The Sinclair QL Forum

Hi

I have completed the date of the QDOS Companion into a ebook PDF format.

I have submitted the draft copy to Andrew Pennell for comment, just awaiting a response.

I think the book is many years out of date.

I will try to determine whether Andy wants to upate the book.

When I get an update from Andy, I will post the PDF file here.

Nice work Derek.

Cheers,
Norm.

Thank you very much for all of your effort, Derek. That is a really valuable contribution to the community! As you said, it's outdated but I think it's still very relevant in many ways. The book has been very helpful in helping create a device driver since I think there was an explanation of some bits and pieces that were not available in an easily digestible form elsewhere.

Hi,

Whether there are problems with the QDOS Companion text, it has renewd my interest in Assembley Language, I forgot about a the neat System Function:Trap and Vector, that can produce some neat effects that other Computers have to have them programmed like: Horizontal & Vertical Scrolling, scrolling of indivdual screen lines.

The only ommission is hardware sprites, then the QL would a games computer like the C64, but faster.

Myother hope is to get all these QL.book authors to regain interest in the QL.

Derek_Stewart wrote:Hi,<>
Myother hope is to get all these QL.book authors to regain interest in the QL.

A pious thought, indeed

IIRC there was some confusion in that book about the RI stack. (Mine only lasted a few days before it split into numerous pieces, so I dont have it.) Perhaps others can confirm?

A general question: Would it be an idea to swap out all the old nomenclature for the new? Now would be the chance!

My main reasons are:
1. to prevent confusion and prevent the evolution of two separate nomenclatures
2. to create a seamless transition once someone wanted to move on to more advanced topics and/or SMSQ/E
3. to enable folk to more easily read and understand SMSQ/E sources, which, after all, are the only properly documented and annotated sources for a "QL" OS we have
4. to enable folk to more easily read and understand the Qdos/SMSQ Reference Manual, the most accurate and current documentation available.

Just my penny's worth

pjw wrote:

Derek_Stewart wrote:Hi,<>
Myother hope is to get all these QL.book authors to regain interest in the QL.

A pious thought, indeed

IIRC there was some confusion in that book about the RI stack. (Mine only lasted a few days before it split into numerous pieces, so I dont have it.) Perhaps others can confirm?

A general question: Would it be an idea to swap out all the old nomenclature for the new? Now would be the chance!

My main reasons are:
1. to prevent confusion and prevent the evolution of two separate nomenclatures
2. to create a seamless transition once someone wanted to move on to more advanced topics and/or SMSQ/E
3. to enable folk to more easily read and understand SMSQ/E sources, which, after all, are the only properly documented and annotated sources for a "QL" OS we have
4. to enable folk to more easily read and understand the Qdos/SMSQ Reference Manual, the most accurate and current documentation available.

Just my penny's worth

Possibly Derek could put the source files in github then people can make their own editions with ease!

I have a copy of both books (Assembly Language Programming and QDOS Companion) and still find them useful as reference material. They are somewhat outdated though as they are based on QDOS 1.03 (JM ROM) and thus miss a lot of developments since early 1985. There are also a number of errors (some of which I have annotated).

Having written thousands of lines of code based on the old nomenclature, I guess I will probably never get accustomed to the SMSQ/E conventions

Derek_Stewart wrote: IIRC there was some confusion in that book about the RI stack. (Mine only lasted a few days before it split into numerous pieces, so I dont have it.) Perhaps others can confirm?

IIRC the only confusion was about the order in which the CA.GTxxx vectors evaluate parameters - the book states that they are evaluated from first to last, while the actual order is the reverse. So the first argument of a procedure or function will be actually at the top of the stack pointed to by BV.RIP - and since the RI stack is upside down this will have the lowest address.

The same paragraph also mentions that these routines do not support coercion, which is not true. They do support coercion, but not using a name as string (e.g. DIR mdv1_ rather than DIR 'mdv1_').

I love that this has sparked something super productive.

XorA wrote:

pjw wrote:

Derek_Stewart wrote:Hi,<>
Myother hope is to get all these QL.book authors to regain interest in the QL.

A pious thought, indeed

IIRC there was some confusion in that book about the RI stack. (Mine only lasted a few days before it split into numerous pieces, so I dont have it.) Perhaps others can confirm?

A general question: Would it be an idea to swap out all the old nomenclature for the new? Now would be the chance!

My main reasons are:
1. to prevent confusion and prevent the evolution of two separate nomenclatures
2. to create a seamless transition once someone wanted to move on to more advanced topics and/or SMSQ/E
3. to enable folk to more easily read and understand SMSQ/E sources, which, after all, are the only properly documented and annotated sources for a "QL" OS we have
4. to enable folk to more easily read and understand the Qdos/SMSQ Reference Manual, the most accurate and current documentation available.

Just my penny's worth

Possibly Derek could put the source files in github then people can make their own editions with ease!

I will do this, once I have the author's OK.