1 Release: version 0.13 - 2008/04/??
2 ------------------------------------------------------------------------------
4 - removed TimeZones from TODO list, as it is done
10 - gui/src/BackupWindow.cc
14 - opensync-plugin/src/barry_sync.cc
15 - added GCC visibility support, limiting what symbols are visible
16 in the shared library, libbarry... this makes a drastic
17 reduction in the symbol table, and should avoid future
18 name clashes with other libraries such as opensync
19 See: http://gcc.gnu.org/wiki/Visibility
21 Release: version 0.12 - 2008/03/21
22 ------------------------------------------------------------------------------
24 - changed maintainer script to use CVS export instead of checkout,
25 to avoid the CVS directories
26 - added support for building Debian source packages with proper
28 - fixed find warning in maintainer/make-rpm.sh
29 - added support for Ubuntu binary packages to the maintainer build
32 - renamed all vformat_ and VFormat symbols in opensync plugin's
33 copy of vformat code, so they don't conflict with the
34 real opensync libraries
36 - updated the install.php web page with Paul Dugas's notes on
37 building RPM's from CVS
38 - fixed problem in tools/Makefile.am where bs11nread was included
39 in build, even if --with-boost was not specified
40 on ./configure command line
42 - changed format of Boost serialization file output from btool,
43 so that the database name is stored as the first line
45 - added tools/bs11nread.cc to read Boost serialization file data
47 - minor adjustments to r_message.h and r_timezone.h
48 - updated s11n-boost.h header to match record classes
50 - applied Time Zone parsing patches from Brian Edginton
51 - added Time Zones to list of supported parsers in btool.cc
53 - changed configure.ac to use bzip2 for "make dist".. thanks to
54 Paul Dugas for the tip
56 - committed Chris Burgess's update to the roadmap document
57 - applied Paul Dugas's x86-64 build patch, except the
59 - updated AUTHORS file
61 - fixed configure.ac's --with-boost option, so that without args,
62 it defaults to using --includedir and --libdir settings,
63 instead of hardcoding /usr/include and /usr/lib... this
64 should allow smarter compiling on 64 bit systems
65 - updated RPM spec file to use new --with-boost behaviour
66 - ahem, fixed configure.ac again... don't use --includedir and
67 --libdir, as those are install targets... oops.
68 Don't add path overrides at all if not specified.
70 - fixed locale issues in src/data.cc
71 - updated copyright statments for 2008
73 - applied wording correction patch to btool from Matt LaPlante
75 - changed hard coded databases names in btool to calls to
76 <recordclass>::GetDBName()
81 - gui/src/BackupWindow.cc
85 - opensync-plugin/src/barry_sync.cc
88 Release: version 0.11 - 2007/12/01
89 ------------------------------------------------------------------------------
91 - fixed null pointer bug in Probe class (bug #1842407)
92 - added pause to maintainer release script
98 - gui/src/BackupWindow.cc
102 - opensync-plugin/src/barry_sync.cc
104 Release: version 0.10 - 2007/11/30
105 ------------------------------------------------------------------------------
107 - added support to the usbwrap Match class to limit its
108 search by busname / devname
109 - added support to the Probe class to limit its search for
110 Blackberry devices to a given bus or device name
111 - added command line options to btool for the above Probe
112 bus/device names, and updated the man page
113 - updated the www docs with regard to the new usb_storage
114 behaviour and the new blacklist files... less
115 concern about berry_charge and usb_storage conflicts
116 - added bidentify tool and accompanying man page
117 - fixed bug in rpm maintenance script that used the wrong
118 spec file for opensuse
120 - added better error checking to breset
121 - commented the sleep() for the 8830
122 - added code to handle the race condition where usb_storage
123 claims the device before bcharge arrives on the scene,
124 by asking the kernel to detach the driver from
125 the Mass Storage interface... this change makes it
126 possible to run bcharge, moving from 0006 to 0004
127 to 0001 and back to 0004, all *while* usb_storage
129 - changed Probe class behaviour so that it will not throw an
130 exception on Device Busy (-EBUSY) but instead
131 log the error in an internal array and keep going.
132 This makes it possible to use btool and barrybackup
133 with two devices plugged in at the same time.
134 BACKWARD COMPATIBILITY NOTE: this changes initializaion
135 behaviour, but in practice, it should only
136 make applications look more capable, not less.
137 - updated btool to show errors logged by Probe
138 - added message to specify which PIN device is being used by btool
139 - added explanatory comments to udev rules files, about why
142 - added ClearHalt() calls to probe and controller classes, just
143 before any communication happens... the 8830 seems
146 - added usbwrap.cc:GetConfiguration()
147 - added code to check whether the configuration is the one
148 we require, before calling SetConfiguration()... and not
149 to call it if unnecessary. This allows us to run
150 at the same time as usb_storage.
151 Thanks to Martin Owens for not letting this issue go. :-)
152 - added sleep to bcharge, before calling usb_reset()... the
153 Blackberry 8830 behaves like a Pearl device in that
154 it has mass storage support, but behaves like a Classic
155 in that it resets itself... the usb_reset() seems
156 to reset it back to 0006. The sleep allows the device
157 to disappear via its own reset, and let usb_reset()
158 fail naturally on these devices.
159 This is not an ideal solution... mainly for testing.
161 - added modprobe/blacklist-berry_charge as per Simon Ruggier's
162 suggestion. Thanks also to Niels de Vos for a similar
164 - fixed udev rules to work on both old and new distros
165 Thanks to Duncan Mak for the bug report.
167 - fixed string comparison to use case insensitive compare in
169 - updated Debian package to include modprobe blacklist file
170 - updated rpm spec file to include modprobe blacklist file
171 - split the maintenance scripts into release and test, so
172 that release package builds can be done faster
174 - opensync plugin fixes:
175 - fixed memory leak in vBase::GetAttr()
176 - added support in vformat parser to parse "blocks"
177 and keep track of the hierarchy when
178 parsing an iCalendar, etc.
179 - added support for searching for attributes only
180 inside certain blocks (i.e. find first
181 DTSTART in VEVENT block, not VTIMEZONE block)
183 - fixed bug in Calendar::Clear, missing the Free / Class
184 member variables, leaving them uninitialized
185 - added examples/addcalendar.cc
186 - added more logging to the opensync plugin, logging the
187 ToBarry conversion results
188 - added code to force struct tm.tm_isdst to -1 before calling
190 - fixed typo in previous fix
192 - renamed ktrans and translate to bktrans and btranslate to avoid
193 file name clashes with libtranslate of Gnome
194 Closes bug #1812410 reported by Troy Engel.
199 - gui/src/BackupWindow.cc
203 - opensync-plugin/src/barry_sync.cc
204 - fixed missing default config file in the debian package for
206 Closes bug #1812224 reported by Ariel (adonatti).
207 - added debian dbg packages for tracing segfaults
208 - modified maintainer scripts so debug packages would reference
209 source code located in /usr/src/barry-0.9, etc.
211 Release: version 0.9 - 2007/10/12
212 ------------------------------------------------------------------------------
214 - adjusted autosuspend disabling code in bcharge, based on new
215 information on how kernels 2.6.21 / 2.6.22 work.
216 See bcharge.cc for comments. If device/power/level or
217 device/power/autosuspend exist, charging should now work
218 without recompiling the kernel.
219 - updated website docs with autosuspend info
221 - large update to the doc/www web pages, moving some of the
222 information in README files to html format, so that it
223 is commonly available on the web
224 - added step in maintenance scripts to automatically build the
225 .html files in doc/www when making a tarball release
227 - added libopensync 0.22 version requirement to barry.spec file,
228 but it doesn't seem to work on OpenSuSE 10.2, where it's
231 - updated README to mention new tree build configure options
232 - fixed the configure scripts and makefiles so that when building
233 with custom CXXFLAGS and LDFLAGS environment variables,
234 it doesn't break the full-tree build mode, when
235 using --enable-gui and/or --enable-opensync-plugin
236 - added config.h.in~ to buildgen.sh clean target
237 - converted Debian build scripts to use cdbs
238 - added opensync plugin binary package for Debian
239 - added "Getting Started" section to web pages, as well as an
240 explanatory document for various system requirements
241 - turned on opensync build in RPM maintainer scripts
243 - added filename to power/state error message in bcharge.cc
244 - adjusted maintainer scripts to automatically adjust barry.spec
245 for opensuse rpm builds
247 - fixed some of the release build scripts and added support for
248 automated opensuse 10.2 builds
249 - fixed build target to i386
250 - changed scripts to add to PKG_CONFIG_PATH variable instead
253 - added website docs to doc/www... these are out of date at the
254 moment, and hopefully by including them here, they
255 will be updated more frequently
257 - applied Simon Ruggier's build system patch that allowed a build
258 of all 3 components from the root directory and fixed the
259 configure scripts to support external build directories
260 - reordered clean commands in the buildgen.sh scripts, and removed
261 the INSTALL files, since they are auto-generated by autoconf
262 - removed extraneous buildgen.sh calls in maintainer scripts
263 - added Simon Ruggier to AUTHORS
264 - added explanatory note to 3 stage test build process in
267 - fixed Category handling in the Contact record class, to
268 parse the comma separated string into an array,
270 Note: this involves a change to the boost serialization
271 and will affect backward compatibility with
272 old serialization datafiles
273 - added Category to examples/addcontact.cc
274 - added Category support to the opensync plugin
276 - fixed wrong read/write in btool output (oops)
277 - added BadSize exception, and changed CheckSize() function
279 - changed backup GUI to catch BadSize exceptions on initial
280 connect, and try a reset if that error occurs, since
281 that likely means the device didn't shutdown properly
282 - fixed in opensync plugin:
283 - some vcard data contains types codes in capital letters,
284 so added lower case conversion routine
285 - changed the order of address lines, so that address1
286 is the street, address2 is the extended address,
287 and address3 is the PO Box. The Blackberry
288 doesn't seem to have any PO Box field.
290 - added -e endpoint override option to btool
291 - added status message to endpoint override
293 - added configure check for strnlen, and implementation for
294 systems that don't have it, or that have it implemented
295 incorrectly. Went a little paranoid here, since
296 AC_FUNC_STRNLEN doesn't set a define in config.h,
297 so I had to do it manually, and check in case autoconf
298 changes in the future... ugh. See src/strnlen.h for
300 - added code in probe.cc to fetch the device's description.
301 The description is now part of the probe results
302 along with the PIN. Updates to protocol.h, adding
303 a new attribute definition.
304 Thanks to Rick Scott for pointing out the code in
305 XmBlackberry where he implemented this first.
306 - fixed header dependency in record-internal.h
307 - updated the "required autoconf" version to 2.61 for root build
308 - moved nested structs out of the record classes that contain them,
309 in order to better support SWIG and Python
311 - changed bcharge to use getopt() for its argument processing
312 - added options -p and -s to bcharge, to adjust usb suspend
313 settings automatically where possible
314 - updated the udev rules to use the new bcharge settings
315 - updated the bcharge man page with the new arguments
317 - cleaned up Usb exception handling, moving libusb error code
318 processing into the Usb::Error exception class.
319 error.cc is now an empty shell
320 - added return codes to the Usb::Error exceptions
321 - removed aboutdialog.{h,cc} from the backup GUI since even
322 Debian stable now has up-to-date gtkmm libraries
323 - added more user friendly error message for the "device busy"
324 case, recommending 'rmmod usb_storage'.
326 - added better error checking to bcharge.cc
331 - gui/src/BackupWindow.cc
335 - opensync-plugin/src/barry_sync.cc
336 - removed usb_set_configuration() check from bcharge.cc, since
337 that may fail under normal operation, in this case
339 Release: version 0.8 - 2007/08/03
340 ------------------------------------------------------------------------------
342 - moved Interface implementation to usbwrap.cc file
343 - added dout debugging messages to usbwrap.cc
344 - clarified unhandled packet error message in socket.cc
345 - added temporary workaround in opensync-plugin environment's
346 Reconnect(), for odd message seen on newer Blackberry
348 - fixing issues with Brian Edginton's MessageRecord patch (2007/06/29)
349 - fixed missing size check in message oriented ParseHeader()
350 - fixed endian conversion issues (also found one pre-existing
351 endian bug in the process)
352 - moved message timestamp converter code to time.cc
353 - added Barry version number to opensync plugin name
354 - changed maintainer scripts and rpm.spec to use bzip2 instead of gzip
355 - added opensync-plugin to configure script build
356 - added -S option to btool, to list supported parsers and builders
357 - slight change to opensync-plugin/buildgen.sh to cleanup config.h.in~
358 - opensync-plugin/README update
359 - added opensync-plugin/patches for others who may be building and
360 testing the plugin with OpenSync 0.22
361 - added debian package build scripts to maintainer/
363 - added entry to doc/TroubleShooting.txt
365 - added doc/TroubleShooting.txt
368 - fixed bug in vcard.cc that was using the FN field
369 for the broken-down name instead of N
370 - added dbId and dbName to DatabaseSyncState
371 - added Reconnect() functionality to BarryEnvironment class
372 and moved connect logic out of barry_sync.cc
373 - changed the batch commit functionality in barry_sync.cc
374 to single-change-commit. Why? Because opensync
375 0.22 groups all batch changes into one group
376 and calls one batch commit function, even if you
377 registered two. Moving to single change commit
378 mode removes a chunk of unneccessary code
379 from the plugin as well, and can in theory
380 ease memory pressure in the library.
381 - added FinishSync() to perform actions needed on successful
383 - disconnect on successful sync completion, and reconnect,
384 since the device's dirty flags don't seem to
385 be updated until a disconnect occurs.
387 - added ReadDataArray() to data.h, using an istream instead of
389 - added raw data version of BuildField() to record-internal.h
390 This is so that btool doesn't add extra null terminators
391 to unknown data, when restoring with option -s
392 - fixed bugs in r_contact.cc:
393 - fixed bug that caused a LastName/FirstName swap
394 if FirstName was empty and you wrote a contact
395 record into the device.... First/Last Name
396 fields are position sensitive, and the code
397 was missing the empty FirstName case on writing
398 - fixed bug in BuildFields() and Dump() where they
399 were not updated to handle the new field
400 table added for the postal address changes
403 - changed the plugin config file, so it is possible to
404 turn debug output on/off without recompiling
405 - moved the standalone Map2Uid() function into
406 the DatabaseSyncState class, and changed the
407 code to produce "contact-#" and "calendar-#"
408 UID strings, instead of just numbers
409 - changed default PIN to -1, so that if the user doesn't
410 configure properly, it will fail... this is a
411 safety measure to avoid syncing with different
412 devices and corrupting the cache/idmap...
413 this should be fixed someday, to handle this
415 On second thought... is this right? Perhaps
416 it is better to fail if the PIN is not what
417 is expected, so a user doesn't accidentally
418 sync the wrong device and lose data.
419 Actions to take on new PIN:
420 - force slow_sync, or
422 - incorporated the new VCardConverter code into the
423 main barry_sync.cc plugin code
424 - fixed bug that did not clear the devices dirty flags
425 in the case where a sync only read data from it,
427 - changed PIN error message number to hex
429 - minor wording change in backup GUI prompt: "Backup working..." to
430 "Backup in progress..."
431 - renamed Contact class's Title field to JobTitle for clarity
432 - added record function ParseFieldString(), which properly handles
433 strings from the device that have multiple null terminators,
434 and changed the record parsing code to use this new function
435 - moved FullName() function from ldif.cc to r_contact.cc since it
436 is generally useful in areas outside of LDIF processing
437 - refactored postal addresses in the Contact record class into
438 a new class called PostalAddress
439 - renamed Address class in record.h to EmailAddress
440 - moved Contact::GetPostalAddress() to record.cc as part of the
441 new PostalAddress class and renamed it
442 PostalAddress::GetLabel()
444 - refactored vformat helpers into separate base class
446 - added vCard converter class
447 - commented out the code that skipped over the adding
448 of empty attribute values and parameters,
449 since VCARD depends on some of this behaviour
450 - movified vformat.c:vformat_find_attribute() to take
451 an extra nth argument, in order to search
452 for multiple attributes with the same name...
453 the way this is coded is a bit inefficient...
454 but we're moving away from vformat when we move
455 to OpenSync 0.3x / 0.40, so this is just temporary
458 - added SUBSYSTEM=="usb_device" to example udev rules, in order
459 to avoid running bcharge for each endpoint that is
460 added by udev, when first plugged in. Running bcharge
461 back-to-back multiple times can crash some
465 - fixed ID mapping bug in the case of a non-numeric UID
466 from syncing external -> Barry, and then the
467 new Barry ID would look like a new change
468 when syncing back from Barry -> external.
469 - added support for slow sync
470 - added clear() to idmap class
471 - added vformat attr parameter extraction class
472 - added more trace logging during vformat/Barry conversions
473 - added Notification time support
474 - handled "all day" vformat special case, when DTEND
475 does not exist in a vformat data block
476 - changed the min1900_t time conversion functions to handle
477 the 0xffffffff special case, when notifications
478 are turned off on a calendar event (src/time.cc)
479 - also made the calendar data dump output clearer for
480 disabled timestamps, and added debug output
481 for the raw min1900 value, in case there are more
482 special cases in the future
484 - fixed null pointer access in opensync's HasMultipleVEvents()
486 - upper management directive: revert all OpenSync 0.3x changes
487 and proceed with OpenSync 0.22. Reverting the
488 opensync-plugin/src tree to Barry 0.7 status.
490 - reorganized exception and trace code in opensync plugin
491 - added XmlToCalendar() function, based on unfinished opensync
492 library API that is not yet finished... will be
493 committed to opensync svn tree when complete
494 - added OSyncXMLField logger to plugin trace class
496 - applied email MessageRecord patch from Brian Edginton (thanks!)
497 - updated Message::Dump() to output more valid mbox data...
498 the goal is to be able to do:
499 btool -d "Messages" > email.mbox
501 - changed some TODO's to FIXME's to keep it grep-consistent
502 - applied Folder support patch from Brian Edginton
503 - fixed pedantic compile warnings, and changed order in src/Makefile.am
504 - removed unneeded cast in src/r_folder.cc
506 - minor adjustment to ctags building in buildgen.sh
507 - applied patch from Brian Edginton for better boost detection
509 - added boost usage message to btool -h output
511 - more porting of OpenSync module to 0.30
512 - added ctags build for opensync source tree to buildgen.sh
514 - changed PKG_CONFIG_PATH setting in rpm script, so that systems
515 like SuSE that have an extensive PKG_CONFIG_PATH
516 in general use will not break on build
517 - added PKG_CONFIG_PATH to debian script too, just in case :-)
518 - added debian/changelog to doc/ReleaseChecklist.txt
519 - added comment to rpm/barry.spec regarding SuSE RPM builds
521 - applied patch from Niels de Vos fixing debian package build
523 - applied Brian Edginton's Saved Email Messages patch
524 - updated r_saved_messages.cc comment to match header
525 - added GetRecType()/GetUniqueId() to r_saved_messages.h, since
526 SetIds() is already there
527 - removed Address version of operator<<() from message related headers
528 and moved into record.h
529 - added std::hex to message related dumps, and changed output to
531 - updated doc/CodingStyle.txt to cover switch() statements
532 - opensync compiles again, with 0.30! (not functional yet)
533 - moving closer to more modular and reusable plugin functions
534 - updated release checklist for opensync version number
536 - added doc/CodingStyle.txt... preliminary version
538 - applied Brian Edginton's PIN message patch
539 - changes to PIN message patch:
540 - fixed missing backslash in src/Makefile.am
541 - fixed tab alignment throughout
542 - removed typedef from Address struct
543 - added clear() to Address... might as well, if it it is
544 going to be a standalone type
545 - fixed missing comma in PINMessageFieldLinks table
546 - changed MessageRecordId in PINMessage to load from a new
547 uint32_t field in CommonField, instead of piggy
548 backing on a field for other use
549 - did some testing and found that PINMessage's MessageRecordId
550 is the same as the RecordId given through the SetIds()
551 API... but these Id's are duplicated in the protocol...
552 RecordId is stored in the protocol header, while
553 MessageRecordId is stored in the record it self as
554 a common field... added RecType and RecordId to
555 PIN Message so that we store both of these
558 - added item to gui/TODO
559 - applied Brian Edginton's s11n-boost.h patch
561 - renamed template values to more consistent naming scheme
562 - applied memos and tasks patch from Brian Edginton, slightly
563 modified to split into the new separate files system
564 - changes to memos and tasks classes:
565 - reformatted the code to match coding style
567 - added copyright notice
568 - changed class to singular (Memo, not Memos)
569 - placed in Barry namespace
570 - added #ifdef header protectors
571 - added new files to Makefile.am
572 - added new r_*.h headers to record.h
573 - added support for Memos and Tasks to tools/btool.cc, based on
574 Brian Edginton's patch
575 (currently commented out, for Boost library reasons):
576 - added doc/CommitPolicy.txt
579 - added ctags generation to buildgen.sh
580 - applied record.{h,cc} fix patch from Brian Edginton
584 - fixed broken formatting in above patch
585 - refactored record parser classes into separate files for each
586 - changed the ConvertHtoB<> template to inline, as it caused
587 linker errors with the new file layout
589 - fixed bug in Debian packaging that put util commands in
590 /bin and /sbin instead of /usr/bin and /usr/sbin
591 - commited byteswap.h replacement code from gm2net
593 - removed unneeded debug messages on stdout
594 - changed the default for debug output to "off"
595 and added a command line option to turn
596 it on if necessary (--debug-output)
597 - command line option --help now also works
598 - fixed bug in Calendar class where it didn't initialize
599 RecType properly... and cleaned up other RecType
600 code to use the GetDefaultRecType() functions,
601 localizing the constants in one place
603 - applied patches from Brian Edginton, parsing more
604 fields in Contact and Calendar records
605 - added Brian Edginton to AUTHORS
606 - reformatted record.{h,cc} to match the rest of the code,
607 removed commented out code, and removed mention of
608 Outlook from the comments.
609 - updated udev/10-blackberry.rules.Debian for Debian Etch
610 - updated debian/rules to use new udev rules file
611 - removed debian/postinst, as stable versions of Debian now
612 have a recent udev, and don't need the special handling
613 - updated Debian build scripts to build separate packages:
618 - added menu support to Debian barrybackup-gui package
619 - made the buttons bigger in the backup GUI, as they were too
622 - more progress in porting to opensync 0.30, heavily
623 based on the example plugin in the opensync
624 svn tree... this is a work in progress, and so far
628 - partial porting to opensync 0.30 (in progress!)
629 - fixed bug in sync_done where it reported context error
630 and then success afterward
631 - checking in new INSTALL files based on Etch's autoconf run
632 - added item to gui's todo list
633 - bumped version numbers to 0.8 in:
637 - gui/src/BackupWindow.cc (about dialog)
641 Release: version 0.7 - 2007/05/02
642 ------------------------------------------------------------------------------
644 - added rpm building to maintainer/release scripts
645 - added "HEAD" feature to tarball creation script, for easier testing
646 - updated maintainer/README
647 - updated maintainer/release.sh to use new HEAD feature too
649 - added pppob to rpm barry.spec
650 - bumped release number in barry.spec
653 - fixed size check bugs (strlen() == 0 when empty)
654 - finished refactoring calendar conversion routines
655 - started timezone parsing, but this may go away
656 if we use opensync 0.30's time functions and
658 - added development note to README
660 - updated README and opensync-plugin/README with more current
663 - refactored some calendar conversion routines
664 - added Barry->vCalendar conversion for recurring appointments
667 - removed some unneeded trace logs
668 - fixed VEventConverter constructor bug, with uninitialized
670 - eased up the requirement that every calendar event
671 must have a summary... defaults to "<blank subject>"
673 - added more VCALENDAR fields to match Barry::Calendar record
674 - fixed case sensitive compares in vformat.c
676 - opensync module work:
677 - added id map class to map string uid's from OpenSync to
678 numeric ID's from the Blackberry
679 - fixed bug in CommitChange() where it was still using
680 a hard coded calendar record state table
681 - added lots of logging throughout
682 - fixed bug where newly added records didn't use the
683 change object's uid, but created a new ID,
684 causing another sync on the next run
685 i.e. (CommitData_t was missing an argument)
686 - fixed bug where uninitialized notification time was
687 written to the Blackberry
688 - more opensync module work:
689 - refactored the environment class into another subclass
690 called DatabaseSyncState, and moved some stand-
691 alone functions from barry_sync.cc into
692 the environment classes
693 - added a tagged constructor to the trace class
694 - fixed idmap::Load() to clear map before loading
696 - renamed barry-config to barry-sync, the default configuration
697 file for the opensync module... and filled it with
698 a sample configuration
699 - added code to update the calendar cache, and write it to disk
700 on success (opensync module)
701 - added code to clear dirty flags in the device (opensync module)
702 - allowed comments in the config file
703 - added vformat.c from libopensync 0.22, since the vformat
704 routines are not available from the plugin side
707 - added -d switch to bcharge to switch back from 0001 to 0004 mode,
708 changed the auto-detection logic to ignore iProduct and
709 only use idProduct, and updated man page
710 - updated breset to recognize the Pearl in 00004 mode as well
711 - changed src/probe.cc logic to assume that the second endpoint
712 pair is the one needed for database communication, instead
713 of doing a reverse search from the end... newer Pearls
714 in Dual mode seem to have at least 3 or 4 endpoint pairs,
715 and the old search didn't always work
716 - applied sourceforge patch #1696884 from Peter Silva, updating
717 the Debian build for Etch, and including all files in one deb
718 - added Peter Silva to AUTHORS file
719 - uncommented the "make distclean" in debian/rules, and removed
720 the reliance on /tmp when installing 99-barry-perms.rules
721 - applied bb_task_format.txt patch from Peter Silva
722 - fixed each misspelling of "recurrance" to "recurrence"
723 - tweaked src/probe.cc again for paranoia's sake... start at
724 offset 1 if more than 1 endpoint pair, otherwise start
726 - split up opensync module into multiple source files, and started
727 using the opensync vformat routines for parsing and building
728 the vevent20 data (experimental)
730 - more incremental work on the opensync plugin, adding ADDED
731 and MODIFIED functionality... (experimental!)
733 - added code to dump extra descriptors during device discovery,
735 - bumped opensync version number in configure.ac, and made
736 the configure scripts more consistent with the rest
738 - fixed printf-format size bug in opensync module
740 - applied patch from Niels de Vos, fixing deb udev rules for Pearl
742 - fixed extra space bug in LDIF output (added space when creating
743 FullName, even if there was no last name)
744 - fixed Address Book record parsing in Contact class... the
745 Blackberry database uses field id 0x20 for both the
746 first and last names, and if only a last name exists,
747 the first name will be empty. Fixed parser to handle
748 this special case, without relying on string length.
749 Thanks to Troy Engel for reporting above two LDIF bugs.
750 - added a few more constants to bcharge.cc
751 - renamed opensync's configure.in to configure.ac
753 - fixed bcharge to avoid endless reset loop... oops
755 - reorganized bcharge.cc to make it possible to switch between
756 Pearl modes 0004 and 0001
757 - added flag to BadPassword exception class, to signal whether
758 the library considers the device "out of retries" or not
759 - updated socket.cc for new BadPassword exception
760 - added special case to Controller class destructor to reset the
761 device if being destroyed in a half-open state... this is
762 a temporary fix until we understand how to handle
763 the "already open" message we get when opening the Desktop
765 - added password prompt dialog to the backup GUI
766 - fixed possible memory leak in GUI's DeviceInterface::Connect()
768 - committed Troy Engel's bbrules_mass_symlinks.patch, fixing
769 udev rules for the Pearl
770 - updated 10-blackberry.rules.Debian to match 10-blackberry.rules's
772 - applied Troy Engel's barry.spec patch to generate all required
773 RPM's... Thank you! Notes in the README taken from
774 the sourceforge tracker entry containing the patch.
775 - added libusb lib flags to tools/ and examples/ makefiles,
776 and tweaked configure.ac with a better default,
777 to build cleanly on RHEL 4
778 - modified rpm/barry.spec:
779 - does not create doc-only barry base package anymore
780 - fixed "barry-backup" typo
781 - removed the patch step, as version 0.7 shouldn't need it
782 - added license file to each package
783 - commented out console perms scripts, pending mailing list
786 - applied Troy Engel's barry.spec patch to fix RHEL4 compile issues
787 - fixed 80 column wrapping error in barry.spec, thanks to Troy Engel
789 - applied Troy Engel's udev permissions patch for Redhat/Fedora
790 systems, which automatically set the ownership of a
791 newly plugged in device to the currently logged in console
793 - added udev/README to document what all those files are for...
794 part of the text comes from the sourceforge comment
795 on the above udev permissions patch from Troy Engel
796 - uncommented the console perms scripts for rpm/barry.spec
797 - added libglademm-2.4 to gui/README's dependency list
798 - made usbwrap.* calls more threadsafe
799 - added Usb::Timeout exception for timeout errors
800 - added preliminary UsbSerData support to library for GPRS modem
801 functionality, using Rick Scott's XmBlackBerry's serdata.c
803 - added overloaded Socket::Send() for sending without receiving
804 - added preliminary pppob command line tool
810 - gui/src/BackupWindow.cc (about dialog)
813 - build system tweaks for opensync-plugin, and added a buildgen.sh
815 - added better debug messages in probe.cc and usbwrap.cc's
816 device discovery code
817 - changed error handling for null interface->altsetting pointers
818 when doing device discovery in usbwrap.cc. Seems that
819 some devices report a higher number in bNumInterfaces
820 than are really available... in this case, let's
821 just ignore the missing data and assume success
822 - applied udev script patch #1663986 from Troy Engel
823 - applied barry-bcharge.spec patch #1672178 from Troy Engel
824 - ReleaseChecklist update
825 - added rpm/README based on Troy Engel's patch notes
826 - updated new barry-bcharge.spec for version 0.7, removing patch
829 Release: version 0.6 - 2007/02/28
830 ------------------------------------------------------------------------------
832 - fixed size checks for device database records larger than 64K
833 Thanks to Michael Brown for the bug report.
834 - added initial release maintainer scripts
835 - fixed odd backup success message in GUI when doing a backup with
836 no databases selected in configuration
837 Thanks to Michael Brown for the bug report.
838 - added udev permissions file patch from Michael Brown
839 - added Michael Brown to AUTHORS
840 - updated README, gui/README, and doc/USB-capture.txt
841 - added -o command line switch to bcharge.cc, and updated man page
843 - changed library behaviour so it only turns on libusb debug output
844 if in data dump mode. In btool, -v controls this output.
845 - minor stream formatting fix in btool
846 - added OpenSSL dependency to build for password hash support
847 - made debug dout() output conditional on Init() flag... In btool,
848 -v now controls this as well
849 - reverse engineered more of the initial probe packets, and
850 implemented probing in struct based code... changes include:
851 - added support for "attribute fetch" packets, used at
852 the beginning of USB conversations
853 - defined object/attribute id codes for the PIN number
854 - made Protocol::SocketCommand more robust, and implemented
855 its sequence number behaviour
856 - moved ModeSelectCommand into SocketCommand, as it is a
857 socket 0 level command
858 - added initial protocol struct: PasswordChallenge
859 - renamed Barry::Packet to DBPacket
860 - added ZeroPacket for socket 0 packet creation and analysis
861 - Barry::Packet is now the base class for the above
862 - implemented socket 0 behaviour in socket.cc
863 - changed return values for Send/Receive/Packet/NextRecord
864 members in the Socket class from bool to void, as they
865 don't generate errors themselves, but only can
866 have an error if Usb::Device throws an exception.
867 Cleaned up all code that called these members as well.
868 - removed "last status" from Socket class as it was unused
869 - updated probe.cc to use new socket 0 implementation
870 - removed two of the hard coded packet handshakes in favour
872 - updated controller.cc:
873 - uses new socket 0 implementation
874 - removed "flag" hack when opening sockets and
875 selecting modes, as it should use the
876 socket 0 sequence number
877 - added initial, incomplete prototype header for the C API
878 - updated some .cvsignore files
880 - updated ReleaseChecklist.txt, adding rpm spec file
881 - added the following to ZeroPacket:
886 - moved Command() to base Packet class
887 - added comments to protostructs.h and size constants for the
888 new PasswordChallenge struct
889 - added password support to Socket and Controller classes
890 Thanks to Rick Scott's XmBlackBerry for the openssl
891 password hashing logic.
892 - updated btool to use new password support, and updated its man page
894 - updated Doxygen input files and version, and added to ReleaseChecklist
895 - fixed btool PIN output for LDIF mode, now commented in output
896 - major LDIF overhaul:
897 - split LDIF specific code out of the record class and
899 - fixed "full name" behaviour that added an extra space
900 in dn attributes, as reported by Troy Engel
901 - added support for mapping of LDIF attributes to Barry
903 - fixed base64 handling, so attributes are only so encoded
905 - fixed base64 attribute reading inconsistency... some
906 notes fields were missed in old version if not
908 - added support for specifying alternate attribute for
909 constructing a FQDN, instead of just "cn"
910 - updated btool and upldif to use new features
911 - large update to btool manpage
912 - fixed initialization bug in Contact constructor (missed RecType)
914 - added ktrans.cc to make reading of kernel usbfs_snoop logs easier
915 - removed commented code from translate.cc
916 - added Pearl handshake support to bcharge.cc, which resets the
917 Pearl so that the vendor specific 0xFF class exists.
918 This should allow use of btool on the Pearl.
919 - added Data::AppendHexString to data.cc
920 - added better probing support for the Pearl, so it detects the
921 Product ID 0x0004 mode, and does a search for class 0xFF
922 instead of hardcoding it.
924 - applied README patch from Ian Darwin
925 - minor tweaks to README
926 - added missing errno.h header in src/usbwrap.cc for AMD64
927 Thanks to Jonathan Hudson for reporting the bug.
928 - tightened up size checks in Data::ReleaseBuffer()
929 - stable libusb *does* return the actual length of the USB packet...
930 fixed this in usbwrap.cc to set the actual size of read data,
931 and removed the size hacks elsewhere in the code:
932 probe.cc, protocol.cc, socket.cc, usbwrap.cc
934 - added proper return codes for error conditions in btool.cc
935 - clarified some stream output code in the library, for hex/dec numbers
936 - removed old, non-working connect.cc test program
937 - added iomanip to debug.h
938 - changed usbwrap to call libusb again if EINTR and EAGAIN
939 This should fix one of the OpenBSD issues
942 - fixed bug in barrybackup GUI that didn't create the target path
943 if the user didn't do Edit | Config first.
944 Thanks to Ian Darwin for reporting the bug.
945 - applied gui/src/ConfigFile.cc header fix patch for OpenBSD
948 - added initial btool.1 man page from Ian Darwin
949 - updated man/Makefile.am for new btool.1
951 - applied btool -h help clarification patch from Ian Darwin
953 - reworked exception hierarchy, so applications can handle all
954 Barry related exceptions, including Usb errors, with
955 one base class: Barry::Error
956 - fixed all the ripple changes this caused throughout the codebase
957 - added BadPassword exception
963 - gui/src/BackupWindow.cc (about dialog)
964 - updated doc/ReleaseChecklist.txt
965 - added usb-level breset.cc command line tool
966 - fixed bug in gui/src/tarfile.* for systems where
967 sizeof(int) != sizeof(void*)
968 - TarFile class now uses a plugin style for the compression
969 operations, to allow for threadsafe versions if needed
971 - also fixed leaked file handle bug in open_compressed()
972 - Thanks to Jonathan Hudson for reporting this for 64bit systems
973 - minor signed comparison fixes in the library
974 - added better protocol data dumping code in probe.cc
975 - added Usb::Device::BulkDrain(), to prevent Barry from hanging
978 Release: version 0.5 - 2007/01/26
979 ------------------------------------------------------------------------------
981 - added missing errno.h header to gui/src/util.cc
982 - lots more endian fixes after testing on iMac
983 - iostream output size fixes
984 - added option to btool to reset the device via software
985 - beefed up exception handling in barrybackup
986 - checked in opensync work-in-progress: trace logging, pointer bugfix
987 this is still experimental
988 - added aboutdialog.* files from gtkmm24 2.10.6, so we have a dialog
989 class for systems running older versions of gtkmm, like
991 - uncommented the about box menu handler
992 - added "cleanall" option to root buildgen.sh script
993 - update ReleaseChecklist.txt
994 - removed CVS auto-history strings from aboutdialog.*
996 - found a link between the unknown field in tagged protocol headers
997 and the unknown field in RecordStateTable entries. This
998 field matters when uploading certain databases, such as
999 Browser Options. It is unknown what this field really
1000 means, but we're calling it RecType. Updated the API
1001 to use this in all parser objects and builder objects
1002 and record objects that use it.
1003 - changed some uses of uint64_t to uint32_t, in Contact and
1004 Calendar record objects... unique IDs only seem to be
1006 - added new static function to record classes: GetDefaultRecType()
1007 This returns the default rectype that should work
1008 for that record. Ideally this should be retrieved from
1009 the device, but in the case where a database has no
1010 records, this info would be good to know.
1011 - updated convenience template function Controller::AddRecordByType()
1012 to use GetDefaultRecType()
1013 - documented the recurrence data and how to use it in Calendar objects
1014 - added data debug output for incoming USB packets in the
1015 socket class... when we ported to the stable libusb
1016 we lost some of the verbose output... this should get it
1017 back. (In the official release, you need to uncomment
1018 ddout() in debug.h for the full blast of debug output).
1019 - backup GUI updates:
1020 - changed error_done signal into a pure error signal,
1021 and changed the logic so the restore continues
1022 even if there was a protocol error on one database
1023 - changed status bar messages to show what's happening
1024 i.e. Backup or Restore
1025 - default to not restore the "Handheld Agent" database as
1026 it appears to be read-only on the device
1027 - added temporary Restore & Backup mode for debugging
1028 - added extra hyphen to backup files to show date/time better
1029 - updated GUI to save and restore the new RecType field
1030 supported now by the library
1032 - added future considerations to doc/ReleaseChecklist.txt
1033 - applied big endian patch from Jonathan Hudson
1035 - removed endian.h include from barry.h, since applications
1036 using the library should not care about endianness
1037 - hard coded the path to config.h in endian.h
1038 so that endian.h never makes it into the final
1040 - removed DEFAULT_INCLUDES setting from src/Makefile.am
1042 - removed config.h dependency in btool.cc and used library
1043 Version call instead
1044 - added big/little endian text to the Barry::Version call
1046 - fixed install location of bcharge when building with ./configure;
1047 now goes to sbin/ for both source compile and binary packages
1048 Thanks to Jonathan Hudson for reporting this bug.
1050 - added smarter timeout values to controller, and allowed
1051 timeout overrides in the packet, socket, and usbwrap code...
1052 extra timeouts are needed for erasing databases for
1054 - added an explicit DataDumpParser class to btool.cc, as the
1055 library doesn't always have extreme verbose debugging on...
1056 this makes sure anyone can always get raw database packets
1057 through btool, for development
1058 - added thread callback signals for error finishes and erase_db
1059 operations, so the GUI is updated properly
1060 - more GUI cleanups (status bar updates, etc)
1061 - finished implementation of GUI restore
1062 - fixed gtkmm/glibmm exception handling when the exceptions
1063 occur in sigc signal handlers... in that case,
1064 exceptions need to be handled with
1065 Glib::add_exception_handler()
1066 - added filename-only reading support to TarFile class
1067 - added pending gui/TODO items
1069 - backup GUI now properly shows progress
1070 - the buttons become insensitive when backup is busy
1071 - refactored directory check code, and added mkdir check
1072 when to the backup setup (so the directory exists
1073 when we try to write to it)
1074 - updated code to use Barry::Data in new namespace
1076 - moved Data class into Barry namespace
1078 - updated copyright dates for the new year
1079 - clarified exception message in controller.cc
1080 - fixed pointer initialization bug in controller.cc
1081 - added documentation comments to parser.h and builder.h
1082 - updated ReleaseChecklist.txt
1084 - changed Barry library configure scripts to make better
1085 use of pkg-config for autodetecting dependencies
1086 - added initial Barry Backup GUI (gui/ directory), capable of making
1087 backups only at this point
1089 - added doc/ReleaseChecklist.txt
1090 - added version API, returning version number and string
1091 - added Troy Engel to AUTHORS
1092 - added special case to bcharge.cc for the Blackberry Pearl,
1093 which doesn't reset itself after the charge handshake
1094 Thanks to Troy Engel for testing.
1096 - added pkg-config support for the library (libbarry-0.pc.in)
1097 - fixed compile error on g++ 3.3 systems (missing stdint.h
1099 - fixed protocol hang at end of Controller operation... in
1100 controller destructor, it was deleting the USB interface
1101 before closing the socket at the BlackBerry level, thereby
1102 leaving the device in Desktop mode
1104 - support the newer udev packages on Ubuntu, which need slightly
1106 - modified bcharge.cc to search for Pearl devices too.
1108 - added Pearl product ID to probe code, but likely not functional
1109 yet until we find someone able to get a USB capture for us
1111 - version bump in configure.ac and src/Makefile.am
1113 Release: version 0.4 - 2006/12/29
1114 ------------------------------------------------------------------------------
1116 - added bcharge.cc to the tools/ directory (whoohoo!)
1117 - added initial release version of rpm/barry.spec (aka bcharge.spec)
1118 This begins the barry binary release, currently only
1119 including the bcharge utility, but possibly more in
1120 the future. Future binary packages will likely include
1121 libbarry, libbarry-devel, etc, which barry will then
1123 - added udev rule script for BlackBerry devices
1124 - added man/ directory and bcharge.1 manpage
1125 - added debian/ directory for building deb packages, with initial
1127 - added rpm/make-rpm-tarball.sh for building RPMs
1128 - split out technical docs into separate libbarry.docs file for
1129 a future Debian install
1130 - added more cleaning to buildgen.sh
1131 - added new udev rule file for Debian stable
1132 - hardcoded g++ to 3.3 in debian/rules as I have multiple versions
1133 on my system and need to compile releases with stable
1135 - ported Barry to use the stable branch of libusb
1136 NOTE: no longer depends on the devel libusb tarball!
1137 Just use your distro's libusb packages.
1138 - fixed more endian issues
1139 - minor update to TODO
1141 - moved Exceptions, Hacking, and VersionNotes to doc/
1142 - added doc/USB-capture.txt
1144 - added Controller::AddRecord() and template helper
1145 - added beginning version of generic RecordFetch<> template
1146 - set default of Barry::Init() to false (no data dump mode)
1147 - added examples/ directory, with first addcontact.cc example
1149 - added autoconf support based on Peter McAlpine's patch
1150 - renamed Makefiles to Makefile.orig and added script to turn
1151 on old build system if wanted
1152 - fixed boost-specific error path in btool
1153 - moved platform specific code out of time.h and into time.cc
1154 - moved btool.cc, translate.cc, and upldif.cc out of src/ and
1155 into their own tools/ subdirectory
1156 - added NEWS file to make autotools happy
1157 - added a "clean" option to the buildgen.sh script... maintainer's
1159 - reworked autoconf and automake files, to support
1160 barry as a library, to move the programs into their
1161 own tools directory, and support a proper installation
1162 - fixed headers in parser.h, btool.cc, and upldif.cc to work
1163 properly when installed standalone
1164 - added tools/Makefile.plain for testing compiling against
1165 installed barry library
1166 - added commentary on version number system
1167 - removed Makefile.conf and Makefile.orig, as new autoconf
1168 system supercedes it
1169 - added support for boost serialization library in autoconf build
1170 - added library version number to autoconf build
1171 - changed boost flag to __BARRY_BOOST_MODE__ to avoid collisions
1172 - removed debug.h dependency from btool.cc
1173 - removed libusb check in favour of specific --with-libusb
1174 option... may need to revisit this later
1175 - quick installation update in README
1177 - fixed some missed endian conversions in record.cc and added
1179 - fixed endian issues in time.h
1180 - added 2 more functions to the time zone API, and adjusted
1181 the time zone table for better defaults in North America
1182 - added btool and upldif to the make install sequence
1183 - added support for Calendar recurrence data
1184 - small doc/TimeZones.txt update
1186 - added BlackBerry time zone code list, reverse engineered from a 7750
1187 - added GetTimeZone() to convert device time zone codes to useful data
1189 Release: version 0.0.3 - 2006/11/24
1190 ------------------------------------------------------------------------------
1192 - added return code check to Controller::SaveDatabase()
1193 - fixed Data() class constructor, so it doesn't auto-convert
1195 - fixed size checking bug in Packet::ReturnCode() that caused
1196 uploading to device to fail
1198 - added Exceptions doc file, documenting the C++ exceptions used,
1200 - fixed small bug in 'make install' which didn't build first
1201 - minor documentation and cleanup in upldif.cc
1202 - added utility functions to the RecordStateTable class,
1203 GetIndex() and MakeNewRecordId()
1204 - changed btool so that the -d command modifiers can be
1205 specified multiple times to work with multiple records
1207 - added Calendar sync code to the opensync module (not complete)
1208 - cleaned up exception handling in the opensync module, making
1209 more use of the base exception to make sure no
1210 exceptions leak into the C library code
1212 Release: version 0.0.2 - 2006/10/12
1213 ------------------------------------------------------------------------------
1215 - fixed casting error in opensync module
1216 - minor updates to documentation to prepare for release
1218 - added stdint.h to parser.h for uint32_t. Thanks Ron Gage for
1221 - large refactoring of the packet parsing and building code,
1222 splitting the code into 3 groups:
1223 - record: handles record specific parsing and building...
1224 this is mostly subfields with CommonField structs
1225 as their headers, but sometimes has a specific
1226 header of its own, in the case of email
1227 Note: email's header is as yet undecoded.
1228 Record does not know the details of the packet
1229 formats, only the formats of its own record data.
1230 - packet: handles building of complete command packets
1231 and handles parsing of commonly used header values,
1232 providing an API for it... packet does not know
1233 the record format details, but does know the
1234 details of each Database Operation (protocol.h)
1236 So far, it seems that there is a real separation
1237 of operation formats and record formats, with the
1238 record formats being the same even with different
1239 database operation codes (GET_RECORDS /
1240 OLD_GET_RECORDS), which was not confirmed before.
1241 See the header size of the email record after
1243 - controller: handles device state, and the management of
1244 sockets, commands, and checking for response values.
1245 The Controller object does not know the format
1246 of either the record or the packet it is sending,
1247 but does know the protocol handshake logic used
1248 to talk to the device.
1249 - implemented SetRecord (SET_RECORD_BY_INDEX)... syncing, here
1251 - spelling correction in opensync-plugin/AUTHORS
1253 - added sample hotplug scripts
1254 - added .cvsignore files for doxygen and opensync directories
1255 - added src/endian.h... still need to add configure support to
1256 handle non-little-endian machines
1257 - added exception handler in opensync plugin commit_change()
1258 - removed the device reset code in probe.cc, as resetting a device
1259 can renumber the devices in /proc on kernel 2.4.x,
1260 and cause a failure to reach the device...
1261 FIXME - this should be changed someday to reset if probing
1262 fails, and restart the probe on reset, as sometimes
1263 the blackberry devices respond differently when not reset
1264 Ideally, the probe protocol should be properly reverse
1266 - added support for retrieving record state table, which is needed
1268 - added support for retrieving, deleting, and clearing the dirty
1269 flags of individual blackberry device records
1270 - added endian macros throughout the code
1272 - added opensync-plugin to root Makefile's clean
1273 - updated time conversion calls to match opensync's latest SVN
1274 - added opensync_change_set_changetype() (only add for now,
1276 - removed slow sync functions for now
1278 - added quick "install" makefile target, to src/install by default
1279 - added static library target... this is temporary until autoconf
1280 stuff is added properly... needed for the opensync plugin
1281 - documented in comments that Barry::BError is intended to be
1282 the base classes for all exceptions
1283 - added first round of implementation of an opensync plugin, based
1284 on the example plugin code from the opensync sources
1286 - added to Contact record class:
1288 - ReadLdif() to load and parse LDAP data
1289 - added fragmented send support to Socket class
1290 - added Probe::FindActive() to make client programs more streamlined
1291 - fixed hex/decimal printing of record counts in btool
1292 - added upldif.cc, a utility to take ldapsearch output on stdin
1293 and upload it to a blackberry
1294 - updated src/.cvsignore
1296 - added parsing of service book fields:
1298 - old and new unique ID (UID)
1300 - old and new description
1302 - bes domain / domain
1303 - date change in legal.txt
1304 - added initial parsing of ServiceBook field 0x09, and calling it
1305 ServiceBookConfig for now, for lack of a better name
1306 - added parsing for ServiceBook field 0x02 on 72xx Blackberries...
1307 seems to be a name or description, but doesn't appear
1308 in the device's own GUI... called "Hidden Name" for now
1310 - added ServiceBook record class and parser code, based on some
1311 of Ron Gage's reverse engineering... code is by Chris Frey
1312 - added AUTHORS file
1313 - fixed the libusb set_configuration/claim_interface order
1314 bug in controller.{h,cc}... already fixed in probe,
1315 but not controller. Thanks Ron Gage for finding this.
1317 - renamed Changelog to ChangeLog in preparation for autoconf
1318 - renamed Todo to TODO in preparation for autoconf
1320 - make clean cleans up tests properly now
1322 - turned source code browsing on in Doxyfile
1323 - added LoadDatabaseByName() and LoadDatabaseByType() template members
1324 to the Controller class
1325 - new file: controllertmpl.h
1326 - documentation fixes in builder.h
1327 - added static GetDBName() functions to all record classes, in
1328 support of the template Controller functions
1329 - fleshed out DatabaseDatabase::GetDBNumber() and GetDBName(),
1330 returning errors properly
1331 - documented sample btool.cc code better, and added example
1332 of Controller template member usage
1334 - copyright dates updated for 2006
1336 - backed out update to latest libusb DEVEL tree, since it is buggy.
1337 Stick with libusb 2005/11/26
1339 - added clean target to convenience Makefile
1340 - added Boost to list of dependencies in README
1341 - fixed minor compile warnings in base64.cc
1342 - fixed new USB discovery code, and changed endpoint usage to use
1343 it, instead of the READ_ENDPOINT and WRITE_ENDPOINT
1344 constants... constants are now removed
1345 common.h, controller.cc, probe.cc, probe.h, usbwrap.cc
1346 - added ClearHalt() to Usb::Device (usbwrap.cc)
1347 - updated usbwrap.h to match latest libusb CVS devel tree
1348 - updated README to refer to 2005/12/30 libusb CVS devel tree
1349 - added roadmap to Todo list
1351 - added convenience Makefile in root directory
1352 - added Hacking document
1353 - minor touchups to README
1354 - added caution warning to README, since this release will support
1355 delete / overwrite / upload functionality
1356 - minor compile order change in Makefile (speed reasons)
1357 - added dynamic endpoint discovery to probe.cc
1358 - changed claim_interface/set_configuration() function call order in
1359 probe.cc based on feedback on the libusb mailing list
1360 - added USB discovery code to usbwrap.{h,cc}
1362 - fixed packing bug in new upload header struct
1363 - renamed CommonField union from data to u to be consistent
1364 - added return_code field to DBAccess struct, for error checking
1365 - added response checking during upload: if response is non-zero
1366 throw exception in controller.cc
1367 - added sorting operators to record parser classes... when writing
1368 contact data to the device, it may be important to write
1369 group link items last, since they reference other contacts...
1370 This needs to be tested, and if not required, removed.
1371 - added GROUP_FLAG field code to Contact parser class
1372 - fixed offset bug in Contact and Calendar Build() functions
1373 - fixed name bug in Contact::Build(), now output only with data
1374 - added appointment type flag support to Calendar record parser
1375 class, and "all day event" flag support
1377 - added autoconf to Todo list
1378 - refactored the protocol structures so that record structs can
1379 be reused for upload and download
1380 - with new protostructs, more header control is handled by
1381 controller.cc instead of record.cc parsers and builders
1382 - parser code now accepts an offset, to govern start of record data
1383 - changed Data sizes from int to size_t and removed casts
1384 - removed unneeded Parser::GetHeaderSize() (new protostructs make
1386 - added Calendar::Build() for calendar uploads
1388 - added s11n-boost.h serialization templates for boost::serialization
1389 - added boost::serialization to build
1390 - minor change in usbwrap.cc to display error codes in decimal
1391 - added SizePacket special case code to Socket::Send()... this is
1392 required when uploaded packet size is a multiple of 0x40,
1393 a special 3 byte size packet is sent
1394 - moved some record class variables to public: so they can be
1396 - fixed variable name but in RecordBuilder template
1397 - added -f switch to btool, for saving and loading of record data
1398 - added -s switch to btool, for uploading saved data back to handheld
1399 - added Controller::SaveDatabase()
1401 - fixed potential casting / pointer bug in controller.cc
1402 - added builder.h, to support uploading
1403 - minor comment fixes in parser.h
1404 - added preliminary Contact protocol record building support
1405 - fixed field order bug in ContactRecord protocol struct
1406 - Contact record.h class now saves unknown Group Link field,
1409 - added -fno-strict-aliasing to compiler flags so gcc-4.0.x won't
1410 optimize away protocol struct casts
1412 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25235
1413 http://mail-index.netbsd.org/tech-kern/2003/08/11/0001.html
1414 http://groups.google.ca/group/comp.lang.c++.moderated/ \
1415 browse_thread/thread/e7bf096832526f8e/5714701b \
1416 02a2a3cc?hl=en#5714701b02a2a3cc
1417 - changed headers so that any low level protocol-specific sizes and
1418 structs are completely hidden in the library...
1419 this adds library-only header: protostructs.h
1421 Release: version 0.0.1 - 2005/11/25
1422 ------------------------------------------------------------------------------
1424 - added barry.h for documentation purposes and application usage
1425 - updated README for release
1426 - renamed SBError exception class to BError
1427 - updated Doxyfile to version 1.4.5
1428 - added more doxygen comments
1430 - added conversion routines (and tests) between time_t and min1900_t
1431 - added -v option to btool to dynamically control protocol dumping
1432 - removed duplicated database database table display in controller.cc
1433 - minor spelling corrections
1434 - added copyright to usbwrap.{h,cc} as per earlier meeting
1435 - added support for Old protocol Calendar entries
1437 - added legal.txt, to save typing when adding new source files
1438 - ignore test-base64 file (cvsignore)
1439 - added contact-to-ldif generation, and updated supporting code
1440 - added argument to btool to support contact-to-ldif operation
1441 - added Contact::GetPostalAddress()
1442 - added README documentation in preparation for future release
1444 - added parser.{h,cc} virtual wrapper to generalize LoadDatabase()
1445 - added code to socket.cc's Packet() command to re-receive on
1446 reception of empty packets... this seems to be a signal
1447 from the device that it needs more time for long transfers
1448 Current limit: 10 blank packets before exception is thrown
1449 Observed maximum blank count: 1
1450 - added MessageRecord and OldMessageRecord to protocol.h
1451 - changed common record field structs to one CommonField struct
1452 - increased default USB timeout to 10 seconds (some loaded devices
1453 seem to need more time...)
1454 - reorganized record.{h,cc} to try to reduce the code duplication
1455 - record.h classes now "know" the sizes of their corresponding
1456 protocol structs... this may cause trouble down the line
1457 with Python, if used(?)
1458 - added Message class for Email records
1459 - removed the "Get*" database functions and replaced with
1460 general LoadDatabase() call, which loads any database
1461 available, and parses it with a Parser object
1462 - added GetDBID() for searching for database numbers via name
1463 - added test code to retrieve and display all email
1464 - cleaned up comments
1465 - renamed blackberry.{h,cc} to controller.{h,cc}
1466 - renamed bbtool.cc to btool.cc
1467 - added Makefile.conf
1468 - added copyright notices to source files
1469 - added COPYING file with GPL
1470 - put parser in Barry namespace, and added auto_ptr-like mode for
1471 store object pointers
1472 - added doxygen docs
1473 - added header comments to convo.awk script
1474 - removed Test() function in Controller object, and made class
1475 better suited to its general API purpose
1476 - changed OpenMode() so it loads command table and DBDB... it is now
1477 a public API function
1478 - moved Controller test code to btool.cc and made tool more
1479 generically useful... can now download any DB by name on
1481 - added mode checks to Controller
1482 - added -t and -d command line options to btool
1483 - added base64 routines from John Walker / Fourmilab and Citadel/UX
1484 - converted base64 routines into an API, instead of hardcoded
1487 - changed project name to Barry, including namespaces
1488 - renamed sbcommon.* to common.*
1489 - increased USB default timeout to 2 seconds
1490 - fixed formatting bug in Data operator<< output
1491 - cleaned up the protocol packet structs, getting rid of the
1492 confusing param/simple split... param is now the tableCmd
1493 from the Command Table
1494 - moved protocol structs from record.h to protocol.h
1495 - added support for older version of the USB protocol, which matches
1496 the Cassis spec more closely
1497 - fixed the hardcoded offsets in blackberry.cc's protocol parsing
1498 - fixed Socket::Close() so on error, socket flags are reset.
1499 This stops Close() from being called twice on exception
1502 - added Doxyfile to the project
1504 - added mode selection support to Blackberry class (RIM Desktop,
1505 RIM Bypass, and RIM_JavaLoader)
1506 Mode selection governs which socket is used in subsequent
1508 - added Command Table parsing and protocol support
1509 Blackberry class, and CommandTable record class
1510 - added Database database parsing and protocol support
1511 Blackberry class and DatabaseDatabase record class
1512 - added Address book retrieval support - dumps data to stdout for now
1513 - fixed bug in CopyOnWrite mode in Data class, where could allocate
1514 less buffer than data copied
1515 - added "easy exception" debug output macro
1516 - fixed minor spelling error in protocol.cc
1517 - fixed ModeSelectCommand protocol structure - it is not a parameter
1518 command, but uses the unknown slots for socket and flag
1520 - removed hardcoded socket numbers from protocol.h
1521 - fixed bug in socket.cc:Open() where packet size was not properly
1523 - implemented socket::Close()
1524 - fixed sequence number checks, so it is reset properly on socket 0
1525 - added Socket::NextRecord() for multiple record data transfers
1526 - fixed input handling bug in translate.cc
1528 - added .cvsignore files to project
1529 - added README notes to document the source architecture
1530 - added dbsetup data script, which shows how to get into database
1531 mode, using the RIM Desktop mode, and open a socket
1532 - added blackberry.{h,cc} class
1533 - added command line parameters to bbtool.cc
1535 -p = specify pin of blackberry device
1536 - added blackberry test code in bbtool.cc - debugging the initial
1537 opening protocol: sockets, modes, etc... see dbsetup for
1539 - added copy-on-write functionality to Data class, so it can be used
1540 with plain data buffers with little performance impact
1541 - added eout() and ddout() debugging - to separate:
1542 - regular debug output
1543 - exception debug output (should probably be logged to
1544 syslog someday, and never turned off)
1545 - raw data dump output
1546 - changed errno SBError exception handler to try to decode the error
1547 numbers into english (limited success)
1548 - fixed probe and socket USB Bulk message handling, so they don't
1549 use PollCompletions(), but use IO::Wait() (libusb_io_wait())
1550 - added common packet size checker in protocol.cc
1551 - fixed unpacked protocol packet structs... using gcc's
1552 __attribute__ ((packed)) ... see protocol.h
1553 - added ModeSelectCommand structure (protocol.h)
1554 - finished Socket::Open() call (Close() not yet implemented)
1555 - moved Socket::Send/Receive() out of private, as they can be used
1556 for socket 0 operations.
1557 - cleaned up pointer management with macros:
1561 - added IO::Wait() and IO::Cancel() to correspond with libusb API
1562 - added datadumping in usbwrap for easy protocol capture
1563 - added preliminary record.{h,cc} classes
1564 - added Contact record parsing, and test code
1566 - syncberry tree started