On Thu, Jun 28, 2012 at 5:36 AM, Nicolaas Matthijs
<[email protected]> wrote:
> I'm with Ray on this one. If we assume that the user.exists service
> just tells us whether or not a username is available, then doing a 204
> does seem to be pretty accurate and the existing red cross should be
> sufficient for that.

I agree that the red cross is sufficient. I don't like the idea of the
same http code meaning different things though.
>
> Given that the Sign Up page is almost never used in production, it
> doesn't seem like the best use of time to change the UI code to catch
> the other status code and display an additional warning.

We (rSmart) ue the register page in production.
The UX patch was a one-liner.
https://github.com/sakaiproject/3akai-ux/commit/b22d27912662bd8030b11a30d14aed9bf756af7b

>
> Hope that helps,
> Nicolaas
>
>
>
> On 27 Jun 2012, at 17:14, Ray Davis wrote:
>
>> Personally I'd prefer to see something like that specified and
>> prioritized as a new feature. But the specific bug you're talking
>> about
>> -- not being able to log into the "admin" account that currently
>> actually exists out of the box -- seems like a narrower problem, and a
>> 204 status does seem to describe the real situation. Anyway, I'll step
>> back now & let others express opinions.
>>
>> Best,
>> Ray
>>
>> On 6/27/12 9:00 AM, Erik Froese wrote:
>>> What about something like "Administrator"? or "ViceProvost"?
>>>
>>> 204 would suggest that an account exists by that username, which is
>>> not true.
>>> 404 would tell the UX to go ahead and submit that username for
>>> creation.
>>>
>>> If other services start using this endpoint for other reasons we
>>> could
>>> be sending mixed messages.
>>> 409 seems clearer. "An error occurred because the name you submitted
>>> conflicts with the internal state of the system"
>>>
>>> Erik
>>>
>>> On Wed, Jun 27, 2012 at 11:53 AM, Bert Pareyn
>>> <[email protected]> wrote:
>>>> Hey Eric,
>>>>
>>>> The UI needs to get an OK back from the server to know that the
>>>> username has
>>>> been taken.
>>>> If it returns an error code the username has not been taken and is
>>>> up for
>>>> grabs.
>>>> If you can extend user.exists to check for a list of usernames
>>>> that aren't
>>>> allowed, like admin, that wouldn't require any change in the UI.
>>>>
>>>> Something like
>>>>
>>>> - UI does request for username 'Admin'
>>>> - Server checks if username is blacklisted, if it is return an 'OK'
>>>> - If it isn't, checks if username is already taken, if it is
>>>> return an 'OK'
>>>>
>>>> - Bert
>>>>
>>>> On 27 Jun 2012, at 16:40, Erik Froese wrote:
>>>>
>>>> Would the 409 error code is enough to trigger the default error
>>>> behavior in the UX?
>>>> The same red X icon would probably suffice to let users know we
>>>> don't
>>>> like their choice.
>>>>
>>>> e
>>>>
>>>> On Wed, Jun 27, 2012 at 11:38 AM, Erik Froese
>>>> <[email protected]> wrote:
>>>>
>>>> Hey all,
>>>>
>>>>
>>>> We've recently found that creating a user with the name Admin causes
>>>>
>>>> some bad stuff to happen, namely, you can't log in through the UX as
>>>>
>>>> admin anymore.
>>>>
>>>> I think it might be time to add some logic in order to prevent bad
>>>>
>>>> usernames from being created in the system.
>>>>
>>>>
>>>> Could we add this to the user.exists.html endpoint?
>>>>
>>>>
>>>> /system/userManager/user.exists.html?userid=$username
>>>>
>>>>
>>>> HTTP responses:
>>>>
>>>> 204 (OK No Content): A User with that username (or case-variant)
>>>>
>>>> exists on the system.
>>>>
>>>> 404 (Not found): User name is valid and can be used to create a new
>>>>
>>>> user authorizable.
>>>>
>>>> ** 409 (Conflict): The user name is used or conflicts with internal
>>>>
>>>> configuration (aka evil names regex or list)
>>>>
>>>>
>>>> My hand it up to JIRA and implement this. It shouldn't take very
>>>> long
>>>>
>>>> to add a regex property to the LiteUserExistsServlet.java
>>>>
>>>>
>>>> Erik
>>>>
>>>> _______________________________________________
>>>>
>>>> oae-dev mailing list
>>>> [email protected]
>>>> http://collab.sakaiproject.org/mailman/listinfo/oae-dev
>>>>
>>>>
>>> _______________________________________________
>>> oae-dev mailing list
>>> [email protected]
>>> http://collab.sakaiproject.org/mailman/listinfo/oae-dev
>>>
>>
>>
>> _______________________________________________
>> oae-dev mailing list
>> [email protected]
>> http://collab.sakaiproject.org/mailman/listinfo/oae-dev
>
> _______________________________________________
> oae-dev mailing list
> [email protected]
> http://collab.sakaiproject.org/mailman/listinfo/oae-dev
_______________________________________________
oae-dev mailing list
[email protected]
http://collab.sakaiproject.org/mailman/listinfo/oae-dev

Reply via email to