I must begin this month's Forum with an apology to any members who have written to me between late March and early May and found that a reply was rather slow in arriving. During this period the mainframe system I had been working on for the past three years had just gone live, with the result that everything else was put on 'hold', as my wife would confirm most forcefully if asked.
By the time you read this normal service will have been resumed, and hopefully in catching up no-one will be missed out.
A few months ago I reviewed BBCBASIC(86), the MS-DOS package published by M-TEC which provides a high degree of source program and operational compatibility with the BBC micro's Basic II and Basic IV.
Shortly after the review was published I received a 'phone call from a quite irate Dutch member about the review. Let me say at once that Beebug do not normally give out my phone number, but this member was obviously genuine, and since he had troubled to call from the Netherlands, in this case an exception was made.
I have since been expecting a letter from him, the content of which I had agreed to air in the Forum, but since I have not heard anything more I'll deal with the matter before any more time passes. If I omit any points he made perhaps I'll be excused on the grounds that I'm recalling a three month old telephone conversation.
He has himself been using BBCBASIC on the 512 for two or three years, and felt that the conclusions I reached didn't do the package justice, specifically in the area of execution speed. In short his own experiences were that the package did offer a gain in speed of two or three times the native BBC mode for most of the applications for which he used it. Although he agreed that my tests were (probably) accurate, and that screen output was slow, he felt the tests were un-representative and hence misleading.
I can only say that I had hoped that it was clear from the review that the speed comparison table was not supposed to represent a typical, across the board benchmark. The text made it quite clear that the FOR-NEXT loops and screen output had been specially isolated for the tests, and that for these particular operations the results were clearly disappointing. I hope that it was also clear that the overall results you would achieve in real programs would depend on what the program spends most of its time doing.
I gather that this member's main use of the package involves large amounts of calculation, and it's well known that Richard Russel prides himself, quite rightly, on his implementation of mathematical and trigonometric functions. However, I don't believe that the bulk of the average user's programs consist mainly of mathematical functions, while I do believe that screen output is a pretty common need.
It's always a risk in any review that some points may be overlooked and others overstressed, probably both by the reviewer and the reader (although with unlimited space and time perhaps both could be avoided). While I am completely happy to publicise the views of any member (even when they don't agree with me), I also stand by the contents of the review as published and believe that, in context, it was fair, balanced and accurate.
All the results were genuine, were carefully checked and can be reproduced. I should also say that no-one else has taken me to task about this subject, though presumably amongst the readership there are others who use the package.
I can only round off this item by re-stating the main conclusions again. In some areas the MS-DOS version of BBCBASIC(86) is much slower than I had expected, and in the area of screen output to an unacceptable degree. However, on the converse, its source and syntax compatibility coupled with its ease of operation are beyond reproach. As ever, only you can decide which of the strengths or weaknesses of a package are of relevance to you, and which should influence your decision.
The next item fits in quite neatly with the above mention of BBCBASIC(86) and last month's item on public domain (also sometimes called 'freeware', which I forgot to mention) and shareware software.
Richard Russel, as mentioned before, is the author of BBCBASIC(86), and he also wrote the version of Basic supplied with Acorn's Z80 second processor. What is not so well known is that he also wrote a version of BBC Basic exclusively for the 512, but unlike the previous two packages this one is not commercially marketed.
Having spoken to Richard, I now know a little more about the history of this software. It was originally offered to Acorn with the idea that it would be supplied with the 512, but for whatever reason Acorn declined to take up their offer. I can only assume that their financial position and general lack of interest in supporting the 512 extended to ready-made products like this too. I wonder how many more 512s there might be now if Acorn had included BBC Basic.
The fact that it's not commercially marketed is a two edged sword. On the one hand it means that you probably haven't heard about it because it isn't advertised (though I understand that it has been publicised on some bulletin boards), while on the other it means that if you can get hold of a copy it costs very little.
As I said, it was written for the 512, not for PCs running MS-DOS, so if you need a language for the 512 at home and a PC at work this won't fit the bill. It should also be borne in mind that if your use is in any way commercial, you should always buy your software and pay any multiuser licences as appropriate.
If, however, your need is for a language for the 512 for your own use then this one should suit you very well. The major difference between this and the MS-DOS version is that, as it was written for the 512/BBC micro combination and DOS Plus, with no compromises forced by alien PC hardware or MS-DOS, it provides native BBC screen modes from DOS Plus, including a proper mode 7 display. A further beneficial implication of this is that the output to screen is very much faster than in the MS-DOS version.
The software is not supported, and there are a couple of limitations specific to this version, though these are explained and quite easy to avoid. There is no printed documentation supplied with it either, though not surprisingly it's at least as easy to use as the commercial MS-DOS package, so that shouldn't deter you at all.
The final point about this software is rather good news for Forum readers. If you would like to acquire a copy all you need to do is to write to me at BEEBUG and mark the envelope '512 BBC BASIC'. You must enclose a stamped self-addressed envelope with a (512) formatted 5.25" disc. In return you will receive a free copy of the 512 version of BBC Basic, and it will only have cost you the price of the postage.
I've had several queries about Problem Solver, a product which claims to improve the PC compatibility of the 512 at a cost of £25.00. The reason that I have previously failed to mention it is that I have not been able to test it myself. The following is not a review, but rather is just the story so far.
I have two copies of Problem Solver both of which are version 1.0, but my tests have as yet been completely inconclusive. First impressions are that documentation is absolutely minimal, and very general in its claims for improved compatibility, since there is no list of software for which problems should be cured.
As you may recall, I use the 512 with a model B, but in spite of trying all three versions of Problem Solver (from both copies) with each of the two main DOS Plus versions (1.2 and 2.1), all I have achieved so far is to hang the machine, with seemingly no way out except rebooting.
I have also tried both copies in two Masters, one with an internally fitted 512, one external. I can then get it to load, and it does not hang the machine or interfere with normal DOS commands, but it doesn't do much else either, except that it prevents some software running that did work without it. By the way, in case you think this is just me, Andy Smith at Acorn reached precisely the same conclusions.
For the moment I'm taking the view that both copies in my possession are faulty. I certainly know of one member who had a very similar experience with his first copy, though a replacement was subsequently supplied which did work.
I have written to Shibumi Soft and for the moment I'll go no further, except to suggest that if any of you have this software and would like to supply me with your findings and comments I'll be pleased to include them in a later issue, when hopefully we can publish a proper review.
If you are able to contribute information about Problem Solver, I'd be interested to hear about specific packages for which it does (or doesn't) cure compatibility problems, but please include details of your hardware and software (i.e. Model B, Master, internal /external tube, version of DOS used, applications etc.).
Whether I am, or am not, finally able to successfully test this product myself, I would hope that we can at least draw up a table of conclusions showing which hardware/software configurations it helps (or hinders). After all, while £25.00 isn't a fortune, if it's unsuccessful it is a fairly expensive experiment.
A few letters have again mentioned the RS232 problem when using DOS Plus, so I assume that some readers would like to use the 512 to access PC bulletin boards, of which there are quite a number. I conclude that perhaps some of you missed my earlier comments about comms software compatibility.
In simple terms there is no problem with accessing the serial port from DOS, and this is easy to prove by connecting two 512s (or your 512 and a PC) 'back to back' by means of a serial cable. This I have done several times with an Amstrad portable belonging to a friend, and we have successfully sent files in both directions.
Of course, if you can copy files to a common disc format it's probably a better way to transport data. It's faster and of course it's more convenient, as you don't need both machines in the same place at the same time. However, many machines (like the Amstrad portable) rule this out because they use 3.5" discs, and I have only 5.25" drives, so this isn't possible.
Accessing bulletin boards is of course a completely different matter, and you must use suitable software which must be capable both of driving the modem and 'talking' to the board when you connect. The problem is that most comms software expects to be able to 'illegally' access the serial port on a PC, and hence does not work in the 512. This is because, in many PCs, the legal calls simply operate too slowly for any but the slowest data transfer rates.
I now have an evaluation copy of COMM+, mentioned in a previous Forum, and will be including a report within the next couple of issues. This package is the only comms software I know which uses legal calls and therefore does work in the 512. Due to lack of time I haven't yet investigated it in any detail, but my impressions are that it is a very complete and sophisticated package.
For anyone who can't wait for the review, contact: Margolis & Co., *******. Mention the fact that you have a 512, as there's a special offer for 512 users. You get BOTH the PC-DOS and the DOS Plus versions for £75.00 plus £2.00 P&P. This is a reduction of almost 25% on the normal retail price.