Ticket #1698 (closed unplanned_task: obsolete)

Opened 10 years ago

Last modified 10 years ago

GROUP_APP_SERVER_RESOURCE_ACCESS_R0

Reported by: kyli Owned by: kyli
Priority: major Milestone: M09_BETA1
Component: uncategorized Version: 2.0
Keywords: Cc:
Category: unknown Effort: 0
Importance: Ticket_group:
Estimated Number of Hours: Add Hours to Ticket:
Billable?: Total Hours:
Analysis_owners: kyli,boyan Design_owners: kyli,boyan,kyli
Imp._owners: kyli, tsachev Test_owners:
Analysis_reviewers: deyan,deyan Changelog:
Design_reviewers: meddle, meddle Imp._reviewers: meddle
Test_reviewers: Analysis_score: 3
Design_score: 3 Imp._score: 3
Test_score: 0

Change History

comment:1 Changed 10 years ago by boyan

  • Owner set to boyan
  • Status changed from new to s1a_analysis_started
  • Analysis_owners set to kyli,boyan

starting analysis

comment:2 Changed 10 years ago by boyan

  • Status changed from s1a_analysis_started to s1b_analysis_finished

finished

comment:3 Changed 10 years ago by deyan

  • Status changed from s1b_analysis_finished to new
  • Analysis_reviewers set to deyan
  • Analysis_score changed from 0 to 2.5
  • Some of the task requirements are design-oriented - an analysis should not contain what will be in which module, but what is the benefit of this. You can keep these things in Implementation idea, but stating them as requirements makes the analysis implementation dependent.
  • Since this is first revision of this task, it is a good idea to make an overview of the whole task (ideas about next releases).

comment:4 Changed 10 years ago by deyan

  • Milestone set to M09_BETA1

It is also important to add a Milestone for the group ticket, otherwise the ticket is not shown in the iteration page.

comment:5 Changed 10 years ago by kyli

  • Owner changed from boyan to kyli
  • Status changed from new to s1a_analysis_started

comment:6 Changed 10 years ago by kyli

  • Status changed from s1a_analysis_started to s1b_analysis_finished

Finished again. The overview could not be expanded, since there is no specification about connectivity (at least I couldn't find), maybe it will be generated with the other APP_CONNECTIVITY tasks.

comment:7 Changed 10 years ago by deyan

  • Status changed from s1b_analysis_finished to s1c_analysis_ok
  • Analysis_reviewers changed from deyan to deyan,deyan
  • Analysis_score changed from 2.5 to 3

The design still looks too technical. The overview still does not contain overview of the whole task, which is wrong in my opinion - features need to be designed thinking for all of the features.

comment:8 Changed 10 years ago by kyli

  • Design_owners set to kyli
  • Status changed from s1c_analysis_ok to s2a_design_started

comment:9 Changed 10 years ago by kyli

  • Design_owners changed from kyli to kyli,boyan,kyli

finished design

comment:10 Changed 10 years ago by kyli

  • Status changed from s2a_design_started to s2b_design_finished

comment:11 Changed 10 years ago by meddle

  • Cc kyli, boyan added
  • Design_score changed from 0 to 2.5
  • Design_reviewers set to meddle
  • Status changed from s2b_design_finished to s1c_analysis_ok
  • Important design notes:
    • First of all the Facade definition (the interfaces of the facade) must be moved to the org.sophie2.bas.connectivity package because despite the server core it too provides a facade implementation and it's more appropriate the server.core package to depend to a "base" package.
    • About the Resource facade. I agree at this moment the too methods work for the requirements but:
      • The getResourceList() method returns ResourceRef statement is not full it returns "absolute" ResourceRefs, without entity ids, created on the server with ResourceRef#getAbsoluteReference(String absoluteLocation). Only that is needed for cathing a resource from the server
      • getResource(String entityId); That is wrong, on the server we work with URLs mostly, thats why use getResource(String absoluteLocation).
    • All the methods in the ResourceServeice is appropriate to accept absoluteLocation and not entityId or Resource/ResourceRef.
    • deleteBook is left? Why not deleteResource? Ahaaa in the code is renamed :) But not in the design documentation...
    • Use the Persistence for serializing the ResourceRefs when you sending them to the client as XML, and then on the client use the persistence again to deserialize them.
    • Change the Daos with my notes about the services :)
  • Other ideas:
    • Maybe soon will be time to rename the Services, Managers, Daos to more meaningful things, that is old names that I thought at some point...
  • For the code at this moment I have no special notes, at the implementation review it will be checked in details...

comment:12 Changed 10 years ago by boyan

  • Cc boyan removed

removing cc

comment:13 Changed 10 years ago by kyli

  • Status changed from s1c_analysis_ok to s2a_design_started

comment:14 Changed 10 years ago by kyli

  • Cc kyli removed
  • Status changed from s2a_design_started to s2b_design_finished

Thanks for the review. The following things were changed:

  • A brief explanation of the current facade location has been added.
  • entityIds have been replaced by absoluteLocation in the design.
  • For bookService and BookDao, a requirement has been added to rename the *book* methods to *resource*.
  • Usage of XMLManagementService for resourceRef sending was replaced by the current persistence.

comment:15 Changed 10 years ago by meddle

  • Status changed from s2b_design_finished to s2c_design_ok
  • Design_score changed from 2.5 to 3
  • Design_reviewers changed from meddle to meddle, meddle
  • Notes:
    • Hmm the absolute ResourceRef has only the absolute location, but form it you can get the name of course.
    • And NO, the server core and the server facade are two different things, in the implementation don't merge them in one. I'll not pass such an implementation and Pap will not pass it too. The core is the main logic in the server, the ServerFacade is simple implementation of the Facade that uses part of this core logic over the core!
    • You will change the MemoryDatabasemanager... So don't broke the user/group logic that works now and make sure the web interface for working with books/resources work without exceptions, synchronize with Vlado and the server MUST run with the True run configuration.

3p (20m)

comment:16 Changed 10 years ago by kyli

  • Status changed from s2c_design_ok to s3a_implementation_started
  • Imp._owners set to kyli

Actually started implementation 2 days ago.

comment:17 Changed 10 years ago by kyli

  • Status changed from s3a_implementation_started to s3b_implementation_finished
  • Imp._owners changed from kyli to kyli, tsachev

comment:18 Changed 10 years ago by meddle

  • Status changed from s3b_implementation_finished to s3c_implementation_ok
  • Imp._score changed from 0 to 3
  • Imp._reviewers set to meddle

The task will be passed to the trunk.

  • For the next task -> http://sophie2.org/trac/ticket/1214 these thing must be fixed:
    • Write the JavaDoc of all the server classes (in all the modules) like described in GoodCodeExamples. Especially for these that have "TODO" for that.
    • Replace all the e.printStackTrace(); lines with appropriate log.
    • Fix all the AssertationErrors that are logged when running some of the tests.
  • Bad thing in the implementation:

3p (230m)

comment:19 Changed 10 years ago by deyan

  • Status changed from s3c_implementation_ok to closed
  • Resolution set to obsolete

Batch update from file query-obsoleted.csv

Note: See TracTickets for help on using tickets.