I have a couple of different apps hosted on netlify using google APIs that have broken now because netlify changed their hosting to point to xxx.netlify.app instead of xxx.netlify.com
When I attempt to add the new domain to my oauth client IDs under "Credentials" I get an infinitely hanging wheel of death when I click "save"
I thought that wheel of death is caused by the new URL not being registered for the app, so I went to the settings for my application and tried to add the new domain, but it is failing as well.
I go to OAuth Consent Screen -> Edit App which then has another list of "authorized domain."
In the list, I see my old domain name:
foo.netlify.com
I try to add the new name:
foo.netlify.app
However, that auto-corrects to just plain netlify.app
I try saving that and get the following inscrutable error
Error
The request failed because one of the field of the resource is
invalid.
Tracking Number: c4522674593257747
I don't see any way to read more about the error or to use the tracking number to find out more information.
Any troubleshooting pointers would be appreciated. I'd rather not start over from scratch, but at this point it looks like my best bet would be creating a whole new application and starting the whole credential/API Key process over for all of my apps that were hosted on netlify.
I got the same problem and posted it in the google group. I have a reply from google cloud support which states that only top private domain is allowed as stated here Setting Up OAuth2.0
As the error clearly denotes and explained in this document, you can use any Top Private Domain as an Authorized Domain. So it seems to be the intended behavior.
I suppose that means only the domain owner can get a credential now.
Related
I've been attempting to utilize the Yahoo API to connect to fantasy sports data. I don't have a web URL to connect to when I attempt to create an app, and everything I've read has mentioned that I should select "Installed Application" to avoid using a URL.
This option is not available to me.
Yahoo Developer Screenshot
On the official documentation this is what I see here
Yahoo Developer Documentation
Not sure if it makes a difference but I'm based out of Canada.
Thanks!
It's hard to find the documentation for it, but what you want is called oob (out of bounds, out of band). To get this to work, the steps are a little different (there is no such thing as installed application from what I'm aware of):
Add oob as one of the redirect uris
Make your normal request to the auth/login uri
This will give you a new page that provides you with a token. At this point, you need to manually use this token when you request the /auth/token uri
This last step just replaces the redirect to https://your-redirect.com/token and you will end up in the same position as before.
Now, you still shouldn't put your client_secret in your application if you plan on releasing it, but if you're using it internally or letting others provide their own secret then you should be good.
If you find the original documentation with the PHP source code, you can follow this flow.
I am trying to enroll a user on my Moodle site via using the Moodle API.
My Moodle instance is hosted on AWS and all relevant ports are open and listening. So, from the network perspective, I can commit that is all ok.
The steps I have already done based on Moodle Documentation:
I have enabled web services on Administration > Mobile app >Mobile settings
I have gone through the 10 steps on the overview of allowing an external system to control Moodle as explained in the documentation (shown also in this Youtube video)
For testing purposes, I am using Postman. Some requests are going through (e.g. getting the token for a certain user, getting the list of all courses, etc.)
Example:
But when I try to i.e. create a user or enroll a user in an existing course I am getting this error:
{
"exception": "webservice_access_exception",
"errorcode": "accessexception",
"message": "Access control exception"
}
The way I am trying to i.e. create the user is as follows:
In the body section I am sending the following data:
users[0][username]
users[0][email]
users[0][lastname]
users[0][firstname]
users[0][password]
Based on my research, most of the contributors suggested enabling web services, but as mentioned above I have enabled them but the problem persists.
Can someone help me solve the issue here or maybe suggest a way of debugging it?
Fortunately, I managed to solve the issue for both user creation and user enrollment.
Here is a great guide that helped me. In addition, you need to add some additional functions to the web service (roles wary based on what you want to do in Moodle) and also you need to alter the permissions of the new user (again depending on what you want to do)...
We have 2 domain in salesforce:
1-) https://gablesinsurancerecovery.my.salesforce.com
2-) https://gableinsurancerecovery.force.com
and we have 2 user:
developer#cloudspade.com
communitytest#cloudspade.com
developer#cloudspade.com mail work with success on 1. domain but not work on 2. domain.
Likewise, communitytest#cloudspade.com mail work with success on 2. domain but not work on 1.domain.
Its not a problem. Problem is:
2.domain is important to us. Because we have a react-native project by forcereact and we want work with community domain. ( 2. domain ). We setup with successfully this link: https://www.youtube.com/watch?v=9zxMUrayFZ8&t=1634s
We was work with perfectly with the video in the link for login.salesforce.com or 1. domain but doesn't work for 2. domain.
Error is:
error image
As seen in the photo we was see a page. But our expectation was that it would working we send redirect_uri. The redirect_uri we sent was to return "girApp://success" with an access_token and instance_url at the end. But we encounter a screen as you can see in the photo. Redirect is not working. Our goal is to access the access_token with redirection within the application after login and authorize.
Lots to unpack here.
***.my.salesforce.com is your main domain, for internal users. ***.force.com is for Customer/Partner Experience Cloud (formerly known as Communities, formerly known as Portal).
developer#cloudspade.com mail work with success on 1. domain but not
work on 2. domain
Out of the box Salesforce is perfectly fine with internal users logging in to community or even 1-click switching over from internal SF to community. Collaboration and all that. Your administrator probably marked only certain profiles / permission sets as community members, you'd need to check config. But it's possible to use the community login page, you guys just chose not to.
communitytest#cloudspade.com mail work with success on 2. domain but
not work on 1.domain
Yes. Community members must use community login page. They can't use generic login.salesforce.com, test.salesforce.com or your branded ***.my.salesforce.com
works perfectly with (...) login.salesforce.com or 1. domain but
doesn't work for 2. domain
That's because most of the time the community login url must be full. Just the domain might not be enough for login because you can have up to 100 communities under same domain. You probably saw the example when you were enabling communities (Setup -> Digital Experiences -> Settings)
Go to Setup -> Digital Experiences -> All sites and write down the url you'll see there. It'll probably be something like ***.force.com/myportal. That means that for API login you might have more luck with ***.force.com/myportal than ***.force.com.
Stop reading this answer now and go read Sitecore - How to get User ID if the user was logged in using external identity provider (Salesforce SSO). Play with that OpenId Heroku app, once you get this to work with community user in the browser - you'll know which url to put in your react app. React developer might "like" this link too: https://gablesinsurancerecovery.force.com/.well-known/openid-configuration
It's kind of written in this article's footer: https://help.salesforce.com/s/articleView?id=sf.remoteaccess_oauth_endpoints.htm&type=5
Instead of using login.salesforce.com, you can also use the My Domain,
Experience Cloud site, or test.salesforce.com (sandbox) domain in
these endpoints. For hostname, use the My Domain, Experience Cloud
site, or custom URL
I'm setting up an OAuth login flow.
https://accounts.google.com/o/oauth2/auth first told me the client was deleted - which is fine - I created a new one.
Now, accounts.google returns 400 redirect_uri_mismatch (which is true if you count a trailing '?' that doesn't exist in the new client settings.
Additionally, the error page displays the OLD client project name.
It may be a propagation issue, but if anyone else has seen this behaviour, I'd like to know.
PS the target user is a GFW account.
Turns out the user was still attached to the old/deleted client, although that client had been deleted. The old and new client shared a Google API project.
I have an app that already successfully uses google oauth, but now I am trying to setup a staging deployment. It is a rails app with devise and omniauth, but I think there might be a problem with how I configured Google.
In the google api admin panel (https://code.google.com/apis/console) I can see the existing app. I created another one with a different callback url (because it is staging). Using Postman (a fancy way to edit the url parameters) I can send a get request to google with the current production client_id and redirect_url and it works fine. When I copy and paste in the new clients (staging's) client_id and redirect_url I always get the error Error: invalid_client.
I'm sure where to start with trying to the figure out the problem, but I've tried a lot of different steps, renaming the urls, changing the client secret, or recreating the client in the admin panel. Any ideas? This error is rather cryptic.
We had a temporary issue with a small set of client ids. The issue should be resolved. If you are still having problems please follow up.