Extended BASICs - Ownership & Use

[Omega-TI]

There are multiple versions of Extended BASIC for the TI-99/4A, many of these versions are owned by different individuals. Some people may have written programs for their own use using these variants, but may never have released them for public consumption due to their lack of adoption by the greater community as a whole. So, while many of the Extended BASIC versions are actually BETTER than the TI version, most programs were still released in TI Extended BASIC to reach a wider audience.

Times change, with the advent of new cartridges, and the ability to burn our own, a greater percentage of remaining TI’ers (like Atari Age users) may now be able to exploit new and improved XB programs. For instance a couple of years ago the late Gazoo sold quite a few XB 2.7S cartridges and with them, their enhanced capabilities.

This poll is to determine how many of you actually own, use, program or simply let these different versions collect dust on a shelf.

Please, feel free to add any comments or suggestions in the message area below.

[Airshack]

The compiler for TI's regular XB (w/XB256 extensions) seems to make it stand well above all other versions of BASIC for writing games.

 

Not sure why anyone would consider a BASIC without a compiler?

[sometimes99er]

The compiler for TI's regular XB (w/XB256 extensions) seems to make it stand well above all other versions of BASIC for writing games.

 

Not sure why anyone would consider a BASIC without a compiler?

 

Yes. Why limit yourself. On the other hand, why do we limit ourselves and play with old computers. Nostalgia !? XB256 is a rather new thing. You could twist and turn it and say the TI-99/4A does something that new computers doesn't, and that's probably true ...

[+InsaneMultitasker]

I use the standard TI XB cartridge mostly because it works and does what I need it to do. When I program in XB, it is on real hardware, either with the TI or Geneve. I even prefer the line numbers and editing it natively, though I have on occasion used some of the SXB style tools to resequence or delete chunks of code. PreScanit! comes in handy, though you can do the same work manually by keeping track of your variable names.

[Airshack]

Certainly! Point well taken.

 

From my experiences with the old TI-99/4(no-A) days, I'd have to say I probably abandoned the platform back in 1981 primarily due to the slow performance of TI BASIC and XB, and I couldn't afford the PEB as a gameroom attendant at Chuck E Cheese's Pizza Time Theater.

 

The BASIC compiler Harry Wilhelm has developed is significant because it removes the main flaw with TI XB -- speed. A tremendous flaw.

 

Ever try to use CALL COINC with fast moving Sprites? The technology and capability of TI's extended BASIC is arguably head and shoulders above most BASICs of its era, with the exception of some third-party variants on the ATARI 800.

 

Speed killed TI BASIC and speed kills TI Extended BASIC when trying to build dynamic games.

 

The whole 16-bit CPU shoe-horned into an 8-bit data bus is a problem, right?

 

So my enthusiasm for the compiled XB with XB256 extensions is there because the compiler removes the speed issue! That's a quantum leap for BASIC programming!

 

So that's why the other interpreted BASICs seem not only obsolete, but doubly-obsolete due to the speed of Wilhelm's compiled BASIC.

 

 

 

Sent from my iPhone using Tapatalk

[RXB]

You know the solution to this is what I would like to do if the community would help me pull it off.

 

Rewrite the XB ROM's and remove Strings and Variables in VDP and put them in RAM using the SAMS instead.

String space could be increased to 64K and VDP Stack could be replaced with a RAM STACK instead.

 

Look most of the speed lost in XB is mostly due to all the Strings/Variables/Stack is all in VDP, moving all of this to RAM would be a huge speed increase.

Results would be:

64K of String space instead of under 12K currently.

VDP Stack would be in RAM Stack increasing speed as RAM is much faster than VDP.

Variable space would increase so we could actually use a ARRAY of 7 dimensions instead of the current limit of 4 at best due to lack of memory in 12K VDP space.

XB pays a price in GPL space and ROM space as it has to check to see if using VDP or RAM, this change would be RAM only and 1 Meg of SAMS would really give a bang for the buck.

SPEED OF XB!

[sometimes99er]

Certainly! Point well taken.

 

From my experiences with the old TI-99/4(no-A) days, I'd have to say I probably abandoned the platform back in 1981 primarily due to the slow performance of TI BASIC and XB, and I couldn't afford the PEB as a gameroom attendant at Chuck E Cheese's Pizza Time Theater.

 

The BASIC compiler Harry Wilhelm has developed is significant because it removes the main flaw with TI XB -- speed. A tremendous flaw.

 

Ever try to use CALL COINC with fast moving Sprites? The technology and capability of TI's extended BASIC is arguably head and shoulders above most BASICs of its era, with the exception of some third-party variants on the ATARI 800.

 

Speed killed TI BASIC and speed kills TI Extended BASIC when trying to build dynamic games.

 

The whole 16-bit CPU shoe-horned into an 8-bit data bus is a problem, right?

 

So my enthusiasm for the compiled XB with XB256 extensions is there because the compiler removes the speed issue! That's a quantum leap for BASIC programming!

 

So that's why the other interpreted BASICs seem not only obsolete, but doubly-obsolete due to the speed of Wilhelm's compiled BASIC.

 

Yep.

[Meddler]

... "On the other hand, why do we limit ourselves and play with old computers. Nostalgia !?"...

 

 

... "Speed killed TI BASIC and speed kills TI Extended BASIC when trying to build dynamic games.

 

The whole 16-bit CPU shoe-horned into an 8-bit data bus is a problem, right?" ...

 

Says it all really, advertised as the first 16 bit personal computer and hobbled at the gate!

Which is why Assembler is the only way to go, which is actually true of any computer for speed needs.

 

Anyway, this is not really the right post for all these musings, perhaps it could be continued elsewhere?

[+mizapf]

The whole 16-bit CPU shoe-horned into an 8-bit data bus is a problem, right?

 

The 9995 also has 8 data bus lines but is much more efficient than the 9900 with its 16 lines, as it does not require the read-before-write procedure. The impact on performance is less obvious than it seems. Remember that the Intel 8088 was a 16 bit CPU with only 8 data bus lines as well, so this was not a TI-specific idea.

[RXB]

Says it all really, advertised as the first 16 bit personal computer and hobbled at the gate!

Which is why Assembler is the only way to go, which is actually true of any computer for speed needs.

 

Anyway, this is not really the right post for all these musings, perhaps it could be continued elsewhere?

Do you have any idea how much work it would be to make a Assembly only version of XB?

GPL which is more compact then any language even possibly Forth and Source listing alone is 4 books 4 inches thick each. And I also install EA too in same Cartridge.

It would be a insane amount of work to make the entire thing Assembly and only in the SAMS and then you have to problem of spending an insane amount of time switching out banks.

 

But you are fully welcome to try:

 

 

RXB TEXT ZIP.zip

Edited January 23, 2017 by RXB

[Meddler]

Do you have any idea how much work it would be to make a Assembly only version of XB?

GPL which is more compact then any language even possibly Forth and Source listing alone is 4 books 4 inches thick each. And I also install EA too in same Cartridge.

It would be a insane amount of work to make the entire thing Assembly and only in the SAMS and then you have to problem of spending an insane amount of time switching out banks.

 

But you are fully welcome to try:

 

 

Hi there and Woa down a bit, that it not what I meant at all:

 

I only meant that it is desireable to develop programmes in Assembler to avoid the BASIC interpreter.

[RXB]

Hi there and Woa down a bit, that it not what I meant at all:

 

I only meant that it is desirable to develop programmes in Assembler to avoid the BASIC interpreter.

Yea Basic is twice as slow as Extended Basic.

 

RXB has a built loader for SAMS Assembly memory, only XB that has this built in.

[Omega-TI]

Okay, after a couple of weeks and 32 replies, the results seem to be taking some form...

 

Good old TI Extended is leading the pack with 48% of the respondents saying they actually use it. That's not bad!

My take from this? After all these years, if you want to be able to share a plain Extended BASIC program to the widest audience, the original cartridge is still the route to go.

 

When it comes to using BASIC extensions, only eight people or 25% of respondents say they use them. Now here is where it starts to get interesting... only 6 people or about 19% say they compile their programs into E/A5 code afterwards.... BUT half of the community said they would write programs if they could be of some use to someone. One person even stated, "Not sure why anyone would consider a BASIC without a compiler." -- Which picked up a couple of likes.

 

 

So, TI Extended BASIC with the XB256 extensions seems like a nice route for some budding programmers to explore.

 

Among other things, if everything works out, there is a planned demonstration for Fest West this year that will show newbies and others the steps to take from TI-Extended BASIC through compiled format and into FlashROM 99 format. In the near future you may also see something FANTASTIC using those steps.