That's pretty poor judgement to not even communicate this to affected users
Thanks for the clarification, but imagining it's pretty easy that this could just be a bug, e.g. they knew they wanted to prevent future usernames from having "clyde", and maybe even realized some existing clydes would be grandfathered in, but didn't realize it would prevent existing clydes from updating anything else on their profile.
Amazon: Alexa
Microsoft: Cortana
Apple: Siri
Discord: Clyde
Google: Google
The ideal would be for the voice assistant to have some completely unique word, ideally with sound combinations that are very rare in your native language. And if for some reason, that still doesn't work for you, have a way to change it. Or maybe just not have a default, and require users to pick their own keyword.
Siri defaults to male in some countries and languages.
https://www.thestar.com.my/tech/tech-news/2019/07/03/google-...
https://news.yahoo.com/parents-forced-change-name-6-18055441...
Tragic.
I disagree. The username can be clyde_discord/clyde_bot and the display name could just be "Clyde". Then the username Clyde wouldn't be taken, and the users would be able to still message the bot and not be confused.
Though I personally really dislike simple human names for bots because it sounds cheesy and dumb.
I'm obliged to link you to 'falsehoods programmers believe about names' now. I'm not happy about it, but those are the rules.
https://www.kalzumeus.com/2010/06/17/falsehoods-programmers-...
Also names are part of human culture, and Alexa for example has been in use for a very long time. There are valid reasons for names to die off in a single generation (eg adolf) but this is not a reason that our culture gains anything from imo.
I'm not happy about it, but those are the rules.
..."then change them", as someone might say. But that article seems to be mostly about the edge-cases. I don't think someone named Clyde is an edge-case, nor would someone named Bonnie, for that matter.
If your plan in choosing a name for your bot is to choose one that won't collide with a human, you are going to run into the fact that humans have a lot of very different names. Hence the link to the 'falsehoods' piece.
The US Census of 1950, for example, has a lot of Descords, Descards, Dicords, Dioscords, and... in PA, a Rosa B Discord, in NYC, a Mr Discord.
In general, all I'm saying is, it is far better if your intention is to allow people to choose their name, to not put any arbitrary restriction in place with the justification that 'it probably won't affect anyone'. If there's one thing people really care about it's their name. And no matter how unlikely you think it is that someone will run into your restriction... you're almost certainly wrong.
There was a Roman goddess named Discordia (where "discord" came from), and in the 90s Hercules and Xena TV series the Greek goddess used that Roman name instead of Eris, but simplified to just Discord.
Those two series were pretty popular, so I can see it being more likely recently than in that 1950s census sibling comment found. On the flipside, that character is probably not something parents would want for their daughter.
In this post-Brony society, I’m sure of it.
(Discord was the name of the villain on the show… I think. Never seen more than half an episode myself.)
Here's an easy lower bound, for the context of avoiding collisions: The name has to have been used by at least two different humans.
Or to get closer to a reasonable bound, 100 humans.
We don't even have a database for all the humans alive today, much less one for the last 400 000 years...
I wonder whether there are constraints that made it difficult to implement this functionality in another way, or this is just a bad design decision.
The problem isn't that the (fake) bot is in the same namespace as users; if there were any problem like that at all, it would be a shortage of discriminators.
However, that's not what this message is saying. It's saying that the name cannot contain "Clyde". In other words, you wouldn't be able to use any name that had the word "Clyde" in it, taken or not.
The fact that they haven't attempted to block usage of "CIyde" and similar does suggest they might not have thought it through, but it's verifiably not a namespace issue.
It's not always the wrong choice to mingle identifiers of different types, but I think often people err on the side of convenience (/ laziness) instead of thinking through all of the potential issues.
So it seems almost impossible that this is some technical limitation but rather Discord cracking down on a common deception tactic.
It reminded me of the "our database schema doesn't allow people with the name Jeffrey": https://twitter.com/yephph/status/1249246702126546944
I’ve got to imagine that some people named Alexa find it unfortunate and invasive that suddenly their name is used everywhere. At least “Siri” and “Hey Google” are less popular names for people.
I had heard that some airlines may have tried to discourage the practice but this is the first I heard about such measures.
Seems common for some travelers to book tickets ie. Tuba, Lastname. https://www.flyertalk.com/forum/united-mileage-plus-pre-merg...
Such lists are usually kept secret, but there are a few open source word lists that people can adopt for their services, for example, https://github.com/shouldbee/reserved-usernames.
Why do their messages have to appear the same as humans?
If bots should all operate in a distinctive space, give it that space.
Mind you architectural foresight is hard so maybe they'll get there.
Reserved words in SQL sometimes get blocked as well as things like too many singles quotes or an odd number of single quotes. I've seen the name "Walter" blocked on many forms over the years.
The worst part is that the sender just gets sent to a security or generic page with no idea why, often losing the message they created.
https://crzysdrs.net/dv/victim/223-its-a-damn-shame-what-the...
Poor Clint.
"Name on cake must not be Clint"
> The Scunthorpe problem is unintentional blocking [...] because [...] text contains a string (or substring) of letters that appear to have an [...] unacceptable meaning. Names, abbreviations, and technical terms are most often cited as being affected by the issue.