It is possible I did a version of IS-BASIC with German error messages but I can't remember doing it.

Your memory is good! The German messages are in the language expansion ROM, this also contains a new KEYBOARD device and character set, and EXOS command and variables for switching the language. Later Hungarian, Spanish and Danish versions were made.

But what your team did: built in the capabilities into the EXOS and BASIC for the extension programs can handle the messages.
This is the miracle of the Enterprise: everything can be expanded!

 After the long delays this 50% extra speed was some compensation and some advantages over competitors.

but at the same time!

You may well be right about the border colour. Another possibility is either a different batch of Nick chips with the first batch happening to power up with the register at zero by accident, but not the later batch. Another possibility is just slightly different conditions as the Enterprise comes out of reset eg. different capacitor values on the reset line or different properties of the power supply. I have seen problems with this sort of thing since - it's easy to write a program that forgets to initialise a variable because it always powers up as zero, only for a production change to change that and break the sw.

I would probably need a manual for my own FORTH because I cannot remember enough about it from all those years ago! It's a very efficient language and in the hands of a skilled programmer can produce very readable code, but in the hands of an unskilled one produces a complete unintelligible mess! Large amounts of the US Space Shuttle was apparently written in FORTH (not mine!)

I can't answer your question about the Z80B. I can only imagine that it was either a cost or availability issue. As you say it might have partially compensated for all the delays. I am continually amazed and flattered by what you have done with the Enterprise and in how much depth you have researched it!

but at the same time!

For the EXDOS bug?
It is at the boot sector verification. At one point if the disk parameters are invalid, returned with error flag (NZ), but not loaded error code to the A register, it is contain garbage from the previous calculations. Later this garbage handled as error code. If you select right values for the invalid boot sector then you can select any error message

Another possibility is just slightly different conditions as the Enterprise comes out of reset eg. different capacitor values on the reset line or different properties of the power supply.

At the final Enterprises the reset not reset the Nick only the Dave and Z80.

I would probably need a manual for my own FORTH because I cannot remember enough about it from all those years ago!

I got the original English user manul few months ago (previously we never have it ), I will scan it shortly! Today I leafed through the book, it is look a very interesting programing language!

First question: it is talking separately about the IS-FORTH and then the Enterprise specific FORTH things.
The IS-FORTH is released for other systems? Or it is planned for later but not happened because the collapse?

I can't answer your question about the Z80B. I can only imagine that it was either a cost or availability issue.

I hope one day somebody from the Enterprise Computers will find us and can answer for this and many other questions!

there were UK government restrictions on exports of high technology to Hungary

It is not only the UK, generally the West countries made a technology embargo against the Eastern block.
We called it as COCOM lista. For example the C64, IBM PC, etc also couldn't be exported to Hungary in these years. These were  bought in Wien and hidden in the car and hoping the Customs officer won't find it...

Low-tech was ok, so a 6845 video display chip would have been ok

Up to now we know the story: The Videoton bought the early version of the Enterprise.
You say this is false story (probably at that time the Hungarian management told this), it is not an early version, it is a cut version which meet the technical embargo?

How did the TVC developement happen at the IS? Is it another wired prototype? Or you got a real machine?

Good evening, Bruce! It's really cool to have someone onboard who contributed as much to the Enterprise computer as you have! Your posts are quite entertaining, and they bring back a lot of fond memories... for all of us, as it seems. The Enterprise was so much better than any of the toys of that era... I mean... it really was intelligent. I'm still laughing at console windows these days that don't wrap around words so nicely at the end of a line as the Enterprise editor did... sure, CP/M and MS-DOS have existed quite a few years before _our_ computer, but come on, this is the 2010s now, and they still don't do it - or am I missing something?

Could you share some information about EXOS versions prior to 2.x? Or was 1.x simply what we would call an 0.x alpha/beta build today?

Thank you; now you mention it I do remember having long discussions about word wrapping and what a "word" is, and yes I agree with you there are many editors and word processors that still do not get it right, it drives me mad too! It's not difficult, it just needs thought and attention to detail.

Unfortunately my memory is rather hazy on how EXOS went from 1.x to 2.x but I will read zozo's documents and see if it jogs any memories.

In Hungary all sold machine is Enterprise 128 with EXOS 2.1 and IS-BASIC 2.1.
We were thinking about for a long time why 2.1?
Later at the internet time when firstly contacted with English users get the surprise: the Enterprise 64 have a EXOS 2.0 and IS-BASIC 2.0.
Ok, it is a earlier version... but why 2.0??, where is the 1.x?


In the last few years I got some documents which are from the pre 2.0 times!
Sorted by release date:
- 1983 Danish Setup Guide for ELAN(!) Enterprise 64
- 1984 March Technical Answers from the Enterprise Computers

- 1984 Sept EXOS 1 Technical Manual (subversion unknow ) originaly come with the 000043. machine to a sw developer. But the machine have a 2.0 ROMs (but in EPROM not a ROM).

Early Function key labels, this is shown on many books on the cover Enteprise 64 photo.
[ Guests cannot view attachments ]


The ELAN Setup Guide from 1983 is very amazing! The Enterprise Computers thought the launch is very near and printed out the books? And later the English ELAN books smashed, but the somebody thinked the Danish version not too important, enought good the old version. (I asked the owner, the machine and the Programming Guide is a normal English version.)


The interesting things from this book:
No cartridge at the accessories!
[ Guests cannot view attachments ]
Later it is talking about the cartridge as expansion capatibilities.


Startup screen is very different than the well known!
[ Guests cannot view attachments ]
This is the normal, from the final Setup Guide:
[ Guests cannot view attachments ]


The Technical Answers 19. also say 58K free memory what we see at the ELAN startup screen.

On the early Function Label we see a very interesting commands ECHO, COPY...
The answer 42. talking about these: this is a word processing function.


And differences from the EXOS 1 Technical manual:
- no time/date functions
- some EXOS variables are missing. The most important the CRDISP_FLAG (this can enable/disable the blinking ENTERPRISE start up screen)
- no EXOS file format. This means machine code program can't loaded directly. Some very early machine code games (for example Beatcha) using BASIC loaders for loading the code. I think this is the reason.
The loadable EXOS extensions are also missing.
- for the ROM extension no RAM area allocating, and no initialization CALL
- video device using 256 characters set, the 128 characters are doubled as default. Later 128 characters set used, this is reduced the memory usage with 1152 bytes.


What we think about these:
Originaly the EXOS and the BASIC still in one ROM chip on the motherboard, no cartridge needed! The machine is directly started with the BASIC, no ENTERPRISE copyright display.
Programs can be LOAD-ed by the BASIC such as other machines.
No WP, but the BASIC have some primitive word processing (type and print) capatibilities.


When the release is delayed because the custom chip problems, the programmers used the time for made whole software better and better and better...
When run out the ROM space, then the decision: BASIC will going to the cartridge! Then the EXOS got a lot of free space, added the EXOS module format, loadable system extensions, etc. Plus added the normal Word Processor program, and in the remaining space the blinking ENTERPRISE startup screen.

Probably the BASIC is also extended, but currently don't know about it
Some parts of the BASIC still remain in the onboard ROM, if I know right then the mathematical routines. Other languages (LISP, FORTH) use these routines?



And it is very interesting: the EXOS 1 manual from 1984 September. The EXOS 2.0 from 1984 November.
All of these enhacements writed in these few months?

What we think about these:
Originaly the EXOS and the BASIC still in one ROM chip on the motherboard, no cartridge needed! The machine is directly started with the BASIC, no ENTERPRISE copyright display.
Programs can be LOAD-ed by the BASIC such as other machines.
No WP, but the BASIC have some primitive word processing (type and print) capatibilities.


When the release is delayed because the custom chip problems, the programmers used the time for made whole software better and better and better...
When run out the ROM space, then the decision: BASIC will going to the cartridge! Then the EXOS got a lot of free space, added the EXOS module format, loadable system extensions, etc. Plus added the normal Word Processor program, and in the remaining space the blinking ENTERPRISE startup screen.

Probably the BASIC is also extended, but currently don't know about it
Some parts of the BASIC still remain in the onboard ROM, if I know right then the mathematical routines. Other languages (LISP, FORTH) use these routines?



And it is very interesting: the EXOS 1 manual from 1984 September. The EXOS 2.0 from 1984 November.
All of these enhacements writed in these few months?

Once again zozo I think you have worked out the situation perfectly. So am I right in thinking 1.x is Elan, 2.x is Enterprise? I think we just carried on developing until the hw was nearly ready to ship! And you are quite right the original intention was for everything to fit in the internal ROM with no BASIC cartridge, but we ended up having to split BASIC between a cartridge and the internal ROM. There was no plan to allow other cartridges to access the BASIC floating point in the internal ROM - I don't know if it would work or not and I wouldn't be surprised if it didn't! I'm not even sure what would happen if you used the "wrong" version of the BASIC cartridge for the internal ROM.

I'm not sure about those manual dates - possibly we just edited the EXOS 1 manual a bit to make it suitable for external publication, and changed the date at the same time!