Changes between Version 63 and Version 64 of CategoriesRoadMap


Ignore:
Timestamp:
01/21/11 01:49:22 (2 years ago)
Author:
nthiery
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • CategoriesRoadMap

    v63 v64  
    170170- Roadmap for after: 
    171171  * Add more examples of parents (HELP!): 
    172     Ultimately, this should be for all categories, 
     172    Ultimately, this should be for all categories. 
    173173    Those are needed urgently: 
    174174    * Modules (for doctests for __mul__ / __rmul__ / ...) 
     
    178178    e.g.: Semigroups??, Semigroups().example()?? 
    179179    Semigroups().parent_class?, Semigroups.element_class? 
    180   * Add ObjectMethods for category objects that are not parents. 
     180  * Add ObjectMethods for category objects that are not parents? 
    181181  * Choose, implement, and use systematically a mantra for testing if P is an algebra (over whatever ring): 
    182182    - P in Algebras 
     
    199199  * __init_add__ 
    200200  * Writing and linking to generic documentation 
    201   * Standardize the functorial constructions, and streamline with 
    202     variants (finite / finite dimensional / graded / abstract / ...) 
     201 
     202  * Add support for automatic join for full subcategories (finite / 
     203    finite dimensional / graded / ...). E.g. 
     204    join(FiniteSets() and Groups()) should be FiniteGroups(). 
     205 
     206  * join and meet are misnomers: since we speak of subcategories, then 
     207    the implicit order is A <= B if A is a subcategory of B. So join 
     208    and meet should be exchanged. Maybe it's too late for backward 
     209    compatibility. Then maybe we could use "intersect" instead of 
     210    join. For meet, there does not seem to be an obvious name, but 
     211    it's also very seldom used at the user interface, so the backward 
     212    compatibility issue is not that relevant. 
     213 
    203214  * Hom is *not* a functorial construction. This works for now, but 
    204215    the design and user interface needs to be discussed for the long 
     
    208219    where the specializations (A x b) -> C and and (a x B) -> C are 
    209220    morphisms for possibly different categories (see also #8900) 
     221 
    210222  * Defining new inline operators, at least within the sage interpreter 
    211223  * Convention for the order in super_categories (see primer)