1 Release: version 0.10 - 2007/10/12
2 ------------------------------------------------------------------------------
4 - added better error checking to breset
5 - commented the sleep() for the 8830
6 - added code to handle the race condition where usb_storage
7 claims the device before bcharge arrives on the scene,
8 by asking the kernel to detach the driver from
9 the Mass Storage interface... this change makes it
10 possible to run bcharge, moving from 0006 to 0004
11 to 0001 and back to 0004, all *while* usb_storage
13 - changed Probe class behaviour so that it will not throw an
14 exception on Device Busy (-EBUSY) but instead
15 log the error in an internal array and keep going.
16 This makes it possible to use btool and barrybackup
17 with two devices plugged in at the same time.
18 BACKWARD COMPATIBILITY NOTE: this changes initializaion
19 behaviour, but in practice, it should only
20 make applications look more capable, not less.
21 - updated btool to show errors logged by Probe
22 - added message to specify which PIN device is being used by btool
23 - added explanatory comments to udev rules files, about why
26 - added ClearHalt() calls to probe and controller classes, just
27 before any communication happens... the 8830 seems
30 - added usbwrap.cc:GetConfiguration()
31 - added code to check whether the configuration is the one
32 we require, before calling SetConfiguration()... and not
33 to call it if unnecessary. This allows us to run
34 at the same time as usb_storage.
35 Thanks to Martin Owens for not letting this issue go. :-)
36 - added sleep to bcharge, before calling usb_reset()... the
37 Blackberry 8830 behaves like a Pearl device in that
38 it has mass storage support, but behaves like a Classic
39 in that it resets itself... the usb_reset() seems
40 to reset it back to 0006. The sleep allows the device
41 to disappear via its own reset, and let usb_reset()
42 fail naturally on these devices.
43 This is not an ideal solution... mainly for testing.
45 - added modprobe/blacklist-berry_charge as per Simon Ruggier's
46 suggestion. Thanks also to Niels de Vos for a similar
48 - fixed udev rules to work on both old and new distros
49 Thanks to Duncan Mak for the bug report.
51 - fixed string comparison to use case insensitive compare in
53 - updated Debian package to include modprobe blacklist file
54 - updated rpm spec file to include modprobe blacklist file
55 - split the maintenance scripts into release and test, so
56 that release package builds can be done faster
58 - opensync plugin fixes:
59 - fixed memory leak in vBase::GetAttr()
60 - added support in vformat parser to parse "blocks"
61 and keep track of the hierarchy when
62 parsing an iCalendar, etc.
63 - added support for searching for attributes only
64 inside certain blocks (i.e. find first
65 DTSTART in VEVENT block, not VTIMEZONE block)
67 - fixed bug in Calendar::Clear, missing the Free / Class
68 member variables, leaving them uninitialized
69 - added examples/addcalendar.cc
70 - added more logging to the opensync plugin, logging the
71 ToBarry conversion results
72 - added code to force struct tm.tm_isdst to -1 before calling
74 - fixed typo in previous fix
76 - renamed ktrans and translate to bktrans and btranslate to avoid
77 file name clashes with libtranslate of Gnome
78 Closes bug #1812410 reported by Troy Engel.
83 - gui/src/BackupWindow.cc
87 - opensync-plugin/src/barry_sync.cc
88 - fixed missing default config file in the debian package for
90 Closes bug #1812224 reported by Ariel (adonatti).
91 - added debian dbg packages for tracing segfaults
92 - modified maintainer scripts so debug packages would reference
93 source code located in /usr/src/barry-0.9, etc.
95 Release: version 0.9 - 2007/10/12
96 ------------------------------------------------------------------------------
98 - adjusted autosuspend disabling code in bcharge, based on new
99 information on how kernels 2.6.21 / 2.6.22 work.
100 See bcharge.cc for comments. If device/power/level or
101 device/power/autosuspend exist, charging should now work
102 without recompiling the kernel.
103 - updated website docs with autosuspend info
105 - large update to the doc/www web pages, moving some of the
106 information in README files to html format, so that it
107 is commonly available on the web
108 - added step in maintenance scripts to automatically build the
109 .html files in doc/www when making a tarball release
111 - added libopensync 0.22 version requirement to barry.spec file,
112 but it doesn't seem to work on OpenSuSE 10.2, where it's
115 - updated README to mention new tree build configure options
116 - fixed the configure scripts and makefiles so that when building
117 with custom CXXFLAGS and LDFLAGS environment variables,
118 it doesn't break the full-tree build mode, when
119 using --enable-gui and/or --enable-opensync-plugin
120 - added config.h.in~ to buildgen.sh clean target
121 - converted Debian build scripts to use cdbs
122 - added opensync plugin binary package for Debian
123 - added "Getting Started" section to web pages, as well as an
124 explanatory document for various system requirements
125 - turned on opensync build in RPM maintainer scripts
127 - added filename to power/state error message in bcharge.cc
128 - adjusted maintainer scripts to automatically adjust barry.spec
129 for opensuse rpm builds
131 - fixed some of the release build scripts and added support for
132 automated opensuse 10.2 builds
133 - fixed build target to i386
134 - changed scripts to add to PKG_CONFIG_PATH variable instead
137 - added website docs to doc/www... these are out of date at the
138 moment, and hopefully by including them here, they
139 will be updated more frequently
141 - applied Simon Ruggier's build system patch that allowed a build
142 of all 3 components from the root directory and fixed the
143 configure scripts to support external build directories
144 - reordered clean commands in the buildgen.sh scripts, and removed
145 the INSTALL files, since they are auto-generated by autoconf
146 - removed extraneous buildgen.sh calls in maintainer scripts
147 - added Simon Ruggier to AUTHORS
148 - added explanatory note to 3 stage test build process in
151 - fixed Category handling in the Contact record class, to
152 parse the comma separated string into an array,
154 Note: this involves a change to the boost serialization
155 and will affect backward compatibility with
156 old serialization datafiles
157 - added Category to examples/addcontact.cc
158 - added Category support to the opensync plugin
160 - fixed wrong read/write in btool output (oops)
161 - added BadSize exception, and changed CheckSize() function
163 - changed backup GUI to catch BadSize exceptions on initial
164 connect, and try a reset if that error occurs, since
165 that likely means the device didn't shutdown properly
166 - fixed in opensync plugin:
167 - some vcard data contains types codes in capital letters,
168 so added lower case conversion routine
169 - changed the order of address lines, so that address1
170 is the street, address2 is the extended address,
171 and address3 is the PO Box. The Blackberry
172 doesn't seem to have any PO Box field.
174 - added -e endpoint override option to btool
175 - added status message to endpoint override
177 - added configure check for strnlen, and implementation for
178 systems that don't have it, or that have it implemented
179 incorrectly. Went a little paranoid here, since
180 AC_FUNC_STRNLEN doesn't set a define in config.h,
181 so I had to do it manually, and check in case autoconf
182 changes in the future... ugh. See src/strnlen.h for
184 - added code in probe.cc to fetch the device's description.
185 The description is now part of the probe results
186 along with the PIN. Updates to protocol.h, adding
187 a new attribute definition.
188 Thanks to Rick Scott for pointing out the code in
189 XmBlackberry where he implemented this first.
190 - fixed header dependency in record-internal.h
191 - updated the "required autoconf" version to 2.61 for root build
192 - moved nested structs out of the record classes that contain them,
193 in order to better support SWIG and Python
195 - changed bcharge to use getopt() for its argument processing
196 - added options -p and -s to bcharge, to adjust usb suspend
197 settings automatically where possible
198 - updated the udev rules to use the new bcharge settings
199 - updated the bcharge man page with the new arguments
201 - cleaned up Usb exception handling, moving libusb error code
202 processing into the Usb::Error exception class.
203 error.cc is now an empty shell
204 - added return codes to the Usb::Error exceptions
205 - removed aboutdialog.{h,cc} from the backup GUI since even
206 Debian stable now has up-to-date gtkmm libraries
207 - added more user friendly error message for the "device busy"
208 case, recommending 'rmmod usb_storage'.
210 - added better error checking to bcharge.cc
215 - gui/src/BackupWindow.cc
219 - opensync-plugin/src/barry_sync.cc
220 - removed usb_set_configuration() check from bcharge.cc, since
221 that may fail under normal operation, in this case
223 Release: version 0.8 - 2007/08/03
224 ------------------------------------------------------------------------------
226 - moved Interface implementation to usbwrap.cc file
227 - added dout debugging messages to usbwrap.cc
228 - clarified unhandled packet error message in socket.cc
229 - added temporary workaround in opensync-plugin environment's
230 Reconnect(), for odd message seen on newer Blackberry
232 - fixing issues with Brian Edginton's MessageRecord patch (2007/06/29)
233 - fixed missing size check in message oriented ParseHeader()
234 - fixed endian conversion issues (also found one pre-existing
235 endian bug in the process)
236 - moved message timestamp converter code to time.cc
237 - added Barry version number to opensync plugin name
238 - changed maintainer scripts and rpm.spec to use bzip2 instead of gzip
239 - added opensync-plugin to configure script build
240 - added -S option to btool, to list supported parsers and builders
241 - slight change to opensync-plugin/buildgen.sh to cleanup config.h.in~
242 - opensync-plugin/README update
243 - added opensync-plugin/patches for others who may be building and
244 testing the plugin with OpenSync 0.22
245 - added debian package build scripts to maintainer/
247 - added entry to doc/TroubleShooting.txt
249 - added doc/TroubleShooting.txt
252 - fixed bug in vcard.cc that was using the FN field
253 for the broken-down name instead of N
254 - added dbId and dbName to DatabaseSyncState
255 - added Reconnect() functionality to BarryEnvironment class
256 and moved connect logic out of barry_sync.cc
257 - changed the batch commit functionality in barry_sync.cc
258 to single-change-commit. Why? Because opensync
259 0.22 groups all batch changes into one group
260 and calls one batch commit function, even if you
261 registered two. Moving to single change commit
262 mode removes a chunk of unneccessary code
263 from the plugin as well, and can in theory
264 ease memory pressure in the library.
265 - added FinishSync() to perform actions needed on successful
267 - disconnect on successful sync completion, and reconnect,
268 since the device's dirty flags don't seem to
269 be updated until a disconnect occurs.
271 - added ReadDataArray() to data.h, using an istream instead of
273 - added raw data version of BuildField() to record-internal.h
274 This is so that btool doesn't add extra null terminators
275 to unknown data, when restoring with option -s
276 - fixed bugs in r_contact.cc:
277 - fixed bug that caused a LastName/FirstName swap
278 if FirstName was empty and you wrote a contact
279 record into the device.... First/Last Name
280 fields are position sensitive, and the code
281 was missing the empty FirstName case on writing
282 - fixed bug in BuildFields() and Dump() where they
283 were not updated to handle the new field
284 table added for the postal address changes
287 - changed the plugin config file, so it is possible to
288 turn debug output on/off without recompiling
289 - moved the standalone Map2Uid() function into
290 the DatabaseSyncState class, and changed the
291 code to produce "contact-#" and "calendar-#"
292 UID strings, instead of just numbers
293 - changed default PIN to -1, so that if the user doesn't
294 configure properly, it will fail... this is a
295 safety measure to avoid syncing with different
296 devices and corrupting the cache/idmap...
297 this should be fixed someday, to handle this
299 On second thought... is this right? Perhaps
300 it is better to fail if the PIN is not what
301 is expected, so a user doesn't accidentally
302 sync the wrong device and lose data.
303 Actions to take on new PIN:
304 - force slow_sync, or
306 - incorporated the new VCardConverter code into the
307 main barry_sync.cc plugin code
308 - fixed bug that did not clear the devices dirty flags
309 in the case where a sync only read data from it,
311 - changed PIN error message number to hex
313 - minor wording change in backup GUI prompt: "Backup working..." to
314 "Backup in progress..."
315 - renamed Contact class's Title field to JobTitle for clarity
316 - added record function ParseFieldString(), which properly handles
317 strings from the device that have multiple null terminators,
318 and changed the record parsing code to use this new function
319 - moved FullName() function from ldif.cc to r_contact.cc since it
320 is generally useful in areas outside of LDIF processing
321 - refactored postal addresses in the Contact record class into
322 a new class called PostalAddress
323 - renamed Address class in record.h to EmailAddress
324 - moved Contact::GetPostalAddress() to record.cc as part of the
325 new PostalAddress class and renamed it
326 PostalAddress::GetLabel()
328 - refactored vformat helpers into separate base class
330 - added vCard converter class
331 - commented out the code that skipped over the adding
332 of empty attribute values and parameters,
333 since VCARD depends on some of this behaviour
334 - movified vformat.c:vformat_find_attribute() to take
335 an extra nth argument, in order to search
336 for multiple attributes with the same name...
337 the way this is coded is a bit inefficient...
338 but we're moving away from vformat when we move
339 to OpenSync 0.3x / 0.40, so this is just temporary
342 - added SUBSYSTEM=="usb_device" to example udev rules, in order
343 to avoid running bcharge for each endpoint that is
344 added by udev, when first plugged in. Running bcharge
345 back-to-back multiple times can crash some
349 - fixed ID mapping bug in the case of a non-numeric UID
350 from syncing external -> Barry, and then the
351 new Barry ID would look like a new change
352 when syncing back from Barry -> external.
353 - added support for slow sync
354 - added clear() to idmap class
355 - added vformat attr parameter extraction class
356 - added more trace logging during vformat/Barry conversions
357 - added Notification time support
358 - handled "all day" vformat special case, when DTEND
359 does not exist in a vformat data block
360 - changed the min1900_t time conversion functions to handle
361 the 0xffffffff special case, when notifications
362 are turned off on a calendar event (src/time.cc)
363 - also made the calendar data dump output clearer for
364 disabled timestamps, and added debug output
365 for the raw min1900 value, in case there are more
366 special cases in the future
368 - fixed null pointer access in opensync's HasMultipleVEvents()
370 - upper management directive: revert all OpenSync 0.3x changes
371 and proceed with OpenSync 0.22. Reverting the
372 opensync-plugin/src tree to Barry 0.7 status.
374 - reorganized exception and trace code in opensync plugin
375 - added XmlToCalendar() function, based on unfinished opensync
376 library API that is not yet finished... will be
377 committed to opensync svn tree when complete
378 - added OSyncXMLField logger to plugin trace class
380 - applied email MessageRecord patch from Brian Edginton (thanks!)
381 - updated Message::Dump() to output more valid mbox data...
382 the goal is to be able to do:
383 btool -d "Messages" > email.mbox
385 - changed some TODO's to FIXME's to keep it grep-consistent
386 - applied Folder support patch from Brian Edginton
387 - fixed pedantic compile warnings, and changed order in src/Makefile.am
388 - removed unneeded cast in src/r_folder.cc
390 - minor adjustment to ctags building in buildgen.sh
391 - applied patch from Brian Edginton for better boost detection
393 - added boost usage message to btool -h output
395 - more porting of OpenSync module to 0.30
396 - added ctags build for opensync source tree to buildgen.sh
398 - changed PKG_CONFIG_PATH setting in rpm script, so that systems
399 like SuSE that have an extensive PKG_CONFIG_PATH
400 in general use will not break on build
401 - added PKG_CONFIG_PATH to debian script too, just in case :-)
402 - added debian/changelog to doc/ReleaseChecklist.txt
403 - added comment to rpm/barry.spec regarding SuSE RPM builds
405 - applied patch from Niels de Vos fixing debian package build
407 - applied Brian Edginton's Saved Email Messages patch
408 - updated r_saved_messages.cc comment to match header
409 - added GetRecType()/GetUniqueId() to r_saved_messages.h, since
410 SetIds() is already there
411 - removed Address version of operator<<() from message related headers
412 and moved into record.h
413 - added std::hex to message related dumps, and changed output to
415 - updated doc/CodingStyle.txt to cover switch() statements
416 - opensync compiles again, with 0.30! (not functional yet)
417 - moving closer to more modular and reusable plugin functions
418 - updated release checklist for opensync version number
420 - added doc/CodingStyle.txt... preliminary version
422 - applied Brian Edginton's PIN message patch
423 - changes to PIN message patch:
424 - fixed missing backslash in src/Makefile.am
425 - fixed tab alignment throughout
426 - removed typedef from Address struct
427 - added clear() to Address... might as well, if it it is
428 going to be a standalone type
429 - fixed missing comma in PINMessageFieldLinks table
430 - changed MessageRecordId in PINMessage to load from a new
431 uint32_t field in CommonField, instead of piggy
432 backing on a field for other use
433 - did some testing and found that PINMessage's MessageRecordId
434 is the same as the RecordId given through the SetIds()
435 API... but these Id's are duplicated in the protocol...
436 RecordId is stored in the protocol header, while
437 MessageRecordId is stored in the record it self as
438 a common field... added RecType and RecordId to
439 PIN Message so that we store both of these
442 - added item to gui/TODO
443 - applied Brian Edginton's s11n-boost.h patch
445 - renamed template values to more consistent naming scheme
446 - applied memos and tasks patch from Brian Edginton, slightly
447 modified to split into the new separate files system
448 - changes to memos and tasks classes:
449 - reformatted the code to match coding style
451 - added copyright notice
452 - changed class to singular (Memo, not Memos)
453 - placed in Barry namespace
454 - added #ifdef header protectors
455 - added new files to Makefile.am
456 - added new r_*.h headers to record.h
457 - added support for Memos and Tasks to tools/btool.cc, based on
458 Brian Edginton's patch
459 (currently commented out, for Boost library reasons):
460 - added doc/CommitPolicy.txt
463 - added ctags generation to buildgen.sh
464 - applied record.{h,cc} fix patch from Brian Edginton
468 - fixed broken formatting in above patch
469 - refactored record parser classes into separate files for each
470 - changed the ConvertHtoB<> template to inline, as it caused
471 linker errors with the new file layout
473 - fixed bug in Debian packaging that put util commands in
474 /bin and /sbin instead of /usr/bin and /usr/sbin
475 - commited byteswap.h replacement code from gm2net
477 - removed unneeded debug messages on stdout
478 - changed the default for debug output to "off"
479 and added a command line option to turn
480 it on if necessary (--debug-output)
481 - command line option --help now also works
482 - fixed bug in Calendar class where it didn't initialize
483 RecType properly... and cleaned up other RecType
484 code to use the GetDefaultRecType() functions,
485 localizing the constants in one place
487 - applied patches from Brian Edginton, parsing more
488 fields in Contact and Calendar records
489 - added Brian Edginton to AUTHORS
490 - reformatted record.{h,cc} to match the rest of the code,
491 removed commented out code, and removed mention of
492 Outlook from the comments.
493 - updated udev/10-blackberry.rules.Debian for Debian Etch
494 - updated debian/rules to use new udev rules file
495 - removed debian/postinst, as stable versions of Debian now
496 have a recent udev, and don't need the special handling
497 - updated Debian build scripts to build separate packages:
502 - added menu support to Debian barrybackup-gui package
503 - made the buttons bigger in the backup GUI, as they were too
506 - more progress in porting to opensync 0.30, heavily
507 based on the example plugin in the opensync
508 svn tree... this is a work in progress, and so far
512 - partial porting to opensync 0.30 (in progress!)
513 - fixed bug in sync_done where it reported context error
514 and then success afterward
515 - checking in new INSTALL files based on Etch's autoconf run
516 - added item to gui's todo list
517 - bumped version numbers to 0.8 in:
521 - gui/src/BackupWindow.cc (about dialog)
525 Release: version 0.7 - 2007/05/02
526 ------------------------------------------------------------------------------
528 - added rpm building to maintainer/release scripts
529 - added "HEAD" feature to tarball creation script, for easier testing
530 - updated maintainer/README
531 - updated maintainer/release.sh to use new HEAD feature too
533 - added pppob to rpm barry.spec
534 - bumped release number in barry.spec
537 - fixed size check bugs (strlen() == 0 when empty)
538 - finished refactoring calendar conversion routines
539 - started timezone parsing, but this may go away
540 if we use opensync 0.30's time functions and
542 - added development note to README
544 - updated README and opensync-plugin/README with more current
547 - refactored some calendar conversion routines
548 - added Barry->vCalendar conversion for recurring appointments
551 - removed some unneeded trace logs
552 - fixed VEventConverter constructor bug, with uninitialized
554 - eased up the requirement that every calendar event
555 must have a summary... defaults to "<blank subject>"
557 - added more VCALENDAR fields to match Barry::Calendar record
558 - fixed case sensitive compares in vformat.c
560 - opensync module work:
561 - added id map class to map string uid's from OpenSync to
562 numeric ID's from the Blackberry
563 - fixed bug in CommitChange() where it was still using
564 a hard coded calendar record state table
565 - added lots of logging throughout
566 - fixed bug where newly added records didn't use the
567 change object's uid, but created a new ID,
568 causing another sync on the next run
569 i.e. (CommitData_t was missing an argument)
570 - fixed bug where uninitialized notification time was
571 written to the Blackberry
572 - more opensync module work:
573 - refactored the environment class into another subclass
574 called DatabaseSyncState, and moved some stand-
575 alone functions from barry_sync.cc into
576 the environment classes
577 - added a tagged constructor to the trace class
578 - fixed idmap::Load() to clear map before loading
580 - renamed barry-config to barry-sync, the default configuration
581 file for the opensync module... and filled it with
582 a sample configuration
583 - added code to update the calendar cache, and write it to disk
584 on success (opensync module)
585 - added code to clear dirty flags in the device (opensync module)
586 - allowed comments in the config file
587 - added vformat.c from libopensync 0.22, since the vformat
588 routines are not available from the plugin side
591 - added -d switch to bcharge to switch back from 0001 to 0004 mode,
592 changed the auto-detection logic to ignore iProduct and
593 only use idProduct, and updated man page
594 - updated breset to recognize the Pearl in 00004 mode as well
595 - changed src/probe.cc logic to assume that the second endpoint
596 pair is the one needed for database communication, instead
597 of doing a reverse search from the end... newer Pearls
598 in Dual mode seem to have at least 3 or 4 endpoint pairs,
599 and the old search didn't always work
600 - applied sourceforge patch #1696884 from Peter Silva, updating
601 the Debian build for Etch, and including all files in one deb
602 - added Peter Silva to AUTHORS file
603 - uncommented the "make distclean" in debian/rules, and removed
604 the reliance on /tmp when installing 99-barry-perms.rules
605 - applied bb_task_format.txt patch from Peter Silva
606 - fixed each misspelling of "recurrance" to "recurrence"
607 - tweaked src/probe.cc again for paranoia's sake... start at
608 offset 1 if more than 1 endpoint pair, otherwise start
610 - split up opensync module into multiple source files, and started
611 using the opensync vformat routines for parsing and building
612 the vevent20 data (experimental)
614 - more incremental work on the opensync plugin, adding ADDED
615 and MODIFIED functionality... (experimental!)
617 - added code to dump extra descriptors during device discovery,
619 - bumped opensync version number in configure.ac, and made
620 the configure scripts more consistent with the rest
622 - fixed printf-format size bug in opensync module
624 - applied patch from Niels de Vos, fixing deb udev rules for Pearl
626 - fixed extra space bug in LDIF output (added space when creating
627 FullName, even if there was no last name)
628 - fixed Address Book record parsing in Contact class... the
629 Blackberry database uses field id 0x20 for both the
630 first and last names, and if only a last name exists,
631 the first name will be empty. Fixed parser to handle
632 this special case, without relying on string length.
633 Thanks to Troy Engel for reporting above two LDIF bugs.
634 - added a few more constants to bcharge.cc
635 - renamed opensync's configure.in to configure.ac
637 - fixed bcharge to avoid endless reset loop... oops
639 - reorganized bcharge.cc to make it possible to switch between
640 Pearl modes 0004 and 0001
641 - added flag to BadPassword exception class, to signal whether
642 the library considers the device "out of retries" or not
643 - updated socket.cc for new BadPassword exception
644 - added special case to Controller class destructor to reset the
645 device if being destroyed in a half-open state... this is
646 a temporary fix until we understand how to handle
647 the "already open" message we get when opening the Desktop
649 - added password prompt dialog to the backup GUI
650 - fixed possible memory leak in GUI's DeviceInterface::Connect()
652 - committed Troy Engel's bbrules_mass_symlinks.patch, fixing
653 udev rules for the Pearl
654 - updated 10-blackberry.rules.Debian to match 10-blackberry.rules's
656 - applied Troy Engel's barry.spec patch to generate all required
657 RPM's... Thank you! Notes in the README taken from
658 the sourceforge tracker entry containing the patch.
659 - added libusb lib flags to tools/ and examples/ makefiles,
660 and tweaked configure.ac with a better default,
661 to build cleanly on RHEL 4
662 - modified rpm/barry.spec:
663 - does not create doc-only barry base package anymore
664 - fixed "barry-backup" typo
665 - removed the patch step, as version 0.7 shouldn't need it
666 - added license file to each package
667 - commented out console perms scripts, pending mailing list
670 - applied Troy Engel's barry.spec patch to fix RHEL4 compile issues
671 - fixed 80 column wrapping error in barry.spec, thanks to Troy Engel
673 - applied Troy Engel's udev permissions patch for Redhat/Fedora
674 systems, which automatically set the ownership of a
675 newly plugged in device to the currently logged in console
677 - added udev/README to document what all those files are for...
678 part of the text comes from the sourceforge comment
679 on the above udev permissions patch from Troy Engel
680 - uncommented the console perms scripts for rpm/barry.spec
681 - added libglademm-2.4 to gui/README's dependency list
682 - made usbwrap.* calls more threadsafe
683 - added Usb::Timeout exception for timeout errors
684 - added preliminary UsbSerData support to library for GPRS modem
685 functionality, using Rick Scott's XmBlackBerry's serdata.c
687 - added overloaded Socket::Send() for sending without receiving
688 - added preliminary pppob command line tool
694 - gui/src/BackupWindow.cc (about dialog)
697 - build system tweaks for opensync-plugin, and added a buildgen.sh
699 - added better debug messages in probe.cc and usbwrap.cc's
700 device discovery code
701 - changed error handling for null interface->altsetting pointers
702 when doing device discovery in usbwrap.cc. Seems that
703 some devices report a higher number in bNumInterfaces
704 than are really available... in this case, let's
705 just ignore the missing data and assume success
706 - applied udev script patch #1663986 from Troy Engel
707 - applied barry-bcharge.spec patch #1672178 from Troy Engel
708 - ReleaseChecklist update
709 - added rpm/README based on Troy Engel's patch notes
710 - updated new barry-bcharge.spec for version 0.7, removing patch
713 Release: version 0.6 - 2007/02/28
714 ------------------------------------------------------------------------------
716 - fixed size checks for device database records larger than 64K
717 Thanks to Michael Brown for the bug report.
718 - added initial release maintainer scripts
719 - fixed odd backup success message in GUI when doing a backup with
720 no databases selected in configuration
721 Thanks to Michael Brown for the bug report.
722 - added udev permissions file patch from Michael Brown
723 - added Michael Brown to AUTHORS
724 - updated README, gui/README, and doc/USB-capture.txt
725 - added -o command line switch to bcharge.cc, and updated man page
727 - changed library behaviour so it only turns on libusb debug output
728 if in data dump mode. In btool, -v controls this output.
729 - minor stream formatting fix in btool
730 - added OpenSSL dependency to build for password hash support
731 - made debug dout() output conditional on Init() flag... In btool,
732 -v now controls this as well
733 - reverse engineered more of the initial probe packets, and
734 implemented probing in struct based code... changes include:
735 - added support for "attribute fetch" packets, used at
736 the beginning of USB conversations
737 - defined object/attribute id codes for the PIN number
738 - made Protocol::SocketCommand more robust, and implemented
739 its sequence number behaviour
740 - moved ModeSelectCommand into SocketCommand, as it is a
741 socket 0 level command
742 - added initial protocol struct: PasswordChallenge
743 - renamed Barry::Packet to DBPacket
744 - added ZeroPacket for socket 0 packet creation and analysis
745 - Barry::Packet is now the base class for the above
746 - implemented socket 0 behaviour in socket.cc
747 - changed return values for Send/Receive/Packet/NextRecord
748 members in the Socket class from bool to void, as they
749 don't generate errors themselves, but only can
750 have an error if Usb::Device throws an exception.
751 Cleaned up all code that called these members as well.
752 - removed "last status" from Socket class as it was unused
753 - updated probe.cc to use new socket 0 implementation
754 - removed two of the hard coded packet handshakes in favour
756 - updated controller.cc:
757 - uses new socket 0 implementation
758 - removed "flag" hack when opening sockets and
759 selecting modes, as it should use the
760 socket 0 sequence number
761 - added initial, incomplete prototype header for the C API
762 - updated some .cvsignore files
764 - updated ReleaseChecklist.txt, adding rpm spec file
765 - added the following to ZeroPacket:
770 - moved Command() to base Packet class
771 - added comments to protostructs.h and size constants for the
772 new PasswordChallenge struct
773 - added password support to Socket and Controller classes
774 Thanks to Rick Scott's XmBlackBerry for the openssl
775 password hashing logic.
776 - updated btool to use new password support, and updated its man page
778 - updated Doxygen input files and version, and added to ReleaseChecklist
779 - fixed btool PIN output for LDIF mode, now commented in output
780 - major LDIF overhaul:
781 - split LDIF specific code out of the record class and
783 - fixed "full name" behaviour that added an extra space
784 in dn attributes, as reported by Troy Engel
785 - added support for mapping of LDIF attributes to Barry
787 - fixed base64 handling, so attributes are only so encoded
789 - fixed base64 attribute reading inconsistency... some
790 notes fields were missed in old version if not
792 - added support for specifying alternate attribute for
793 constructing a FQDN, instead of just "cn"
794 - updated btool and upldif to use new features
795 - large update to btool manpage
796 - fixed initialization bug in Contact constructor (missed RecType)
798 - added ktrans.cc to make reading of kernel usbfs_snoop logs easier
799 - removed commented code from translate.cc
800 - added Pearl handshake support to bcharge.cc, which resets the
801 Pearl so that the vendor specific 0xFF class exists.
802 This should allow use of btool on the Pearl.
803 - added Data::AppendHexString to data.cc
804 - added better probing support for the Pearl, so it detects the
805 Product ID 0x0004 mode, and does a search for class 0xFF
806 instead of hardcoding it.
808 - applied README patch from Ian Darwin
809 - minor tweaks to README
810 - added missing errno.h header in src/usbwrap.cc for AMD64
811 Thanks to Jonathan Hudson for reporting the bug.
812 - tightened up size checks in Data::ReleaseBuffer()
813 - stable libusb *does* return the actual length of the USB packet...
814 fixed this in usbwrap.cc to set the actual size of read data,
815 and removed the size hacks elsewhere in the code:
816 probe.cc, protocol.cc, socket.cc, usbwrap.cc
818 - added proper return codes for error conditions in btool.cc
819 - clarified some stream output code in the library, for hex/dec numbers
820 - removed old, non-working connect.cc test program
821 - added iomanip to debug.h
822 - changed usbwrap to call libusb again if EINTR and EAGAIN
823 This should fix one of the OpenBSD issues
826 - fixed bug in barrybackup GUI that didn't create the target path
827 if the user didn't do Edit | Config first.
828 Thanks to Ian Darwin for reporting the bug.
829 - applied gui/src/ConfigFile.cc header fix patch for OpenBSD
832 - added initial btool.1 man page from Ian Darwin
833 - updated man/Makefile.am for new btool.1
835 - applied btool -h help clarification patch from Ian Darwin
837 - reworked exception hierarchy, so applications can handle all
838 Barry related exceptions, including Usb errors, with
839 one base class: Barry::Error
840 - fixed all the ripple changes this caused throughout the codebase
841 - added BadPassword exception
847 - gui/src/BackupWindow.cc (about dialog)
848 - updated doc/ReleaseChecklist.txt
849 - added usb-level breset.cc command line tool
850 - fixed bug in gui/src/tarfile.* for systems where
851 sizeof(int) != sizeof(void*)
852 - TarFile class now uses a plugin style for the compression
853 operations, to allow for threadsafe versions if needed
855 - also fixed leaked file handle bug in open_compressed()
856 - Thanks to Jonathan Hudson for reporting this for 64bit systems
857 - minor signed comparison fixes in the library
858 - added better protocol data dumping code in probe.cc
859 - added Usb::Device::BulkDrain(), to prevent Barry from hanging
862 Release: version 0.5 - 2007/01/26
863 ------------------------------------------------------------------------------
865 - added missing errno.h header to gui/src/util.cc
866 - lots more endian fixes after testing on iMac
867 - iostream output size fixes
868 - added option to btool to reset the device via software
869 - beefed up exception handling in barrybackup
870 - checked in opensync work-in-progress: trace logging, pointer bugfix
871 this is still experimental
872 - added aboutdialog.* files from gtkmm24 2.10.6, so we have a dialog
873 class for systems running older versions of gtkmm, like
875 - uncommented the about box menu handler
876 - added "cleanall" option to root buildgen.sh script
877 - update ReleaseChecklist.txt
878 - removed CVS auto-history strings from aboutdialog.*
880 - found a link between the unknown field in tagged protocol headers
881 and the unknown field in RecordStateTable entries. This
882 field matters when uploading certain databases, such as
883 Browser Options. It is unknown what this field really
884 means, but we're calling it RecType. Updated the API
885 to use this in all parser objects and builder objects
886 and record objects that use it.
887 - changed some uses of uint64_t to uint32_t, in Contact and
888 Calendar record objects... unique IDs only seem to be
890 - added new static function to record classes: GetDefaultRecType()
891 This returns the default rectype that should work
892 for that record. Ideally this should be retrieved from
893 the device, but in the case where a database has no
894 records, this info would be good to know.
895 - updated convenience template function Controller::AddRecordByType()
896 to use GetDefaultRecType()
897 - documented the recurrence data and how to use it in Calendar objects
898 - added data debug output for incoming USB packets in the
899 socket class... when we ported to the stable libusb
900 we lost some of the verbose output... this should get it
901 back. (In the official release, you need to uncomment
902 ddout() in debug.h for the full blast of debug output).
903 - backup GUI updates:
904 - changed error_done signal into a pure error signal,
905 and changed the logic so the restore continues
906 even if there was a protocol error on one database
907 - changed status bar messages to show what's happening
908 i.e. Backup or Restore
909 - default to not restore the "Handheld Agent" database as
910 it appears to be read-only on the device
911 - added temporary Restore & Backup mode for debugging
912 - added extra hyphen to backup files to show date/time better
913 - updated GUI to save and restore the new RecType field
914 supported now by the library
916 - added future considerations to doc/ReleaseChecklist.txt
917 - applied big endian patch from Jonathan Hudson
919 - removed endian.h include from barry.h, since applications
920 using the library should not care about endianness
921 - hard coded the path to config.h in endian.h
922 so that endian.h never makes it into the final
924 - removed DEFAULT_INCLUDES setting from src/Makefile.am
926 - removed config.h dependency in btool.cc and used library
928 - added big/little endian text to the Barry::Version call
930 - fixed install location of bcharge when building with ./configure;
931 now goes to sbin/ for both source compile and binary packages
932 Thanks to Jonathan Hudson for reporting this bug.
934 - added smarter timeout values to controller, and allowed
935 timeout overrides in the packet, socket, and usbwrap code...
936 extra timeouts are needed for erasing databases for
938 - added an explicit DataDumpParser class to btool.cc, as the
939 library doesn't always have extreme verbose debugging on...
940 this makes sure anyone can always get raw database packets
941 through btool, for development
942 - added thread callback signals for error finishes and erase_db
943 operations, so the GUI is updated properly
944 - more GUI cleanups (status bar updates, etc)
945 - finished implementation of GUI restore
946 - fixed gtkmm/glibmm exception handling when the exceptions
947 occur in sigc signal handlers... in that case,
948 exceptions need to be handled with
949 Glib::add_exception_handler()
950 - added filename-only reading support to TarFile class
951 - added pending gui/TODO items
953 - backup GUI now properly shows progress
954 - the buttons become insensitive when backup is busy
955 - refactored directory check code, and added mkdir check
956 when to the backup setup (so the directory exists
957 when we try to write to it)
958 - updated code to use Barry::Data in new namespace
960 - moved Data class into Barry namespace
962 - updated copyright dates for the new year
963 - clarified exception message in controller.cc
964 - fixed pointer initialization bug in controller.cc
965 - added documentation comments to parser.h and builder.h
966 - updated ReleaseChecklist.txt
968 - changed Barry library configure scripts to make better
969 use of pkg-config for autodetecting dependencies
970 - added initial Barry Backup GUI (gui/ directory), capable of making
971 backups only at this point
973 - added doc/ReleaseChecklist.txt
974 - added version API, returning version number and string
975 - added Troy Engel to AUTHORS
976 - added special case to bcharge.cc for the Blackberry Pearl,
977 which doesn't reset itself after the charge handshake
978 Thanks to Troy Engel for testing.
980 - added pkg-config support for the library (libbarry-0.pc.in)
981 - fixed compile error on g++ 3.3 systems (missing stdint.h
983 - fixed protocol hang at end of Controller operation... in
984 controller destructor, it was deleting the USB interface
985 before closing the socket at the BlackBerry level, thereby
986 leaving the device in Desktop mode
988 - support the newer udev packages on Ubuntu, which need slightly
990 - modified bcharge.cc to search for Pearl devices too.
992 - added Pearl product ID to probe code, but likely not functional
993 yet until we find someone able to get a USB capture for us
995 - version bump in configure.ac and src/Makefile.am
997 Release: version 0.4 - 2006/12/29
998 ------------------------------------------------------------------------------
1000 - added bcharge.cc to the tools/ directory (whoohoo!)
1001 - added initial release version of rpm/barry.spec (aka bcharge.spec)
1002 This begins the barry binary release, currently only
1003 including the bcharge utility, but possibly more in
1004 the future. Future binary packages will likely include
1005 libbarry, libbarry-devel, etc, which barry will then
1007 - added udev rule script for BlackBerry devices
1008 - added man/ directory and bcharge.1 manpage
1009 - added debian/ directory for building deb packages, with initial
1011 - added rpm/make-rpm-tarball.sh for building RPMs
1012 - split out technical docs into separate libbarry.docs file for
1013 a future Debian install
1014 - added more cleaning to buildgen.sh
1015 - added new udev rule file for Debian stable
1016 - hardcoded g++ to 3.3 in debian/rules as I have multiple versions
1017 on my system and need to compile releases with stable
1019 - ported Barry to use the stable branch of libusb
1020 NOTE: no longer depends on the devel libusb tarball!
1021 Just use your distro's libusb packages.
1022 - fixed more endian issues
1023 - minor update to TODO
1025 - moved Exceptions, Hacking, and VersionNotes to doc/
1026 - added doc/USB-capture.txt
1028 - added Controller::AddRecord() and template helper
1029 - added beginning version of generic RecordFetch<> template
1030 - set default of Barry::Init() to false (no data dump mode)
1031 - added examples/ directory, with first addcontact.cc example
1033 - added autoconf support based on Peter McAlpine's patch
1034 - renamed Makefiles to Makefile.orig and added script to turn
1035 on old build system if wanted
1036 - fixed boost-specific error path in btool
1037 - moved platform specific code out of time.h and into time.cc
1038 - moved btool.cc, translate.cc, and upldif.cc out of src/ and
1039 into their own tools/ subdirectory
1040 - added NEWS file to make autotools happy
1041 - added a "clean" option to the buildgen.sh script... maintainer's
1043 - reworked autoconf and automake files, to support
1044 barry as a library, to move the programs into their
1045 own tools directory, and support a proper installation
1046 - fixed headers in parser.h, btool.cc, and upldif.cc to work
1047 properly when installed standalone
1048 - added tools/Makefile.plain for testing compiling against
1049 installed barry library
1050 - added commentary on version number system
1051 - removed Makefile.conf and Makefile.orig, as new autoconf
1052 system supercedes it
1053 - added support for boost serialization library in autoconf build
1054 - added library version number to autoconf build
1055 - changed boost flag to __BARRY_BOOST_MODE__ to avoid collisions
1056 - removed debug.h dependency from btool.cc
1057 - removed libusb check in favour of specific --with-libusb
1058 option... may need to revisit this later
1059 - quick installation update in README
1061 - fixed some missed endian conversions in record.cc and added
1063 - fixed endian issues in time.h
1064 - added 2 more functions to the time zone API, and adjusted
1065 the time zone table for better defaults in North America
1066 - added btool and upldif to the make install sequence
1067 - added support for Calendar recurrence data
1068 - small doc/TimeZones.txt update
1070 - added BlackBerry time zone code list, reverse engineered from a 7750
1071 - added GetTimeZone() to convert device time zone codes to useful data
1073 Release: version 0.0.3 - 2006/11/24
1074 ------------------------------------------------------------------------------
1076 - added return code check to Controller::SaveDatabase()
1077 - fixed Data() class constructor, so it doesn't auto-convert
1079 - fixed size checking bug in Packet::ReturnCode() that caused
1080 uploading to device to fail
1082 - added Exceptions doc file, documenting the C++ exceptions used,
1084 - fixed small bug in 'make install' which didn't build first
1085 - minor documentation and cleanup in upldif.cc
1086 - added utility functions to the RecordStateTable class,
1087 GetIndex() and MakeNewRecordId()
1088 - changed btool so that the -d command modifiers can be
1089 specified multiple times to work with multiple records
1091 - added Calendar sync code to the opensync module (not complete)
1092 - cleaned up exception handling in the opensync module, making
1093 more use of the base exception to make sure no
1094 exceptions leak into the C library code
1096 Release: version 0.0.2 - 2006/10/12
1097 ------------------------------------------------------------------------------
1099 - fixed casting error in opensync module
1100 - minor updates to documentation to prepare for release
1102 - added stdint.h to parser.h for uint32_t. Thanks Ron Gage for
1105 - large refactoring of the packet parsing and building code,
1106 splitting the code into 3 groups:
1107 - record: handles record specific parsing and building...
1108 this is mostly subfields with CommonField structs
1109 as their headers, but sometimes has a specific
1110 header of its own, in the case of email
1111 Note: email's header is as yet undecoded.
1112 Record does not know the details of the packet
1113 formats, only the formats of its own record data.
1114 - packet: handles building of complete command packets
1115 and handles parsing of commonly used header values,
1116 providing an API for it... packet does not know
1117 the record format details, but does know the
1118 details of each Database Operation (protocol.h)
1120 So far, it seems that there is a real separation
1121 of operation formats and record formats, with the
1122 record formats being the same even with different
1123 database operation codes (GET_RECORDS /
1124 OLD_GET_RECORDS), which was not confirmed before.
1125 See the header size of the email record after
1127 - controller: handles device state, and the management of
1128 sockets, commands, and checking for response values.
1129 The Controller object does not know the format
1130 of either the record or the packet it is sending,
1131 but does know the protocol handshake logic used
1132 to talk to the device.
1133 - implemented SetRecord (SET_RECORD_BY_INDEX)... syncing, here
1135 - spelling correction in opensync-plugin/AUTHORS
1137 - added sample hotplug scripts
1138 - added .cvsignore files for doxygen and opensync directories
1139 - added src/endian.h... still need to add configure support to
1140 handle non-little-endian machines
1141 - added exception handler in opensync plugin commit_change()
1142 - removed the device reset code in probe.cc, as resetting a device
1143 can renumber the devices in /proc on kernel 2.4.x,
1144 and cause a failure to reach the device...
1145 FIXME - this should be changed someday to reset if probing
1146 fails, and restart the probe on reset, as sometimes
1147 the blackberry devices respond differently when not reset
1148 Ideally, the probe protocol should be properly reverse
1150 - added support for retrieving record state table, which is needed
1152 - added support for retrieving, deleting, and clearing the dirty
1153 flags of individual blackberry device records
1154 - added endian macros throughout the code
1156 - added opensync-plugin to root Makefile's clean
1157 - updated time conversion calls to match opensync's latest SVN
1158 - added opensync_change_set_changetype() (only add for now,
1160 - removed slow sync functions for now
1162 - added quick "install" makefile target, to src/install by default
1163 - added static library target... this is temporary until autoconf
1164 stuff is added properly... needed for the opensync plugin
1165 - documented in comments that Barry::BError is intended to be
1166 the base classes for all exceptions
1167 - added first round of implementation of an opensync plugin, based
1168 on the example plugin code from the opensync sources
1170 - added to Contact record class:
1172 - ReadLdif() to load and parse LDAP data
1173 - added fragmented send support to Socket class
1174 - added Probe::FindActive() to make client programs more streamlined
1175 - fixed hex/decimal printing of record counts in btool
1176 - added upldif.cc, a utility to take ldapsearch output on stdin
1177 and upload it to a blackberry
1178 - updated src/.cvsignore
1180 - added parsing of service book fields:
1182 - old and new unique ID (UID)
1184 - old and new description
1186 - bes domain / domain
1187 - date change in legal.txt
1188 - added initial parsing of ServiceBook field 0x09, and calling it
1189 ServiceBookConfig for now, for lack of a better name
1190 - added parsing for ServiceBook field 0x02 on 72xx Blackberries...
1191 seems to be a name or description, but doesn't appear
1192 in the device's own GUI... called "Hidden Name" for now
1194 - added ServiceBook record class and parser code, based on some
1195 of Ron Gage's reverse engineering... code is by Chris Frey
1196 - added AUTHORS file
1197 - fixed the libusb set_configuration/claim_interface order
1198 bug in controller.{h,cc}... already fixed in probe,
1199 but not controller. Thanks Ron Gage for finding this.
1201 - renamed Changelog to ChangeLog in preparation for autoconf
1202 - renamed Todo to TODO in preparation for autoconf
1204 - make clean cleans up tests properly now
1206 - turned source code browsing on in Doxyfile
1207 - added LoadDatabaseByName() and LoadDatabaseByType() template members
1208 to the Controller class
1209 - new file: controllertmpl.h
1210 - documentation fixes in builder.h
1211 - added static GetDBName() functions to all record classes, in
1212 support of the template Controller functions
1213 - fleshed out DatabaseDatabase::GetDBNumber() and GetDBName(),
1214 returning errors properly
1215 - documented sample btool.cc code better, and added example
1216 of Controller template member usage
1218 - copyright dates updated for 2006
1220 - backed out update to latest libusb DEVEL tree, since it is buggy.
1221 Stick with libusb 2005/11/26
1223 - added clean target to convenience Makefile
1224 - added Boost to list of dependencies in README
1225 - fixed minor compile warnings in base64.cc
1226 - fixed new USB discovery code, and changed endpoint usage to use
1227 it, instead of the READ_ENDPOINT and WRITE_ENDPOINT
1228 constants... constants are now removed
1229 common.h, controller.cc, probe.cc, probe.h, usbwrap.cc
1230 - added ClearHalt() to Usb::Device (usbwrap.cc)
1231 - updated usbwrap.h to match latest libusb CVS devel tree
1232 - updated README to refer to 2005/12/30 libusb CVS devel tree
1233 - added roadmap to Todo list
1235 - added convenience Makefile in root directory
1236 - added Hacking document
1237 - minor touchups to README
1238 - added caution warning to README, since this release will support
1239 delete / overwrite / upload functionality
1240 - minor compile order change in Makefile (speed reasons)
1241 - added dynamic endpoint discovery to probe.cc
1242 - changed claim_interface/set_configuration() function call order in
1243 probe.cc based on feedback on the libusb mailing list
1244 - added USB discovery code to usbwrap.{h,cc}
1246 - fixed packing bug in new upload header struct
1247 - renamed CommonField union from data to u to be consistent
1248 - added return_code field to DBAccess struct, for error checking
1249 - added response checking during upload: if response is non-zero
1250 throw exception in controller.cc
1251 - added sorting operators to record parser classes... when writing
1252 contact data to the device, it may be important to write
1253 group link items last, since they reference other contacts...
1254 This needs to be tested, and if not required, removed.
1255 - added GROUP_FLAG field code to Contact parser class
1256 - fixed offset bug in Contact and Calendar Build() functions
1257 - fixed name bug in Contact::Build(), now output only with data
1258 - added appointment type flag support to Calendar record parser
1259 class, and "all day event" flag support
1261 - added autoconf to Todo list
1262 - refactored the protocol structures so that record structs can
1263 be reused for upload and download
1264 - with new protostructs, more header control is handled by
1265 controller.cc instead of record.cc parsers and builders
1266 - parser code now accepts an offset, to govern start of record data
1267 - changed Data sizes from int to size_t and removed casts
1268 - removed unneeded Parser::GetHeaderSize() (new protostructs make
1270 - added Calendar::Build() for calendar uploads
1272 - added s11n-boost.h serialization templates for boost::serialization
1273 - added boost::serialization to build
1274 - minor change in usbwrap.cc to display error codes in decimal
1275 - added SizePacket special case code to Socket::Send()... this is
1276 required when uploaded packet size is a multiple of 0x40,
1277 a special 3 byte size packet is sent
1278 - moved some record class variables to public: so they can be
1280 - fixed variable name but in RecordBuilder template
1281 - added -f switch to btool, for saving and loading of record data
1282 - added -s switch to btool, for uploading saved data back to handheld
1283 - added Controller::SaveDatabase()
1285 - fixed potential casting / pointer bug in controller.cc
1286 - added builder.h, to support uploading
1287 - minor comment fixes in parser.h
1288 - added preliminary Contact protocol record building support
1289 - fixed field order bug in ContactRecord protocol struct
1290 - Contact record.h class now saves unknown Group Link field,
1293 - added -fno-strict-aliasing to compiler flags so gcc-4.0.x won't
1294 optimize away protocol struct casts
1296 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25235
1297 http://mail-index.netbsd.org/tech-kern/2003/08/11/0001.html
1298 http://groups.google.ca/group/comp.lang.c++.moderated/ \
1299 browse_thread/thread/e7bf096832526f8e/5714701b \
1300 02a2a3cc?hl=en#5714701b02a2a3cc
1301 - changed headers so that any low level protocol-specific sizes and
1302 structs are completely hidden in the library...
1303 this adds library-only header: protostructs.h
1305 Release: version 0.0.1 - 2005/11/25
1306 ------------------------------------------------------------------------------
1308 - added barry.h for documentation purposes and application usage
1309 - updated README for release
1310 - renamed SBError exception class to BError
1311 - updated Doxyfile to version 1.4.5
1312 - added more doxygen comments
1314 - added conversion routines (and tests) between time_t and min1900_t
1315 - added -v option to btool to dynamically control protocol dumping
1316 - removed duplicated database database table display in controller.cc
1317 - minor spelling corrections
1318 - added copyright to usbwrap.{h,cc} as per earlier meeting
1319 - added support for Old protocol Calendar entries
1321 - added legal.txt, to save typing when adding new source files
1322 - ignore test-base64 file (cvsignore)
1323 - added contact-to-ldif generation, and updated supporting code
1324 - added argument to btool to support contact-to-ldif operation
1325 - added Contact::GetPostalAddress()
1326 - added README documentation in preparation for future release
1328 - added parser.{h,cc} virtual wrapper to generalize LoadDatabase()
1329 - added code to socket.cc's Packet() command to re-receive on
1330 reception of empty packets... this seems to be a signal
1331 from the device that it needs more time for long transfers
1332 Current limit: 10 blank packets before exception is thrown
1333 Observed maximum blank count: 1
1334 - added MessageRecord and OldMessageRecord to protocol.h
1335 - changed common record field structs to one CommonField struct
1336 - increased default USB timeout to 10 seconds (some loaded devices
1337 seem to need more time...)
1338 - reorganized record.{h,cc} to try to reduce the code duplication
1339 - record.h classes now "know" the sizes of their corresponding
1340 protocol structs... this may cause trouble down the line
1341 with Python, if used(?)
1342 - added Message class for Email records
1343 - removed the "Get*" database functions and replaced with
1344 general LoadDatabase() call, which loads any database
1345 available, and parses it with a Parser object
1346 - added GetDBID() for searching for database numbers via name
1347 - added test code to retrieve and display all email
1348 - cleaned up comments
1349 - renamed blackberry.{h,cc} to controller.{h,cc}
1350 - renamed bbtool.cc to btool.cc
1351 - added Makefile.conf
1352 - added copyright notices to source files
1353 - added COPYING file with GPL
1354 - put parser in Barry namespace, and added auto_ptr-like mode for
1355 store object pointers
1356 - added doxygen docs
1357 - added header comments to convo.awk script
1358 - removed Test() function in Controller object, and made class
1359 better suited to its general API purpose
1360 - changed OpenMode() so it loads command table and DBDB... it is now
1361 a public API function
1362 - moved Controller test code to btool.cc and made tool more
1363 generically useful... can now download any DB by name on
1365 - added mode checks to Controller
1366 - added -t and -d command line options to btool
1367 - added base64 routines from John Walker / Fourmilab and Citadel/UX
1368 - converted base64 routines into an API, instead of hardcoded
1371 - changed project name to Barry, including namespaces
1372 - renamed sbcommon.* to common.*
1373 - increased USB default timeout to 2 seconds
1374 - fixed formatting bug in Data operator<< output
1375 - cleaned up the protocol packet structs, getting rid of the
1376 confusing param/simple split... param is now the tableCmd
1377 from the Command Table
1378 - moved protocol structs from record.h to protocol.h
1379 - added support for older version of the USB protocol, which matches
1380 the Cassis spec more closely
1381 - fixed the hardcoded offsets in blackberry.cc's protocol parsing
1382 - fixed Socket::Close() so on error, socket flags are reset.
1383 This stops Close() from being called twice on exception
1386 - added Doxyfile to the project
1388 - added mode selection support to Blackberry class (RIM Desktop,
1389 RIM Bypass, and RIM_JavaLoader)
1390 Mode selection governs which socket is used in subsequent
1392 - added Command Table parsing and protocol support
1393 Blackberry class, and CommandTable record class
1394 - added Database database parsing and protocol support
1395 Blackberry class and DatabaseDatabase record class
1396 - added Address book retrieval support - dumps data to stdout for now
1397 - fixed bug in CopyOnWrite mode in Data class, where could allocate
1398 less buffer than data copied
1399 - added "easy exception" debug output macro
1400 - fixed minor spelling error in protocol.cc
1401 - fixed ModeSelectCommand protocol structure - it is not a parameter
1402 command, but uses the unknown slots for socket and flag
1404 - removed hardcoded socket numbers from protocol.h
1405 - fixed bug in socket.cc:Open() where packet size was not properly
1407 - implemented socket::Close()
1408 - fixed sequence number checks, so it is reset properly on socket 0
1409 - added Socket::NextRecord() for multiple record data transfers
1410 - fixed input handling bug in translate.cc
1412 - added .cvsignore files to project
1413 - added README notes to document the source architecture
1414 - added dbsetup data script, which shows how to get into database
1415 mode, using the RIM Desktop mode, and open a socket
1416 - added blackberry.{h,cc} class
1417 - added command line parameters to bbtool.cc
1419 -p = specify pin of blackberry device
1420 - added blackberry test code in bbtool.cc - debugging the initial
1421 opening protocol: sockets, modes, etc... see dbsetup for
1423 - added copy-on-write functionality to Data class, so it can be used
1424 with plain data buffers with little performance impact
1425 - added eout() and ddout() debugging - to separate:
1426 - regular debug output
1427 - exception debug output (should probably be logged to
1428 syslog someday, and never turned off)
1429 - raw data dump output
1430 - changed errno SBError exception handler to try to decode the error
1431 numbers into english (limited success)
1432 - fixed probe and socket USB Bulk message handling, so they don't
1433 use PollCompletions(), but use IO::Wait() (libusb_io_wait())
1434 - added common packet size checker in protocol.cc
1435 - fixed unpacked protocol packet structs... using gcc's
1436 __attribute__ ((packed)) ... see protocol.h
1437 - added ModeSelectCommand structure (protocol.h)
1438 - finished Socket::Open() call (Close() not yet implemented)
1439 - moved Socket::Send/Receive() out of private, as they can be used
1440 for socket 0 operations.
1441 - cleaned up pointer management with macros:
1445 - added IO::Wait() and IO::Cancel() to correspond with libusb API
1446 - added datadumping in usbwrap for easy protocol capture
1447 - added preliminary record.{h,cc} classes
1448 - added Contact record parsing, and test code
1450 - syncberry tree started