|
50 | 50 | * {@link bdv.ViewerImgLoader} in order to use the bigdataviewer playground
|
51 | 51 | * global cache. Two implementations are provided: - a Caffeine backed cache
|
52 | 52 | * {@link CaffeineGlobalCache} - and a LinkedHashMap cache
|
53 |
| - * {@link BoundedLinkedHashMapGlobalCache} The |
54 |
| - * {@link GlobalCacheBuilder} object can be serialized to store the |
55 |
| - * cache configuration The global caching allows to bound the memory used when |
56 |
| - * many sources are potentially accessed in a random manner by the program. The |
57 |
| - * reason to use many sources is if the user wants to work on different spim |
58 |
| - * data objects at the same time, work with resampled sources, etc Global |
59 |
| - * caching allow to bound the memory used at the JVM level, instead of a per |
60 |
| - * source level or per spimdata level. Each value of the value is weighted by |
61 |
| - * its memory footprint, assuming it is a {@link Cell} object. If the value is |
62 |
| - * not a Cell object, an error is logged. This allows to bound memory correctly |
63 |
| - * even when sources have very different block size (setting a fixed number of |
64 |
| - * items maintained in cache would not be precise enough). |
| 53 | + * {@link BoundedLinkedHashMapGlobalCache} The {@link GlobalCacheBuilder} object |
| 54 | + * can be serialized to store the cache configuration The global caching allows |
| 55 | + * to bound the memory used when many sources are potentially accessed in a |
| 56 | + * random manner by the program. The reason to use many sources is if the user |
| 57 | + * wants to work on different spim data objects at the same time, work with |
| 58 | + * resampled sources, etc Global caching allow to bound the memory used at the |
| 59 | + * JVM level, instead of a per source level or per spimdata level. Each value of |
| 60 | + * the value is weighted by its memory footprint, assuming it is a {@link Cell} |
| 61 | + * object. If the value is not a Cell object, an error is logged. This allows to |
| 62 | + * bound memory correctly even when sources have very different block size |
| 63 | + * (setting a fixed number of items maintained in cache would not be precise |
| 64 | + * enough). |
65 | 65 | *
|
66 | 66 | * @author Nicolas Chiaruttini
|
67 | 67 | */
|
|
0 commit comments