resurrection of Brix

[daniel simko]

Hi everyone,

I wanted to let you know that I have decided to resurrect Brix CMS. I tried some other projects from WordPress to Liferay but when one start with Wicket nothing is comparable then. Brix is also amazing piece of software and is big shame that it didn't gain more popularity. I would like to discuss my plans:

1) Upgrade Brix to Wicket 8 (This seems to be done. All features in demo application seems to be working.)

2) Merge changes to the master branch (PR sent - https://github.com/brix-cms/brix-cms/pull/176)

3) Testing and bug fixing.

4) AdminPanel and demo app redesign using Bootstrap (I am not sure about this step).

5) Creating a new User&Role management plugin.

6) Creating a new plugin for easy content management.

Any thoughts are very welcome.

Thanks,

Dan

[Dan Simko]

[Tomas Baca]

I know Jackrabbit is a perfect solution, but to increase the popularity of the project would be interesting to do a JPA storrage

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to brix-cms-discuss@googlegroups.com.
Visit this group at https://groups.google.com/group/brix-cms-discuss.
For more options, visit https://groups.google.com/d/optout.

[Cserveny Tamas]

Hi,

With the advent of node.js and javascript developers bacame "full-stack", in order to brix to be successful it will need to be in-par with the productivity of hapi or express. These folks started server side prerendering of javascript laden pages and other funk.

I've tried hapi out as a java guy who thinkered a bit with javascript: I was able to create something useful within a week (just goofing around after work). Within this week I was able to create a basic cms with twitter bootstrap css, + facebook and google auth, as user management.

With brix, I spent more time just to understand what is it, and why it is good for me. I still think brix is/was a good idea, Wicket is the king for server side web applications in the java landscape.

In summary, in order to be successful, the following topics needs to be addressed:

- Documentation is next to nothing (at the last time I checked),
- Productivity is lower than node.js, do not underestimate this!

- Too few reusable components (again, something with productivity)

A reasonable target for brix could be developers with less client side knowledge, so having some templates makes sense.

Of course brix has its merits which should not be lost:

- it is kind of a lib instead of a framework, so it is easy to embed into an existing (wicket?) application.

Cheers,

Tamas

On 2017. Jan 9., Mon at 8:30, Tomas Baca <bt...@baca.sro.sk> wrote:

I know Jackrabbit is a perfect solution, but to increase the popularity of the project would be interesting to do a JPA storrage

On Sat, Jan 7, 2017 at 9:20 AM, Dan Simko <dan....@gmail.com> wrote:

Hi everyone,

I wanted to let you know that I have decided to resurrect Brix CMS. I tried some other projects from WordPress to Liferay but when one start with Wicket nothing is comparable then. Brix is also amazing piece of software and is big shame that it didn't gain more popularity. I would like to discuss my plans:

1) Upgrade Brix to Wicket 8 (This seems to be done. All features in demo application seems to be working.)

2) Merge changes to the master branch (PR sent - https://github.com/brix-cms/brix-cms/pull/176)

3) Testing and bug fixing.

4) AdminPanel and demo app redesign using Bootstrap (I am not sure about this step).

5) Creating a new User&Role management plugin.

6) Creating a new plugin for easy content management.

Any thoughts are very welcome.

Thanks,

Dan

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discu...@googlegroups.com.
To post to this group, send email to brix-cms...@googlegroups.com.

Visit this group at https://groups.google.com/group/brix-cms-discuss.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discu...@googlegroups.com.
To post to this group, send email to brix-cms...@googlegroups.com.

[Martin Grigorov]

Re-sending with my @gmail.com account.

On Mon, Jan 9, 2017 at 8:02 PM, Martin Grigorov <mgri...@apache.org> wrote:

Hi Dan,

Just merged your Pull Request. Great work!

Added Korbinian Bachl to CC. I know he still uses Brix and was interested to migrate it to Wicket 7.x recently.

Probably you may join forces and revive this project!

Martin Grigorov

Wicket Training and Consulting

https://twitter.com/mtgrigorov

--

You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to brix-cms-discuss@googlegroups.com.

[daniel simko]

Hi,

Thank you guys for your responses! My points are below.

2017-01-09 12:17 GMT+01:00 Cserveny Tamas <cserven...@gmail.com>:

Hi,

With the advent of node.js and javascript developers bacame "full-stack", in order to brix to be successful it will need to be in-par with the productivity of hapi or express. These folks started server side prerendering of javascript laden pages and other funk.

I've tried hapi out as a java guy who thinkered a bit with javascript: I was able to create something useful within a week (just goofing around after work). Within this week I was able to create a basic cms with twitter bootstrap css, + facebook and google auth, as user management.

With brix, I spent more time just to understand what is it, and why it is good for me. I still think brix is/was a good idea, Wicket is the king for server side web applications in the java landscape.

In summary, in order to be successful, the following topics needs to be addressed:

- Documentation is next to nothing (at the last time I checked),
- Productivity is lower than node.js, do not underestimate this!

- Too few reusable components (again, something with productivity)

A reasonable target for brix could be developers with less client side knowledge, so having some templates makes sense.

Of course brix has its merits which should not be lost:

- it is kind of a lib instead of a framework, so it is easy to embed into an existing (wicket?) application.

 

Yes. I would like to improve productivity by implementing some basic plugins for common CMS functionality. And if one need some special feature he can use wicket to implement it. Wicket is IMHO the most productive tool for creating web applications in the Java world. I have also some experience with javascript SPA & REST and I was enthusiastic at the beginning but at the end of the day I returned to the Wicket. I realised that I am more productive and I have much more fun with it. I have also invested so much time to the jvm ecosystem that I'm not willing to switch to Node.js. I can imagine that e.g. hapi can be super productive but the other side of the coin could be maintainability..?
 

Cheers,

Tamas

On 2017. Jan 9., Mon at 8:30, Tomas Baca <bt...@baca.sro.sk> wrote:

I know Jackrabbit is a perfect solution, but to increase the popularity of the project would be interesting to do a JPA storrage

Yes I think that jackrabbit (or JCR generally) is great for content like text and pictures. But another stuff like users and roles belong to database.
 

On Sat, Jan 7, 2017 at 9:20 AM, Dan Simko <dan....@gmail.com> wrote:

Hi everyone,

I wanted to let you know that I have decided to resurrect Brix CMS. I tried some other projects from WordPress to Liferay but when one start with Wicket nothing is comparable then. Brix is also amazing piece of software and is big shame that it didn't gain more popularity. I would like to discuss my plans:

1) Upgrade Brix to Wicket 8 (This seems to be done. All features in demo application seems to be working.)

2) Merge changes to the master branch (PR sent - https://github.com/brix-cms/brix-cms/pull/176)

3) Testing and bug fixing.

4) AdminPanel and demo app redesign using Bootstrap (I am not sure about this step).

5) Creating a new User&Role management plugin.

6) Creating a new plugin for easy content management.

Any thoughts are very welcome.

Thanks,

Dan

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to brix-cms-discuss@googlegroups.com.

Visit this group at https://groups.google.com/group/brix-cms-discuss.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to brix-cms-discuss@googlegroups.com.

Visit this group at https://groups.google.com/group/brix-cms-discuss.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to brix-cms-discuss@googlegroups.com.

[Korbinian]

Hello Dan,

its nice to see that you want to resurrect brix. As one of the last guys working on it and still using it today productive I really like to join forces. I wonder for what usage you need brix and how dedicated you might be to it, as my last experience in this regards was not that well. Thats also the reason I didnt really publish anything to github for a long time.

Regarding your bullet points:

1. perfectly fine and necessary
2. done by marting
3. brix IMHO needs a lot of automated testing added - so that changes cant easily break it. To understand that one needs to remember that brix is "wicket on the fly", meaning you can create a wicket app by just using brix cms and adding some tiles to it. While the JS may work, no one can be sure that the JCR-fed part wont break on execution.
4. Basic idea sounds good, I would go for foundation css and not boostrap as foundation seems closer to the pluggable approach one needs; You need to remember that the backend can get extended by custom tiles as well (management tiles) and that changes here might break backends that are already existing;
5. I ended up using apache shiro for that, but an easy embedded one might be interesting; However, if you store any data of that kind in the JCR you need to be careful as it might break security quite fast....
6. Not sure what you mean here - brix CMS was meant to be a "developers" CMS, so if you expect an enduser to make WYSIWYG stuff then a content plugin would be the way to go :)

Looking forward to hear from you and to joing forces :)

Best,

Korbinian

PS: using brix on wicket 1.4 with CDI and custom tiles + custom fixes on ModeShape 4.x as JCR backend, see it live under: https://www.whiskyworld.de

[Korbinian]

Am Montag, 9. Januar 2017 20:43:42 UTC+1 schrieb Dan Simko:

Hi,

Thank you guys for your responses! My points are below.

2017-01-09 12:17 GMT+01:00 Cserveny Tamas <cserven...@gmail.com>:

Hi,

With the advent of node.js and javascript developers bacame "full-stack", in order to brix to be successful it will need to be in-par with the productivity of hapi or express. These folks started server side prerendering of javascript laden pages and other funk.

I've tried hapi out as a java guy who thinkered a bit with javascript: I was able to create something useful within a week (just goofing around after work). Within this week I was able to create a basic cms with twitter bootstrap css, + facebook and google auth, as user management.

With brix, I spent more time just to understand what is it, and why it is good for me. I still think brix is/was a good idea, Wicket is the king for server side web applications in the java landscape.

In summary, in order to be successful, the following topics needs to be addressed:

- Documentation is next to nothing (at the last time I checked),
- Productivity is lower than node.js, do not underestimate this!

- Too few reusable components (again, something with productivity)

A reasonable target for brix could be developers with less client side knowledge, so having some templates makes sense.

Of course brix has its merits which should not be lost:

- it is kind of a lib instead of a framework, so it is easy to embed into an existing (wicket?) application.

 

Yes. I would like to improve productivity by implementing some basic plugins for common CMS functionality. And if one need some special feature he can use wicket to implement it. Wicket is IMHO the most productive tool for creating web applications in the Java world. I have also some experience with javascript SPA & REST and I was enthusiastic at the beginning but at the end of the day I returned to the Wicket. I realised that I am more productive and I have much more fun with it. I have also invested so much time to the jvm ecosystem that I'm not willing to switch to Node.js. I can imagine that e.g. hapi can be super productive but the other side of the coin could be maintainability..?
 

Cheers,

Tamas

On 2017. Jan 9., Mon at 8:30, Tomas Baca <bt...@baca.sro.sk> wrote:

I know Jackrabbit is a perfect solution, but to increase the popularity of the project would be interesting to do a JPA storrage

Yes I think that jackrabbit (or JCR generally) is great for content like text and pictures. But another stuff like users and roles belong to database.

I hope its ok to go on here: JCR is basically nothing but a object oriented database - but can also be used as a connector; Dont underestimate it!
 

 

On Sat, Jan 7, 2017 at 9:20 AM, Dan Simko <dan....@gmail.com> wrote:

Hi everyone,

I wanted to let you know that I have decided to resurrect Brix CMS. I tried some other projects from WordPress to Liferay but when one start with Wicket nothing is comparable then. Brix is also amazing piece of software and is big shame that it didn't gain more popularity. I would like to discuss my plans:

1) Upgrade Brix to Wicket 8 (This seems to be done. All features in demo application seems to be working.)

2) Merge changes to the master branch (PR sent - https://github.com/brix-cms/brix-cms/pull/176)

3) Testing and bug fixing.

4) AdminPanel and demo app redesign using Bootstrap (I am not sure about this step).

5) Creating a new User&Role management plugin.

6) Creating a new plugin for easy content management.

Any thoughts are very welcome.

Thanks,

Dan

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discu...@googlegroups.com.
To post to this group, send email to brix-cms...@googlegroups.com.

Visit this group at https://groups.google.com/group/brix-cms-discuss.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discu...@googlegroups.com.
To post to this group, send email to brix-cms...@googlegroups.com.

Visit this group at https://groups.google.com/group/brix-cms-discuss.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discu...@googlegroups.com.
To post to this group, send email to brix-cms...@googlegroups.com.

[daniel simko]

Hello Korbinian,

there are many reasons for which I need Brix:

- It happens to me quite often that someone is asking me for creating some website. I used to use Wordpress for that purpose because users can easily modify and create content. One of them is e.g. http://www.petr-mazal.cz. But when they want something special I am not able/willing to do it by PHP/Wordpress.

- I don't want always start from scratch when I am creating a new project (e.g. https://bucksfortime.com) and Brix has excellent pluggable architecture which allows great reusability.

- I also have some sites running on Brix. I have one instance of Brix which serves e.g.:
  http://sportave.com (there is cca 300GB in Jacrabbit datastore and cca 2GB in MySQL)
  http://www.cyklo-ski-policka.cz
  http://www.wickeria.com
  http://www.casovabanka.cz
  http://www.kadernictviolina.cz

- At last but not least I really like this project and enjoy working on it.

I agree with you that automated tests are necessary and I am going to write some later. I am not sure if we could use WicketTester or rather something like http://arquillian.org or Selenium. Probably some combination of them.

I prefer Bootstrap because I have very good experience with it and I also think that it is more common than Foundation. And I am afraid that it is not possible to go forward without any backward compatibility breaks.

I was looking at shiro few years ago and it seemed to me fairly complicated but I'll definitely look at it again. I was thinking about just using standard "wicket-auth-roles” combined with “org.brixcms.auth.AuthorizationStrategy”. So no extra dependency were needed for brix-core. User management plugin could use only "org.springframework.security.core.userdetails.UserDetails" and new “Extended” demo application will contain all implementation. I am also thinking about using Spring Boot for this new demo. I have also used CDI but nowadays Spring Boot is maybe better choice. It has integration for everything and good support for testing.

Yes I mean CMS for end users.

Looking forward to cooperate with you!

Best,

Dan

P.S. I was looking at https://www.whiskyworld.de and It is really nice site!

--

You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to brix-cms-discuss@googlegroups.com.

[daniel simko]

2017-01-10 12:49 GMT+01:00 Korbinian <korbini...@gmail.com>:

Am Montag, 9. Januar 2017 20:43:42 UTC+1 schrieb Dan Simko:

Hi,

Thank you guys for your responses! My points are below.

2017-01-09 12:17 GMT+01:00 Cserveny Tamas <cserven...@gmail.com>:

Hi,

With the advent of node.js and javascript developers bacame "full-stack", in order to brix to be successful it will need to be in-par with the productivity of hapi or express. These folks started server side prerendering of javascript laden pages and other funk.

I've tried hapi out as a java guy who thinkered a bit with javascript: I was able to create something useful within a week (just goofing around after work). Within this week I was able to create a basic cms with twitter bootstrap css, + facebook and google auth, as user management.

With brix, I spent more time just to understand what is it, and why it is good for me. I still think brix is/was a good idea, Wicket is the king for server side web applications in the java landscape.

In summary, in order to be successful, the following topics needs to be addressed:

- Documentation is next to nothing (at the last time I checked),
- Productivity is lower than node.js, do not underestimate this!

- Too few reusable components (again, something with productivity)

A reasonable target for brix could be developers with less client side knowledge, so having some templates makes sense.

Of course brix has its merits which should not be lost:

- it is kind of a lib instead of a framework, so it is easy to embed into an existing (wicket?) application.

 

Yes. I would like to improve productivity by implementing some basic plugins for common CMS functionality. And if one need some special feature he can use wicket to implement it. Wicket is IMHO the most productive tool for creating web applications in the Java world. I have also some experience with javascript SPA & REST and I was enthusiastic at the beginning but at the end of the day I returned to the Wicket. I realised that I am more productive and I have much more fun with it. I have also invested so much time to the jvm ecosystem that I'm not willing to switch to Node.js. I can imagine that e.g. hapi can be super productive but the other side of the coin could be maintainability..?
 

Cheers,

Tamas

On 2017. Jan 9., Mon at 8:30, Tomas Baca <bt...@baca.sro.sk> wrote:

I know Jackrabbit is a perfect solution, but to increase the popularity of the project would be interesting to do a JPA storrage

Yes I think that jackrabbit (or JCR generally) is great for content like text and pictures. But another stuff like users and roles belong to database.

I hope its ok to go on here: JCR is basically nothing but a object oriented database - but can also be used as a connector; Dont underestimate it!

Yes and when you use proper TM (e.g. http://narayana.io) you can enlist more resources (e.g. DB and JCR) to one XA transaction.
 

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to brix-cms-discuss@googlegroups.com.

[Korbinian]

Hello Dan,

there are many reasons for which I need Brix:

- It happens to me quite often that someone is asking me for creating some website. I used to use Wordpress for that purpose because users can easily modify and create content. One of them is e.g. http://www.petr-mazal.cz. But when they want something special I am not able/willing to do it by PHP/Wordpress.

- I don't want always start from scratch when I am creating a new project (e.g. https://bucksfortime.com) and Brix has excellent pluggable architecture which allows great reusability.

*thats* exactly the reason I use it - I dont know any kind of stable, fast and flexible CMS that can be easily extended in a programmatic way while still using the full JEE power
 

 

- I also have some sites running on Brix. I have one instance of Brix which serves e.g.:
  http://sportave.com (there is cca 300GB in Jacrabbit datastore and cca 2GB in MySQL)
  http://www.cyklo-ski-policka.cz
  http://www.wickeria.com
  http://www.casovabanka.cz
  http://www.kadernictviolina.cz

Interesting - when I evaluated Jackrabbit back in 2012 it had the bad habbit to break itself during the bootstrap when coming under immediate load.... thats why I moved to ModeShape; My content however is more small with not much more than 20MB in the JCR itself;
 

- At last but not least I really like this project and enjoy working on it.

I agree with you that automated tests are necessary and I am going to write some later. I am not sure if we could use WicketTester or rather something like http://arquillian.org or Selenium. Probably some combination of them.

I think we need a JCR test-workspace where all kind of relations and link tests are set so that a "test"-worker can check them one by one; WicketTester may work, but since we need to bootstrap it all and have a real http request scenario something like selenium might be better suited;
 

I prefer Bootstrap because I have very good experience with it and I also think that it is more common than Foundation. And I am afraid that it is not possible to go forward without any backward compatibility breaks.

Maybe I didnt express myself right: I think we can chose now anything, but we mustnt change this anytime soon later on as this will break stuff - so even a full release would be a bad idea to have a change there as it ends up with work in areas where no real benefit can be seen;

 

I was looking at shiro few years ago and it seemed to me fairly complicated but I'll definitely look at it again. I was thinking about just using standard "wicket-auth-roles” combined with “org.brixcms.auth.AuthorizationStrategy”. So no extra dependency were needed for brix-core. User management plugin could use only "org.springframework.security.core.userdetails.UserDetails" and new “Extended” demo application will contain all implementation. I am also thinking about using Spring Boot for this new demo. I have also used CDI but nowadays Spring Boot is maybe better choice. It has integration for everything and good support for testing.

For a simple Demo the wicket-auth-roles is fine - I thought you wanted to have something for real world usage out of the box;

 

Yes I mean CMS for end users.

Looking forward to cooperate with you!

Me, too. I've noted your pages are on czech and polish - may I ask where your from? I'm located in bavaria, near the czech border so maybe we can meet in person. Feel free to PM me.

Best
 

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discu...@googlegroups.com.
To post to this group, send email to brix-cms...@googlegroups.com.

[daniel simko]

Hello Korbinian,

I was thinking about "security & user management" little bit and it seems to me quite challenging. Here are my thoughts:

1) We are happy with wicket-auth-roles because we need only one thing from it - not allow to access AdminPage to users who are not authenticated. Everything else will be managed by org.brixcms.auth.AuthorizationStrategy.

2) New "user management plugin" will provide UI for managing Permissions and Roles. Permission is just String and Role is group of permissions. It will be possible to create any Permission and any Role. Roles will be possible to assign to Users.

3) Here comes the tricky part. How to map org.brixcms.auth.Action to Permission? Action can be dynamic (e.g. SiteNodeAction contains BrixNode). And new Actions also comes with new plugins. The best solution (which I found on Brix mailing list archive - kudos to Igor) is: "Create a new extension point called authorizationvoter and have a strategy that delegates to registered voters. That way the plugin that provides the action also provides a voter and is notified."

Voter interface could look something like this - https://gist.github.com/dsimko/b120dc6c5ad3d51f4330637611671d33 And its implementation cau use e.g. shiro or spring security.

I think that implementation of Voter could be much simpler when we add method:

String createPermissionName();

to the org.brixcms.auth.Action interface. This method could be also beneficial for "user management plugin" which could allow to register static Actions and pre generate Permissions for them.

What do you think?

P.S. I am from Czech republic and I am located mainly in Brno so we are not so far and we definitely could meet.

To unsubscribe from this group and stop receiving emails from it, send an email to brix-cms-discuss+unsubscribe@googlegroups.com.
To post to this group, send email to brix-cms-discuss@googlegroups.com.

[Korbinian]

Hello Dan,

What do you think?

I've looked at your gist. I think the idea might work, but the naming scheme is very irritationg for me. I think we need to carefully work on this as 1) is quite different than the needs of 2) and 3);

So making the AdminPage - or better said, the AdminWebPage protected by default using wicket-auth-roles sounds quite well. This way brix has a initial security delivered. The need on the CMS part however is very subject to the wanted usage later on. I need to dig in the code a bit and get a closer look at the authAction; In simple cases the permission(s) could be tied to an authAction by a direct relation, but what happens when you need finer grained controls? - Even wordpress gives some kind of Author, Redact etc; - but does brix need this in its core? Isn't this better part of an WebCMS-Plugin for brix? there we could deliver the more fine grained security model tied to the plugin while still alowing flexibility for other usages. Plus the developement of the plugin isn't then locked to the brix core itself.

At the current stage I would first advise on woring to make brix running with wicket 8 and do the automated testing thing, especially things like

1. protocol-scheme resolving and redirecting
2. Stateless tiles
3. Stateful tiles
4. AJAX testing
5. index page testing
6. error page testing

That way we then got a working foundation and this shouldnt break so easy. Then we could embedd the protection to the AdminPage. After that the wysiwyg CMS plugin with basic auth- and role model similar to wordpress logic (I refer to that since majority knows this and seems to work quite well);

Your thoughts on that?

 

P.S. I am from Czech republic and I am located mainly in Brno so we are not so far and we definitely could meet.

I just checked it - we're 315km out, so meeting would be no problem. We can discuss the details off-list :)

 

[daniel simko]

Hello Korbinian,

thank you for your response,

2017-01-12 18:11 GMT+01:00 Korbinian <korbini...@gmail.com>:

Hello Dan,

What do you think?

I've looked at your gist. I think the idea might work, but the naming scheme is very irritationg for me.

Could you please be more concrete, what exactly irritate you?
 

I think we need to carefully work on this as 1) is quite different than the needs of 2) and 3);

I don't get you. I am afraid that I didn't explain my ideas clearly. What I mean is that wicket-auth-roles (or implementation of org.apache.wicket.authorization.IAuthorizationStrategy) will be used only for *authentication* and org.brixcms.auth.AuthorizationStrategy for *authorization*. And I don't mean authorization only for CMS part but for whole Brix (e.g. whether SiteTab or GlobalTilesTab are visible).
 

So making the AdminPage - or better said, the AdminWebPage protected by default using wicket-auth-roles sounds quite well. This way brix has a initial security delivered. The need on the CMS part however is very subject to the wanted usage later on. I need to dig in the code a bit and get a closer look at the authAction; In simple cases the permission(s) could be tied to an authAction by a direct relation, but what happens when you need finer grained controls? - Even wordpress gives some kind of Author, Redact etc;

It shouldn't be problem. You can create a permission called Redactor and implement AccessDecisionVoter which can easily decide if current user with this permission can edit some content, because it gets instance of Action (please see e.g. org.brixcms.plugin.site.auth.SiteNodeAction) with all needed information (who is author, what is status, ...)
 

- but does brix need this in its core?

Brix already has this in its core, please see all implementations of org.brixcms.auth.Action. But nobody checks these actions by default. But we need to add some more actions to brix-core, e.g. "can see SiteTab". And implementation of "Voters" won't be in brix core but in plugins.

 

Isn't this better part of an WebCMS-Plugin for brix? there we could deliver the more fine grained security model tied to the plugin while still alowing flexibility for other usages. Plus the developement of the plugin isn't then locked to the brix core itself.

At the current stage I would first advise on woring to make brix running with wicket 8 and do the automated testing thing, especially things like

1. protocol-scheme resolving and redirecting
2. Stateless tiles
3. Stateful tiles
4. AJAX testing
5. index page testing
6. error page testing

I fully understand you and agree that this would be the best start. But I need basic end-user CMS quickly so I will start to prototype my idea and we will see if it will work. Maybe before I start to work on plugin for permissions/roles/users I will try to switch AdminPanel to bootstrap.
 

That way we then got a working foundation and this shouldnt break so easy. Then we could embedd the protection to the AdminPage. After that the wysiwyg CMS plugin with basic auth- and role model similar to wordpress logic (I refer to that since majority knows this and seems to work quite well);

Your thoughts on that?

 

P.S. I am from Czech republic and I am located mainly in Brno so we are not so far and we definitely could meet.

I just checked it - we're 315km out, so meeting would be no problem. We can discuss the details off-list :)

Yes I agree :)

Best,

Dan

 

 

--
You received this message because you are subscribed to the Google Groups "brix-cms-discuss" group.