7:42 a.m.
Picketlink is dead?
The last commit in the project repo was in 9 july..
Is there a schedule for the new version or something like that?
at.,
*Arthur P. Gregório*
*+55 45 9958-0302 <%2B55%2045%209958-0302>*
@gregorioarthur
www.arthurgregorio.eti.br
7:47 a.m.
It certainly appears that everything has moved to key-cloak but I am unsure
that keycloak is ready to take the burden of current Picketlink
implementations. Nor am I sure how the migration process would occur. The
abruptness of the change is a bit disconcerting. Having said that
Picketlink is working fine save for one defect that which I requested that
is on the git HEAD but not in any particular release.
On Mon, Nov 23, 2015 at 8:43 AM Arthur Gregório <arthurshakal(a)gmail.com>
wrote:
7:54 a.m.
And KC does not have the same purpose as the PL.
In short, I have no reason to migrate from one to the other, I use PL or go
back to Spring Security.
But it seems that there has not been any development in PL, at least in
recent months, in short, it seems that the project is dying and all that
were used for its own account.
And with bugs like this https://developer.jboss.org/thread/266387, it's not
cool to let the project stalled...
*Arthur P. Gregório*
*+55 45 9958-0302*
@gregorioarthur
www.arthurgregorio.eti.br
2015-11-23 11:47 GMT-02:00 Stephen Agneta <sagneta(a)gmail.com>:
9:04 a.m.
I'll share what I know with you in the hopes that it will help somehow.
Well KC (keycloak) is a super-set of the PL (PicketLink) functionality thus
in theory it ought to work fine once it is ready and once some sort of
migration path is known. You may not wish to move to KC due to the
additional functionality which may be off-putting for lite applications but
KC will perform everything PL did and more and will do so in VM memory if
you so choose.
Essentially KC is a real federated authentication and authorization service
with identity management that can run standalone or in-VM within a WildFly
cluster. Although a Java implementation it works with other systems and
languages out of process. It does integrate with Spring which may interest
you.
The following link provides information for Wildfly 9 clustered
installation:
http://keycloak.github.io/docs/userguide/keycloak-server/html/server-inst...
Thus you should be able to have your authorization demands met _in VM_ as
opposed to over-the-wire for performance reasons if necessary.
IMOP I think the KC project is the right move. They are fixing the big
issue which is the lack of an opensource Federated Identity Management
System. They also fixed little things such as Composite Roles which are
missing from PL.
I merely disliked the abrupt change-over. I also can't move to keycloak
until I have more of an idea how the migration would work.
For example, how different is the default KC relational schema from the
default basic PL schema:
http://keycloak.github.io/docs/userguide/keycloak-server/html/server-inst...
It is also not clear if keycloak has a CDI demand system ready like
PicketLink. They only hint at it. Also it runs in-cluster on Wildfly 9 and
I am on 8. Nothing huge but issues that will need to be addressed.
Hope that helps.
On Mon, Nov 23, 2015 at 8:54 AM Arthur Gregório <arthurshakal(a)gmail.com>
wrote:
9:07 a.m.
Please take a look at
http://picketlink.org/news/2015/03/10/PicketLink-and-Keycloak-project-merge/
I think this post answers your question.
On Mon, Nov 23, 2015 at 1:05 PM Stephen Agneta <sagneta(a)gmail.com> wrote:
10:19 a.m.
I see this post, and i know what KC do..
What I mean is that I do not need all the things that KC does, I want
simple with the something like PL.
I posted in a thread about it on the same topic "continuity of PL" on the
dev list of KC and the same answer was given.
PL is such a cool framework, I refuse to believe that only I use it or only
I noticed this deep sleep that the project came...
Finally, the fact is that PL is like Spring Security, a swatter convenient
and fast flies. KC is already like a cannon, large and meaningless to the
context of solving a simple problem like killing a single mosquito.
But if so, the business is to make a project fork and working on my own
version.
at.,
*Arthur P. Gregório*
*+55 45 9958-0302*
@gregorioarthur
www.arthurgregorio.eti.br
2015-11-23 13:07 GMT-02:00 Bruno Oliveira <bruno(a)abstractj.org>:
4:40 p.m.
This is a disappointing situation.
PL should have been continued and then consumed by KC.
I will not be pulling in KC in its entirely in order to do SAML SP
implementations - we will need to move to something else.
I suggest that a PL module be published from KC that has minimal
dependencies.
You can migrate the PL functionality to KC this way but not force all of
the new dependencies on consumers.
On Mon, Nov 23, 2015 at 11:19 AM, Arthur Gregório <arthurshakal(a)gmail.com>
wrote:
1:46 a.m.
Keycloak now supports SAML SP implementation, which doesn't require KC
server. It can talk to any other SAML Idp. The docs is here
http://keycloak.github.io/docs/userguide/saml-client-adapter/html/index.html
. For the future, we will mainly focus on improve/maintain the Keycloak
SAML SP rather than Picketlink.
Also there is no need to fork the Picketlink project to your own, you
can still propose and send PR to Picketlink . This will allow that more
people from the community can suffer from your work.
Marek
On 23/11/15 23:40, larry mccay wrote:
4:50 a.m.
I only use JPA/LDAP authentication, simple and not need to mess with XML
and have only a single Jar in my classpath without relying on my aplication
server or something.
In short, KC can be very good at it, but for those who already have
something done and solid in the PL, it is impossible to migrate.
Until someone says I can use the KC to do this [1] with the same effort
that I have with the PL, I will continue thinking that the framework should
receive attention again.
[1]
https://github.com/jboss-developer/jboss-picketlink-quickstarts/tree/mast...
*Arthur P. Gregório*
*+55 45 9958-0302*
@gregorioarthur
www.arthurgregorio.eti.br
2015-11-24 5:46 GMT-02:00 Marek Posolda <mposolda(a)redhat.com>:
4:29 a.m.
On 24/11/15 11:50, Arthur Gregório wrote:
With Keycloak, you also don't need to mess with
any XML. The idea is,
that there is minimal code needed on application side. All the work like
authentication, identity management, LDAP integration, social
integration etc is done on Keycloak server. Your application just needs
to know how to talk to Keycloak server, so you need to add keycloak.json
file with some "adapter configuration", which points how your
application can talk to Keycloak server (it uses OpenID Connect or SAML2
protocols for communication with server) . The amount of dependencies on
application side is also quite minimal.
The amount of effort on application side is even less than with
picketlink. The thing is that you need keycloak server. It can be
executed either in same JVM ( Wildfly/EAP instance) like your
application or on completely different server.
The best is to start with keycloak documentation and possibly
screencasts and try examples.
Marek
5:25 a.m.
The amount of effort on application side is even less than with
picketlink. *The
thing is that you need keycloak server*. It can be executed either in same
JVM ( Wildfly/EAP instance) like your application or on completely
different server.
The best is to start with keycloak documentation and possibly screencasts
and try examples.
This is the "problem". Think about it:
I develop an open source system that uses a secure system and in theory,
can run on any application server compatible with JEE7+
My "users" will download the war, put on WF them still have to make a
number of settings on the server to be able to type username and password,
login, make control access, create groups, give permission ... Anyway,
unnecessary for this scenario.
What I know is that KC is a more business world, where we have a more
robust infrastructure. PL can be used in that case or something heavier
integrated with KC.
Well, what I am saying since I wrote my first e-mail is that technology
does not replace the other, but can complement each other...
But okay. For now, I would just someone could generate versions of PL,
snapshots, see and apply PR, make snapshots and publish to the maven, this
would be a great help.
*Arthur P. Gregório*
*+55 45 9958-0302*
@gregorioarthur
www.arthurgregorio.eti.br
2015-11-25 8:29 GMT-02:00 Marek Posolda <mposolda(a)redhat.com>:
1:12 p.m.
Correct. Keycloak SAML SP adapter is completely independent of Keycloak
IDP and should be usable with any SAML identity provider. If not, we
are more than happy to quickly fix and release to make that happen.
Keycloak SP is actually derived from PL SAML SP adapter. Think of it as
PL SAML SP version 3.0.
To be honest, PL SAML SP code was becoming a maintenance problem. It
was a mess mainly because the PL team had to move so fast and quickly
and had little time to make things right. It exposed too many
implementation details and IDP and SP code were intermixed and there was
a lot of code copied between different adapter types (filter, tomcat,
jetty, jboss eap, wildfly) making it hard to maintain. The new adapter
shares most code in a common cross-platform module with only minimal
platform specific code. We're also sharing a bit of common code with
our OpenID Connect adapters.
On 11/24/2015 2:46 AM, Marek Posolda wrote:
--
Bill Burke
JBoss, a division of Red Hat
http://bill.burkecentral.com
3405
days inactive
3412
days old
11 comments
6 participants
participants (6)
-
Arthur Gregório -
Bill Burke -
Bruno Oliveira -
larry mccay -
Marek Posolda -
Stephen Agneta