Hacker News Reader

Comment by thekaleb | Hacker News Reader

The username is malware. Github does not namespace user profiles, so the link to the account would be https://github.com/malware . There is no Github user with the name "about". If you go to https://github.com/about it is not a profile page, but a page describing github itself. I assume that github may want to make a page describing malware.

View on HN

It reminds me of the list of names that all SaaS companies should reserve project I did a while ago. Maybe I should add "malware" on it :)

https://zimbatm.github.io/hostnames-and-usernames-to-reserve...

What a great list! I’d recommend adding “secure” too.

Ah, great list!

The "CA ownership verification" ones are a bit excessive.. 'admin', 'administrator', 'webmaster', 'hostmaster', or 'postmaster' are the currently permitted localparts for domain validation. (From 3.2.2.4.4 of https://cabforum.org/wp-content/uploads/CA-Browser-Forum-BR-...). It was more 'CAs make up their own scheme' in the past though but thankfully that's been reigned in.

If my username is preventing them from creating a /malware page, I'd have no problem at all with that explanation; their lack of namespacing is convenient, but can cause these problems to occur.

But why hide behind the shroud of "privacy and security" if that is the case? I'd happily relinquish (not that I have much choice in the matter) the name if it was a namespace clash.

Perhaps because they aren't ready to announce what they're working on yet? Just one of a few examples I can think of why they don't owe you a reason.

It sucks that they're making you change your name and it's obviously inconvenient. They've at least offered some kind of compensation.

It looks like you lucked out that there's no activity or repositories linked to it that would break from the change.

This seems like a lot of fuss for nothing.

> Just one of several examples why they don't owe you a reason. Of course they don't owe me a reason; when does a corporation ever owe you a reason for most things it does? It doesn't stop us from asking for answers.

> These seems like a lot of fuss for nothing.

I wrote as much at the end of the post.

That makes sense but why cant they tell him that? It's innocent and easy enough to explain and for the user to accept.

Is it too much to ask the article's author to assume the bloody obvious?

> I just want to know why.

I'd wager more than a few people at Github are scratching their heads, wondering if the article's author's IQ is higher than a shoe size.

or you know, they could tell him why...

The chap is savvy enough to use Github. "Don't feed the troll" isn't just for online forums.

> Is it too much to ask the article's author to assume the bloody obvious?

Their intentions are not obvious to me, and I have asked both publicly and privately (individuals that I know within github).

> wondering if the article's author's IQ is higher than a shoe size.

Wow, really?

That crosses into personal attack, which is not ok here. Please read https://news.ycombinator.com/newsguidelines.html and don't do this again.

This is almost certainly the answer. The author is overthinking this.

> This is almost certainly the answer.

I agree. I just wanted to know! But, as the replies on this post have shown, the only thing you get for asking questions that you don't have the answer to is vitriol and ridicule on HN.

I write technical posts that require lots of experimentation, time and research, and no one gives a shit. I write a rant-y blog post on Friday night about something trivial that I found frustrating and suddenly I'm a moron with an IQ the size of my shoe that is a waste of human life.

Hacker News is a fucked up place.

Hacker News is a fucked up place.

Used to be nice around these parts. Unfortunately, it is rapidly going the way of Reddit. I have yet to find a community that allows downvoting to remain a nice place. Sooner or later, the quality of conversation goes downhill.

Seems like a stupid move from github not to prefix usernames from the beginning or at least start doing it later like linkedin did.

It is very possible that one day there will be another URL they need which belongs to a large account that can not be renamed just like that.

Seems like a stupid move? This is a pretty common pattern and we rarely ever hear of stories like this.

They can always prefix their page urls instead, especially if it's not something too important.

I don't think I'll follow Linkedin lead on software design.

I do like the medium.com approach of the usernames as /@username similar to the apache user dirs /~username

Any sane company would use something like

domain.com/user/foo or user.domain.com/foo or even a different domain for user accounts. Github put themselves into that position.

> Github put themselves into that position.

Does it matter now?

Obviously, what if tomorrow they need github.com/foobarbaz as well? github.com/foobarqux ? github.com/foobarbiz ?

This is my bet, they are trying to namespace their site. They may not have EULA rule they can cite, so hand-wave security and confidentiality.

They don't need a specific EULA for this. They can terminate your account at any time, for any reason [0]. The downside of them giving easy, memorable /paths is that you're subject to this sort of thing on occasion.

Yelp has something even more uncomfortable to maintain - custom subdomains [1]

[0] https://help.github.com/articles/github-terms-of-service/#m-...

[1] https://food.yelp.com/

And they would likely do so if he continues to refuse. I'm not saying they're doing anything wrong other than not give a reason.

Oh, not a bad hypothesis! I hadn't thought about that.

Seems like 100% of Githubs problem. Solution: either use /github/about or /profile/name

Or just tell him and be done with it.

Github seems pretty open about stuff, so they should be honest about their eff up.

Sounds like a plausible reason, but what about: https://github.com/privacy https://github.com/faq https://github.com/software

Reply on news.ycombinator.com