| 1 |
$MODULE = "Tie::Cache"; $VERSION = .21; $DATE = '05/19/2014'; |
| 2 |
|
| 3 |
+ fixed internal $VERSION number |
| 4 |
|
| 5 |
$MODULE = "Tie::Cache"; $VERSION = .19; $DATE = '02/23/2012'; |
| 6 |
|
| 7 |
+ added cleaner ref() checking for data size calculations when using a MaxBytes cache, to address bug |
| 8 |
https://rt.cpan.org/Public/Bug/Display.html?id=69976 |
| 9 |
|
| 10 |
+ updated test.pl to "use Tie::Cache" to prevent test conflicts with Cache.pm |
| 11 |
|
| 12 |
$MODULE = "Tie::Cache"; $VERSION = .17; $DATE = '07/03/2002'; |
| 13 |
|
| 14 |
+ added tied(%cache)->flush() method to sync dirty writes |
| 15 |
that have been cache with WriteSync => 0 config for a true |
| 16 |
cache inherited object. Added testing to cover in test.pl |
| 17 |
|
| 18 |
$MODULE = "Tie::Cache"; $VERSION = .15; $DATE = '03/20/2001'; |
| 19 |
|
| 20 |
+ Better test.pl timing output, also differentiate between |
| 21 |
cache with MaxBytes setting and one with just MaxCount, |
| 22 |
as MaxBytes is a bit slower for data size calculations. |
| 23 |
|
| 24 |
+ Better Tie::Cache options / object config error checking, |
| 25 |
die() on obvious misconfigurations that will get developers |
| 26 |
in trouble. |
| 27 |
|
| 28 |
- MaxSize only set if MaxSize or MaxBytes are defined, |
| 29 |
was defaulting to 1 otherwise, killing basic MaxCount config. |
| 30 |
|
| 31 |
+ Optimizations for common use where Tie::Cache is not |
| 32 |
subclassed for write()/read() API |
| 33 |
|
| 34 |
$MODULE = "Tie::Cache"; $VERSION = .11; $DATE = '03/18/2001'; |
| 35 |
|
| 36 |
+ use of arrays for internal nodes structure instead of |
| 37 |
hashes for savings on memory and lookup speed. |
| 38 |
|
| 39 |
+ MaxBytes now more accurate with recursive length calculations |
| 40 |
on reference keys & values, and added estimates of 240 bytes |
| 41 |
per key/value pair, + 16 bytes per reference found, in addition |
| 42 |
to the length of the data being stored. |
| 43 |
|
| 44 |
A MaxBytes setting of 1000000 would then try to really limit |
| 45 |
total memory consumption of the entire cache to 1M. This is |
| 46 |
relevant when being used in high process environments like mod_perl |
| 47 |
where memory savings are critical. |
| 48 |
|
| 49 |
+ MaxSize has default of MaxBytes/10, so large cache entries don't |
| 50 |
purge many others. If cache entry exceeds MaxSize in bytes, then |
| 51 |
that value won't be cached. |
| 52 |
|
| 53 |
$MODULE = "Tie::Cache"; $VERSION = .10; $DATE = '11/22/00'; |
| 54 |
|
| 55 |
- Allow for undef values to be stored like $cache{1} = undef |
| 56 |
Really doesn't affect cache, but got rid off error message |
| 57 |
in this case. |
| 58 |
|
| 59 |
$MODULE = "Tie::Cache"; $VERSION = .09; $DATE = 'TBA'; |
| 60 |
|
| 61 |
+ Added benchmark section comparing Tie::Cache & Tie::Cache::LRU |
| 62 |
|
| 63 |
$MODULE = "Tie::Cache"; $VERSION = .08; $DATE = '12/14/99'; |
| 64 |
|
| 65 |
- FETCH, or %cache reads like $cache{key} would returned |
| 66 |
a defined value, null, when it should have been undefined. |
| 67 |
This would breaks tests that test for defined($cache{$key}), |
| 68 |
and is a big problem. I introduced this bug with my |
| 69 |
tuning in .07 :( |
| 70 |
|
| 71 |
$MODULE = "Tie::Cache"; $VERSION = .07; $DATE = '12/13/99'; |
| 72 |
|
| 73 |
+ Improved test suite to do some benchmarking, and |
| 74 |
real interpreted tests. |
| 75 |
|
| 76 |
+ Performance tuning making the cache about 5-10% faster. |
| 77 |
|
| 78 |
+ MaxSize setting makes it so that key/value pair lengths |
| 79 |
of MaxSize or greater will not be stored in the cache. |
| 80 |
This prevents odd large entries from flushing the cache |
| 81 |
of many smaller entries. |
| 82 |
|
| 83 |
- setting Debug for one Tie::Cache hash is independent of |
| 84 |
another's Debug setting, before one setting would affect |
| 85 |
all the rest. To set Debug value for all Tie::Cache |
| 86 |
hashes, set $Tie::Cache::Debug. |
| 87 |
|
| 88 |
$MODULE = "Tie::Cache"; $VERSION = .06; $DATE = '2/16/99'; |
| 89 |
|
| 90 |
+ WriteSynch config option for TRUE CACHE. WriteSynch => 0 |
| 91 |
will have the dirty data be written back as late as possible. |
| 92 |
WriteSynch => 1 is immediate write-through for data dirtied. |
| 93 |
|
| 94 |
- STORE returns value stored; $cache{$key} = $value returns $value now |
| 95 |
|
| 96 |
+ decomped FETCH better so it doesn't use STORE internally |
| 97 |
necessary for new WriteSynch functionality |
| 98 |
|
| 99 |
+ WriteSynch => 0 config set in test.pl to demostrate use. |
| 100 |
|
| 101 |
+ Optimizations, especially for refreshing entries in cache on FETCH |
| 102 |
|
| 103 |
$MODULE = "Tie::Cache"; $VERSION = .05; |
| 104 |
|
| 105 |
+ Keep track of hits / misses even without debug option set. |
| 106 |
|
| 107 |
$MODULE = "Tie::Cache"; $VERSION = .04; |
| 108 |
|
| 109 |
- Get rid of -w warnings in test.pl. |
| 110 |
|
| 111 |
$MODULE = "Tie::Cache"; $VERSION = .031; |
| 112 |
|
| 113 |
- Removed test2.pl from installation as it was getting installed |
| 114 |
with Tie::Cache :( |
| 115 |
|
| 116 |
$MODULE = "Tie::Cache"; $VERSION = .03; |
| 117 |
|
| 118 |
- Global destructor bug fixed. Wasn't flushing cache consistently. |
| 119 |
|
| 120 |
$MODULE = "Tie::Cache"; $VERSION = .02; |
| 121 |
|
| 122 |
+ MaxBytes config option is new. It allows cache size to |
| 123 |
be based on the bytes the cache holds. |
| 124 |
|
| 125 |
+ test.pl output has Debug set to 2, so full debugging output is |
| 126 |
displayed. This allows a new user to see how cache works. |
| 127 |
|
| 128 |
$MODULE = "Tie::Cache"; $VERSION = .01; |
| 129 |
|
| 130 |
First release of module. |
| 131 |
|
| 132 |
|
| 133 |
|
| 134 |
|
| 135 |
|
| 136 |
|
| 137 |
|