* History
* ¸=====¹
*  ¸===¹

* Letters now being used rather than a full stop until final release version
* to stop the version number becoming very big.

* Version 1h09
* ~~~~~~~~~~~~
* "Are You Sure" prompt no longer has too many spaces in it. Problem caused
* by using an early version of NeatAsm that didn't handle quoted strings
* on the source file.
*

* Version 1j09
* ~~~~~~~~~~~~
* ROM version improved with common MEGA_EXT command and no requirement for
* separate assembler module ROM_asm.
* Extra facilities added to GRAB and RELEASE along with ISGRAB, GRABOWNER and
* GRABLEN functions added.
* ISROM and ROM_NAME$ now both default to 49152 (i.e. The ROM port)

* Version 1k09
* ~~~~~~~~~~~~
* FREESEC now accepts a device name as a parameter. Like TK2 it requires a
* backslash in front of the device name to distinguish it from a channel
* number. Why TK2 requires this, I don't know, it (and I) could just check
* for the presence of a #. Would it be better to do this rather than stick
* to the TK2 method?

* Version 1l09
* ~~~~~~~~~~~~
* FREESEC no longer takes \ before device name. It checks for a '#' instead.
* MED_NAME$ also altered to work like FREESEC. _ERROR now sets the top bit
* of an error code to allow user defined error codes more easily. e.g.
* a=GRAB(64)
* a$='A very serious error'
* POKE_W a,LEN(a$)
* POKE_$ a+2,a$
* _ERROR a
* RELEASE a

* Version 1m09
* ~~~~~~~~~~~~
* Wotsit system implemented. GRAB_WOTSIT, FIND_WOTSIT and KILL_WOTSIT
* commands added to control WOTSITS.

* Version 1n09
* ~~~~~~~~~~~~
* Wotsit commands now working. New command LIST_WOTSITS added for listing
* wotsits. Currently only lists the names but later versions may give the
* address too.

* Version 1p09
* ~~~~~~~~~~~~
* The plunge is taken, routines no longer use sv.mega, they use the new
* Wotsit approach. Here goes.....

* Version 1q09
* ~~~~~~~~~~~~
* It took me less time than expected, only a couple of hours. Now then, lets
* hope it works properly. This version will also have any long branches that
* could be short sorted out.

* Version 1p09
* ~~~~~~~~~~~~
* Toolkit workspace now streamlined removing now redundant items. It is now
* 296 bytes + QDOS rounding.

* Version 1r09
* ~~~~~~~~~~~~
* Various miscellaneous changes

* Version 1s09
* ~~~~~~~~~~~~
* Runtime version cleaned up. No longer contains the workspace or many
* commands such as _1 to _8 and WINS. This means that attempting to configure
* the runtime toolkit will fail.

* Version 1s09 II
* ~~~~~~~~~~~~~~~
* Nasty bug found by Dilwyn fixed. BPEEK/_W/_L functions used a word offset
* into the basic area. This was due to them originally being modified copies
* of the SPEEK functions and the system variables are always less than 32K.

* Version 1t09
* ~~~~~~~~~~~~
* _ERROR function now only sets the top bit of the error code if the error
* code is non zero. This bug was located by a horrible stream of characters
* appearing on #0 when using _ERROR OPEN_DATA. Luckily the bug is in _ERROR
* and not OPEN_DATA.

* Version 1t09 II
* ~~~~~~~~~~~~~~~
* MEGA_EXT was not initialising the ROM version correctly due to the wotsit
* routine automatically assuming an error if the MTK? wotsit was not present.

* Version 1u09
* ~~~~~~~~~~~~
* CPEEK error trapping was failing, if a non console/screen channel or a
* non-existent channel was supplied then a nonsensical result was returned.

* Version 1v09
* ~~~~~~~~~~~~
* QPIPE command removed. I thought I'd done it ages ago! It is no longer
* required as OPEN_DATA/PROG do the job much better.

* Version 1w09
* ~~~~~~~~~~~~
* OPEN_DATA/PROG commands weren't enlarging the channel table by quite enoough
* (40 bytes!!!). This was causing the basic return stack to be corrupted and
* therefore spurious PROC/FN cleared errors. I hope that sorts out OPEN_PROG
* and OPEN_DATA for good!

* Version 1x09
* ~~~~~~~~~~~~
* At DJC's request GET_DATA, SET_DATA and FILE_LEN commands added. GET_TYPE
* and SET_TYPE also altered to default to channel #3 and are now actually
* smaller!
* Guess what, another bug in OPEN_DATA, as a result of the previous bug fix
* if the channel number used was the highest so far in that basic then if
* a channel was already in use then it wasn't closed properly before being
* opened for the new device.

* Version 1y09
* ~~~~~~~~~~~~
* FREESEC modified to support level 2 device drivers with very large capacities
* (e.g. Miracle hard drive) by using the Level 2 fs.xinf call.
* IS_LEV2 function added to check if a drive supports level 2 facilites.

* Version 1.10 - Release 1
* ~~~~~~~~~~~~~~~~~~~~~~~~
* Letter in version number replaced by point for first release version!!!!!

* Version 1a10
* ~~~~~~~~~~~~
* Dilwyn has found some bugs! Due to the pipes being wraparound the QFREE and
* QUSED functions were failing when the front of the queue wrapped around
* behind the back. Now sorted thanks to doing wraparound queues in A Level
* computing a couple of weeks ago!
* Now to take a look at _LBYTES.

* Version 1b10
* ~~~~~~~~~~~~
* OPEN_DATA/PROG and _P$/_D$ use both the data and program defaults. No
* extensions have been provided to set these independently. (_D sets both
* to the same as do _1 to _8). The solution? _PROG and _DATA.

* 26/11/92
* Version 1.11 - Release 1 (for real maybe!)
* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
* Sent Thursday 26th November to DJC.