GROUP_PERSISTENCE_R0

[boyan]

All tasks related to persistence are grouped in GROUP_PERSISTENCE_R0

Ticket	Summary	Effort	Status
#327	BASE_PERSISTENCE_COMMONS_R0	1	closed
#328	BASE_PERSISTENCE_COMMONS_R1	1	closed
#331	BASE_PERSISTENCE_INTERMEDIATE_STORAGE_R0	1	closed
#334	BASE_PERSISTENCE_FORMAT_REGISTRY_R0	1	closed
#337	BASE_PERSISTENCE_PROPERTY_FORMAT_R0	1	closed
#340	BASE_PERSISTENCE_IMMUTABLES_R0	1	closed
#341	BASE_PERSISTENCE_IMMUTABLES_R1	1.5	closed
#343	BASE_PERSISTENCE_FORMAT_POINT_R0	1	closed

[boyan]

taken for analyzing

[boyan]

done in 1h

[pav]

Full enough for me. 30m

[boyan]

taken for designing

[pap]

fixing trac ticket query

[boyan]

dropping the design

[deyan]

Considered the analysis obsolete as it is more than a month old

[deyan]

added owners and started the analysis2

[deyan]

fixing

[deyan]

design is started, this task is in progress

[meddle]

I'll fail this design... I tried many things for something like 3 hours, but you have some major problems...

• First of all the tests:
  • Read how to write tests! In the assertations there are rules for writing, expected and actual order and etc...
  • Spacing!
  • There have to be tests for all the persister implementations with provided revisions.
  • The provided revisions will be merged to the trunk after Design OK, so no warning and no errors, if you must, you can provide all your changesets up to now.
• For the code quality
  • Spacing!
  • Accurate javadoc
  • And again spacing!
• For the builds
  • I was fighting building your code to run with the True author/reader runner, you have to give me buildable code with maven.
  • Configure accurately your pom, the bundle activator is missing, the package export was no right, you have some ..*.persistence.storage package provided.

The code for the design is important and it have to be right, I wanted to pass that design and the design itself works, I gave too many time trying to fix the problems with the build. Please fix the problems, fix your tests, they are important, test all the new Resource creation methods, In one of the revisions with the new implementation I provided you such a test, write tests for the other methods.

2.5 (3h)

[meddle]

My notes:

• That is the design. When you add new methods to the Storage for example, write their accurate names...
• When you write new logic, you describe it in the design, I don't see the new resource related methods from the ResourceSpace and the ResourceLocalCache described in the design, make sure at the implementation phase you describe them!
• The file format : I think you should describe more of the attributes/tags in it, for example what is the meaning of the "index" attribute of the lists and why it is necessary can't we use the order of the tags instead, such things...
• I think tests for this "refs" is needed, you use them in the higher-level tests, but you don't have their own test, besides that, you can test strange use cases like overriding their "get" method.
• Other thing is that when you are saving the ImmSize for example you are using saving of Integers, so don't you thing the method for saving the size must be preceded by one for the integers in this case?
• The use case in which you are saving text frame is important. I think in the BookPersistenceTest, you can add to the page of the book a text frame to clean this case.
• I can see absolute paths in you tests like "C:
  blq.sjrb", fix them for the implementation phase!!!

But as a whole I can pass the design, the problems are mainly for the implementation. The persistence will run under the true author module, and the code quality is good, but I can not merge design related code at this state, and having in mind that the persistence task is a big one and the design and the implementation code are merged in one pile of code :D, I'll commit all of the persistence logic in the trunk after the implementation review passes.

3p (30m)

[meddle]

With the little fixes everything look right and works with the runners.

3.5p (100m)

[deyan]

Batch update from file query-obsoleted.csv