0.6.0 #293

aeneasr · 2016-10-09T19:46:42Z

bc breaks:

removed connection
mysql

andrewmcnamara · 2016-10-11T02:13:01Z

The DATABASE_URL environment variable doesn't seem to be picked up anymore when starting under this branch.

aeneasr · 2016-10-11T08:10:52Z

@andrewmcnamara this branch is wip :)

…ript compatibility - closes #277

* warden: make it clear that ladon.Request.Subject is not required or break bc and remove it - closes #270

…#285

* docs: improve introduction * docs: improve introduction

* connections: remove connections API - closes #265 * oauth2: token revocation endpoint - closes #233 * vendor: update to fosite 0.5.0

aeneasr · 2016-10-18T12:25:14Z

@janekolszak check out 1d20c6e

janekolszak · 2016-10-18T17:19:06Z

client/manager_sql.go

-	ID      string `db:"id"`
-	Version int    `db:"version"`
-	Client  []byte `db:"client"`
+	ID                string `db:"id"`


Now it looks much simpler and better.

Plain SQL will be a great advantage. It's fast and predictable.

janekolszak · 2016-10-18T17:19:19Z

client/manager_sql.go

+	ResponseTypes     string `db:"response_types"`
+	Scope             string `db:"scope"`
+	Owner             string `db:"owner"`
+	PolicyURI         string `db:"policy_uri"`


I'd divide this table into two - one with OIDC specific data and one with the ClientInfo.
I'm not sure Hydra should handle ClientInfo part at all. It looks IdP specific and I know I will have to add some fields in my service and not use others. I might as well store my own version of ClientInfo

It's the default data that makes hydra compliant with https://tools.ietf.org/html/rfc7592 - you can still use your own metadata storage

…oses #296

coveralls · 2016-10-24T11:08:32Z

Coverage increased (+10.9%) to 68.758% when pulling c51397f on prepare-0.6.0 into f5299a1 on master.

waynerobinson · 2016-10-24T21:02:47Z

Our use case is preventing issues with backwards compatibility based on what we've implemented.

The existing version uses aud, not client_id as they are practically the same thing.

And it's the way we identify a client credentials token from a standard auth one, because these fields match.

And what spec?

If you're going to use the JWT fields of iat, exp, etc. Mixing in the non-JWT client_id when aud represents the same thing is very strange.

aeneasr · 2016-10-24T21:44:35Z

https://tools.ietf.org/html/rfc7662#section-2.2

client_id
OPTIONAL. Client identifier for the OAuth 2.0 client that
requested this token.

aud
OPTIONAL. Service-specific string identifier or list of string
identifiers representing the intended audience for this token, as
defined in JWT [RFC7519].

We could include both, though. Btw BC breaks are part of the minor versioning and the reason why Hydra is not marked as stable (major version 1) yet. :)

waynerobinson · 2016-10-24T21:47:54Z

But you use the Client ID as the AUD in the OpenID token now right? So the behaviour is inconsistent. And you really need some way of marking an access token as a client credentials (subject-free) version.

aeneasr · 2016-10-24T21:50:24Z

And you really need some way of marking an access token as a client credentials (subject-free) version.

I don't think that the subject of a client credentials token should be empty. It should be the client id. In which case it's pretty easy to figure out by checking if client_id == subject.

But you use the Client ID as the AUD in the OpenID token now right?

Yes, this is why I offered two times to include the aud claim as well!

waynerobinson · 2016-10-24T21:51:50Z

Adding the aud claim would be great!

Sorry, didn't notice that above.

…295

waynerobinson · 2016-10-25T18:22:31Z

Our use case is preventing issues with backwards compatibility based on
what we've implemented.

The existing version uses aud, not client_id as they are practically the
same thing.

And it's the way we identify a client credentials token from a standard
auth one, because these fields match.

On Monday, 24 October 2016, Aeneas notifications@github.com wrote:

@andrewmcnamara https://github.com/andrewmcnamara per spec, client_id
is the right value. we could add the aud claim too though. What's your use
case?

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#293 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAG4poWcAlC_GF-gBSu_a6T4-Njawyilks5q3FJsgaJpZM4KSFxV
.

waynerobinson · 2016-10-25T18:23:30Z

And what spec?

If you're going to use the JWT fields of iat, exp, etc. Mixing in the
non-JWT client_id when aud represents the same thing is very strange.

On Monday, 24 October 2016, Wayne Robinson wayne.robinson@gmail.com wrote:

Our use case is preventing issues with backwards compatibility based on
what we've implemented.

The existing version uses aud, not client_id as they are practically the
same thing.

And it's the way we identify a client credentials token from a standard
auth one, because these fields match.

On Monday, 24 October 2016, Aeneas <notifications@github.com
javascript:_e(%7B%7D,'cvml','notifications@github.com');> wrote:

@andrewmcnamara https://github.com/andrewmcnamara per spec, client_id
is the right value. we could add the aud claim too though. What's your use
case?

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#293 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAG4poWcAlC_GF-gBSu_a6T4-Njawyilks5q3FJsgaJpZM4KSFxV
.

waynerobinson · 2016-10-25T18:23:32Z

Is the scope/s key also different?

On Monday, 24 October 2016, Andrew McNamara notifications@github.com
wrote:

@arekkas https://github.com/arekkas Token introspection no longer
returns the "aud" field which is part of the JWT spec. Instead it returns a
client_id field. Can the introspect response be change to include "aud"?

Current Hydra (version 5)

hydra token validate TBpyZhWoJGB7xvuyPzL2jZPuEazYnmdm7wYfG32k3qc.mr3OtSkj90d_lf584s4dktwlO7BB0oNnkq6FNI072zk
{
"sub": "f13d49c3-b61d-4fbc-b0a7-048a5f138115",
"scopes": [
"hydra"
],
"iss": "hydra.localhost",
"aud": "f13d49c3-b61d-4fbc-b0a7-048a5f138115",
"iat": "2016-10-24T14:23:57.897+11:00",
"exp": "2016-10-24T15:23:57.897+11:00",
"ext": null
}

New Hydra

hydra token validate "qLmMAnvRtXeGs3fvKYqtsv9XqKeTe3ORZ0aDmZ6PKN4.fMIFVa7LpFYXwZ61CI_WsWNEbrLlKBsnOIgBCI6IT5g"
{
"active": true,
"scope": "hydra",
"client_id": "828c4958-4359-49d6-83be-02e235841c64",
"sub": "828c4958-4359-49d6-83be-02e235841c64",
"exp": 1477280498,
"iat": 1477276899
}

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#293 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAG4psqUOWje53jMYJ7WncyIu0ux-DCRks5q3CmJgaJpZM4KSFxV
.

aeneasr mentioned this pull request Oct 16, 2016

docs: fix typo in consent.md #294

Closed

Aeneas Rekkas (arekkas) and others added 9 commits October 16, 2016 18:58

oauth2: scopes should be separated by %20 and not +, to ensure javasc…

bc9e614

…ript compatibility - closes #277

* oauth2/introspect: make endpoint rfc7662 compatible - closes #289

547558c

* warden: make it clear that ladon.Request.Subject is not required or break bc and remove it - closes #270

* travis: execute gox build only when new commit is a new tag - closes …

8daa8a2

…#285

gofmt

35c18d4

glide lock

cc03f46

docs: improve introduction (#267)

0fb8c41

* docs: improve introduction * docs: improve introduction

unstaged docs

95306ea

docs

31e24b6

core: (health) monitoring endpoint - closes #216

0c930c0

aeneasr force-pushed the prepare-0.6.0 branch from d72e407 to 0c930c0 Compare October 16, 2016 16:59

aeneasr self-assigned this Oct 16, 2016

aeneasr added bug Something is not working. feat New feature or request. documentation breaking change Changes behavior in a breaking manner. labels Oct 16, 2016

aeneasr added this to the 0.6.0 milestone Oct 16, 2016

aeneasr mentioned this pull request Oct 16, 2016

core: (health) monitoring endpoint #216

Closed

Aeneas Rekkas (arekkas) added 7 commits October 17, 2016 00:19

token revocation / introspection

fc6c8f2

* oauth2/introspect: make endpoint rfc7662 compatible - closes #289

cde615d

* connections: remove connections API - closes #265 * oauth2: token revocation endpoint - closes #233 * vendor: update to fosite 0.5.0

core: add sql support #292

719ad74

connections: remove connections API - closes #265

ffd5907

connections: remove connections API - closes #265

bc7f2b6

connections: remove connections API - closes #265

a56f15d

update client sql

1d20c6e

janekolszak reviewed Oct 18, 2016

View reviewed changes

Aeneas Rekkas (arekkas) added 8 commits October 24, 2016 10:40

coverage

b09baaf

coverage

60c3e10

coverage

ccdc6fe

coverage

af22a36

fix deadlock

573ab64

coverage

ccd95dc

coverage

c51397f

all: coverage report is missing covered lines of nested packages - cl…

97f2b5d

…oses #296

Aeneas Rekkas (arekkas) added 2 commits October 24, 2016 15:01

fix travis

1fbddce

fix travis

0deeb42

Aeneas Rekkas (arekkas) added 9 commits October 25, 2016 09:11

cmd: prettify the hydra token user output - closes #281

86f023a

tests

dbc295e

tests

b3904f5

tests

f53e8a8

tests

dd1ac96

tests

5e27f3c

patreon

ce57469

tests

b5cd3b4

travis: make it possible for travis-ci to build forked repos - closes #…

7a60a9a

…295

aeneasr merged commit 8256356 into master Oct 25, 2016

aeneasr deleted the prepare-0.6.0 branch October 25, 2016 10:01

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

0.6.0 #293

0.6.0 #293

aeneasr commented Oct 9, 2016 •

edited

Loading

andrewmcnamara commented Oct 11, 2016

aeneasr commented Oct 11, 2016

aeneasr commented Oct 18, 2016

janekolszak Oct 18, 2016 •

edited

Loading

janekolszak Oct 18, 2016

aeneasr Oct 18, 2016

coveralls commented Oct 24, 2016

waynerobinson commented Oct 24, 2016

aeneasr commented Oct 24, 2016

waynerobinson commented Oct 24, 2016

aeneasr commented Oct 24, 2016

waynerobinson commented Oct 24, 2016

waynerobinson commented Oct 25, 2016

waynerobinson commented Oct 25, 2016

waynerobinson commented Oct 25, 2016

0.6.0 #293

0.6.0 #293

Conversation

aeneasr commented Oct 9, 2016 • edited Loading

andrewmcnamara commented Oct 11, 2016

aeneasr commented Oct 11, 2016

aeneasr commented Oct 18, 2016

janekolszak Oct 18, 2016 • edited Loading

Choose a reason for hiding this comment

janekolszak Oct 18, 2016

Choose a reason for hiding this comment

aeneasr Oct 18, 2016

Choose a reason for hiding this comment

coveralls commented Oct 24, 2016

waynerobinson commented Oct 24, 2016

aeneasr commented Oct 24, 2016

waynerobinson commented Oct 24, 2016

aeneasr commented Oct 24, 2016

waynerobinson commented Oct 24, 2016

waynerobinson commented Oct 25, 2016

waynerobinson commented Oct 25, 2016

waynerobinson commented Oct 25, 2016

aeneasr commented Oct 9, 2016 •

edited

Loading

janekolszak Oct 18, 2016 •

edited

Loading