| 1 |
$Id$ |
| 2 |
|
| 3 |
Changes after 1.16 are listed in the CHANGES file. |
| 4 |
|
| 5 |
*** 1.16 2005-09-17 |
| 6 |
|
| 7 |
Applied Moshe Kaminsky's patch to support utf-8 submissions. |
| 8 |
|
| 9 |
*** 1.15 2004-09-08 |
| 10 |
|
| 11 |
Tilman Sauerbeck updated the ID calculations to work with FreeDB's |
| 12 |
reference CD. Previously they were slightly off, but the protocol's |
| 13 |
fuzzy matching had been compensating. |
| 14 |
|
| 15 |
*** 1.14 2004-09-01 |
| 16 |
|
| 17 |
Applied Steve James' patch to improve dead connection detection. It |
| 18 |
prevents harmless yet annoying warnings. |
| 19 |
|
| 20 |
Applied Ronan Waide's patch to expose the XMCD file parsing code. The |
| 21 |
new method is parse_xmcd_file(). |
| 22 |
|
| 23 |
*** 1.12 2004-02-12 |
| 24 |
|
| 25 |
Applied Duncan Martin's patch for submitting year and long genre data. |
| 26 |
|
| 27 |
Performed some long overdue tweaking in the tests. The CD databases |
| 28 |
are constantly moving targets, and the freedb servers recently |
| 29 |
prohibited multiple queries for the same disc in the same connection. |
| 30 |
Upshot: The tests pass once again. |
| 31 |
|
| 32 |
Cleaned up the code a little in CDDB.pm and cddb.t. |
| 33 |
|
| 34 |
*** 1.11 2003-08-13 |
| 35 |
|
| 36 |
Version 1.10 was an empty distribution, because I mistakenly deleted |
| 37 |
the MANIFEST. This fixes the last release. |
| 38 |
|
| 39 |
*** 1.10 2003-08-13 |
| 40 |
|
| 41 |
Applied Albrecht Kleine's patch to make the xmcd parser more robust. |
| 42 |
Some software and/or devices submit records that are slightly off and |
| 43 |
broke CDDB.pm in the past. |
| 44 |
|
| 45 |
*** 1.09 2003-07-31 |
| 46 |
|
| 47 |
Added Christopher Hartmann's patches to the 1.08 changes, and set the |
| 48 |
1.08 release date. Whoops! I was in a hurry to release that one. |
| 49 |
|
| 50 |
Colin Meyer pointed out some confusing language in the get_discs() |
| 51 |
documentation. Rocco fixed it. |
| 52 |
|
| 53 |
Added Ron Grabowski's get_discs_by_query() function. This fetches |
| 54 |
discs by a query string, rather than a table of contents. |
| 55 |
|
| 56 |
Added Michael Jung's parse_cdinfo() function. This parses the output |
| 57 |
of cdinfo into a table of contents suitable for calculate_id(). |
| 58 |
|
| 59 |
Michael Jung suggested that the module display a diagnostic if all the |
| 60 |
known CDDB protocol servers are unavailable. That seemed reasonable, |
| 61 |
so CDDB.pm warns if it can't contact a server. |
| 62 |
|
| 63 |
*** 1.08 2002.01.19 |
| 64 |
|
| 65 |
Applied henke's patch to support CDDB protocol level 5, if available. |
| 66 |
|
| 67 |
Applied Yannick Le Saint's patch to try localhost:8880. Made |
| 68 |
localhost the first server to try. |
| 69 |
|
| 70 |
Applied Yannick Le Saint's patch to add a missing =back to the POD. |
| 71 |
|
| 72 |
Jonathan Stowe and Yannick Le Saint supplied patches to avoid dying if |
| 73 |
no cddbp servers are available. Rocco applied something different to |
| 74 |
achieve the same goal. |
| 75 |
|
| 76 |
Applied Christopher Hartmann's patch to let programmers specif new |
| 77 |
cddbp versions. |
| 78 |
|
| 79 |
Applied Christopher Hartmann's patch to return the entire unparsed |
| 80 |
xmcd record. |
| 81 |
|
| 82 |
*** 1.07 2001.11.18 |
| 83 |
|
| 84 |
Fixed t/cddb.t to not care *what* it gets as long as it gets |
| 85 |
something. |
| 86 |
|
| 87 |
*** 1.06 2001.07.17 |
| 88 |
|
| 89 |
Added get_mail_host() and get_mail_address() methods to query what the |
| 90 |
module thinks your SMTP host and e-mail address are. Their purpose is |
| 91 |
to fetch values for confirmation or correction before submitting disc |
| 92 |
records. |
| 93 |
|
| 94 |
FreeDB servers sometimes emit error 530 on connection instead of the |
| 95 |
regular 201 banner. The CDDB module wouldn't handle this very well; |
| 96 |
now it closes the connection and tries again automatically. |
| 97 |
|
| 98 |
Disc submissions had hardcoded revisions of 1. They should be 0 for |
| 99 |
new discs and one more than the last for updates. Fixed submit_disc() |
| 100 |
to honor its documented Revision parameter. |
| 101 |
|
| 102 |
Fixed submit_disc() to transmit e-mail submissions using ISO-8859-1 |
| 103 |
(Latin 1) and quoted-printable. What's this mean? Basically: Western |
| 104 |
European characters won't get mangled by intermediate mail servers. |
| 105 |
|
| 106 |
Drew Taylor discovered that the tests fail on MSWin32 because |
| 107 |
getpwuid() isn't implemented on that platform. Worked around its |
| 108 |
absence on MSWin32 systems, and documented this bit of CDDB more |
| 109 |
carefully. |
| 110 |
|
| 111 |
*** 1.05 2001.05.03 |
| 112 |
|
| 113 |
Christoph Lorenz pointed out that the module is noisy even when |
| 114 |
debugging is turned off. I made sure that all the carp calls are |
| 115 |
behind debuging checks. |
| 116 |
|
| 117 |
*** 1.04 2001.03.24 |
| 118 |
|
| 119 |
Switched from cddb.com servers to freedb.org servers. cddb.com has |
| 120 |
been denying CDDB.pm users. |
| 121 |
|
| 122 |
Made some of the tests even fuzzier. |
| 123 |
|
| 124 |
*** 1.03 2000.03.27 |
| 125 |
|
| 126 |
Perl no longer likes defined(@array); tweak, tweak, tweak. |
| 127 |
|
| 128 |
CDDB returns a different code for "Achtung, Baby"; tweak, tweak, |
| 129 |
tweak. |
| 130 |
|
| 131 |
Updated the default hosts with the most recent CDDB protocol sites |
| 132 |
list; tweak, tweak, tweak. |
| 133 |
|
| 134 |
*** 1.02 1999.08.13 |
| 135 |
|
| 136 |
Once again, someone has changed the compact disc record that CDDB |
| 137 |
tests against. This broke tests 16 (disc title) and 26 (track |
| 138 |
titles). Once again I have made the tests fuzzier, so they will pass |
| 139 |
again. |
| 140 |
|
| 141 |
This version's CDDB.pm is identical to 1.01's, except for the version |
| 142 |
number. |
| 143 |
|
| 144 |
*** 1.01 1999.07.16 |
| 145 |
|
| 146 |
Just as I was getting ready to celebrate a bug-free CDDB module, along |
| 147 |
comes some Macintosh freak with spaces in his login ID (Hi, Schwern!). |
| 148 |
So now the module squashes spaces and replaces them with underscores. |
| 149 |
And if that isn't enough, CDDB's constructor now accepts an optional |
| 150 |
Login parameter. |
| 151 |
|
| 152 |
*** 1.00 1999.07.16 |
| 153 |
|
| 154 |
Whenever someone updates a CDDB record that cddb.t tests against, it |
| 155 |
breaks the test. It's a fairly harmless problem, but it does cause |
| 156 |
CDDB not to pass its tests. I'm fed up with it after a few |
| 157 |
iterations, so I changed the hard "retrieved track offset != stored |
| 158 |
one" tests with a more lenient "retrieved track offset within 5% of |
| 159 |
stored one". |
| 160 |
|
| 161 |
While I'm at it, make the title tests approximate by normalizing |
| 162 |
whitespace and only comparing consonants. Sort of an approximate |
| 163 |
metaphone match. |
| 164 |
|
| 165 |
Still to do: The genres and track-lengths tests haven't been made |
| 166 |
fuzzy in this version. |
| 167 |
|
| 168 |
*** 0.07-beta 1998.11.27 |
| 169 |
|
| 170 |
Added a list of CDDB servers inside CDDB.pm. Added code to cycle |
| 171 |
through the servers upon receipt of "417 Database access limit |
| 172 |
reached" or connection errors. |
| 173 |
|
| 174 |
Made t/cddb.t a little lamer for the sake of passing the fuzzy matches |
| 175 |
tests. |
| 176 |
|
| 177 |
Added information about the CDDB submission test to README. |
| 178 |
|
| 179 |
Chris Nandor suggested a get_discs_after_calculating_id() macro |
| 180 |
function. Added it under the name get_discs_by_toc(). |
| 181 |
|
| 182 |
Whoops! CDDB::message() was missing, but it was never called during |
| 183 |
testing. Until today. Added CDDB::message(). |
| 184 |
|
| 185 |
*** 0.06-beta |
| 186 |
|
| 187 |
Fixed newlines in submit_disc so they meet Net::Cmd expectations |
| 188 |
|
| 189 |
Added the HISTORY file to track changes. |
| 190 |
|
| 191 |
-EOF- |