Format:

<string in about requester> = <version string(s)>
  - <List of changes>

MakeCD V3.0 = MakeCD 3.22 (19.12.97)
              makecdromfs.module V10.2 (20-12-1997)
              ReadWrite.module 11.23 (07.01.98)
              
              CDR_JVC_Teac.driver 10.24 (20.12.97)
              CDR_MMC_ATAPI.driver 10.24 (20.12.97)
              CDR_Panasonic.driver 10.24 (20.12.97)
              CDR_Philips_2000.driver 10.24 (20.12.97)
              CDR_Philips_2600.driver 10.24 (20.12.97)
              CDR_Plextor_Ricoh.driver 10.24 (20.12.97)
              CDR_Ricoh_6200.driver 10.24 (20.12.97)
              CDR_Sony.driver 10.24 (20.12.97)
              CDR_Yamaha_10x.driver 10.24 (20.12.97)
              CD_ATAPI.driver 10.24 (20.12.97)
              CD_NEC.driver 10.24 (20.12.97)
              CD_Plextor.driver 10.24 (20.12.97)
              CD_ROM.driver 10.24 (20.12.97)
              CD_Sony.driver 10.24 (20.12.97)
              CD_Toshiba.driver 10.24 (20.12.97)

 - disc-at-once
 - indices (not always possible in track-at-once)
 - rewrote and enhanced GUI
 - renamed SCSI drivers
 - new settings
 - AHI support
 - Formating a CD-RW is much more flexible now: can delete session only,
   remove fixation, ...
 - ISO image creation: a file with a list of filenames can be used to
   specify files that are included in the image
 - MS-DOS boot options: MakeCD can create ISO images for CD-Rs booting MS-DOS
   on a PC now.
 - JVC/Teac driver: mode 2 writing (without form) might work now
 - CD_ATAPI.driver: Some drives incorrectly reported "wrong track type" errors
   when reading CDDA.
 - Bugfix: Left and right audio channel should no longer be swapped when
   using Amiga audio.
 - Bugfix: User aborts were ignored during writing if the IO error handling
   was set to "Ignore" or "Pad".


MakeCD V3.1: MakeCD 3.27 (23.01.98)
             makecdromfs.module V10.2 (20-12-1997)
             ReadWrite.module 12.26 (23.01.98)

             CDR_JVC_Teac.driver 11.31 (23.01.98)
             CDR_Panasonic.driver 11.31 (23.01.98)
             CDR_Philips_2000.driver 11.31 (23.01.98)
             CDR_Philips_2600.driver 11.31 (23.01.98)
             CDR_Plextor_Ricoh.driver 11.31 (23.01.98)
             CDR_Ricoh_6200.driver 11.31 (23.01.98)
             CDR_SCSI3_ATAPI.driver 11.31 (23.01.98)
             CDR_Sony.driver 11.31 (23.01.98)
             CDR_Yamaha_10x.driver 11.31 (23.01.98)
             CD_ATAPI.driver 11.31 (23.01.98)
             CD_NEC.driver 11.31 (23.01.98)
             CD_Plextor.driver 11.31 (23.01.98)
             CD_ROM.driver 11.31 (23.01.98)
             CD_Sony.driver 11.31 (23.01.98)
             CD_Toshiba.driver 11.31 (23.01.98)

 ***************** SCSI and IO changes **********************
 - Read speed can be reduced for data tracks, too.
 - slightly improved generic SCSI handling
 - SCSI driver now choose supported write command automatically. This
   makes writing with Panasonic 7502 and HP 7500+ possible.
 - recognize more drives
 - MMC driver supports Panasonic/Matsushita 7502 now. However,
   this drive has several problems (at least with firmware V3.02 -
   V3.05 is available, but untested):
   * It often forgets the correct contents of the CD-R. E.g. "Update"
     after writing and fixation in test mode makes it forget the
     fixation, but not the tracks. Reinsert the CD-R in such a case.
     This prevents "write after test" writing.
   * It sometimes reads wrong track types from a CD-R (e.g. mode 2 
     instead of audio).
   * ISRCs in DAO don't work at the moment. Make sure your projects
     don't contain any, otherwise you will get an "Illegal field in
     command descriptor error". This might be fixed soon.
   * In DAO mixed CDs have to be copied with raw data. The drive
     cannot handle the different block sizes MakeCD uses. If you try
     to use cooked data you might get a "buffer underrun" between data
     and audio tracks. V3.1 will select raw data if you instruct MakeCD
     to choose automatically by using data mode "writer".
   * Mode 2, form 1 does not work in TAO.
   * In TAO gaps between tracks of different types are too short.
 - Bugfix: TOC type was chosen incorrectly for DAO if raw data was written,
   resulting in wrong medium type and not working PSX copies. This also
   caused Philips/Ricoh/Plextor writers to report a wrong mode for the
   mode 2 data track.
 - Bugfix: although none given, Philips/Ricoh wrote medium catalog number 
   both in TAO  and DAO. This made PSX copies unusable in a PlayStation.
 - Bugfix: data encoding for CDR_Plextor_Ricoh.driver was wrong, resulting
   in trashed DAO CD-Rs.
 - Bugfix (Philips, Plextor): didn't work with all controllers
 - Bugfix: removing the source CD while MakeCD was preparing for
   index scanning trapped MakeCD in a loop until the CD was inserted
   again.
 - Bugfix: CDR_Sony.driver couldn't write audio tracks.
 - SCSI error detection failed for JVC/Teac drives when connected to some
   controllers. This prevented writing to empty CD-Rs.
 - Yamaha CDR10x couldn't always write all tracks with
   cybppc.device V44.2, because this device doesn't report
   some essential sense codes. Enhanced workaround.
 - Bugfix: priority of conversion process was not always increased
 - Bugfix: MMC driver still knew Ricoh MP6200S.
 - MMC/ATAPI: support for writing mode 2 with block size 2332; enabled
   only for Sony 928E, because mode 2 with 2336 didn't work there. Not
   tested completely with Sony 928E yet.
   If only mode 2 writing fails for a MMC/ATAPI writer than this optional
   mode is not implemented. For known drives like the Sony 928E MakeCD selects
   a substitute automatically, but not generally, because PSX copies might
   be less compatible in this mode. You can enforce this workaround by
   setting a environment variable with "setenv MAKECD_MODE2_MIXED". If it
   works only this way, please tell Patrick.
 - MMC/ATAPI: $30 and $C are recognizes as buffer underruns, too.
 - Bugfix: CDs with two consecutive data tracks of the same type were
   not copied properly (e.g. video CDs).
 - Bugfix: Adding pause blocks to a ISO image that was written online was not
   possible.
 - MakeCD now adds some blocks to a created ISO image to compensate for read
   errors that might occur at the end (but never did so far - only the
   new Panasonic 7502 might trash the last two blocks).
 - Renamed CDR_MMC_ATAPI.driver to CDR_SCSI3_ATAPI.driver because of the
   confusion caused by the "ATAPI" in the name. ATAPI and MMC (Multi Media
   Command set defined by SCSI 3) are so similar that only ome driver is
   sufficient to cover both.

 ************************* GUI changes ********************************
 - Several QuickHelps were rewritten.
 - ISO image source file requester offers pattern gadgets and shows
   .info by default.
 - MakeCD can execute several user selectable commands on different
   occasions, e.g. to play sounds. If a command is not set, MakeCD
   flashes the screen instead to draw the intention to requesters.
 - MakeCD now indicates more clearly when the write window is idle again.
 - Track option gadgets are disabled if no track exists.
 - Track options window is opened after adding a new track.
 - Bugfix: changing tracks while pausing audio playing is possible now.
 - Bugfix: lock on base dir was not freed if the advanced options
   window was open when exiting the program.
 - Bugfix: Pausing audio playing did not always work properly, and volume
   and filter were reset at the beginning of each new track.
 - Bugfix: "-" could be entered in medium catalog number gadget, although it
   is an illegal character


MakeCD V3.1a: MakeCD 3.33 (12.02.98)
              makecdromfs.module V10.2 (20-12-1997)
              ReadWrite.module 13.28 (12.02.98)
              SCSISupport.module 12.21 (12.02.98)

              CDR_JVC_Teac.driver 12.40 (12.02.98)
              CDR_Panasonic.driver 12.40 (12.02.98)
              CDR_Philips_2000.driver 12.40 (12.02.98)
              CDR_Philips_2600.driver 12.40 (12.02.98)
              CDR_Plextor_Ricoh.driver 12.40 (12.02.98)
              CDR_Ricoh_6200.driver 12.40 (12.02.98)
              CDR_SCSI3_ATAPI.driver 12.40 (12.02.98)
              CDR_Sony.driver 12.40 (12.02.98)
              CDR_Yamaha_10x.driver 12.40 (12.02.98)
              CD_ATAPI.driver 12.40 (12.02.98)
              CD_NEC.driver 12.40 (12.02.98)
              CD_Plextor.driver 12.40 (12.02.98)
              CD_ROM.driver 12.40 (12.02.98)
              CD_Sony.driver 12.40 (12.02.98)
              CD_Toshiba.driver 12.40 (12.02.98)

 ***************** SCSI and IO changes **********************
 - Bugfix CDR_Yamaha_CDR10x: DAO writing of cooked data + audio was broken.
 - Bugfix CDR_Panasonic: writing of audio was broken and the changed generic
   SCSI handling might have caused a worse performance.
 - Sony 928E: the drive didn't read mode 2 correctly, but with the new raw
   mode 2 reading it works now. Writing of mode 2 + audio now works, too.
 - Panasonic 7502: the tray is opened and closed to make the drive forget
   tracks written in test mode (necessary because just turning off the test
   mode confused this drive). With the Commodore scsi.device writing was
   aborted due to "bad status" errors although writing seemed to work fine.
   This is ignored now, but not completely tested yet.
 - Bugfix Philips CDD522: choosing speeds > 2 was not caught by the driver
   but caused an error requester.
 - Bugfix (SCSI3): ISRCs were written incorrectly in DAO
 - Bugfix: MakeCD might have got the track starts and lengths wrong by a
   factor of 2 or 4 if the source drive didn't use 2048 bytes as default
   block size or was set to this value by some other application.
 - CD_Toshiba.driver couldn't read audio tracks after reading a raw
   data track (it failed already for the first block).
 - improved index scanning to make it more robust
 - added some more drive IDs (see Compatibility.guide in case your drive
   is still not known)
 - Workaround: mode 2 tracks might have been recognized as mode 1 when
   writing in DAO with raw data. The relevant part of the track is okay
   and a PSX copy actually works in a PlayStation nonetheless. The reason
   are probably the missing checksums in the gap before the data track.
 - Workaround for mode 2 reading problems (see "MAKECD_READ_MODE2_RAW"
   below).
 - SCSI3/ATAPI Workaround "MAKECD_MODE2_MIXED" was renamed to
   "MAKECD_WRITE_MODE2_MIXED".
 - Workaround for PPC SCSI and fixation/repair/format problems:
   cybppc.device with version < 44.36 cancels all SCSI commands automatically
   that are not returned by the drive in a certain amount of time. This is
   done to prevent lockups of the SCSI bus, but it also aborted some of
   the commands that are used for CD writing and really go on for a while,
   the most common one probably being fixation. Formating, repair and DAO
   might also be affected.

   MakeCD now applies a workaround for this problem by making the drive return
   the command immediately and querying the drive until it is ready again.
   However, it depends on the drive and on the command whether this is
   possible or not. For fixation this immediate response is only supported by
   SCSI3/ATAPI, Philips, Plextor/Ricoh and Yamaha drives, but not by JVC/Teac,
   Panasonic/Matsushita or Sony drives (unless they use the SCSI3/ATAPI driver).
   It is not supported by any drive for the repair or format commands.

   This workaround is only used for the cybppc.device, because not all
   other controllers would handle this correctly (the dracoscsi.device
   crashes when fixating with a Philips CDD2600 this way).

 ********************* GUI changes **************************
 - CONTEXT SENSITIVE AMIGAGUIDE: Pressing HELP over a gadget opens the
   AmigaGuide documentation at the corresponding node. However, this is not
   possible with the V2.5 English documentation and the new translation is
   not ready yet.
 - MakeCD now allows the frequency of AIFF to be off by upto 10%, because
   not every system is able to produce exactly 44.1kHz. BE WARNED: MakeCD
   does not adapt the frequency, it just assumes it was 44.1kHz. Playing
   a sample with a frequency that is 10% too high or low already sounds
   quite awful!
 - Bugfix: text for publisher field in ISO image was wrong in some cases
 - Bugfix: manually jumping between tracks during audio playing still forgot
   the current settings for low pass filter and volume.
 - Bugfix: updating from final TAO to DAO was not possible with the menu
   item "Projects/Registration..."
 - MakeCD is more intelligent when choosing the best data format if you don't
   override this choice yourself by choosing "raw" or "cooked" explicitely:
   some drives like the Panasonic 7502 (firmeware 3.02) have problems when
   copying CDs with cooked data, so MakeCD uses raw data by default. However,
   if your source is an ISO image than cooked is used as MakeCD couldn't
   convert this to raw.
 - If speed cannot be set, then the speed gadgets are empty now and no longer
   set to the unused value.
 - Fix for problems with saving the settings. This might have failed after
   fixing the icon position of MakeCD if the Workbench protected the
   MakeCD.info from writing.
 - fixed some English texts


MakeCD V3.1a - update 1: CD_Toshiba 12.41 (16.02.98)
 - Bugfix CD_Toshiba: the fix for the "audio after raw" problem made things
   worse. Audio reading now didn't work at all with at least one drive :-(


MakeCD V3.1a - update 2 (available as public beta only):
   ReadWrite.module 13.32 public beta (18.03.98)
   SCSISupport.module 12.22 public beta (18.03.98)

   CDR_JVC_Teac.driver 12.46 public beta (18.03.98)
   CDR_Panasonic.driver 12.46 public beta (18.03.98)
   CDR_Philips_2000.driver 12.46 public beta (18.03.98)
   CDR_Philips_2600.driver 12.46 public beta (18.03.98)
   CDR_Plextor_Ricoh.driver 12.46 public beta (18.03.98)
   CDR_Ricoh_6200.driver 12.46 public beta (18.03.98)
   CDR_SCSI3_ATAPI.driver 12.46 public beta (18.03.98)
   CDR_Sony.driver 12.46 public beta (18.03.98)
   CDR_Yamaha_10x.driver 12.46 public beta (18.03.98)
   CD_ATAPI.driver 12.46 public beta (18.03.98)
   CD_NEC.driver 12.46 public beta (18.03.98)
   CD_Plextor.driver 12.46 public beta (18.03.98)
   CD_ROM.driver 12.46 public beta (18.03.98)
   CD_Sony.driver 12.46 public beta (18.03.98)
   CD_Toshiba.driver 12.46 public beta (18.03.98)

 - Setting speed with SCSI3/ATAPI is more sophisticated now and also
   checks the real speed used by the drive. This method solves problems
   with Plextor PX-426S and 2x with Ricoh MP6200S. Note that the Ricoh
   seems to be unable to write CD-RW in single speed at all.
 - Added raw encoding: now cooked data can be written to drives only
   supporting raw data in DAO, e.g. an ISO image can be written in DAO
   to a CDD2600.
   However, this conversion demands very much computing power and is even
   on a 060 hardly fast enough to keep up with 4x. The GUI definetely has
   to be changed to warn of that, and as this is not implemented yet the
   encoding feature is still disabled by default. It has to be enabled
   by setting the environment variable MAKECD_ALLOW_ENCODING.
 - The problem with reading the first blocks of a raw DAO copy in some drives
   should be fixed now.
 - Added some drive IDs and disabled session fixation in DAO for Ricoh 1420/
   Plextor PX-24C because these drives don't support that.
 - According to the Pioneer documentation modern Pioneer drives (like DR-U12X
   and DR-U24X) don't support speed selection at all, therefore they are now
   only recognised by the CD_Sony.driver, which doesn't set speed.
 - DAO was disabled for the HP 7100+ because firmware 1.x didn't implement
   it. However, we have been told that it was added with firmware 2.x to
   the HP 7100+ and the similar Philips CDD3610. DAO is now allowed by
   MakeCD again, but it hasn't been tested with either of these drives yet.
 - DAO is disabled for some drives that haven't implemented it. This avoids
   SCSI errors when trying to use it with such drives. However, for tests
   with newer firmware versions DAO can be enabled again by setting the
   environment variable MAKECD_ALLOW_DAO.
 - The priority of read, write and conversion process is increased when doing
   critical work. The default priority is 6, because the console.device is
   doing CPU consuming work at priority 5. These defaults can be changed by
   setting the environment variables MAKECD_INC_PRI_[READ|WRITE|CONV], but
   usually there should be no need to do that.
 - Invalid status errors that are reported by the combination A3000 SCSI +
   Panasonic 7502 are finally ignored properly, so data (= mode 1) works now.
   Audio never caused this invalid status errors. However, mode 2 is not
   working because the SCSI bus locks completely.
 - CD_ATAPI can (and should) be used to read CDs with the Sony 928E.
 - No CD_ driver tries to read session infos any more - this was not used
   for the source drive and sometimes seemed to have caused problems.
 - Automatic "IO Error" handling didn't work properly. Now the real errors
   are always reported, too. Write errors are no longer ignored automatically.
   However, for read errors that lead to an automatic abort the text still
   says "User abort".
 - Bugfix: potential memory trashing with the READ_MODE2_RAW workaround.
 - Bugfix: using delete mode "immediately" with DAO caused an internal error.
 - "Write after test" didn't work properly with drives using the open/close
   tray method (Panasonic 7502).
 - Bugfix: under a certain timing the amount of filled bytes in the
   MakeCD buffer could have been displayed as a large number and 100%.
 - Bugfix: SCSI errors were ignored after one command (only happened with
   Plextor PX-624C in one case).


MakeCD V3.1b - MakeCD 3.33 (12.02.98)
               makecdromfs.module V10.2 (20-12-1997)
               ReadWrite.module 13.33 (08.04.98)
               SCSISupport.module 12.22 (08.04.98)

               CDR_JVC_Teac.driver 12.47 (08.04.98)
               CDR_Panasonic.driver 12.47 (08.04.98)
               CDR_Philips_2000.driver 12.47 (08.04.98)
               CDR_Philips_2600.driver 12.47 (08.04.98)
               CDR_Plextor_Ricoh.driver 12.47 (08.04.98)
               CDR_Ricoh_6200.driver 12.47 (08.04.98)
               CDR_SCSI3_ATAPI.driver 12.47 (08.04.98)
               CDR_Sony.driver 12.47 (08.04.98)
               CDR_Yamaha_10x.driver 12.47 (08.04.98)
               CD_ATAPI.driver 12.47 (08.04.98)
               CD_NEC.driver 12.47 (08.04.98)
               CD_Plextor.driver 12.47 (08.04.98)
               CD_ROM.driver 12.47 (08.04.98)
               CD_Sony.driver 12.47 (08.04.98)
               CD_Toshiba.driver 12.47 (08.04.98)

 all changes from public beta above and:
 - Bugfix: converting from raw to cooked data could cause memory trashing.
 - Added fallback in mode 2 recognition to raw block reading => mode 2
   should be recognized by some more drives.
 - Allowed CD_Plextor.driver for Pioneer drives again, because at least
   one user was able to set the speed with his DR-U12X (firmware 1.06
   1996/10/), although the documentation said that this drive always reads
   at maximum speed.
 - Teac 56S is no longer fixed to CD_ATAPI.driver, because audio reading
   didn't always work.


MakeCD V3.2 - MakeCD 3.59 (10.07.98)
              makecdromfs.module V11.0 (24-05-1998)
              ReadWrite.module 15.53 (10.07.98)
              SCSISupport.module 14.26 (10.07.98)

              CDR_JVC_Teac.driver 13.72 (10.07.98)
              CDR_Panasonic.driver 13.72 (10.07.98)
              CDR_Philips_2000.driver 13.72 (10.07.98)
              CDR_Philips_2600.driver 13.72 (10.07.98)
              CDR_Plextor_Ricoh.driver 13.72 (10.07.98)
              CDR_Ricoh_6200.driver 13.72 (10.07.98)
              CDR_SCSI3_ATAPI.driver 13.72 (10.07.98)
              CDR_Sony.driver 13.72 (10.07.98)
              CDR_Yamaha_10x.driver 13.72 (10.07.98)
              CD_ATAPI.driver 13.72 (10.07.98)
              CD_NEC.driver 13.72 (10.07.98)
              CD_ROM.driver 13.72 (10.07.98)
              CD_Sony_Plextor.driver 13.72 (10.07.98)
              CD_Toshiba.driver 13.72 (10.07.98)


 ***************** SCSI and IO changes **********************

 - *** DAO for JVC/Teac ***
   This is the recommended way of writing CD-Rs with the JVC drives.
   It's less error prone than the stop-and-go writing in TAO, and
   therefore TAO has to be regarded as being unsupported. Mode 2
   writing in TAO has never been tested, but works in DAO. DAO also
   solves the TAO problem of CD players stopping at the gap between
   tracks.
   ISRC/Medium Catalog Number are not supported by at least the JVC
   XR-W2010 with firmware V1.51 and errors when trying to use them
   are silently ignored.
   At least the Philips CDD2600 always only sees the first session
   of a DAO-CD-R written with JVC XR-W2010 V1.51, even though
   "Fixate: Session" was used and other drives (JVC itself, Yamaha
   CDR100) find the next session(s). Could only be fixed by JVC or
   Philips.
 - MakeCD now reads mono and stereo audio data in 8 or 16 bit and with
   any frequency. They can be transformed to stereo, 16 bit, 44.1kHz
   on-the-fly, but frequency resampling requires a fast CPU.
   The following audio file formats are supported:
   * AIFF
   * AIFC (only uncompressed)
   * MAUD (uncompressed, u-law, a-law)
   * WAV (only uncompressed)
   * MPEG-A (mpega.library from Stephane TAVENARD has to be installed)
   * Samplitude HDP (only 44.1kHz!)
 - MPEG-A: recognition of this file type cannot be done with 100% certainty.
   It takes a long time to find that a normal image file is not MPEG-A,
   and sometimes some are even recognized as MPEGA-A although they aren't.
   Therefore MakeCD is no longer testing if a file that has a corresponding
   ".CDT" (i.e. was originally written by MakeCD itself) or has the suffix
   ".cdda" or ".raw" is MPEG-A. You cannot compress the image files any
   more without renaming them or deleting the .CDT file, but otherwise you
   are strongly encouraged to leave the .CDT file as it is or to use one of
   the suffices to mark CDDA!
   If you have problems with MPEG-A recognition, then please contact the
   author of the mpega.library, as the recognition is completely done
   by this library.
 - MakeCD also writes different audio formats. This is selected by
   using appropriate suffices for the image filenames:
   AIFF: .aiff, .aif
   AIFC: .afc, .aifc
   MAUD: .maud
   WAVE: .wav, .wave
   All data is always uncompressed, 16bit, stereo, 44.1kHz.
 - Some drives allow to write more than the nominal capacity of data in DAO.
   This is probably possible by bending some rules in CD-ROM/CD-R standards.
   You may use this "feature" now by ignoring the warning that MakeCD
   issues if space is not sufficient, but be warned: your drive may report
   errors.
   Yamaha CDR10x doesn't support this. CDD2600 allows writing of any amount
   of data, but does no check itself at all. Instead, writing fails at the end.
 - Changed generic SCSI handling: MakeCD was using redundant commands
   that might have avoided certain timing problems with some systems
   (e.g. Philips CDD2600 + Phase 5 SCSI kit + audio reading).
   Unfortunately this also seemed to increase the likelihood of SCSI
   problems on unstable systems (e.g. reported for Fastlane). Now
   these redundant commands are only used when reading audio from
   Philips drives. If you think MakeCD is less stable or audio quality
   decreased in V3.2, then use "setenv MAKECD_COMPLEX_SCSI_IO=1" to
   go back to the previous behaviour and report if this helps. As
   before, "setenv MAKECD_SIMPLE_SCSI_IO=1" avoids unnecessary commands
   completely.
 - CD_Sony and CD_Plextor were merged, because the only difference
   was the ability to set speed with CD_Plextor. This ability is now
   tested automatically, so there's no need anymore to select the
   correct driver. Should this test fail, then you can fix this ability
   with MAKECD_SONY_PLEXTOR_SPEED_NO and MAKECD_SONY_PLEXTOR_SPEED_YES.
 - Raw encoding (calculating error correction codes and checksums) is enabled,
   but there is a warning about the high CPU demands. One choice in this
   requester disables the warning permanently. Raw encoding is necessary to
   write ISO images in DAO to drives like the Philips CDD2600, Ricoh MP6200S
   and Ricoh 1420 that only support raw data in DAO.
 - TAO + same source/target + "Write Tracks": you can avoid swapping
   source and target disc for every track now. If you choose
   "Delete image: after CD/never" instead of "immediately", then MakeCD
   will copy these tracks to images first, because you need the space
   for all images at once anyway. "Delete image: immediately" can still
   be used to copy a CD one track at a time (in case you don't have enough
   HD space).
 - Teac 55S: sometimes the drive reports an empty CD if there are read
   errors when the drive specific commands are used; MakeCD recognizes this
   now and uses the standard command in this case.
 - Plextor PX-412: the BLOCK jumper may be set now. This helped to avoid
   SCSI problems with a CyberSCSI controller and might help on other
   systems, too. This helps because this jumper not only changes the default
   block size but also reduces the peak transfer rate from 10MB/s to 5MB/s.
 - Mitsumi 2600TE/2800TE: the MAKECD_MODE2_WRITE_MIXED workaround is now 
   enabled by default, because the normal method of writing with 2336 bytes
   is not supported by the drive.
   Workaround for crash during fixation. Many thanks to Eyetech for providing
   their EZWriter kit (A1200 + Mitsumi CR-2801TE) as test environment!
 - This workaround will be selected automatically now for all unknown drives.
 - CD_Atapi: workaround for NEC CD-ROM DRIVE:282 V3.05
 - Classic Squirrel: Some SCSI devices don't set scsi_Actual correctly.
   In this case MakeCD is adding warnings to the internal log, which may
   slow down the reading/writing if they cause every read/write command
   to be logged.
   This happened with the Classic Squirrel that seems to return only
   transfer lengths that are a multiple of 512, even if some other amount
   of data was transfered. These warnings are now always turned off for
   squirrelscsi.device. If you have another SCSI device that has this
   problem, then "setenv MAKECD_NO_LENGTH_WARNING=1".
   NTB: not all of the warnings are harmful! When reading the buffer is
   larger than the transferred data.
 - Philips CDD2000/2600: writing indices in TAO is now optional and
   disabled by default, because the command used to setup the indices
   seems to be more prone to a SCSI lock-up. Use
   "setenv MAKECD_ALLOW_TAO_INDICES=1" if you want to use this feature
   again, but be warned...
 - Bugfix CDR_SCSI3_ATAPI: "cooked data" in DAO caused wrong buffer underrun
   messages if a "negative pause" was used to get pregap data from the source
   track. Unfortunately "format of the writer" was "cooked data" for all
   drives but Panasonic 7502.
 - Bugfix: errors after writing the tracks or the session were not
   reported - this could have hidden the fact the drive didn't write
   the CD-R correctly, e.g. due to medium errors.
 - Bugfix CDR_Panasonic: 1x and 2x speed couldn't be set.
 - Bugfix: reading didn't proceed in the special case of direct copying of a
   audio CD in DAO, lead-in data generated for drive (e.g. Philips CDD2600),
   buffer larger than 20MB and "Fill Buffer completely" enabled.
 - Bugfix (at least CDR_Philips_2600): a CD-R that had been filled completely
   (e.g. by ignoring a warning about insufficient space in TAO) was reported
   as being fixated even if it wasn't. Fixation was still possible, though.
 - Bugfix (at least CDR_Philips_2600): "Write after Test" in DAO might have
   failed with a SCSI error after the successful simulated write and before
   starting the real write.
 - Bugfix: "Ignore" in the "Incomplete image file" requester didn't work
   properly when creating several image files at once.
 - Bugfix: reading/writing project and .CDT files with no indices might
   have trashed memory.
 - Bugfix: scanning ISO tracks with a directory with many (>5000) files could
   have failed
 - When writing in TAO without any fixation, MakeCD now assumes you are going
   to fix the CD and not only the session. This avoids an unnecessary warning.
 - Source drive is reset properly after scanning for indices.
 - Added fallback in mode 2 recognition to raw block reading => mode 2
   should be recognized by some more drives.
 - When aborting writing after a read error caused a write error, the
   write error was reported and not the read error (happened only with
   JVC DAO).
 - Read errors at the very end of an audio CD (the last 1/3 second) are
   ignored automatically. Many Toshibas had problems reading the last block.
 - If your writer doesn't update properly after a test write, then
   "setenv MAKECD_UPDATE_EJECT=1" chooses eject/insert as the method
   to do the update.
 - To avoid problems with the maximum filename length of the FS ".CDT"
   files get shorter filenames now if necessary and another suffices are
   replaced.
 - MAKECD_NO_WRITE_LOG disables logging of write process during
   time critical sections.


 ********************* GUI changes **************************

 - ** NEW: AIFF-CD *** - a file format that allows you to store all the
   data defining an audio CD or a mixed mode CD in a single file. Programs
   not knowing the extensions will still be able to use the audio data,
   but this format is especially useful if you have a sound editing
   software that knows the CD specific extensions, like Samplitude.
   Then you can setup your audio CD including indices and tracks with
   the advanced features of the audio editing software and write with
   MakeCD, or copy a CD into a AIFF-CD file with MakeCD and edit this
   before writing it back to CD-R.
 - *** novice mode *** that hides the more advanced gadgets:
   audio tracks: indices, ISRC, Medium Catalog Number
   track options: start and length of source, pause
   project options: start block, audio CD start, DAO data format
   ISO images: all ISO/Rockridge and boot options
   copy/append CD: data and file format, ISRC reading and index scanning
   write window: transfer rates, indices
 - "ISO level: Amiga" is now the default value as this mode is more
   flexibel and CD-ROMs created this way work fine with Amiga and
   other systems, although some ISO 9660 rules are no longer enforced.
 - added track editing function to split the current track at the selected
   indices
 - error requester for invalid audio image length now allows to accept
   such images and to change the corresponding setting
 - If using image files is the default (either because the current track
   uses one or this gadget state was saved), this is not disabled when
   adding a MPEG-A image because decoding MPEG-A might be so slow that
   a temporary image file is required.
 - The MakeCD environment variables can now alternatively also be
   set in the tooltypes instead. They control some internal behaviour
   only and are not ment to be used regularly.
 - MakeCD now supports automatic opening of the tray/disc ejection
   after writing and when asking for a specific disc. This behaviour
   is configurable in the commands settings window.
 - Enhanced "audio CD start" workaround: instead of setting up a project
   that reads audio tracks from different positions the driver itself
   reads from the appropriate position to get the right data. This can
   be selected in the compatibility settings.
 - Made index scanning more robust to compensate for inexact drives.
   However, still the same method is used to get subcodes, so it won't
   solve the problem that some drives fail with a "track following" or
   "positioning" error.
   As some copies of games only work with index pause, a default index
   pause will be set if no scanning was done or failed. This can be
   selected in the "Append/Copy CD" window.
 - Automatically created filenames now also contain the length to get
   different filenames for tracks from different CDs.
 - The following warnings/messages will now be issued:
   + raw data encoding, MPEG-A decoding or frequency resampling
     might be too slow for on-the-fly
   + buffer cannot be allocated completely
   + writing in DAO and trying to delete an image file immediately
   Reactions for each can (optionally) be saved permanently in the
   requester.
 - ISO image path is now always cleared when disabling it and set to
   the path derived from the source when enabling it again.
 - more detailed explanation why type of image file is not suitable
 - MakeCD now saves the settings in the icon it was started with.
   The main program doesn't have to be copied to use different settings for
   different occasions, it's sufficient to create some project icon that
   has MakeCD as default project. If the new icon hasn't got any settings,
   the main program's settings are used as default until they are saved for
   the first time.
 - "CD_COPY" as tooltype or shell switch causes MakeCD to go directly
   to the CD copy window. From there "Exit Window" and "Setup Project"
   still lead back to the main window. This tooltype is used in
   CDCopy.info to get a simple "separate" CD copy program...
 - "USE_MAIN_PREFS" causes MakeCD to change tooltypes in the program's
   icon when saving settings instead of the icon it was started with. This
   is used in CDCopy.info already so that settings only have to be done
   once for the whole program and stay consistent.
 - All other windows are closed while the write window is open.
 - Selection of new audio position during playing was improved (now waits
   until value hasn't been changed for a short while before jumping).
 - MakeCD doesn't require valid base dir to open settings any more.
 - "Copy CD" achieved exact track starts in TAO by truncating audio tracks.
   This was not obvious and not always desired, therefore this behaviour
   is now configurable in the "Copy CD" window.
 - The write window and the abort requester now both remind of the current
   activity (writing for real, in test mode, not to CD-R at all). It is
   also much more verbose about what MakeCD is doing and even predicts
   what actions will happen after "Start" and how much time is still
   required.
 - Write window displays message while unlocking filesystems and before
   closing the window - some filesystems sometimes lock up MakeCD at this
   stage, for unknown reasons.
 - The maximum buffer size is not updated after writing even though the memory
   is already freed - this is not quite correct, but less confusing.
 - Playing a single track now starts immediately because no track
   selection is necessary and the other parameters can be easily changed
   while playing.
 - MakeCD now distinguishes between "user abort" and "automatic abort as
   requested by IO error setting" in messages.
 - amigaguide.library is no longer essential to run MakeCD.
 - Bugfix: file requester for CDTV/CD32 boot file didn't work properly.
 - Bugfix: menu item "Update writer" didn't work.
 - Bugfix: menu item "Fix CD-R" only fixated the session


Known problems:
 - Teac 55S: discs with mode 2 + audio tracks are not written correctly
   in one case - reason unknown, still under investigation.
 - Panasonic 7502 V4.10: audio subcodes are incorrectly written in
   one case, but others were not able to reproduce this - reason unknown,
   still under investigation.
 - Remaining time during writing is sometimes completely wrong due to
   the way this estimation is done. That's just the way it works.
 - JVC XR-W2010 V1.51: ISRC/Medium Catalog Number are not supported,
   therefore errors when trying to use them are silently ignored.
   At least the Philips CDD2600 always only sees the first session
   of a DAO-CD-R written with JVC XR-W2010 V1.51, even though
   "Fixate: Session" was used and other drives (JVC itself, Yamaha
   CDR100) find the next session(s). Could only be fixed by JVC or
   Philips.
 - Before closing the write window MakeCD hangs on some systems. This
   is caused by the filesystem that MakeCD locked to get exclusive
   access to the hardware. Don't mount filesystems on your writer 
   or unmount them before using MakeCD.
 - The Sony 928E sometimes returns garbage values when reading the contents
   of a CD with the CD writer specific commands. This doesn't happen when
   the standard commands are used, so the CD_ATAPI driver should be used for
   reading from the Sony 928E.
 - Not a real problem: ATAPI doesn't specify DAO, so not all ATAPI writer
   support it. But as we are not sure, we haven't disabled DAO in the
   driver for all ATAPI writers unless someone has tested it. If yours fails
   with the error "MODE_SELECT6: Illegal Field in Parameter List" when trying
   to write in DAO, then it doesn't and Patrick would appreciate a short
   notice.
 - workaround for mode 2 (2336 bytes) reading problems:
   Some drives deliver invalid data with this block size (Sony 928E), others
   report checksum errors although they shouldn't use the checksums (Plextor
   PX-20TS). In both cases reading raw data still works, so now MakeCD can
   read mode 2 tracks raw and throws away the unnecessary bytes if necessary.

   This workaround is enabled only for the Sony 928E, but not for any other
   drive, because some drives might read raw data slower and the additional
   processing requires more CPU time. On slow computers with fast drives the
   computer might perhaps even appear to be locked.

   If your drive has similar problems you can enable this workaround by
   creating the environment variable "MAKECD_READ_MODE2_RAW" (e.g. with
   "setenv MAKECD_READ_MODE2_RAW=1"). Please tell Patrick if your drive only
   reads mode 2 properly with this workaround.
 - Panasonic 7501: writing mode 2 fails with an error. As I haven't found a
   way around that mode 2 in spite of several different attempts this will
   stay this way, i.e. writing mode 2 will never be supported by
   CDR_Panasonic.driver.
 - Triton has some problems when you keep the cursor keys pressed
   to move within a listview: it might cause MakeCD to open windows.
   It also doesn't calculate the width of the data mode gadget correctly.
 - Ricoh MP6200S (V1.00): doesn't support medium catalog number in DAO,
                          swallows ISRC of first audio track in DAO
 - Yamaha CDR400 (V1.0i): audio scanning gives indices that are too
                          high by two seconds
 - The amount of free blocks is not perfectly correct, because writers
   report different values depending on what was already written. Sometimes
   gaps are included and sometimes not. This causes an inaccuracy of upto
   3 seconds. The end of CD-R check is not affected, though.
 - The Ricoh MP6200S doesn't simulate formating. Therefore MakeCD doesn't
   issue the formating command at if test mode is enabled.
 - Writing audio with preemphasize is not possible with CDR400 (upto V1.0d).
   Yamaha didn't respond to my bug report, so I implemented a fallback
   to normal audio.
 - Yamaha CDR400 sometimes panics after a buffer underrun and doesn't
   stop writing. Yamaha could not help with this problem yet.
 - Ricoh sometimes doesn't read contents properly. Exact reason/solution
   unknown, but turning disconnect off and SCSI bus reset help.
 - dracoscsi.device SCSI direct still doesn't work properly. You must have
   enough free memory in your first memory module or a 32MB module in the
   first socket, otherwise the dracoscsi.device will fail and probably cause
   crashes or read errors. The same rules apply to every other software doing
   SCSI direct.


MakeCD 3.2a:
   MakeCD 3.62 (12.08.98)
   makecdromfs.module V11.0 (24-05-1998)
   ReadWrite.module 15.57 (12.08.98)
   SCSISupport.module 14.26 (12.08.98)
   
   CDR_JVC_Teac.driver 13.75 (12.08.98)
   CDR_Panasonic.driver 13.75 (12.08.98)
   CDR_Philips_2000.driver 13.75 (12.08.98)
   CDR_Philips_2600.driver 13.75 (12.08.98)
   CDR_Plextor_Ricoh.driver 13.75 (12.08.98)
   CDR_Ricoh_6200.driver 13.75 (12.08.98)
   CDR_SCSI3_ATAPI.driver 13.75 (12.08.98)
   CDR_Sony.driver 13.75 (12.08.98)
   CDR_Yamaha_10x.driver 13.75 (12.08.98)
   CD_ATAPI.driver 13.75 (12.08.98)
   CD_NEC.driver 13.75 (12.08.98)
   CD_ROM.driver 13.75 (12.08.98)
   CD_Sony_Plextor.driver 13.75 (12.08.98)
   CD_Toshiba.driver 13.75 (12.08.98)

 - **8SVX samples*** are now imported in mono and stereo as long as they
   are pure samples, i.e. without repititions.
 - bugfix: "Adapt Audio Start" in the hardware compatibilities was not saved
   correctly.
 - Filtering .CDT in the image file requester is no longer done with a simple
   pattern but with a more intelligent function, because for long file names
   the suffix is truncated and not recognized.
 - Some drives had problems during scanning of the SCSI bus if the
   CD_Sony_Plextor.driver was installed.
 - bugfix: jumping within MPEG-A images during playing was only possible in
   the first ~40min of each image
 - bugfix: getting the length of .WAV files without FACT chunk broke in a
   last minute fix for invalid .WAV headers :-/
 - CD_ATAPI.driver: reading didn't work with drives any more that didn't
   support selecting the block size.
 - ".iso" can also be used to prevent the (false) recognition of images
   as MPEG-A.
 - bugfix CDR_Ricoh_6200/CDR_SCSI3_ATAPI.driver: erasing a track on a CD-RW
   was _clearing_ the track, i.e. overwriting its contents with zero, and
   not actually _removing_ it as documented in MakeCD.
   BTW: if erasing a CD-RW doesn't work, turn off the global test mode again
   as it is by default - the test mode setting in the write window is set
   and saved independently.
 - bugfix DAO + session fixation: MakeCD overestimated the required
   space by 30s and therefore issued a warning in cases where it wasn't 
   necessary.
 - bugfix: conversion of strange frequencies left some random bytes at the
   end of the sample uncleared, which resulted in a short clicking noise.
 - workaround for MAUD: different interpretations are used for the field
   defining the length of the sample. Unfortunately, MakeCD had been
   made compliant with a software that set this field incorrectly and
   therefore read only one half of correct samples. The real length is
   now recognized in both cases and correct MAUD images are generated.

MakeCD V3.2a, Rev II

 ReadWrite.module 15.58 public beta (12.09.98)
 CD_Toshiba.driver 13.78 public beta (12.09.98)
 
 - workaround for 8SVX files with invalid number of samples in header
 - CD_Toshiba: the driver now tries to enable a 4x audio read mode
   that is supported by some Toshiba drives, e.g. XM5701B, XM6201B, and
   XM3801B. Reducing the speed for normal data tracks to 1x is also
   possible.
 - CD_Toshiba/CD_Plextor: the environment variable for enabling/preventing
   speed selection have been changed to MAKECD_SPEED_NO and
   MAKECD_SPEED_YES.
 - bugfix: byte order of raw data files could have been reversed,
   completely trashing data tracks in DAO.
