Do not make the TLD cheap
RobertLowe opened this issue · 19 comments
Do not make the TLD cheap.
The price point should be high to prevent domain parking.
What if instead of raising the price, there was a requirement that in order to renew the domain the project has to be active in development?
@chrisabrams perhaps, but I feel like that will just get worked around somehow
@RobertLowe I concur there is no perfect solution - you can't really define what is "active" vs "inactive" as time is relative, unless you actually defined it in the terms of use. Kind of like how Twitter's TOS on usernames and 9 months of inactivity, or not tweeting. I'm not saying this is a good/bad idea - but was just thinking outside the box.
What about having a "setup" fee, and then cheaper renewals? Kickstarter would get you setup costs and 1 year free? Just a thought.
I have a "in stealth" open source project that i would like to unveil later on. I'm sure others do too. So while i really like proof of development as a prereq to avoid squatting, it needs to work with that use case.
@paulirish Oh jeez, tease us why don't you 😏
About the use case you mean having development as a prereq?
I'm not against it, however in my opinion $50-100 USD is cheap for a domain (not to imply that should be the final price). I'd speculate that $10-20 USD might be too low to keep it managed & in-check.
@paulirish That goes back to the idea I mentioned earlier on renewing the domain, not registering. That way you have a year to get something out and show active development before you can renew it.
I think having a $50+ registration price is prohibitive, and should be avoided. Not all developers are well-off, and many are in countries where $50 US is a lot of money. Plus, many people have multiple libraries, and paying $300/yr for a few domain names to host them on is ridiculous for an open source project.
If somebody makes an awesome open source JS library and wants to share it, they can host it on GitHub for free. They can then, if they choose, grab a .com/.org/.net domain name for ~$10. If we don't compete with that, we won't as many get people choosing .js for their open source JS libraries.
I think we should do a $5-10 registration fee, but have a review process so you can't register or renew a domain until you've gotten approval. Whether this approval is by the governance committee, or by some sort of community voting process can be decided.
Some international domains require proof that the organization that will be using the domain actually exists and is active - why not do the same with the domain? Then @paulirish can be stealth as he doesn't have to reveal his project, just that he is working on it, and people don't park. Also, this prevents domains from needing to be really expensive.
@Kerrick @chrisabrams Who will manage approvals and checking that they stay in order? And if there is an approval process why wouldn't I just get a .com for $10 instantly?
@RobertLowe I think those are fair questions.
I don't want to make assumptions for anyone else, but my viewpoint is that it would be nice, not necessary by any means, but nice to have a TLD that is know for active projects as there is a community of developers making sure that anything on .js is not parked/vacant.It would make it quicker to sift through a Google search knowing that a .js site is an active project. Just a nice thought.
As far as who would oversee it, I think we'd form a community from whom is involved in these current discussions. The hope would be that it is an objective decision - is it being worked on? Not decisions on if it is "good" or "bad" - just simply verify the author has an intention of using the domain for a work they are creating or have created.
I agree with the simple boolean test of "real" project or not. Here's an example of a repeatable way to test against squatters:
- Is there code? (To prevent squatters. It can be the humble beginnings of a project.)
- Is the code primarily written in or for JavaScript? (To prevent mis-categorized code.)
- Is the applicant a primary author or maintainer of the code? (To prevent the wrong person getting the domain.)
- Is the code licensed under an OSI-approved license? (Optional test, depending on the opinion of the governance committee. To encourage free software and code sharing.)
- Is there at least one demonstration of the code in use? (To prevent vaporware as a workaround for domain squatting.)
- If not, is there at least one mockup/writeup of how the code will perform in use? (To allow for projects in the beginning stages.)
- Is the domain being applied for relevant to the chosen name of the code? (To prevent domain squatting on the wrong names for otherwise real projects.)
If the TLD were .src then an OSI-approved license should be on the checklist, but if it were just .js, then maybe optional.
I'd say that delegation of domains should be an approval process, there aren't THAT many projects, and there are plenty of developers willing to give their time to make sure parking is not an issue. A good requirement could be that the source is on github and openly available. The github page MUST point to that domain with good activity. Etc.
Locking it into GitHub is a bad idea. What if people prefer Google Code, BitBucket, or their own hosted solution?
Strong point
@KorvinSzanto - I use BitBucket over Github
@Kerrick - I don't think enforcing the project to be Open Source is the right approach, paid libraries/tools should also be allowed to hold a domain
I think there should be a commercial use registration held at premium price for this purpose. Make that $50-75. Then a general reg. something around $25 for open source projects. And allow students to apply for a reduced price reg. around $15-20.
In my view all of the verification of applications in each category can be very easily automated, and could leverage basic machine learning to do so with precision. If and when someone feels they have been erroneously "evaluated", human intervention may be necessary.
I think the .js idea should really be just a cool TLD with the normal TLD policies.
As in, I could make a website about cooking and have a .js domain, or make a custom url shortner with it.
If you're going to restrict people, then as said before, people could just get an ordinary domain in a perhaps short amount of time without much effort except paying for it.