| 7d42c22 | Hye-Shik Chang | 28 March 2006, 08:36:44 UTC | Relocate the misplaced heading. | 28 March 2006, 08:36:44 UTC |
| 6c52761 | Hye-Shik Chang | 28 March 2006, 08:27:27 UTC | Ubuntu bug #29289: Fixed a bug that the gb18030 codec raises RuntimeError on encoding surrogate pair area on UCS4 build. This is a partial backport of r43320. (Approved by Anthony Baxter) | 28 March 2006, 08:27:27 UTC |
| 653cf02 | Anthony Baxter | 28 March 2006, 08:21:00 UTC | it never ends! more ignoring | 28 March 2006, 08:21:00 UTC |
| 33355ee | Anthony Baxter | 28 March 2006, 08:04:57 UTC | ignore more generated files (*.pyo, libpython.2.4.a) | 28 March 2006, 08:04:57 UTC |
| cb9051a | Anthony Baxter | 28 March 2006, 07:32:36 UTC | after discussions with perky, reverted fix for Bug #1379994: Builtin unicode_escape and raw_unicode_escape codec now encodes backslash correctly. This caused another issue for unicode repr strings being double-escaped (SF Bug #1459029). Correct fix will be in 2.5, but is too risky for 2.4.3. Added a testcase for #1459029. | 28 March 2006, 07:32:36 UTC |
| 51487fe | Anthony Baxter | 28 March 2006, 07:18:33 UTC | news for test_capi change | 28 March 2006, 07:18:33 UTC |
| 1d25000 | Tim Peters | 28 March 2006, 07:09:33 UTC | Merge rev 43181 from the trunk. Try to repair at least one segfault on the Mac buildbot, as diagnosed by Nick Coghlan. test_capi.py: A test module should never spawn a thread as a side effect of being imported. Because this one did, the segfault one of its thread tests caused didn't occur until a few tests after test_regrtest.py thought test_capi was finished. Repair that. Also join() the thread spawned at the end, so that test_capi is truly finished when regrtest reports that it's done. _testcapimodule.c test_thread_state(): this spawns a couple of non-threading.py threads, passing them a PyObject* argument, but did nothing to ensure that those threads finished before returning. As a result, the PyObject* _could_ (although this was unlikely) get decref'ed out of existence before the threads got around to using it. Added explicit synchronization (via a Python mutex) so that test_thread_state can reliably wait for its spawned threads to finish. | 28 March 2006, 07:09:33 UTC |
| e77ff2a | Anthony Baxter | 28 March 2006, 06:32:00 UTC | preparing for 2.4.3 final | 28 March 2006, 06:32:00 UTC |
| 55dd2b4 | Neal Norwitz | 28 March 2006, 06:05:21 UTC | Fix the refleak from test_unicode. Backport 42973 (lots of whitespace changes intermixed): - Reindent a confusingly indented piece of code (no intended code changes there) - Add missing DECREFs of inner-scope 'temp' variable - Add various missing DECREFs by changing 'return NULL' into 'goto onError' - Avoid double DECREF when last _PyUnicode_Resize() fails Coverity found one of the missing DECREFs, but oddly enough not the others. | 28 March 2006, 06:05:21 UTC |
| adbd28f | Martin v. Löwis | 27 March 2006, 21:47:54 UTC | Backport of compile.c part of r41531 (neal.norwitz, 2005-11-24): Fix a ref leak. | 27 March 2006, 21:47:54 UTC |
| 97334ae | Martin v. Löwis | 27 March 2006, 21:38:32 UTC | Backport of r41530 (neal.norwitz, 2005-11-24): Move registration of the codec search function to the module scope so it is only executed once. Otherwise the same search function is repeated added to the codec search path when regrtest is run with -R and leaks are reported. | 27 March 2006, 21:38:32 UTC |
| c47bfb0 | Thomas Wouters | 27 March 2006, 12:50:42 UTC | Backport Ka-Ping Yee's trunk checkin r41400: Fix SF bug #417833 (pydoc HTTP reload failure) by removing from sys.modules all submodules of a the given module/package path when trying to reload a module. | 27 March 2006, 12:50:42 UTC |
| a8de08d | Neal Norwitz | 26 March 2006, 04:10:42 UTC | Backport: Handle sys.getfilesystemencoding() returning None. ascii seems like the safest bet, it should exist. I wonder if utf-8 would be a better choice? This should get test_fileinput passing on OpenBSD. | 26 March 2006, 04:10:42 UTC |
| d9578f9 | Raymond Hettinger | 26 March 2006, 01:43:15 UTC | Fix mark-up. | 26 March 2006, 01:43:15 UTC |
| ee43b6c | Raymond Hettinger | 26 March 2006, 01:33:11 UTC | Guarantee evaluation order for izip(). Document its creative uses and its limitations. | 26 March 2006, 01:33:11 UTC |
| af7bb8e | Neal Norwitz | 25 March 2006, 21:27:08 UTC | Backport of rev 43312: SF bug # 1457358 and patch # 1458419, floor division not documented. Patch by Andy. | 25 March 2006, 21:27:08 UTC |
| 1fcf87f | Neal Norwitz | 24 March 2006, 05:41:48 UTC | Backport: SF bug #1457411, fix errors using variables that don't exist. Rename file -> filename to be clear. | 24 March 2006, 05:41:48 UTC |
| 692b4a1 | Martin v. Löwis | 23 March 2006, 19:16:48 UTC | Add 2.4.3 UUIDs. | 23 March 2006, 19:16:48 UTC |
| a6c0c20 | Tim Peters | 23 March 2006, 19:07:46 UTC | Mechanical whitespace normalization. | 23 March 2006, 19:07:46 UTC |
| e5fc774 | Anthony Baxter | 23 March 2006, 02:50:34 UTC | 2.4.3c1 | 23 March 2006, 02:50:34 UTC |
| 73cdff4 | Anthony Baxter | 23 March 2006, 02:49:35 UTC | take 3 | 23 March 2006, 02:49:35 UTC |
| e311fdc | Anthony Baxter | 23 March 2006, 02:49:17 UTC | aarrrrghhh | 23 March 2006, 02:49:17 UTC |
| 65fa3a0 | Anthony Baxter | 23 March 2006, 02:26:32 UTC | 2.4.3c1 | 23 March 2006, 02:26:32 UTC |
| 022a55c | Anthony Baxter | 23 March 2006, 02:26:07 UTC | update - still some old .cvsignore files lying around | 23 March 2006, 02:26:07 UTC |
| 49cbb50 | Anthony Baxter | 23 March 2006, 02:25:51 UTC | take2... | 23 March 2006, 02:25:51 UTC |
| 0e433a7 | Anthony Baxter | 23 March 2006, 02:22:32 UTC | Tagging for 2.4.3c1 | 23 March 2006, 02:22:32 UTC |
| c52b4ad | Anthony Baxter | 23 March 2006, 02:21:49 UTC | update svn:ignore across the board | 23 March 2006, 02:21:49 UTC |
| e33da3b | Anthony Baxter | 23 March 2006, 02:20:57 UTC | 2.4.3c1 | 23 March 2006, 02:20:57 UTC |
| a87b549 | Anthony Baxter | 23 March 2006, 02:13:43 UTC | 2.4.3c1 | 23 March 2006, 02:13:43 UTC |
| 73c8c2c | Martin v. Löwis | 22 March 2006, 13:55:50 UTC | Change SystemError into SyntaxError, when a Unicode string containing an encoding declaration is compile()d. Fixes #1115379. | 22 March 2006, 13:55:50 UTC |
| 88fd500 | Anthony Baxter | 22 March 2006, 04:00:19 UTC | 2.4.3c1 preparations | 22 March 2006, 04:00:19 UTC |
| 2f58855 | Armin Rigo | 20 March 2006, 09:37:16 UTC | Backport of r41629: hotshot generated half-meaningless numbers since the introduction of C calls profiling events. | 20 March 2006, 09:37:16 UTC |
| 2c20676 | Vinay Sajip | 20 March 2006, 09:22:40 UTC | Catch situations where currentframe() returns None. See SF patch #1447410, this is a different implementation. | 20 March 2006, 09:22:40 UTC |
| 099e793 | Tim Peters | 20 March 2006, 06:06:07 UTC | The new fetch_data_files.py downloads all the input data files used by encoding tests. Fiddled the Windows buildbot helper scripts to invoke this if needed. Note that this isn't needed on the trunk (the encoding tests download input files automatically in 2.5). | 20 March 2006, 06:06:07 UTC |
| 0ca2212 | Tim Peters | 20 March 2006, 02:12:05 UTC | Merge revs 42842, 42844, and part of a whitespace normalization patch from the trunk. This stops test_socket_ssl from dying with: TypeError: 'NoneType' object is not callable in urlparsre.py's urljoin() when the tests are run in this order: test_??? test_importhooks test_socket_ssl "test_???" can be various things, but must be there. For example, test_urllibnet "works" to provoke the failure. Alas, nobody actually understands _why_ test_socket_ssl fails then, or why this hack makes the problem go away. Amazingly, the tests just happened to run in the right order on the 2.4 branch on two Windows buildbot slaves today, causing them both to fail their most recent test runs before this patch. | 20 March 2006, 02:12:05 UTC |
| 5a8605e | Neal Norwitz | 20 March 2006, 02:05:58 UTC | Backport 43147: Fix problem spotted by Coverity that occurs if tzinfo.tzname().replace() returns a non-string when converting %Z. | 20 March 2006, 02:05:58 UTC |
| 210256e | Kurt B. Kaiser | 19 March 2006, 22:12:03 UTC | Source file f.flush() after writing; trying to avoid lossage if user kills GUI. Report from Bruce Sherwood. | 19 March 2006, 22:12:03 UTC |
| c612a5d | Tim Peters | 18 March 2006, 02:47:38 UTC | Merge rev 43091 from the trunk. """ Try to find a host that responds slower from python.org so this test does not fail on macteagle (G4 OSX.4 in buildbot) """ Since testConnectTimeout() frequently fails in the same way in 2.4 branch, and this patch seems to have fixed it on the trunk, it should fix it on the 2.4 branch too. | 18 March 2006, 02:47:38 UTC |
| d9ca66f | Georg Brandl | 17 March 2006, 19:04:15 UTC | Backport: Fix missing NULL checks after PyTuple_New, PyList_New, PyDict_New | 17 March 2006, 19:04:15 UTC |
| d1442b4 | Tim Peters | 16 March 2006, 18:56:35 UTC | Merge rev 43081 from trunk. Set eol-style to native. | 16 March 2006, 18:56:35 UTC |
| d82f391 | Tim Peters | 16 March 2006, 18:52:17 UTC | Merge rev 43079 from the trunk. This should allow test_winsound to pass on a box without a sound card. Update test_winsound to check for a configured sound card (using a VBScript helper written by Roger Upole and Mark Hammond) and adjust the expected PlaySoundTest case results accordingly. | 16 March 2006, 18:52:17 UTC |
| 03de6fd | Tim Peters | 16 March 2006, 02:33:55 UTC | Merge rev 43063 from trunk. Oops! Use python_d.exe _before_ it's destroyed :-) | 16 March 2006, 02:33:55 UTC |
| b6b7a3a | Tim Peters | 16 March 2006, 01:56:34 UTC | Merge rev 43061 from the trunk. Change the Windows buildbot "clean" step to remove stale .pyc files. | 16 March 2006, 01:56:34 UTC |
| 854e918 | Neal Norwitz | 15 March 2006, 05:44:35 UTC | Backport 43037: Move test code out of xxmodule and into _testcapimodule.c where it belongs. | 15 March 2006, 05:44:35 UTC |
| 8a2bffc | Neal Norwitz | 14 March 2006, 06:47:07 UTC | Backport 42932: Try to be a bit more consistent on all platforms: python . python < . both print a message, return non-zero and do not core dump. This hopefully fixes the failure on Solaris. | 14 March 2006, 06:47:07 UTC |
| 9165680 | Neal Norwitz | 14 March 2006, 06:10:22 UTC | Backport 43022: Fix and test (manually w/xx module) passing NULLs to PyObject_Str() and PyObject_Unicode(). This problem was originally reported from Coverity and addresses mail on python-dev "checkin r43015". This inlines the conversion of the string to unicode and cleans up/simplifies some code at the end of the PyObject_Unicode(). We really need a complete C API test module for all public APIs and passing good and bad parameter values. | 14 March 2006, 06:10:22 UTC |
| 2f014f8 | Georg Brandl | 13 March 2006, 22:22:15 UTC | Fix bug found by Coverity: don't allow NULL argument to PyUnicode_CheckExact (backport from rev. 43014) | 13 March 2006, 22:22:15 UTC |
| 6b3cf5e | Martin v. Löwis | 13 March 2006, 13:08:41 UTC | Backport of 42994: Let the buildbot make a single pass in the test suite only. | 13 March 2006, 13:08:41 UTC |
| 25aeec5 | Martin v. Löwis | 13 March 2006, 13:06:46 UTC | Backport of 42993: Update to bzip2 1.0.3 Make buildbot slaves automatically fetch bzip2 1.0.3. | 13 March 2006, 13:06:46 UTC |
| b0c3bdd | Hye-Shik Chang | 13 March 2006, 10:24:31 UTC | Backport from trunk r42989: Bug #1448490: Fixed a bug that ISO-2022 codecs could not handle SS2 (single-shift 2) escape sequences correctly. | 13 March 2006, 10:24:31 UTC |
| ef63e9f | Tim Peters | 13 March 2006, 05:53:04 UTC | Merge rev 42963 from the trunk. When the new -w option (yay! great idea) reruns a failed test, first display the name of the test (else it's not always clear from the output which test is getting run). | 13 March 2006, 05:53:04 UTC |
| 1baff3c | Tim Peters | 13 March 2006, 04:50:34 UTC | Merge rev 41859 from the trunk. test_main(): Restore the original root logger level after running the tests. This stops the confusing/annoying: No handlers could be found for logger "cookielib" message we got whenever some test running after test_logging happened to use cookielib.py (when not using regrtest's -r, this happened during test_urllib2; when using -r, it varied). | 13 March 2006, 04:50:34 UTC |
| 9104048 | Tim Peters | 13 March 2006, 04:17:40 UTC | Merge rev 42984 from trunk. Adding the /useenv means that one's PATH actually gets through. This is important for the _ssl.vproj build. It calls build_ssl.py which tries to find a Perl to use. Without "/useenv" Visual Studio is getting a PATH from somewhere else (presumably from its internal environment configuration). The result is that build_ssl.py fallsback to its "well-known" locations for a Perl install. | 13 March 2006, 04:17:40 UTC |
| 7a0109f | Martin v. Löwis | 12 March 2006, 09:51:52 UTC | Move buildbot scripts to Tools/buildbot. | 12 March 2006, 09:51:52 UTC |
| 9eee616 | Tim Peters | 12 March 2006, 07:05:34 UTC | Add a do-nothing "clean" operation to the _ssl project, so the buildbot's "clean" step doesn't fail due to _ssl whining that it can't spawn an empty string. | 12 March 2006, 07:05:34 UTC |
| c8bbcc3 | Tim Peters | 12 March 2006, 07:01:31 UTC | Copied from the trunk for buildbot use. | 12 March 2006, 07:01:31 UTC |
| c7ddece | Tim Peters | 12 March 2006, 06:49:18 UTC | Copy from the trunk, for the buildbot. | 12 March 2006, 06:49:18 UTC |
| 8d51d55 | Tim Peters | 12 March 2006, 06:48:57 UTC | Copy from the trunk, for the buildbot. | 12 March 2006, 06:48:57 UTC |
| 82e8f38 | Tim Peters | 11 March 2006, 04:55:45 UTC | Merge rev 42904 from the trunk. _hotshot hotshot_profiler(): If write_header() returned an error code, this let `self` leak. This is a disaster on Windows, since `self` already points to a newly-opened file object, and it was impossible for Python code to close the thing since the only reference to it was in a blob of leaked C memory. test_hotshot test_bad_sys_path(): This new test provoked the C bug above. This test passed, but left an open "@test" file behind, which caused a massive cascade of bogus test failures in later, unrelated tests on Windows. Changed the test code to remove the @test file it leaves behind, which relies on the change above to close that file first. | 11 March 2006, 04:55:45 UTC |
| fea70fe | Martin v. Löwis | 10 March 2006, 22:56:20 UTC | Backport of 42551 (EXTRATESTOPTS). | 10 March 2006, 22:56:20 UTC |
| eed78ef | Martin v. Löwis | 10 March 2006, 22:15:48 UTC | Backport of 42960, to support the options buildbot uses. | 10 March 2006, 22:15:48 UTC |
| ea901a1 | Georg Brandl | 08 March 2006, 20:59:12 UTC | Bug #1445901: os.path.realpath() is available on Win/Mac too. (backport from rev. 42923) | 08 March 2006, 20:59:12 UTC |
| 865ddfb | Georg Brandl | 08 March 2006, 09:34:57 UTC | Fix pydoc.synopsis() so that it doesn't error out with an unreadable module. (backport from rev. 42912) | 08 March 2006, 09:34:57 UTC |
| 360641c | Neal Norwitz | 08 March 2006, 06:39:50 UTC | Backport fixes reported by Coverity. | 08 March 2006, 06:39:50 UTC |
| 5f2ad0b | Andrew M. Kuchling | 07 March 2006, 16:17:09 UTC | Typo fix | 07 March 2006, 16:17:09 UTC |
| 361cd4b | Hye-Shik Chang | 07 March 2006, 15:59:09 UTC | Backport r42894: SF #1444030 Fix several potential defects found by Coverity. | 07 March 2006, 15:59:09 UTC |
| 82735da | Thomas Wouters | 07 March 2006, 14:16:02 UTC | Backport trunk's r42890 (thomas.wouters): Coverity found bug: test result of PyTuple_New() against NULL before use. and r42891 (thomas.wouters): Fix gcc 4.0.x warning about use of uninitialized value. | 07 March 2006, 14:16:02 UTC |
| d704935 | Thomas Wouters | 07 March 2006, 14:06:31 UTC | Backport trunk's r42888 (thomas.wouters): Coverity found refleak: need to free 'v' after calling Tkinter_Error(). | 07 March 2006, 14:06:31 UTC |
| 4fbad67 | Thomas Wouters | 07 March 2006, 13:39:26 UTC | Backport trunk's r42885 (thomas.wouters): Coverity-found bug: don't use temp->next *before* checking it for NULL. Also return rather than use it again. | 07 March 2006, 13:39:26 UTC |
| e34fc70 | Thomas Wouters | 07 March 2006, 12:08:42 UTC | Backport trunk's r42878 (neal.norwitz): Thanks to Coverity, these were all reported by their Prevent tool. and r42881 (thomas.wouters): Don't DECREF a borrowed reference. | 07 March 2006, 12:08:42 UTC |
| a4b4dc3 | Walter Dörwald | 06 March 2006, 22:44:03 UTC | Backport revision 42872: If size is specified, try to read at least size characters. This is a alternative version of patch #1379332. | 06 March 2006, 22:44:03 UTC |
| 39be806 | Barry Warsaw | 06 March 2006, 00:55:25 UTC | Skip codecs tests on Python 2.3. | 06 March 2006, 00:55:25 UTC |
| a10a09d | Neal Norwitz | 03 March 2006, 23:13:19 UTC | Backport: SF #1442767, docs for os.statvfs miss f_bsize parameter. | 03 March 2006, 23:13:19 UTC |
| 518f2a5 | Martin v. Löwis | 01 March 2006, 21:36:32 UTC | Backport of memory leak fixes. | 01 March 2006, 21:36:32 UTC |
| 68e4cbb | Brett Cannon | 27 February 2006, 23:15:56 UTC | Return value off PyErr_Warn() for raising string exceptions was not being checked. Problem when 'warnings' was set to "error" and thus would re-raise a new exception. | 27 February 2006, 23:15:56 UTC |
| 9e777e5 | Tim Peters | 27 February 2006, 17:49:41 UTC | do_mkvalue(): Squash compiler warnings about mixing signed and unsigned types in comparison. | 27 February 2006, 17:49:41 UTC |
| 91e742f | Tim Peters | 27 February 2006, 17:47:02 UTC | Merge rev 42607 from the trunk. Patch 1413181, by Gabriel Becedillas. PyThreadState_Delete(): if the auto-GIL-state machinery knows about the thread state, forget it (since the thread state is being deleted, continuing to remember it can't help, but can hurt if another thread happens to get created with the same thread id). | 27 February 2006, 17:47:02 UTC |
| 05b9254 | Tim Peters | 27 February 2006, 17:22:16 UTC | Trimmed trailing whitespace. | 27 February 2006, 17:22:16 UTC |
| aefbf72 | Skip Montanaro | 25 February 2006, 21:04:03 UTC | backport | 25 February 2006, 21:04:03 UTC |
| 5f5124e | Neal Norwitz | 21 February 2006, 09:19:45 UTC | Backport 41753: Bug #1378022, UTF-8 files with a leading BOM crashed the interpreter. Also bug #1435487 (dup). | 21 February 2006, 09:19:45 UTC |
| e8cebba | Georg Brandl | 20 February 2006, 21:26:29 UTC | Bug #947571: By default, urllib.urlopen() does not raise IOErrors on server error codes. Make this clear. | 20 February 2006, 21:26:29 UTC |
| fea6059 | Georg Brandl | 20 February 2006, 20:29:56 UTC | Bug #1210377: close bsddb cursor correctly after NotFoundError. | 20 February 2006, 20:29:56 UTC |
| e5db6b2 | Georg Brandl | 20 February 2006, 17:37:39 UTC | Patch #931938: prevent setting sys.prefix to "" | 20 February 2006, 17:37:39 UTC |
| f1a408d | Georg Brandl | 20 February 2006, 13:12:07 UTC | Bug #1102649: add a note to pickle docs that pickle files should be opened in binary mode. | 20 February 2006, 13:12:07 UTC |
| 9c65202 | Georg Brandl | 20 February 2006, 12:15:26 UTC | Bug #1323369: document possible values of sys.getwindowsversion() platform field | 20 February 2006, 12:15:26 UTC |
| f07e42c | Georg Brandl | 20 February 2006, 10:32:05 UTC | Bug #1101233: fix test_fcntl on netbsd2 platform. | 20 February 2006, 10:32:05 UTC |
| 0b61a45 | Armin Rigo | 20 February 2006, 10:21:14 UTC | Backported from r41842. | 20 February 2006, 10:21:14 UTC |
| 2fc63f7 | Georg Brandl | 20 February 2006, 09:42:37 UTC | Bug #854823: socketmodule now builds on Sun platforms even when INET_ADDRSTRLEN is not defined. | 20 February 2006, 09:42:37 UTC |
| fc10228 | Georg Brandl | 19 February 2006, 13:22:58 UTC | Make clear that patch #1336582 wasn't a new feature. | 19 February 2006, 13:22:58 UTC |
| 044d2ce | Georg Brandl | 19 February 2006, 09:51:33 UTC | Patch #1337756: fileinput now accepts Unicode filenames. | 19 February 2006, 09:51:33 UTC |
| 4655481 | Georg Brandl | 19 February 2006, 09:39:02 UTC | Patch #1352711: make zipimport raise a complete IOError | 19 February 2006, 09:39:02 UTC |
| 394e4af | Georg Brandl | 19 February 2006, 00:12:46 UTC | Bug #801349: document that start/stop/step slice arguments can be None | 19 February 2006, 00:12:46 UTC |
| 2aaf918 | Georg Brandl | 18 February 2006, 22:56:02 UTC | Patch #1415507: clarify docs on reference stealing | 18 February 2006, 22:56:02 UTC |
| dc2a854 | Georg Brandl | 18 February 2006, 21:57:28 UTC | Bug #1366000: cleanup BZ2File.seek() logic. Fixes the case of whence=2, offset>=0. | 18 February 2006, 21:57:28 UTC |
| f3cba11 | Neal Norwitz | 18 February 2006, 21:25:56 UTC | Backport: This should get test_timeout to pass when running on any python.org host. | 18 February 2006, 21:25:56 UTC |
| b02be7e | Georg Brandl | 18 February 2006, 21:10:59 UTC | Patch #1373643: The chunk module can now read chunks larger than two gigabytes. | 18 February 2006, 21:10:59 UTC |
| f1fadcc | Georg Brandl | 17 February 2006, 11:51:20 UTC | Remove soundex from "undocumented modules". | 17 February 2006, 11:51:20 UTC |
| a957e8f | Georg Brandl | 17 February 2006, 09:53:00 UTC | Bug #1430298: It is now possible to send a mail with an empty return address using smtplib. | 17 February 2006, 09:53:00 UTC |
| d057c91 | Georg Brandl | 17 February 2006, 09:47:47 UTC | Bug #1432260: better handle lambda functions' names in pydoc | 17 February 2006, 09:47:47 UTC |
| fbf5ff3 | Martin v. Löwis | 17 February 2006, 08:39:55 UTC | Patch #1432345: Make python compile on DragonFly. | 17 February 2006, 08:39:55 UTC |
