, I have added more d.b.a.s and a.k.a.s to the "top ten" list for the following reasons:

• Different ranking and reporting organizations use different names for the businesses. Follow the links in the article to see this in action.
• Rankings do not always use the name most commonly recognized by Wikipedia readers -- e.g. INWW.com is more commonly known as Melbourne IT, Schlund is more commonly known as 1&1.
• Gawd I don't want to go through the list of reasons again. Read the rest of this page.

Thirdbeach (talk) 18:21, 11 March 2008 (UTC)

I've left out the links to home pages but restored the a.k.a.s etc. because the different published rankings -- DomainTools, ICANN, WebHosting.info, and RegistrarSTATS -- use different company names. Also, because historically the NameTools list has been least controversial, I've listed the names it uses first. I'm not saying I agree -- NameTools is the least frequently updated -- but it does have the benefit of not being a registrar itself. And at least with links to all four rankings people can evaluate for themselves. Thirdbeach (talk) 20:28, 21 November 2008 (UTC)

Moved. Thirdbeach (talk) 20:57, 21 November 2008 (UTC)

INWW & Melbourne IT

Any proof to back this up? --File:Ottawa flag.png Spinboy 17:36, 5 September 2005 (UTC)

Seems you're right, I created a redirect for it. --File:Ottawa flag.png Spinboy 17:39, 5 September 2005 (UTC)

Thanks, Spinboy. GreenJoe, please type inww.com in a browser; you will be redirected to Melbourne IT -- hence my a.k.a., which I will again replace because it's correct and verifiable. If you wish to change again, please cite your sources. Thirdbeach (talk) 18:44, 11 December 2007 (UTC)

Oops, further information -- since the company is known as "Melbourne IT" rather than "INWW" in the industry, I am re-adding the Melbourne IT text, with links to documentation of the relationship between them. Thirdbeach (talk) 18:54, 11 December 2007 (UTC)

Isn't the link enough? When they click on it, it goes to Melbourne IT. GJ (talk) 00:14, 12 December 2007 (UTC)

I certainly thought it was enough, but you stated you were "confused" so I added links specifically for you. Thirdbeach (talk) 00:43, 13 December 2007 (UTC)

I mean the INWW.com link that exists. Why do we need to blurt out on the same line that they are Melbourne IT? GJ (talk) 01:31, 13 December 2007 (UTC)

There are two reasons. 1) In the ISP/Web hosting/domain name industries, the organization is referred to as "Melbourne IT" far more routinely than "INWW". 2) In the three links to registrar stats that are currently in the article, Name Intelligence/domaintools calls the organization "INWW"; eNom/RegistrarSTATS calls it "Melbourne IT"; Directi/WebHosting.info calls it "Melbourne IT". My sense from a usability standpoint is that it helps when a writer uses words readers are familiar with ("Melbourne IT"), and that clarifying the relationship reduces confusion for those observant readers who actually go to all three links and notice the discrepancy. Thirdbeach (talk) 01:00, 14 December 2007 (UTC)

Ok, but it's already linked to "Melbourne IT", so why do we need the link beside it? GJ (talk) 02:33, 14 December 2007 (UTC)

Are you asking why the list item reads "INWW.com (a.k.a. Melbourne IT)" instead of reducing it to just "INWW.com"? Or are you asking why I placed the two links (DNS lookup and news item) next to "INWW.com (a.k.a. Melbourne IT)"? Not sure I'm understanding what you're suggesting as an alternative. Thirdbeach (talk) 01:03, 15 December 2007 (UTC)

I am asking why it reads "INWW.com (a.k.a. Melbourne IT)" GJ (talk) 01:20, 15 December 2007 (UTC)

What do you think it should read, and why? Thirdbeach (talk) 01:42, 15 December 2007 (UTC)

I don't see it needing the "aka" bit since it already links to Melbourne IT. GJ (talk) 01:44, 15 December 2007 (UTC)

I think that "INWW.com (a.k.a. Melbourne IT)" is the optimal content for several reasons:

1. Provide access: Wikipedia is all about easy access to information. If people can see the relationship between INWW.com and Melbourne IT in one place (rather than expecting them to read our minds and realize that we expect them to click the link), we're better serving Wikipedia's raison d'être.

2. Reduce confusion: There are currently three sets of registar stats linked in the article. Two of them list the organization as "Melbourne IT". The third uses "INWW.com". Readers who go to all three sites may notice the discrepancy; including the "a.k.a." settles the confusion in just twenty characters.

3. Increase retention: "Melbourne IT" is the name that's more familiar to most readers. My sense is that the list in the article uses "INWW.com" out of deference to what's labeled as "ICANN data" on domaintools; however, "INWW.com" is more familiar to Wikipedia's general readership as "Melbourne IT". Posting the "a.k.a" gives readers something familiar to work with, so that they can connect what they're reading to the information that's already in their brains (and therefore raise their chances of understanding and remembering it).

4. Respect history: Posting both names achieves familiarity and augments useful content while simultaneously respecting the history of this article and its contributors. Thirdbeach (talk) 02:48, 15 December 2007 (UTC)

Hi 208.100.248.38 -- please explain your undo. My changes were thoroughly explained and discussed above, sourced on the page, and expand the information in the article. Thirdbeach (talk) 23:52, 27 December 2007 (UTC)

Hi GreenJoe, please explain your removal of "Melbourne IT". How can removing contextual information that's more familiar to users be an improvement? Thirdbeach (talk) 01:29, 29 December 2007 (UTC)

Hi 208.100.249.195 -- Please explain your undo after considering the discussion above. In addition, I'm curious why the contextual information in "Wild West (part of Go Daddy)" does not seem to be controversial, but comparable information "INWW.com (a.k.a. Melbourne IT)" seems to be so troubling. Thirdbeach (talk) 00:36, 1 January 2008 (UTC)

On the links I provided

Perhaps there is an imbalance in the tekst/link ratio. However I think the links I provided are as relevant to the subject as the ones that were there already and are well suited to base new article material on. Gone 20:19, 8 September 2005 (UTC)

I think most of them are irrelevant, and the links section is quite a bit longer with them than the article. --File:Ottawa flag.png Spinboy 21:31, 8 September 2005 (UTC)

Then by your own account you have removed at least some relevant links that I put there... (A state of affairs that is definitely making me not want to contribute any more.) But let's (for argument's sake) just look at the relevance of the first set of links I provided (Lists of registrars):

1. The article itself talks about completeness of the list. How in the world can a link to e.g. the "InterNIC - Registrar List" (which should be complete) (except of course that the registrars that are combined in the predecessor of the ICANN (CORE) are not listed seperately) then be irrelevant?

(Also without it I would never recommend anyone to visit the item.)

2. As for the other links in this section: I recently found a registrar missing from the most commonly used list, but still present on another. My inquiries to Verisign and ICANN resulted in the information on this registrar being put back on the most commonly used list. (Seems to me this is quite relevant with regard to this registrar.)

3. Don't blame me if ICANN puts in multiple lists what could be put in a single one.

4. The last link (on alternative I.D.s): If you search the internet (usenet) for these I.D.s you will see that many people are asking what registrar is meant with such an I.D., which makes it relevant.

...................

Some additional remarks:

I am definitely NOT going to write an article on how to become a registrar.

I was considering some articles on how problem resolution with registrars works out in practice, however all I want to do now is unsubscribe my account at wikipedia. Can you tell me how?

Gone 05:06, 9 September 2005 (UTC)

1. The article is a list based on one source provided. I'm open to another list from another source, we can create sections for each of them. However, if you don't want to put the work into creating that list and it's section, and maintaining it, then please don't add a link for another list. This list is clearly marked as to its source.

2. See #1.

3. I don't blame you, I'm saying that we are working from one source.

4. I see no problem if you wish to re-add that link.

5. If you want to write an article on how to resolve problems with registrars, great. Just remember the NPOV policy. --File:Ottawa flag.png Spinboy 13:21, 9 September 2005 (UTC)

Contentions and Disputes

Factual cases of Contentions and Disputes may be reported here. --Connection 18:41, 12 March 2006 (UTC)

Massive revert and edit

Whoever decided to cut this article to replicate the top 10 list at registrarstats.com made a huge mistake. First of all, that information is copied straight from the website and may fall under copyright violation. And since the article offers no other information other than a top 10 list, its validity as a WP article is also questionable. If I were to nominate the article for deletion I'm guessing that it would probably get deleted. Looking back through the article's history, however, it seems that one editor (whose user page states that he was blocked by an admin and has left WP forever) was responsible for making this a top 10 page. The article used to contain a table of registrars, with pertinent details about their service offerings, which was maintained up until this time last year (see edits by Cburnett in April 2006). Since that table was the only encyclopedic portion of this article, I am going to do a huge revert back to that version and delete the top 10 list. The top 10 list will be linked as an external reference. If anyone objects to this and reverts it back to the top 10 list, please take fair warning that I will be putting this article up for deletion under copyvio and notability. VanishingUser 13:26, 14 April 2007 (UTC)

I'm not going to revert it yet, but I have to disagree with you. We cannot list every last ICANN accredited registrar out there. There are something like 800 of them, and that would be original research. At least with the top 10 list it was cited. GreenJoe 15:11, 14 April 2007 (UTC)

I'm not suggesting that it be a complete list, indeed I removed that terminology from the previous version of the article -- plus, a complete list already exists which is linked to at ICANN. Notability is the criteria for inclusion. The top 10 list was not just cited, it was copied. The problem is that this article contributed nothing above what was copied from its source, so you are basically mirroring data from a copyrighted webpage. Even if you don't like my alternative, the chances that this page would survive as its top 10 incarnation are very small. And what part of the new table do you believe falls under the criteria of OR? VanishingUser 16:16, 14 April 2007 (UTC)

How do you determine who gets included in that "table" and who doesn't? There's long list of ICANN accredited registrars, and that's assuming we're sticking just to those that are ICANN accredited. There are even more resellers. It will be too long if you include all 800 of the ICANN registrars. As for WP:OR, if you're not including all 800, then it's rather subjective, and can't be cited. The old list was cited, it said it came from Registrarstats as I recall. GreenJoe 05:27, 15 April 2007 (UTC)

I ended up reverting it. I couldn't stand to see how bad it was. GreenJoe 15:24, 14 April 2007 (UTC)

You do realize that eNom themselves are replacing the ICANN statistics with their own stats website. ICANN data is likely to be more neutral for such claims. —Preceding unsigned comment added by KillAllSpammers (talk • contribs) 03:22, 6 November 2007 (UTC)

Schlund + Partner

According to this page: 1&1 Facts & figures Schlund + Partner was aquired by 1&1, can anyone confirm/deny this and should the entry on domain name registrars reflect this? Thanks --Campbecf 03:36, 27 July 2007 (UTC)

Yes the above is correct and is reflected in the page. —Preceding unsigned comment added by 201.131.221.141 (talk) 21:01, 3 March 2009 (UTC)

Comparisons

It would be helpful to at least link to good summary comparisons of the top domain registrars, esp. if this article is not going to include comparison data.

A key thing to consider when choosing a registrar would be, How easy is it to transfer out, if and when you ever want/need to? In the case of godaddy, it seems to be "medium" hard to transfer out, just some tedious hassle and confusion and delay -- at least they make the authinfo transfer code readily available. Unless you have "issues" with them, in which case it may be very difficult/expensive to transfer out. In the case of moniker, they seem to have a very strict procedure, that you can not learn about without actually doing a transfer out. The authinfo transfer code does not seem to be available, until it is actually needed. Maybe it is good, and not a problem, that they have strict transfer-out control procedures? -69.87.203.98 (talk) 13:53, 26 January 2008 (UTC)

DomainTools, RegistrarStats, and Sockpuppetry by banned user KillAllSpammers

I've reverted the source of rankings from DomainTools and back to RegistrarStats for several reasons. First, the DomainTools list is over a year old and as a result its rankings are now incorrect. Second, the editor incorrectly removed Moniker, which is #8 on the DomainTools ranking and #9 on the more current RegistrarStats ranking. Third, this IP is a confirmed sockpuppet of banned user KillAllSpammers. Multiple admins have examined this user and have not only permanently banned the logins but also have placed extended semi-protection on eNom because of editor's IP-hopping. The editor's attempt to shift from RegistrarStats to DomainTools is not about accuracy, it's about this editor's abundantly demonstrated hostility to anything eNom (as stated in this article and on the RegistrarStats Web site, RegistrarStats.com is run by eNom). I've followed Aniani1's recommendation of using RegistrarStats because it's updated daily and is therefore the more widely used industry standard. Thirdbeach (talk) 21:42, 12 January 2009 (UTC) Tweaked Thirdbeach (talk) 21:59, 12 January 2009 (UTC)

Page rendering

I revised the tagging of names at the top of the list. Although technically correct, [[Melbourne IT|INWW.com]] renders as INWW.com, which is not used by the company, rather than Melbourne IT, which is. Also, the Wikipedia community appears to be more comfortable using RegistrarStats as the authoritative ranking, and RegistrarStats lists the company as "Melbourne IT". Thirdbeach (talk) 18:11, 7 July 2009 (UTC)

Error in Webhosting.info Statistics WRT Registrarstats and Domaintools

There seems to be a difference between the figures provided by Webhosting.info and the statistics provided by Registrarstats and Domaintools. The error is approximately 2 million domains. Registrarstats and Domaintools are in rough agreement so the error is in how Webhosting.info calculates its figures. Jmccormac (talk) 15:05, 13 August 2009 (UTC)