1 Release: version 0.9 - 2007/08/??
2 ------------------------------------------------------------------------------
4 - added better error checking to bcharge.cc
9 - gui/src/BackupWindow.cc
13 - opensync-plugin/src/barry_sync.cc
14 - removed usb_set_configuration() check from bcharge.cc, since
15 that may fail under normal operation, in this case
17 Release: version 0.8 - 2007/08/03
18 ------------------------------------------------------------------------------
20 - moved Interface implementation to usbwrap.cc file
21 - added dout debugging messages to usbwrap.cc
22 - clarified unhandled packet error message in socket.cc
23 - added temporary workaround in opensync-plugin environment's
24 Reconnect(), for odd message seen on newer Blackberry
26 - fixing issues with Brian Edginton's MessageRecord patch (2007/06/29)
27 - fixed missing size check in message oriented ParseHeader()
28 - fixed endian conversion issues (also found one pre-existing
29 endian bug in the process)
30 - moved message timestamp converter code to time.cc
31 - added Barry version number to opensync plugin name
32 - changed maintainer scripts and rpm.spec to use bzip2 instead of gzip
33 - added opensync-plugin to configure script build
34 - added -S option to btool, to list supported parsers and builders
35 - slight change to opensync-plugin/buildgen.sh to cleanup config.h.in~
36 - opensync-plugin/README update
37 - added opensync-plugin/patches for others who may be building and
38 testing the plugin with OpenSync 0.22
39 - added debian package build scripts to maintainer/
41 - added entry to doc/TroubleShooting.txt
43 - added doc/TroubleShooting.txt
46 - fixed bug in vcard.cc that was using the FN field
47 for the broken-down name instead of N
48 - added dbId and dbName to DatabaseSyncState
49 - added Reconnect() functionality to BarryEnvironment class
50 and moved connect logic out of barry_sync.cc
51 - changed the batch commit functionality in barry_sync.cc
52 to single-change-commit. Why? Because opensync
53 0.22 groups all batch changes into one group
54 and calls one batch commit function, even if you
55 registered two. Moving to single change commit
56 mode removes a chunk of unneccessary code
57 from the plugin as well, and can in theory
58 ease memory pressure in the library.
59 - added FinishSync() to perform actions needed on successful
61 - disconnect on successful sync completion, and reconnect,
62 since the device's dirty flags don't seem to
63 be updated until a disconnect occurs.
65 - added ReadDataArray() to data.h, using an istream instead of
67 - added raw data version of BuildField() to record-internal.h
68 This is so that btool doesn't add extra null terminators
69 to unknown data, when restoring with option -s
70 - fixed bugs in r_contact.cc:
71 - fixed bug that caused a LastName/FirstName swap
72 if FirstName was empty and you wrote a contact
73 record into the device.... First/Last Name
74 fields are position sensitive, and the code
75 was missing the empty FirstName case on writing
76 - fixed bug in BuildFields() and Dump() where they
77 were not updated to handle the new field
78 table added for the postal address changes
81 - changed the plugin config file, so it is possible to
82 turn debug output on/off without recompiling
83 - moved the standalone Map2Uid() function into
84 the DatabaseSyncState class, and changed the
85 code to produce "contact-#" and "calendar-#"
86 UID strings, instead of just numbers
87 - changed default PIN to -1, so that if the user doesn't
88 configure properly, it will fail... this is a
89 safety measure to avoid syncing with different
90 devices and corrupting the cache/idmap...
91 this should be fixed someday, to handle this
93 On second thought... is this right? Perhaps
94 it is better to fail if the PIN is not what
95 is expected, so a user doesn't accidentally
96 sync the wrong device and lose data.
97 Actions to take on new PIN:
100 - incorporated the new VCardConverter code into the
101 main barry_sync.cc plugin code
102 - fixed bug that did not clear the devices dirty flags
103 in the case where a sync only read data from it,
105 - changed PIN error message number to hex
107 - minor wording change in backup GUI prompt: "Backup working..." to
108 "Backup in progress..."
109 - renamed Contact class's Title field to JobTitle for clarity
110 - added record function ParseFieldString(), which properly handles
111 strings from the device that have multiple null terminators,
112 and changed the record parsing code to use this new function
113 - moved FullName() function from ldif.cc to r_contact.cc since it
114 is generally useful in areas outside of LDIF processing
115 - refactored postal addresses in the Contact record class into
116 a new class called PostalAddress
117 - renamed Address class in record.h to EmailAddress
118 - moved Contact::GetPostalAddress() to record.cc as part of the
119 new PostalAddress class and renamed it
120 PostalAddress::GetLabel()
122 - refactored vformat helpers into separate base class
124 - added vCard converter class
125 - commented out the code that skipped over the adding
126 of empty attribute values and parameters,
127 since VCARD depends on some of this behaviour
128 - movified vformat.c:vformat_find_attribute() to take
129 an extra nth argument, in order to search
130 for multiple attributes with the same name...
131 the way this is coded is a bit inefficient...
132 but we're moving away from vformat when we move
133 to OpenSync 0.3x / 0.40, so this is just temporary
136 - added SUBSYSTEM=="usb_device" to example udev rules, in order
137 to avoid running bcharge for each endpoint that is
138 added by udev, when first plugged in. Running bcharge
139 back-to-back multiple times can crash some
143 - fixed ID mapping bug in the case of a non-numeric UID
144 from syncing external -> Barry, and then the
145 new Barry ID would look like a new change
146 when syncing back from Barry -> external.
147 - added support for slow sync
148 - added clear() to idmap class
149 - added vformat attr parameter extraction class
150 - added more trace logging during vformat/Barry conversions
151 - added Notification time support
152 - handled "all day" vformat special case, when DTEND
153 does not exist in a vformat data block
154 - changed the min1900_t time conversion functions to handle
155 the 0xffffffff special case, when notifications
156 are turned off on a calendar event (src/time.cc)
157 - also made the calendar data dump output clearer for
158 disabled timestamps, and added debug output
159 for the raw min1900 value, in case there are more
160 special cases in the future
162 - fixed null pointer access in opensync's HasMultipleVEvents()
164 - upper management directive: revert all OpenSync 0.3x changes
165 and proceed with OpenSync 0.22. Reverting the
166 opensync-plugin/src tree to Barry 0.7 status.
168 - reorganized exception and trace code in opensync plugin
169 - added XmlToCalendar() function, based on unfinished opensync
170 library API that is not yet finished... will be
171 committed to opensync svn tree when complete
172 - added OSyncXMLField logger to plugin trace class
174 - applied email MessageRecord patch from Brian Edginton (thanks!)
175 - updated Message::Dump() to output more valid mbox data...
176 the goal is to be able to do:
177 btool -d "Messages" > email.mbox
179 - changed some TODO's to FIXME's to keep it grep-consistent
180 - applied Folder support patch from Brian Edginton
181 - fixed pedantic compile warnings, and changed order in src/Makefile.am
182 - removed unneeded cast in src/r_folder.cc
184 - minor adjustment to ctags building in buildgen.sh
185 - applied patch from Brian Edginton for better boost detection
187 - added boost usage message to btool -h output
189 - more porting of OpenSync module to 0.30
190 - added ctags build for opensync source tree to buildgen.sh
192 - changed PKG_CONFIG_PATH setting in rpm script, so that systems
193 like SuSE that have an extensive PKG_CONFIG_PATH
194 in general use will not break on build
195 - added PKG_CONFIG_PATH to debian script too, just in case :-)
196 - added debian/changelog to doc/ReleaseChecklist.txt
197 - added comment to rpm/barry.spec regarding SuSE RPM builds
199 - applied patch from Niels de Vos fixing debian package build
201 - applied Brian Edginton's Saved Email Messages patch
202 - updated r_saved_messages.cc comment to match header
203 - added GetRecType()/GetUniqueId() to r_saved_messages.h, since
204 SetIds() is already there
205 - removed Address version of operator<<() from message related headers
206 and moved into record.h
207 - added std::hex to message related dumps, and changed output to
209 - updated doc/CodingStyle.txt to cover switch() statements
210 - opensync compiles again, with 0.30! (not functional yet)
211 - moving closer to more modular and reusable plugin functions
212 - updated release checklist for opensync version number
214 - added doc/CodingStyle.txt... preliminary version
216 - applied Brian Edginton's PIN message patch
217 - changes to PIN message patch:
218 - fixed missing backslash in src/Makefile.am
219 - fixed tab alignment throughout
220 - removed typedef from Address struct
221 - added clear() to Address... might as well, if it it is
222 going to be a standalone type
223 - fixed missing comma in PINMessageFieldLinks table
224 - changed MessageRecordId in PINMessage to load from a new
225 uint32_t field in CommonField, instead of piggy
226 backing on a field for other use
227 - did some testing and found that PINMessage's MessageRecordId
228 is the same as the RecordId given through the SetIds()
229 API... but these Id's are duplicated in the protocol...
230 RecordId is stored in the protocol header, while
231 MessageRecordId is stored in the record it self as
232 a common field... added RecType and RecordId to
233 PIN Message so that we store both of these
236 - added item to gui/TODO
237 - applied Brian Edginton's s11n-boost.h patch
239 - renamed template values to more consistent naming scheme
240 - applied memos and tasks patch from Brian Edginton, slightly
241 modified to split into the new separate files system
242 - changes to memos and tasks classes:
243 - reformatted the code to match coding style
245 - added copyright notice
246 - changed class to singular (Memo, not Memos)
247 - placed in Barry namespace
248 - added #ifdef header protectors
249 - added new files to Makefile.am
250 - added new r_*.h headers to record.h
251 - added support for Memos and Tasks to tools/btool.cc, based on
252 Brian Edginton's patch
253 (currently commented out, for Boost library reasons):
254 - added doc/CommitPolicy.txt
257 - added ctags generation to buildgen.sh
258 - applied record.{h,cc} fix patch from Brian Edginton
262 - fixed broken formatting in above patch
263 - refactored record parser classes into separate files for each
264 - changed the ConvertHtoB<> template to inline, as it caused
265 linker errors with the new file layout
267 - fixed bug in Debian packaging that put util commands in
268 /bin and /sbin instead of /usr/bin and /usr/sbin
269 - commited byteswap.h replacement code from gm2net
271 - removed unneeded debug messages on stdout
272 - changed the default for debug output to "off"
273 and added a command line option to turn
274 it on if necessary (--debug-output)
275 - command line option --help now also works
276 - fixed bug in Calendar class where it didn't initialize
277 RecType properly... and cleaned up other RecType
278 code to use the GetDefaultRecType() functions,
279 localizing the constants in one place
281 - applied patches from Brian Edginton, parsing more
282 fields in Contact and Calendar records
283 - added Brian Edginton to AUTHORS
284 - reformatted record.{h,cc} to match the rest of the code,
285 removed commented out code, and removed mention of
286 Outlook from the comments.
287 - updated udev/10-blackberry.rules.Debian for Debian Etch
288 - updated debian/rules to use new udev rules file
289 - removed debian/postinst, as stable versions of Debian now
290 have a recent udev, and don't need the special handling
291 - updated Debian build scripts to build separate packages:
296 - added menu support to Debian barrybackup-gui package
297 - made the buttons bigger in the backup GUI, as they were too
300 - more progress in porting to opensync 0.30, heavily
301 based on the example plugin in the opensync
302 svn tree... this is a work in progress, and so far
306 - partial porting to opensync 0.30 (in progress!)
307 - fixed bug in sync_done where it reported context error
308 and then success afterward
309 - checking in new INSTALL files based on Etch's autoconf run
310 - added item to gui's todo list
311 - bumped version numbers to 0.8 in:
315 - gui/src/BackupWindow.cc (about dialog)
319 Release: version 0.7 - 2007/05/02
320 ------------------------------------------------------------------------------
322 - added rpm building to maintainer/release scripts
323 - added "HEAD" feature to tarball creation script, for easier testing
324 - updated maintainer/README
325 - updated maintainer/release.sh to use new HEAD feature too
327 - added pppob to rpm barry.spec
328 - bumped release number in barry.spec
331 - fixed size check bugs (strlen() == 0 when empty)
332 - finished refactoring calendar conversion routines
333 - started timezone parsing, but this may go away
334 if we use opensync 0.30's time functions and
336 - added development note to README
338 - updated README and opensync-plugin/README with more current
341 - refactored some calendar conversion routines
342 - added Barry->vCalendar conversion for recurring appointments
345 - removed some unneeded trace logs
346 - fixed VEventConverter constructor bug, with uninitialized
348 - eased up the requirement that every calendar event
349 must have a summary... defaults to "<blank subject>"
351 - added more VCALENDAR fields to match Barry::Calendar record
352 - fixed case sensitive compares in vformat.c
354 - opensync module work:
355 - added id map class to map string uid's from OpenSync to
356 numeric ID's from the Blackberry
357 - fixed bug in CommitChange() where it was still using
358 a hard coded calendar record state table
359 - added lots of logging throughout
360 - fixed bug where newly added records didn't use the
361 change object's uid, but created a new ID,
362 causing another sync on the next run
363 i.e. (CommitData_t was missing an argument)
364 - fixed bug where uninitialized notification time was
365 written to the Blackberry
366 - more opensync module work:
367 - refactored the environment class into another subclass
368 called DatabaseSyncState, and moved some stand-
369 alone functions from barry_sync.cc into
370 the environment classes
371 - added a tagged constructor to the trace class
372 - fixed idmap::Load() to clear map before loading
374 - renamed barry-config to barry-sync, the default configuration
375 file for the opensync module... and filled it with
376 a sample configuration
377 - added code to update the calendar cache, and write it to disk
378 on success (opensync module)
379 - added code to clear dirty flags in the device (opensync module)
380 - allowed comments in the config file
381 - added vformat.c from libopensync 0.22, since the vformat
382 routines are not available from the plugin side
385 - added -d switch to bcharge to switch back from 0001 to 0004 mode,
386 changed the auto-detection logic to ignore iProduct and
387 only use idProduct, and updated man page
388 - updated breset to recognize the Pearl in 00004 mode as well
389 - changed src/probe.cc logic to assume that the second endpoint
390 pair is the one needed for database communication, instead
391 of doing a reverse search from the end... newer Pearls
392 in Dual mode seem to have at least 3 or 4 endpoint pairs,
393 and the old search didn't always work
394 - applied sourceforge patch #1696884 from Peter Silva, updating
395 the Debian build for Etch, and including all files in one deb
396 - added Peter Silva to AUTHORS file
397 - uncommented the "make distclean" in debian/rules, and removed
398 the reliance on /tmp when installing 99-barry-perms.rules
399 - applied bb_task_format.txt patch from Peter Silva
400 - fixed each misspelling of "recurrance" to "recurrence"
401 - tweaked src/probe.cc again for paranoia's sake... start at
402 offset 1 if more than 1 endpoint pair, otherwise start
404 - split up opensync module into multiple source files, and started
405 using the opensync vformat routines for parsing and building
406 the vevent20 data (experimental)
408 - more incremental work on the opensync plugin, adding ADDED
409 and MODIFIED functionality... (experimental!)
411 - added code to dump extra descriptors during device discovery,
413 - bumped opensync version number in configure.ac, and made
414 the configure scripts more consistent with the rest
416 - fixed printf-format size bug in opensync module
418 - applied patch from Niels de Vos, fixing deb udev rules for Pearl
420 - fixed extra space bug in LDIF output (added space when creating
421 FullName, even if there was no last name)
422 - fixed Address Book record parsing in Contact class... the
423 Blackberry database uses field id 0x20 for both the
424 first and last names, and if only a last name exists,
425 the first name will be empty. Fixed parser to handle
426 this special case, without relying on string length.
427 Thanks to Troy Engel for reporting above two LDIF bugs.
428 - added a few more constants to bcharge.cc
429 - renamed opensync's configure.in to configure.ac
431 - fixed bcharge to avoid endless reset loop... oops
433 - reorganized bcharge.cc to make it possible to switch between
434 Pearl modes 0004 and 0001
435 - added flag to BadPassword exception class, to signal whether
436 the library considers the device "out of retries" or not
437 - updated socket.cc for new BadPassword exception
438 - added special case to Controller class destructor to reset the
439 device if being destroyed in a half-open state... this is
440 a temporary fix until we understand how to handle
441 the "already open" message we get when opening the Desktop
443 - added password prompt dialog to the backup GUI
444 - fixed possible memory leak in GUI's DeviceInterface::Connect()
446 - committed Troy Engel's bbrules_mass_symlinks.patch, fixing
447 udev rules for the Pearl
448 - updated 10-blackberry.rules.Debian to match 10-blackberry.rules's
450 - applied Troy Engel's barry.spec patch to generate all required
451 RPM's... Thank you! Notes in the README taken from
452 the sourceforge tracker entry containing the patch.
453 - added libusb lib flags to tools/ and examples/ makefiles,
454 and tweaked configure.ac with a better default,
455 to build cleanly on RHEL 4
456 - modified rpm/barry.spec:
457 - does not create doc-only barry base package anymore
458 - fixed "barry-backup" typo
459 - removed the patch step, as version 0.7 shouldn't need it
460 - added license file to each package
461 - commented out console perms scripts, pending mailing list
464 - applied Troy Engel's barry.spec patch to fix RHEL4 compile issues
465 - fixed 80 column wrapping error in barry.spec, thanks to Troy Engel
467 - applied Troy Engel's udev permissions patch for Redhat/Fedora
468 systems, which automatically set the ownership of a
469 newly plugged in device to the currently logged in console
471 - added udev/README to document what all those files are for...
472 part of the text comes from the sourceforge comment
473 on the above udev permissions patch from Troy Engel
474 - uncommented the console perms scripts for rpm/barry.spec
475 - added libglademm-2.4 to gui/README's dependency list
476 - made usbwrap.* calls more threadsafe
477 - added Usb::Timeout exception for timeout errors
478 - added preliminary UsbSerData support to library for GPRS modem
479 functionality, using Rick Scott's XmBlackBerry's serdata.c
481 - added overloaded Socket::Send() for sending without receiving
482 - added preliminary pppob command line tool
488 - gui/src/BackupWindow.cc (about dialog)
491 - build system tweaks for opensync-plugin, and added a buildgen.sh
493 - added better debug messages in probe.cc and usbwrap.cc's
494 device discovery code
495 - changed error handling for null interface->altsetting pointers
496 when doing device discovery in usbwrap.cc. Seems that
497 some devices report a higher number in bNumInterfaces
498 than are really available... in this case, let's
499 just ignore the missing data and assume success
500 - applied udev script patch #1663986 from Troy Engel
501 - applied barry-bcharge.spec patch #1672178 from Troy Engel
502 - ReleaseChecklist update
503 - added rpm/README based on Troy Engel's patch notes
504 - updated new barry-bcharge.spec for version 0.7, removing patch
507 Release: version 0.6 - 2007/02/28
508 ------------------------------------------------------------------------------
510 - fixed size checks for device database records larger than 64K
511 Thanks to Michael Brown for the bug report.
512 - added initial release maintainer scripts
513 - fixed odd backup success message in GUI when doing a backup with
514 no databases selected in configuration
515 Thanks to Michael Brown for the bug report.
516 - added udev permissions file patch from Michael Brown
517 - added Michael Brown to AUTHORS
518 - updated README, gui/README, and doc/USB-capture.txt
519 - added -o command line switch to bcharge.cc, and updated man page
521 - changed library behaviour so it only turns on libusb debug output
522 if in data dump mode. In btool, -v controls this output.
523 - minor stream formatting fix in btool
524 - added OpenSSL dependency to build for password hash support
525 - made debug dout() output conditional on Init() flag... In btool,
526 -v now controls this as well
527 - reverse engineered more of the initial probe packets, and
528 implemented probing in struct based code... changes include:
529 - added support for "attribute fetch" packets, used at
530 the beginning of USB conversations
531 - defined object/attribute id codes for the PIN number
532 - made Protocol::SocketCommand more robust, and implemented
533 its sequence number behaviour
534 - moved ModeSelectCommand into SocketCommand, as it is a
535 socket 0 level command
536 - added initial protocol struct: PasswordChallenge
537 - renamed Barry::Packet to DBPacket
538 - added ZeroPacket for socket 0 packet creation and analysis
539 - Barry::Packet is now the base class for the above
540 - implemented socket 0 behaviour in socket.cc
541 - changed return values for Send/Receive/Packet/NextRecord
542 members in the Socket class from bool to void, as they
543 don't generate errors themselves, but only can
544 have an error if Usb::Device throws an exception.
545 Cleaned up all code that called these members as well.
546 - removed "last status" from Socket class as it was unused
547 - updated probe.cc to use new socket 0 implementation
548 - removed two of the hard coded packet handshakes in favour
550 - updated controller.cc:
551 - uses new socket 0 implementation
552 - removed "flag" hack when opening sockets and
553 selecting modes, as it should use the
554 socket 0 sequence number
555 - added initial, incomplete prototype header for the C API
556 - updated some .cvsignore files
558 - updated ReleaseChecklist.txt, adding rpm spec file
559 - added the following to ZeroPacket:
564 - moved Command() to base Packet class
565 - added comments to protostructs.h and size constants for the
566 new PasswordChallenge struct
567 - added password support to Socket and Controller classes
568 Thanks to Rick Scott's XmBlackBerry for the openssl
569 password hashing logic.
570 - updated btool to use new password support, and updated its man page
572 - updated Doxygen input files and version, and added to ReleaseChecklist
573 - fixed btool PIN output for LDIF mode, now commented in output
574 - major LDIF overhaul:
575 - split LDIF specific code out of the record class and
577 - fixed "full name" behaviour that added an extra space
578 in dn attributes, as reported by Troy Engel
579 - added support for mapping of LDIF attributes to Barry
581 - fixed base64 handling, so attributes are only so encoded
583 - fixed base64 attribute reading inconsistency... some
584 notes fields were missed in old version if not
586 - added support for specifying alternate attribute for
587 constructing a FQDN, instead of just "cn"
588 - updated btool and upldif to use new features
589 - large update to btool manpage
590 - fixed initialization bug in Contact constructor (missed RecType)
592 - added ktrans.cc to make reading of kernel usbfs_snoop logs easier
593 - removed commented code from translate.cc
594 - added Pearl handshake support to bcharge.cc, which resets the
595 Pearl so that the vendor specific 0xFF class exists.
596 This should allow use of btool on the Pearl.
597 - added Data::AppendHexString to data.cc
598 - added better probing support for the Pearl, so it detects the
599 Product ID 0x0004 mode, and does a search for class 0xFF
600 instead of hardcoding it.
602 - applied README patch from Ian Darwin
603 - minor tweaks to README
604 - added missing errno.h header in src/usbwrap.cc for AMD64
605 Thanks to Jonathan Hudson for reporting the bug.
606 - tightened up size checks in Data::ReleaseBuffer()
607 - stable libusb *does* return the actual length of the USB packet...
608 fixed this in usbwrap.cc to set the actual size of read data,
609 and removed the size hacks elsewhere in the code:
610 probe.cc, protocol.cc, socket.cc, usbwrap.cc
612 - added proper return codes for error conditions in btool.cc
613 - clarified some stream output code in the library, for hex/dec numbers
614 - removed old, non-working connect.cc test program
615 - added iomanip to debug.h
616 - changed usbwrap to call libusb again if EINTR and EAGAIN
617 This should fix one of the OpenBSD issues
620 - fixed bug in barrybackup GUI that didn't create the target path
621 if the user didn't do Edit | Config first.
622 Thanks to Ian Darwin for reporting the bug.
623 - applied gui/src/ConfigFile.cc header fix patch for OpenBSD
626 - added initial btool.1 man page from Ian Darwin
627 - updated man/Makefile.am for new btool.1
629 - applied btool -h help clarification patch from Ian Darwin
631 - reworked exception hierarchy, so applications can handle all
632 Barry related exceptions, including Usb errors, with
633 one base class: Barry::Error
634 - fixed all the ripple changes this caused throughout the codebase
635 - added BadPassword exception
641 - gui/src/BackupWindow.cc (about dialog)
642 - updated doc/ReleaseChecklist.txt
643 - added usb-level breset.cc command line tool
644 - fixed bug in gui/src/tarfile.* for systems where
645 sizeof(int) != sizeof(void*)
646 - TarFile class now uses a plugin style for the compression
647 operations, to allow for threadsafe versions if needed
649 - also fixed leaked file handle bug in open_compressed()
650 - Thanks to Jonathan Hudson for reporting this for 64bit systems
651 - minor signed comparison fixes in the library
652 - added better protocol data dumping code in probe.cc
653 - added Usb::Device::BulkDrain(), to prevent Barry from hanging
656 Release: version 0.5 - 2007/01/26
657 ------------------------------------------------------------------------------
659 - added missing errno.h header to gui/src/util.cc
660 - lots more endian fixes after testing on iMac
661 - iostream output size fixes
662 - added option to btool to reset the device via software
663 - beefed up exception handling in barrybackup
664 - checked in opensync work-in-progress: trace logging, pointer bugfix
665 this is still experimental
666 - added aboutdialog.* files from gtkmm24 2.10.6, so we have a dialog
667 class for systems running older versions of gtkmm, like
669 - uncommented the about box menu handler
670 - added "cleanall" option to root buildgen.sh script
671 - update ReleaseChecklist.txt
672 - removed CVS auto-history strings from aboutdialog.*
674 - found a link between the unknown field in tagged protocol headers
675 and the unknown field in RecordStateTable entries. This
676 field matters when uploading certain databases, such as
677 Browser Options. It is unknown what this field really
678 means, but we're calling it RecType. Updated the API
679 to use this in all parser objects and builder objects
680 and record objects that use it.
681 - changed some uses of uint64_t to uint32_t, in Contact and
682 Calendar record objects... unique IDs only seem to be
684 - added new static function to record classes: GetDefaultRecType()
685 This returns the default rectype that should work
686 for that record. Ideally this should be retrieved from
687 the device, but in the case where a database has no
688 records, this info would be good to know.
689 - updated convenience template function Controller::AddRecordByType()
690 to use GetDefaultRecType()
691 - documented the recurrence data and how to use it in Calendar objects
692 - added data debug output for incoming USB packets in the
693 socket class... when we ported to the stable libusb
694 we lost some of the verbose output... this should get it
695 back. (In the official release, you need to uncomment
696 ddout() in debug.h for the full blast of debug output).
697 - backup GUI updates:
698 - changed error_done signal into a pure error signal,
699 and changed the logic so the restore continues
700 even if there was a protocol error on one database
701 - changed status bar messages to show what's happening
702 i.e. Backup or Restore
703 - default to not restore the "Handheld Agent" database as
704 it appears to be read-only on the device
705 - added temporary Restore & Backup mode for debugging
706 - added extra hyphen to backup files to show date/time better
707 - updated GUI to save and restore the new RecType field
708 supported now by the library
710 - added future considerations to doc/ReleaseChecklist.txt
711 - applied big endian patch from Jonathan Hudson
713 - removed endian.h include from barry.h, since applications
714 using the library should not care about endianness
715 - hard coded the path to config.h in endian.h
716 so that endian.h never makes it into the final
718 - removed DEFAULT_INCLUDES setting from src/Makefile.am
720 - removed config.h dependency in btool.cc and used library
722 - added big/little endian text to the Barry::Version call
724 - fixed install location of bcharge when building with ./configure;
725 now goes to sbin/ for both source compile and binary packages
726 Thanks to Jonathan Hudson for reporting this bug.
728 - added smarter timeout values to controller, and allowed
729 timeout overrides in the packet, socket, and usbwrap code...
730 extra timeouts are needed for erasing databases for
732 - added an explicit DataDumpParser class to btool.cc, as the
733 library doesn't always have extreme verbose debugging on...
734 this makes sure anyone can always get raw database packets
735 through btool, for development
736 - added thread callback signals for error finishes and erase_db
737 operations, so the GUI is updated properly
738 - more GUI cleanups (status bar updates, etc)
739 - finished implementation of GUI restore
740 - fixed gtkmm/glibmm exception handling when the exceptions
741 occur in sigc signal handlers... in that case,
742 exceptions need to be handled with
743 Glib::add_exception_handler()
744 - added filename-only reading support to TarFile class
745 - added pending gui/TODO items
747 - backup GUI now properly shows progress
748 - the buttons become insensitive when backup is busy
749 - refactored directory check code, and added mkdir check
750 when to the backup setup (so the directory exists
751 when we try to write to it)
752 - updated code to use Barry::Data in new namespace
754 - moved Data class into Barry namespace
756 - updated copyright dates for the new year
757 - clarified exception message in controller.cc
758 - fixed pointer initialization bug in controller.cc
759 - added documentation comments to parser.h and builder.h
760 - updated ReleaseChecklist.txt
762 - changed Barry library configure scripts to make better
763 use of pkg-config for autodetecting dependencies
764 - added initial Barry Backup GUI (gui/ directory), capable of making
765 backups only at this point
767 - added doc/ReleaseChecklist.txt
768 - added version API, returning version number and string
769 - added Troy Engel to AUTHORS
770 - added special case to bcharge.cc for the Blackberry Pearl,
771 which doesn't reset itself after the charge handshake
772 Thanks to Troy Engel for testing.
774 - added pkg-config support for the library (libbarry-0.pc.in)
775 - fixed compile error on g++ 3.3 systems (missing stdint.h
777 - fixed protocol hang at end of Controller operation... in
778 controller destructor, it was deleting the USB interface
779 before closing the socket at the BlackBerry level, thereby
780 leaving the device in Desktop mode
782 - support the newer udev packages on Ubuntu, which need slightly
784 - modified bcharge.cc to search for Pearl devices too.
786 - added Pearl product ID to probe code, but likely not functional
787 yet until we find someone able to get a USB capture for us
789 - version bump in configure.ac and src/Makefile.am
791 Release: version 0.4 - 2006/12/29
792 ------------------------------------------------------------------------------
794 - added bcharge.cc to the tools/ directory (whoohoo!)
795 - added initial release version of rpm/barry.spec (aka bcharge.spec)
796 This begins the barry binary release, currently only
797 including the bcharge utility, but possibly more in
798 the future. Future binary packages will likely include
799 libbarry, libbarry-devel, etc, which barry will then
801 - added udev rule script for BlackBerry devices
802 - added man/ directory and bcharge.1 manpage
803 - added debian/ directory for building deb packages, with initial
805 - added rpm/make-rpm-tarball.sh for building RPMs
806 - split out technical docs into separate libbarry.docs file for
807 a future Debian install
808 - added more cleaning to buildgen.sh
809 - added new udev rule file for Debian stable
810 - hardcoded g++ to 3.3 in debian/rules as I have multiple versions
811 on my system and need to compile releases with stable
813 - ported Barry to use the stable branch of libusb
814 NOTE: no longer depends on the devel libusb tarball!
815 Just use your distro's libusb packages.
816 - fixed more endian issues
817 - minor update to TODO
819 - moved Exceptions, Hacking, and VersionNotes to doc/
820 - added doc/USB-capture.txt
822 - added Controller::AddRecord() and template helper
823 - added beginning version of generic RecordFetch<> template
824 - set default of Barry::Init() to false (no data dump mode)
825 - added examples/ directory, with first addcontact.cc example
827 - added autoconf support based on Peter McAlpine's patch
828 - renamed Makefiles to Makefile.orig and added script to turn
829 on old build system if wanted
830 - fixed boost-specific error path in btool
831 - moved platform specific code out of time.h and into time.cc
832 - moved btool.cc, translate.cc, and upldif.cc out of src/ and
833 into their own tools/ subdirectory
834 - added NEWS file to make autotools happy
835 - added a "clean" option to the buildgen.sh script... maintainer's
837 - reworked autoconf and automake files, to support
838 barry as a library, to move the programs into their
839 own tools directory, and support a proper installation
840 - fixed headers in parser.h, btool.cc, and upldif.cc to work
841 properly when installed standalone
842 - added tools/Makefile.plain for testing compiling against
843 installed barry library
844 - added commentary on version number system
845 - removed Makefile.conf and Makefile.orig, as new autoconf
847 - added support for boost serialization library in autoconf build
848 - added library version number to autoconf build
849 - changed boost flag to __BARRY_BOOST_MODE__ to avoid collisions
850 - removed debug.h dependency from btool.cc
851 - removed libusb check in favour of specific --with-libusb
852 option... may need to revisit this later
853 - quick installation update in README
855 - fixed some missed endian conversions in record.cc and added
857 - fixed endian issues in time.h
858 - added 2 more functions to the time zone API, and adjusted
859 the time zone table for better defaults in North America
860 - added btool and upldif to the make install sequence
861 - added support for Calendar recurrence data
862 - small doc/TimeZones.txt update
864 - added BlackBerry time zone code list, reverse engineered from a 7750
865 - added GetTimeZone() to convert device time zone codes to useful data
867 Release: version 0.0.3 - 2006/11/24
868 ------------------------------------------------------------------------------
870 - added return code check to Controller::SaveDatabase()
871 - fixed Data() class constructor, so it doesn't auto-convert
873 - fixed size checking bug in Packet::ReturnCode() that caused
874 uploading to device to fail
876 - added Exceptions doc file, documenting the C++ exceptions used,
878 - fixed small bug in 'make install' which didn't build first
879 - minor documentation and cleanup in upldif.cc
880 - added utility functions to the RecordStateTable class,
881 GetIndex() and MakeNewRecordId()
882 - changed btool so that the -d command modifiers can be
883 specified multiple times to work with multiple records
885 - added Calendar sync code to the opensync module (not complete)
886 - cleaned up exception handling in the opensync module, making
887 more use of the base exception to make sure no
888 exceptions leak into the C library code
890 Release: version 0.0.2 - 2006/10/12
891 ------------------------------------------------------------------------------
893 - fixed casting error in opensync module
894 - minor updates to documentation to prepare for release
896 - added stdint.h to parser.h for uint32_t. Thanks Ron Gage for
899 - large refactoring of the packet parsing and building code,
900 splitting the code into 3 groups:
901 - record: handles record specific parsing and building...
902 this is mostly subfields with CommonField structs
903 as their headers, but sometimes has a specific
904 header of its own, in the case of email
905 Note: email's header is as yet undecoded.
906 Record does not know the details of the packet
907 formats, only the formats of its own record data.
908 - packet: handles building of complete command packets
909 and handles parsing of commonly used header values,
910 providing an API for it... packet does not know
911 the record format details, but does know the
912 details of each Database Operation (protocol.h)
914 So far, it seems that there is a real separation
915 of operation formats and record formats, with the
916 record formats being the same even with different
917 database operation codes (GET_RECORDS /
918 OLD_GET_RECORDS), which was not confirmed before.
919 See the header size of the email record after
921 - controller: handles device state, and the management of
922 sockets, commands, and checking for response values.
923 The Controller object does not know the format
924 of either the record or the packet it is sending,
925 but does know the protocol handshake logic used
926 to talk to the device.
927 - implemented SetRecord (SET_RECORD_BY_INDEX)... syncing, here
929 - spelling correction in opensync-plugin/AUTHORS
931 - added sample hotplug scripts
932 - added .cvsignore files for doxygen and opensync directories
933 - added src/endian.h... still need to add configure support to
934 handle non-little-endian machines
935 - added exception handler in opensync plugin commit_change()
936 - removed the device reset code in probe.cc, as resetting a device
937 can renumber the devices in /proc on kernel 2.4.x,
938 and cause a failure to reach the device...
939 FIXME - this should be changed someday to reset if probing
940 fails, and restart the probe on reset, as sometimes
941 the blackberry devices respond differently when not reset
942 Ideally, the probe protocol should be properly reverse
944 - added support for retrieving record state table, which is needed
946 - added support for retrieving, deleting, and clearing the dirty
947 flags of individual blackberry device records
948 - added endian macros throughout the code
950 - added opensync-plugin to root Makefile's clean
951 - updated time conversion calls to match opensync's latest SVN
952 - added opensync_change_set_changetype() (only add for now,
954 - removed slow sync functions for now
956 - added quick "install" makefile target, to src/install by default
957 - added static library target... this is temporary until autoconf
958 stuff is added properly... needed for the opensync plugin
959 - documented in comments that Barry::BError is intended to be
960 the base classes for all exceptions
961 - added first round of implementation of an opensync plugin, based
962 on the example plugin code from the opensync sources
964 - added to Contact record class:
966 - ReadLdif() to load and parse LDAP data
967 - added fragmented send support to Socket class
968 - added Probe::FindActive() to make client programs more streamlined
969 - fixed hex/decimal printing of record counts in btool
970 - added upldif.cc, a utility to take ldapsearch output on stdin
971 and upload it to a blackberry
972 - updated src/.cvsignore
974 - added parsing of service book fields:
976 - old and new unique ID (UID)
978 - old and new description
980 - bes domain / domain
981 - date change in legal.txt
982 - added initial parsing of ServiceBook field 0x09, and calling it
983 ServiceBookConfig for now, for lack of a better name
984 - added parsing for ServiceBook field 0x02 on 72xx Blackberries...
985 seems to be a name or description, but doesn't appear
986 in the device's own GUI... called "Hidden Name" for now
988 - added ServiceBook record class and parser code, based on some
989 of Ron Gage's reverse engineering... code is by Chris Frey
991 - fixed the libusb set_configuration/claim_interface order
992 bug in controller.{h,cc}... already fixed in probe,
993 but not controller. Thanks Ron Gage for finding this.
995 - renamed Changelog to ChangeLog in preparation for autoconf
996 - renamed Todo to TODO in preparation for autoconf
998 - make clean cleans up tests properly now
1000 - turned source code browsing on in Doxyfile
1001 - added LoadDatabaseByName() and LoadDatabaseByType() template members
1002 to the Controller class
1003 - new file: controllertmpl.h
1004 - documentation fixes in builder.h
1005 - added static GetDBName() functions to all record classes, in
1006 support of the template Controller functions
1007 - fleshed out DatabaseDatabase::GetDBNumber() and GetDBName(),
1008 returning errors properly
1009 - documented sample btool.cc code better, and added example
1010 of Controller template member usage
1012 - copyright dates updated for 2006
1014 - backed out update to latest libusb DEVEL tree, since it is buggy.
1015 Stick with libusb 2005/11/26
1017 - added clean target to convenience Makefile
1018 - added Boost to list of dependencies in README
1019 - fixed minor compile warnings in base64.cc
1020 - fixed new USB discovery code, and changed endpoint usage to use
1021 it, instead of the READ_ENDPOINT and WRITE_ENDPOINT
1022 constants... constants are now removed
1023 common.h, controller.cc, probe.cc, probe.h, usbwrap.cc
1024 - added ClearHalt() to Usb::Device (usbwrap.cc)
1025 - updated usbwrap.h to match latest libusb CVS devel tree
1026 - updated README to refer to 2005/12/30 libusb CVS devel tree
1027 - added roadmap to Todo list
1029 - added convenience Makefile in root directory
1030 - added Hacking document
1031 - minor touchups to README
1032 - added caution warning to README, since this release will support
1033 delete / overwrite / upload functionality
1034 - minor compile order change in Makefile (speed reasons)
1035 - added dynamic endpoint discovery to probe.cc
1036 - changed claim_interface/set_configuration() function call order in
1037 probe.cc based on feedback on the libusb mailing list
1038 - added USB discovery code to usbwrap.{h,cc}
1040 - fixed packing bug in new upload header struct
1041 - renamed CommonField union from data to u to be consistent
1042 - added return_code field to DBAccess struct, for error checking
1043 - added response checking during upload: if response is non-zero
1044 throw exception in controller.cc
1045 - added sorting operators to record parser classes... when writing
1046 contact data to the device, it may be important to write
1047 group link items last, since they reference other contacts...
1048 This needs to be tested, and if not required, removed.
1049 - added GROUP_FLAG field code to Contact parser class
1050 - fixed offset bug in Contact and Calendar Build() functions
1051 - fixed name bug in Contact::Build(), now output only with data
1052 - added appointment type flag support to Calendar record parser
1053 class, and "all day event" flag support
1055 - added autoconf to Todo list
1056 - refactored the protocol structures so that record structs can
1057 be reused for upload and download
1058 - with new protostructs, more header control is handled by
1059 controller.cc instead of record.cc parsers and builders
1060 - parser code now accepts an offset, to govern start of record data
1061 - changed Data sizes from int to size_t and removed casts
1062 - removed unneeded Parser::GetHeaderSize() (new protostructs make
1064 - added Calendar::Build() for calendar uploads
1066 - added s11n-boost.h serialization templates for boost::serialization
1067 - added boost::serialization to build
1068 - minor change in usbwrap.cc to display error codes in decimal
1069 - added SizePacket special case code to Socket::Send()... this is
1070 required when uploaded packet size is a multiple of 0x40,
1071 a special 3 byte size packet is sent
1072 - moved some record class variables to public: so they can be
1074 - fixed variable name but in RecordBuilder template
1075 - added -f switch to btool, for saving and loading of record data
1076 - added -s switch to btool, for uploading saved data back to handheld
1077 - added Controller::SaveDatabase()
1079 - fixed potential casting / pointer bug in controller.cc
1080 - added builder.h, to support uploading
1081 - minor comment fixes in parser.h
1082 - added preliminary Contact protocol record building support
1083 - fixed field order bug in ContactRecord protocol struct
1084 - Contact record.h class now saves unknown Group Link field,
1087 - added -fno-strict-aliasing to compiler flags so gcc-4.0.x won't
1088 optimize away protocol struct casts
1090 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25235
1091 http://mail-index.netbsd.org/tech-kern/2003/08/11/0001.html
1092 http://groups.google.ca/group/comp.lang.c++.moderated/ \
1093 browse_thread/thread/e7bf096832526f8e/5714701b \
1094 02a2a3cc?hl=en#5714701b02a2a3cc
1095 - changed headers so that any low level protocol-specific sizes and
1096 structs are completely hidden in the library...
1097 this adds library-only header: protostructs.h
1099 Release: version 0.0.1 - 2005/11/25
1100 ------------------------------------------------------------------------------
1102 - added barry.h for documentation purposes and application usage
1103 - updated README for release
1104 - renamed SBError exception class to BError
1105 - updated Doxyfile to version 1.4.5
1106 - added more doxygen comments
1108 - added conversion routines (and tests) between time_t and min1900_t
1109 - added -v option to btool to dynamically control protocol dumping
1110 - removed duplicated database database table display in controller.cc
1111 - minor spelling corrections
1112 - added copyright to usbwrap.{h,cc} as per earlier meeting
1113 - added support for Old protocol Calendar entries
1115 - added legal.txt, to save typing when adding new source files
1116 - ignore test-base64 file (cvsignore)
1117 - added contact-to-ldif generation, and updated supporting code
1118 - added argument to btool to support contact-to-ldif operation
1119 - added Contact::GetPostalAddress()
1120 - added README documentation in preparation for future release
1122 - added parser.{h,cc} virtual wrapper to generalize LoadDatabase()
1123 - added code to socket.cc's Packet() command to re-receive on
1124 reception of empty packets... this seems to be a signal
1125 from the device that it needs more time for long transfers
1126 Current limit: 10 blank packets before exception is thrown
1127 Observed maximum blank count: 1
1128 - added MessageRecord and OldMessageRecord to protocol.h
1129 - changed common record field structs to one CommonField struct
1130 - increased default USB timeout to 10 seconds (some loaded devices
1131 seem to need more time...)
1132 - reorganized record.{h,cc} to try to reduce the code duplication
1133 - record.h classes now "know" the sizes of their corresponding
1134 protocol structs... this may cause trouble down the line
1135 with Python, if used(?)
1136 - added Message class for Email records
1137 - removed the "Get*" database functions and replaced with
1138 general LoadDatabase() call, which loads any database
1139 available, and parses it with a Parser object
1140 - added GetDBID() for searching for database numbers via name
1141 - added test code to retrieve and display all email
1142 - cleaned up comments
1143 - renamed blackberry.{h,cc} to controller.{h,cc}
1144 - renamed bbtool.cc to btool.cc
1145 - added Makefile.conf
1146 - added copyright notices to source files
1147 - added COPYING file with GPL
1148 - put parser in Barry namespace, and added auto_ptr-like mode for
1149 store object pointers
1150 - added doxygen docs
1151 - added header comments to convo.awk script
1152 - removed Test() function in Controller object, and made class
1153 better suited to its general API purpose
1154 - changed OpenMode() so it loads command table and DBDB... it is now
1155 a public API function
1156 - moved Controller test code to btool.cc and made tool more
1157 generically useful... can now download any DB by name on
1159 - added mode checks to Controller
1160 - added -t and -d command line options to btool
1161 - added base64 routines from John Walker / Fourmilab and Citadel/UX
1162 - converted base64 routines into an API, instead of hardcoded
1165 - changed project name to Barry, including namespaces
1166 - renamed sbcommon.* to common.*
1167 - increased USB default timeout to 2 seconds
1168 - fixed formatting bug in Data operator<< output
1169 - cleaned up the protocol packet structs, getting rid of the
1170 confusing param/simple split... param is now the tableCmd
1171 from the Command Table
1172 - moved protocol structs from record.h to protocol.h
1173 - added support for older version of the USB protocol, which matches
1174 the Cassis spec more closely
1175 - fixed the hardcoded offsets in blackberry.cc's protocol parsing
1176 - fixed Socket::Close() so on error, socket flags are reset.
1177 This stops Close() from being called twice on exception
1180 - added Doxyfile to the project
1182 - added mode selection support to Blackberry class (RIM Desktop,
1183 RIM Bypass, and RIM_JavaLoader)
1184 Mode selection governs which socket is used in subsequent
1186 - added Command Table parsing and protocol support
1187 Blackberry class, and CommandTable record class
1188 - added Database database parsing and protocol support
1189 Blackberry class and DatabaseDatabase record class
1190 - added Address book retrieval support - dumps data to stdout for now
1191 - fixed bug in CopyOnWrite mode in Data class, where could allocate
1192 less buffer than data copied
1193 - added "easy exception" debug output macro
1194 - fixed minor spelling error in protocol.cc
1195 - fixed ModeSelectCommand protocol structure - it is not a parameter
1196 command, but uses the unknown slots for socket and flag
1198 - removed hardcoded socket numbers from protocol.h
1199 - fixed bug in socket.cc:Open() where packet size was not properly
1201 - implemented socket::Close()
1202 - fixed sequence number checks, so it is reset properly on socket 0
1203 - added Socket::NextRecord() for multiple record data transfers
1204 - fixed input handling bug in translate.cc
1206 - added .cvsignore files to project
1207 - added README notes to document the source architecture
1208 - added dbsetup data script, which shows how to get into database
1209 mode, using the RIM Desktop mode, and open a socket
1210 - added blackberry.{h,cc} class
1211 - added command line parameters to bbtool.cc
1213 -p = specify pin of blackberry device
1214 - added blackberry test code in bbtool.cc - debugging the initial
1215 opening protocol: sockets, modes, etc... see dbsetup for
1217 - added copy-on-write functionality to Data class, so it can be used
1218 with plain data buffers with little performance impact
1219 - added eout() and ddout() debugging - to separate:
1220 - regular debug output
1221 - exception debug output (should probably be logged to
1222 syslog someday, and never turned off)
1223 - raw data dump output
1224 - changed errno SBError exception handler to try to decode the error
1225 numbers into english (limited success)
1226 - fixed probe and socket USB Bulk message handling, so they don't
1227 use PollCompletions(), but use IO::Wait() (libusb_io_wait())
1228 - added common packet size checker in protocol.cc
1229 - fixed unpacked protocol packet structs... using gcc's
1230 __attribute__ ((packed)) ... see protocol.h
1231 - added ModeSelectCommand structure (protocol.h)
1232 - finished Socket::Open() call (Close() not yet implemented)
1233 - moved Socket::Send/Receive() out of private, as they can be used
1234 for socket 0 operations.
1235 - cleaned up pointer management with macros:
1239 - added IO::Wait() and IO::Cancel() to correspond with libusb API
1240 - added datadumping in usbwrap for easy protocol capture
1241 - added preliminary record.{h,cc} classes
1242 - added Contact record parsing, and test code
1244 - syncberry tree started