1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
|
----------------------------------------------------------------------------
libgom TODO
----------------------------------------------------------------------------
* Lazy collections need to be implemented. This can be done in the property
getter to return a new collection with the appropriate query.
* Lazy related objects. These should return an unsaved version of the
object if the source is unsaved as well. If saved, it should return a
lazy version of the related object.
* After all the basics work, we should implement the ability to have a
pool of prepared statements for a given query.
* This assumes we are single threaded currently. That is not good because
objects are tied to their adapter, which means they cannot traverse
between threads. That makes having an adapter for each thread also
un-realistic. Both in total connections to the database and managing
shared state. Having adapters hidden behind a repository object might
help both the threaded-adapter model and the object caching model.
* It is important that profiling can be done on particular sql queries.
However, to do that, we need to know what the query actually is. A way
to trace queries should be added. Something such as:
gom_trace_enable();
gom_trace_disable();
This might be extended to SQL explain as well.
|
