OpenAccess free client

Thread is closed for posting
4 posts, 0 answers
  1. Michael Fuchs
    Michael Fuchs avatar
    3 posts
    Member since:
    Apr 2005

    Posted 01 Apr 2009 Link to this post

    Hello Telerik

    I would like to use Telerik OpenAccess in one of my next projects. The goal is to build a distributed application using WCF, but also allow an installation on a single machine without sending the business objects over the wire.

    I would like to make sure that for both scenarios the client doesn't need to know anything about OpenAccess.

    Using WCF, the client should have no reference to OpenAccess by design. As far as I understand, I can use the "ObjectNetworkAttacher" on the server to update/save an object when the client sends back the object to the server.

    Is it also possible to build a non-distributed application without involving OpenAccess on the client?

    Best regards
    Michael
  2. Jan Blessenohl
    Admin
    Jan Blessenohl avatar
    707 posts

    Posted 03 Apr 2009 Link to this post

    Hello Michael Fuchs,
    You are descriping one way of doing it. Just serialize the objects via WCF and if they are comming back use the ObjectNetworkAttacher to attach them again.

    The problem with OpenAccess free clients is that you loose all change tracking and lazy loading capability. Also the generation of minimal change sets is something you have to do by yourself.

    You have to deliver only the Telerik.OpenAccess.dll to the client if you use our ObjectContainer to do the client side data management. Please have a look at our ObjectContainer example.

    Sincerely yours,
    Jan Blessenohl
    the Telerik team

    Check out Telerik Trainer , the state of the art learning tool for Telerik products.
  3. Kyaw Zayar Han
    Kyaw Zayar Han avatar
    5 posts
    Member since:
    Sep 2009

    Posted 14 Sep 2009 Link to this post

    Hi Telerik,

    Could you give a few answers for this comment you made "The problem with OpenAccess free clients is that you loose all change tracking and lazy loading capability. Also the generation of minimal change sets is something you have to do by yourself."

    Is there any obvious benefit over "change tracking" for disconnected/distributed client or is there any drawback lacking of change tracking at client?

    What role does lazy loading play at client, since we need to load all objects in object graph once we retrieve from server? it would give exception at client if we didn't.

    What do you mean minimal change sets to be done by myself? Could you give a sample?

    I have to get these answers from you, because telerik is influencing over our architecture, unnecessarily, even overwrote WCF architecture, you know telerik always put their binaries in WCF objects, instead of pure serialized xml, so it is not possible to have telerik free service client. And if we force to use additional mapping layer between telerik powered data access layer and service layer, there are no more obvious benefits using telerik. we are just moving object data mapping task to object telerik mapping task.

    Please let us have a few suggestions, it would be most valuable to us.

    Thanks,
    Han
  4. Jan Blessenohl
    Admin
    Jan Blessenohl avatar
    707 posts

    Posted 16 Sep 2009 Link to this post

    Hello Kyaw Zayar Han,
    If you just serialize the object tree and send it over the wire, on the client side the data knows nothing about the database or a management instance like OpenAccess ObjectScope. It is pure data. If you now change a field there is nobody to collect those changes and send them back to the server. You have to know that the tree is changed and send the complete tree back to the server.

    The OpenAccess ObjectContainer is able to track those changes and to send only the changed data back to the server instead of a complete tree. If not the complete tree is in the container there is an event mechanism that informs you about the missing data and you have the chance to request the missing data from the server so that your logic is not broken.

    Minimal change set means that the container does not send the complete tree back to the server but only the changed objects.

    Kind regards,
    Jan Blessenohl
    the Telerik team

    Instantly find answers to your questions on the new Telerik Support Portal.
    Watch a video on how to optimize your support resource searches and check out more tips on the blogs.
Back to Top