1 Mon Apr 25 07:29:14 IST 2005 Abhijit Menon-Sen <ams@wiw.org>
5 1. Store weak references
6 2. Add STORABLE_attach hook.
8 Thu Jun 17 12:26:43 BST 2004 Nicholas Clark <nick@ccl4.org>
12 1. Don't change the type of top level overloaded references to RV -
13 they are perfectly correct as PVMG
14 2. Storable needs to cope with incoming frozen data that happens to be
17 Wed Mar 17 15:40:29 GMT 2004 Nicholas Clark <nick@ccl4.org>
21 1. Add regression tests for the auto-require of STORABLE_thaw
22 2. Add auto-require of modules to restore overloading (and tests)
23 3. Change to no context (should give speedup with ithreads)
25 Sat Mar 13 20:11:03 GMT 2004 Nicholas Clark <nick@ccl4.org>
29 1. Storing restricted hashes in canonical order would SEGV. Fixed.
30 2. It was impossible to retrieve references to PL_sv_no and and
31 PL_sv_undef from STORABLE_thaw hooks.
32 3. restrict.t was failing on 5.8.0, due to 5.8.0's unique
33 implementation of restricted hashes using PL_sv_undef
34 4. These changes allow a space optimisation for restricted hashes.
36 Sat Jan 24 16:22:32 IST 2004 Abhijit Menon-Sen <ams@wiw.org>
40 1. Thread safety: Storable::CLONE/init_perlinterp() now create
41 a new Perl context for each new ithread.
42 (From Stas Bekman and Jan Dubois.)
43 2. Fix a tag count mismatch with $Storable::Deparse that caused
44 all back-references after a stored sub to be off-by-N (where
45 N was the number of code references in between).
47 3. Prevent CODE references from turning into SCALAR references.
50 Sat Jan 3 18:49:18 GMT 2004 Nicholas Clark <nick@ccl4.org>
54 Fix minor problems with the CPAN release
55 1: Make Storable.xs work on 5.8.2 and later (already in the core)
56 2: Ship the linux hints file
57 3: Ship Test::More for the benefit of Perls pre 5.6.2
58 4: Correct Makefile.PL to only install in core for 5.8.0 and later
60 Sat Sep 6 01:08:20 IST 2003 Abhijit Menon-Sen <ams@wiw.org>
64 This release works around a 5.8.0 bug which caused hashes to not
65 be marked as having key flags even though an HEK had HEK_WASUTF8
66 set. (Note that the only reasonable solution is to silently drop
67 the flag from the affected key.)
69 Users of RT 3 who were seeing assertion failures should upgrade.
70 (Perl 5.8.1 will have the bug fixed.)
72 Mon May 5 10:24:16 IST 2003 Abhijit Menon-Sen <ams@wiw.org>
76 Minor bugfixes (self-tied objects are now correctly stored, as
77 are the results of additions larger than INT_MAX).
79 Mon Oct 7 21:56:38 BST 2002 Nicholas Clark <nick@ccl4.org>
83 Remove qr// from t/downgrade.t so that it will run on 5.004
84 Mention $File::Spec::VERSION a second time in t/forgive.t so that it
85 runs without warnings in 5.004 (this may be a 5.00405 bug I'm working
87 Fix t/integer.t initialisation to actually generate 64 bits of 9c
88 Fix comparison tests to use eval to get around 64 bit IV conversion
89 issues on 5.6.x, following my t/integer.t ^ precedence bug found by
91 Alter t/malice.t to work with Test/More.pm in t/, and skip individual
92 subtests that use $Config{ptrsize}, so that the rest of the test can
94 Change t/malice.t and the error message in check_magic in Storable.xs
95 from "Pointer integer size" to "Pointer size"
96 Remove prerequisite of Test::More from Makefile.PL
97 Ship Test::Builder, Test::Simple and Test::More in t
99 Thu Oct 3 08:57:22 IST 2002 Abhijit Menon-Sen <ams@wiw.org>
103 Adds support for CODE references from Slaven Rezic
104 <slaven.rezic@berlin.de>.
106 Fri Jun 7 23:55:41 BST 2002 Nicholas Clark
110 Bug fix from Radu Greab <radu@netsoft.ro> (plus regression test)
111 to fix a recently introduced bug detected by Dave Rolsky.
112 Bug was that for a non threaded build, the class information was
113 being lost at freeze time on the first object with a STORABLE_freeze
114 hook. Consequentially the object was not blessed at all when thawed.
115 (The presence (or lack) of STORABLE_thaw was irrelevant; this was
116 a store-time data lost bug, caused by failure to initialize internal
118 The bug was introduced as development perl change 16442 (on
119 2002/05/07), so has been present since 2.00.
120 Patches to introduce more regression tests to reduce the chance of
121 a reoccurance of this sort of goof are always welcome.
123 Thu May 30 20:31:08 BST 2002 Nicholas Clark <nick@ccl4.org>
125 Version 2.03 Header changes on 5.6.x on Unix where IV is long long
127 5.6.x introduced the ability to have IVs as long long. However,
128 Configure still defined BYTEORDER based on the size of a long.
129 Storable uses the BYTEORDER value as part of the header, but
130 doesn't explicity store sizeof(IV) anywhere in the header.
131 Hence on 5.6.x built with IV as long long on a platform that
132 uses Configure (ie most things except VMS and Windows) headers
133 are identical for the different IV sizes, despite the files
134 containing some fields based on sizeof(IV)
136 5.8.0 is consistent; all platforms have BYTEORDER in config.h
137 based on sizeof(IV) rather than sizeof(long). This means that
138 the value of BYTEORDER will change from (say) 4321 to 87654321
139 between 5.6.1 and 5.8.0 built with the same options to Configure
140 on the same machine. This means that the Storable header will
141 differ, and the two versions will wrongly thing that they are
144 For the benefit of long term consistency, Storable now
145 implements the 5.8.0 BYTEORDER policy on 5.6.x. This means that
146 2.03 onwards default to be incompatible with 2.02 and earlier
147 (ie the large 1.0.x installed base) on the same 5.6.x perl.
149 To allow interworking, a new variable
150 $Storable::interwork_56_64bit is introduced. It defaults to
151 false. Set it to true to read and write old format files. Don't
152 use it unless you have existing stored data written with 5.6.x
153 that you couldn't otherwise read, or you need to interwork with
154 a machine running older Storable on a 5.6.x with long long IVs
155 (i.e., you probably don't need to use it).
157 Sat May 25 22:38:39 BST 2002 Nicholas Clark <nick@ccl4.org>
161 Rewrite Storable.xs so that the file header structure for write_magic
162 is built at compile time, and check_magic attempts to the header in
163 blocks rather than byte per byte. These changes make the compiled
164 extension 2.25% smaller, but are not significant enough to give a
167 Thu May 23 22:50:41 BST 2002 Nicholas Clark <nick@ccl4.org>
171 - New regression tests integer.t
172 - Add code to safely store large unsigned integers.
173 - Change code not to attempt to store large integers (ie > 32 bits)
174 in network order as 32 bits.
176 *Never* underestimate the value of a pathological test suite carefully
177 crafted with maximum malice before writing a line of real code. It
178 prevents crafty bugs from stowing away in your released code.
179 It's much less embarrassing to find them before you ship.
180 (Well, never underestimate it if you ever want to work for me)
182 Fri May 17 22:48:59 BST 2002 Nicholas Clark <nick@ccl4.org>
184 Version 2.0, binary format 2.5 (but writes format 2.4 on pre 5.7.3)
186 The perl5 porters have decided to make sure that Storable still
187 builds on pre-5.8 perls, and make the 5.8 version available on CPAN.
188 The VERSION is now 2.0, and it passes all tests on 5.005_03, 5.6.1
189 and 5.6.1 with threads. On 5.6.0 t/downgrade.t fails tests 34 and 37,
190 due to a bug in 5.6.0 - upgrade to 5.6.1.
192 Jarkko and I have collated the list of changes the perl5 porters have
193 from the perl5 Changes file:
195 - data features of upcoming perl 5.8.0 are supported: Unicode hash
196 keys (Unicode hash values have been supported since Storable 1.0.1)
197 and "restricted hashes" (readonly hashes and hash entries)
198 - a newer version of perl can now be used to serialize data which is
199 not supported in earlier perls: Storable will attempt to do the
200 right thing for as long as possible, croaking only when safe data
201 conversion simply isn't possible. Alternatively earlier perls can
202 opt to have a lossy downgrade data instead of croaking
203 - when built with perls pre 5.7.3 this Storable writes out files
204 with binary format 2.4, the same format as Storable 1.0.8 onwards.
205 This should mean that this Storable will inter-operate seamlessly
206 with any Storable 1.0.8 or newer on perls pre 5.7.3
207 - dclone() now works with empty string scalar objects
208 - retrieving of large hashes is now more efficient
209 - more routines autosplit out of the main module, so Storable should
210 load slightly more quickly
211 - better documentation
212 - the internal context objects are now freed explicitly, rather than
213 relying on thread or process exit
214 - bugs fixed in debugging trace code affecting builds made with 64 bit
216 - code tidy-ups to allow clean compiles with more warning options
217 turned on avoid problems with $@ getting corrupted on 5.005_03 if
218 Carp wasn't already loaded
219 - added &show_file_magic, so you can add to /etc/magic and teach
220 Unix's file command about Storable files
222 We plan to keep Storable on CPAN in sync with the Perl core, so
223 if you encounter bugs or other problems building or using Storable,
224 please let us know at perl5-porters@perl.org
227 Sat Dec 1 14:37:54 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
229 This is the LAST maintenance release of the Storable module.
230 Indeed, Storable is now part of perl 5.8, and will be maintained
231 as part of Perl. The CPAN module will remain available there
232 for people running pre-5.8 perls.
234 Avoid requiring Fcntl upfront, useful to embedded runtimes.
235 Use an eval {} for testing, instead of making Storable.pm
236 simply fail its compilation in the BEGIN block.
238 store_fd() will now correctly autoflush file if needed.
240 Tue Aug 28 23:53:20 MEST 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
242 Fixed truncation race with lock_retrieve() in lock_store().
243 The file has to be truncated only once the exclusive lock is held.
245 Removed spurious debugging messages in .xs file.
247 Sun Jul 1 13:27:32 MEST 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
249 Systematically use "=over 4" for POD linters.
250 Apparently, POD linters are much stricter than would
251 otherwise be needed, but that's OK.
253 Fixed memory corruption on croaks during thaw(). Thanks
254 to Claudio Garcia for reproducing this bug and providing the
255 code to exercise it. Added test cases for this bug, adapted
258 Made code compile cleanly with -Wall (from Jarkko Hietaniemi).
260 Changed tagnum and classnum from I32 to IV in context. Also
263 Thu Mar 15 01:22:32 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
265 Last version was wrongly compiling with assertions on, due
266 to an edit glitch. That did not cause any problem (apart from
267 a slight performance loss) excepted on Win* platforms, where the
268 assertion code does not compile.
270 Sat Feb 17 13:37:37 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
274 Forgot to increase version number at previous patch (there were
275 two of them, which is why we jump from 1.0.8 to 1.0.10).
277 Sat Feb 17 13:35:00 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
279 Version 1.0.8, binary format 2.4.
281 Fixed incorrect error message.
283 Now bless objects ASAP at retrieve time, which is meant to fix
286 * Indirect references to overloaded object were not able to
287 restore overloading if the object was not blessed yet,
288 which was possible since blessing occurred only after the
291 * Storable hooks asking for serialization of blessed ref could
292 get un-blessed refs at retrieval time, for the very same
295 The fix implemented here was suggested by Nick Ing-Simmons.
297 Added support for blessed ref to tied structures. This is the
298 cause for the binary format change.
300 Added EBCDIC version of the compatibility test with 0.6.11,
303 Added tests for the new features, and to make sure the bugs they
304 are meant to fix are indeed fixed.
306 Wed Jan 3 10:43:18 MET 2001 Raphael Manfredi <Raphael_Manfredi@pobox.com>
308 Removed spurious 'clean' entry in Makefile.PL.
310 Added CAN_FLOCK to determine whether we can flock() or not,
311 by inspecting Perl's configuration parameters, as determined
314 Trace offending package when overloading cannot be restored
317 Made context cleanup safer to avoid dup freeing, mostly in the
318 presence of repeated exceptions during store/retrieve (which can
319 cause memory leaks anyway, so it's just additional safety, not a
322 Sun Nov 5 18:23:48 MET 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
326 Fixed severe "object lost" bug for STORABLE_freeze returns,
327 when refs to lexicals, taken within the hook, were to be
328 serialized by Storable. Enhanced the t/recurse.t test to
329 stress hook a little more with refs to lexicals.
331 Thu Oct 26 19:14:38 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
335 Documented that store() and retrieve() can return undef.
336 That is, the error reporting is not always made via exceptions,
337 as the paragraph on error reporting was implying.
339 Auto requires module of blessed ref when STORABLE_thaw misses.
340 When the Storable engine looks for the STORABLE_thaw hook and
341 does not find it, it now tries to require the package into which
342 the blessed reference is.
344 Just check $^O, in t/lock.t: there's no need to pull the whole
345 Config module for that.
347 Mon Oct 23 20:03:49 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
351 Protected calls to flock() for DOS platform: apparently, the
352 flock/fcnlt emulation is reported to be broken on that
355 Added logcarp emulation if they don't have Log::Agent, since
356 we now use it to carp when lock_store/lock_retrieve is used
359 Fri Sep 29 21:52:29 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
363 Avoid using "tainted" and "dirty" since Perl remaps them via
364 cpp (i.e. #define). This is deeply harmful when threading
365 is enabled. This concerned both the context structure and
366 local variable and argument names. Brrr..., scary!
368 Thu Sep 28 23:46:39 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
372 Fixed spelling in README.
374 Added lock_store, lock_nstore, and lock_retrieve (advisory locking)
375 after a proposal from Erik Haugan <erik@solbors.no>.
377 Perls before 5.004_04 lack newSVpvn, added remapping in XS.
379 Fixed stupid typo in the t/utf8.t test.
381 Sun Sep 17 18:51:10 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
383 Version 1.0.1, binary format 2.3.
385 Documented that doubles are stored stringified by nstore().
387 Added Salvador Ortiz Garcia in CREDITS section, He identified
388 a bug in the store hooks and proposed the right fix: the class
389 id was allocated too soon. His bug case was also added to
390 the regression test suite.
392 Now only taint retrieved data when source was tainted. A bug
393 discovered by Marc Lehmann.
395 Added support for UTF-8 strings, a contribution of Marc Lehmann.
396 This is normally only activated in post-5.6 perls.
398 Thu Aug 31 23:06:06 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
400 First official release Storable 1.0, for inclusion in perl 5.7.0.
401 The license scheme is now compatible with Perl's.
403 Thu Aug 24 01:02:02 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
405 ANSI-fied most of the code, preparing for Perl core integration.
406 The next version of Storable will be 0.8, and will be integrated
407 into the Perl core (development branch).
409 Dispatch tables were moved upfront to relieve some compilers,
410 especially on AIX and Windows platforms.
412 Merged 64-bit fixes from perl5-porters.
414 Mon Aug 14 09:22:04 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
416 Added a refcnt dec in retrieve_tied_key(): sv_magic() increases
417 the refcnt on the mg_ptr as well.
419 Removed spurious dependency to Devel::Peek, which was used for
420 testing only in t/tied_items.t. Thanks to Conrad Heiney
421 <conrad@fringehead.org> for spotting it first.
423 Sun Aug 13 22:12:59 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
425 Marc Lehmann kindly contributed code to add overloading support
426 and to handle references to tied variables.
428 Rewrote leading blurb about compatibility to make it clearer what
429 "backward compatibility" is about: when I say 0.7 is backward
430 compatible with 0.6, it means the revision 0.7 can read files
433 Mention new Clone(3) extension in SEE ALSO.
435 Was wrongly optimizing for "undef" values in hashes by not
436 fully recursing: as a result, tied "undef" values were incorrectly
439 Sun Jul 30 12:59:17 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
441 First revision of Storable 0.7.
443 The serializing format is new, known as version 2.0. It is fully
444 backward compatible with 0.6. Earlier formats are deprecated and
445 have not even been tested: next version will drop pre-0.6 format.
447 Changes since 0.6@11:
449 - Moved interface to the "beta" status. Some tiny parts are still
450 subject to change, but nothing important enough to warrant an "alpha"
453 - Slightly reduced the size of the Storable image by factorizing
454 object class names and removing final object storage notification due
455 to a redesign of the blessed object storing.
457 - Classes can now redefine how they wish their instances to be serialized
458 and/or deep cloned. Serializing hooks are written in Perl code.
460 - The engine is now fully re-entrant.
462 Sun Apr 2 23:47:50 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
464 Added provision to detect more recent binary formats, since
465 the new upcoming Storable-0.7 will use a different format.
466 In order to prevent attempting the de-serialization of newer
467 formats by older versions, I'm adding this now to the 0.6 series.
469 I'm expecting this revision to be the last of the 0.6 series.
470 Unless it does not work with perl 5.6, which I don't use yet,
471 and therefore against which I cannot test.
473 Wed Mar 29 19:55:21 MEST 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
475 Added note about format incompatibilities with old versions
476 (i.e. pre 0.5@9 formats, which cannot be understood as there
477 was no versionning information in the file by then).
479 Protect all $@ variables when eval {} used, to avoid corrupting
480 it when store/retrieve is called within an exception handler.
482 Mistakenly included "patchlevel.h" instead of <patchlevel.h>,
483 preventing Perl's patchlevel from being included, which is
484 needed starting from 5.6.
486 Tue May 12 09:15:15 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
488 Fixed shared "undef" bug in hashes, which did not remain shared
489 through store/retrieve.
491 Thu Feb 10 19:48:16 MET 2000 Raphael Manfredi <Raphael_Manfredi@pobox.com>
493 added last_op_in_netorder() predicate
494 documented last_op_in_netorder()
495 added tests for the new last_op_in_netorder() predicate
497 Wed Oct 20 19:07:36 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
499 Forgot to update VERSION
501 Tue Oct 19 21:25:02 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
503 Added mention of japanese translation for the manual page.
505 Fixed typo in macro that made threaded code not compilable,
506 especially on Win32 platforms.
508 Changed detection of older perls (pre-5.005) by testing PATCHLEVEL
509 directly instead of relying on internal symbols.
511 Tue Sep 14 22:13:28 MEST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
513 Integrated "thread-safe" patch from Murray Nesbitt.
514 Note that this may not be very efficient for threaded code,
515 see comment in the code.
517 Try to avoid compilation warning on 64-bit CPUs. Can't test it,
518 since I don't have access to such machines.
520 Mon Jul 12 14:37:19 METDST 1999 Raphael Manfredi <Raphael_Manfredi@pobox.com>
522 changed my e-mail to pobox.
524 mentionned it is not thread-safe.
526 updated version number.
528 uses new internal PL_* naming convention.
530 Fri Jul 3 13:38:16 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
532 Updated benchmark figures due to recent optimizations done in
533 store(): tagnums are now stored as-is in the hash table, so
534 no surrounding SV is created. And the "shared keys" mode for
535 hash table was turned off.
537 Fixed backward compatibility (wrt 0.5@9) for retrieval of
538 blessed refs. That old version did something wrong, but the
539 bugfix prevented correct retrieval of the old format.
541 Mon Jun 22 11:00:48 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
543 Changed benchmark figures.
545 Adjust refcnt of tied objects after calling sv_magic() to avoid
546 memory leaks. Contributed by Jeff Gresham.
548 Fri Jun 12 11:50:04 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
550 Added workaround for persistent LVALUE-ness in perl5.004. All
551 scalars tagged as being an lvalue are handled as if they were
552 not an lvalue at all. Added test for that LVALUE bug workaround.
554 Now handles Perl immortal scalars explicitely, by storing &sv_yes
555 as such, explicitely.
557 Retrieval of non-immortal undef cannot be shared. Previous
558 version was over-optimizing by not creating a separate SV for
559 all undefined scalars seen.
561 Thu Jun 4 17:21:51 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
563 Baseline for Storable-0.6@0.
565 This version introduces a binary incompatibility in the generated
566 binary image, which is more compact than older ones by approximatively
567 15%, depending on the exact degree of sharing in your structures.
569 The good news is that your older images can still be retrieved with
570 this version, i.e. backward compatibility is preserved. This version
571 of Storable can only generate new binaries however.
573 Another good news is that the retrieval of data structure is
574 significantly quicker than before, because a Perl array is used
575 instead of a hash table to keep track of retrieved objects, and
576 also because the image being smaller, less I/O function calls are
579 Tue May 12 09:15:15 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
581 Version number now got from Storable.pm directly.
583 Fixed overzealous sv_type() optimization, which would make
584 Storable fail when faced with an "upgraded" SV to the PVIV
585 or PVNV kind containing a reference.
587 Thu Apr 30 15:11:30 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
589 Extended the SYNOPSIS section to give quick overview of the
590 routines and their signature.
592 Optimized sv_type() to avoid flags checking when not needed, i.e.
593 when their type makes it impossible for them to be refs or tied.
594 This slightly increases throughput by a few percents when refs
595 and tied variables are marginal occurrences in your data.
597 Stubs for XS now use OutputStream and InputStream file types to
598 make it work when the given file is actually a socket. Perl
599 makes a distinction for sockets in its internal I/O structures
600 by having both a read and a write structure, whereas plain files
603 Tue Jun 3 09:41:33 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
605 Thanks to a contribution from Benjamin A. Holzman, Storable is now
606 able to correctly serialize tied SVs, i.e. tied arrays, hashes
609 Thu Apr 9 18:07:51 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
611 I said SvPOK() had changed to SvPOKp(), but that was a lie...
613 Wed Apr 8 13:14:29 METDST 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
615 Wrote sizeof(SV *) instead of sizeof(I32) when portable, which
616 in effect mangled the object tags and prevented portability
617 accross 32/64 bit architectures!
619 Wed Mar 25 14:57:02 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
621 Added code example for store_fd() and retrieve_fd() in the
622 man page, to emphasize that file descriptors must be passed as
623 globs, not as plain strings.
625 Cannot use SV addresses as tag when using nstore() on LP64. This
626 was the cause of problems when creating a storable image on an
627 LP64 machine and retrieving it on an ILP32 system, which is
628 exactly what nstore() is meant for...
630 However, we continue to use SV addresses as tags for plain store(),
631 because benchamarking shows that it saves up to 8% of the store
632 time, and store() is meant to be fast at the expense of lack
635 This means there will be approximately an 8% degradation of
636 performance for nstore(), but it's now working as expected.
637 That cost may vary on your machine of course, since it is
638 solely caused by the memory allocation overhead used to create
639 unique SV tags for each distinct stored SV.
641 Tue Jan 20 09:21:53 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
643 Don't use any '_' in version number.
645 Tue Jan 13 17:51:50 MET 1998 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
647 Updated version number.
649 added binmode() calls for systems where it matters.
651 Be sure to pass globs, not plain file strings, to C routines,
652 so that Storable can be used under the Perl debugger.
654 Wed Nov 5 10:53:22 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
656 Fix memory leaks on seen hash table and returned SV refs.
658 Storable did not work properly when tainting enabled.
660 Fixed "Allocation too large" messages in freeze/thaw and added.
661 proper regression test in t/freeze.t.
663 Tue Jun 3 09:41:33 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
665 Updated version number
667 Added freeze/thaw interface and dclone.
669 Fri May 16 10:45:47 METDST 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
671 Forgot that AutoLoader does not export its own AUTOLOAD.
674 use AutoLoader 'AUTOLOAD';
676 but that would not be backward compatible. So the export is
679 Tue Mar 25 11:21:32 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
681 Empty scalar strings are now "defined" at retrieval time.
683 New test to ensure an empty string is defined when retrieved.
685 Thu Feb 27 16:32:44 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
687 Updated version number
689 Declare VERSION as being used
691 Fixed a typo in the PerlIO_putc remapping.
692 PerlIO_read and perlIO_write inverted size/nb_items.
693 (only relevant for pre-perl5.004 versions)
695 Thu Feb 27 15:58:31 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
697 Updated version number
699 Added VERSION identification
701 Allow build with perl5.003, which is ante perlIO time
703 Mon Jan 13 17:53:18 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
707 Wed Jan 22 15:19:56 MET 1997 Raphael Manfredi <Raphael_Manfredi@grenoble.hp.com>
709 Updated version number in Makefile.PL.
711 Added "thanks to" section to README.
713 Documented new forgive_me variable.
717 Added forgive_me support to allow store() of data structures
718 containing non-storable items like CODE refs.