What do you think?

[Sinphaltimus]

......

 

But I have to ask... has anyone ever contemplated a Geneve replacement? As an add-on PEB card that would hold more interest to me.

That would certainly increase my PEB use.

[Omega-TI]

Currently the community appears to be split on this. Without overwhelming interest, or the numbers to purchase...

 

It's all good though. With items like the F18A, FR99 & FG99, 32K sidecar and possibly the TIPI device, the future look good!

Also, for the hardcore TI'er and those graduating from entry level status, there are always P-Boxes available.

[+adamantyr]

I took a look at the ZX kickstarter and I can see the appeal it would have!

 

I think a big issue with the TI though is it's much more than hardware. You are kind of trapped; If you want backwards comparability, you have to support GPL and other limiting technologies. But if you completely redesign the OS it's not really a TI anymore.

 

I actually think "completing" the TI-99/8 would be a better target. Most of the architecture is known, and you can throw in the F18A and maybe some extra sound support.

[Omega-TI]

I think a big issue with the TI though is it's much more than hardware. You are kind of trapped; If you want backwards comparability, you have to support GPL and other limiting technologies. But if you completely redesign the OS it's not really a TI anymore.

 

Under normal circumstances I would agree with you, but with Tursi Technology^TM, Mike's M.P.D. would allow the user to transform their new TI console into nearly anything they personally wanted. And if we've learned one thing here, it's that people have always chosen different paths when it comes to expanding their TI's. To me, this would be a continuation of that well established trend.

[Bill Loguidice]

To me, the ZX Spectrum Next should be the model for all classic computer recreations to follow. Basically, it provides close to 100% backwards compatibility with the additions of key modern conveniences. The catch is, not every classic gaming community could fund such an effort for their platform of choice. I'm skeptical if the TI community could.

[+mizapf]

I actually think "completing" the TI-99/8 would be a better target. Most of the architecture is known, and you can throw in the F18A and maybe some extra sound support.

 

On that path, I'd rather go for a 99105-based system. *That* would surely be an interesting thing with new challenges and potential.

[Omega-TI]

 

On that path, I'd rather go for a 99105-based system. *That* would surely be an interesting thing with new challenges and potential.

 

If it could be 100% backwards compatible with the 99/4A I'd say sure... at least for me. I, like many others, do not have any nostalgic ties to other systems, for example I never had a Geneve, so I'm simply not interested in one. Now If a system appeared to be a 99/4A until 'unlocked' by the end user's choice that might be a different matter. For example I was really impressed with the way Matthew180 did the F18A, it's a straight 9918 replacement... until the user decides he wants to use the 'extra stuff'. In essence it satisfies most people in both camps.

[+mizapf]

The 99xxx processors offer all features of the original TI-990, and their instruction set is a superset of the TMS9900. So indeed, it can execute the 99/4A programs. But beyond that, it offers a kernel mode, so you get a true operating system support with privileged commands, system calls, and hence true virtualization, multi-user support and so on ... So this, plus a good video controller (F18A, or a v9938-compatible one), some system memory with a reasonable mapper, and there you go. I can imagine that in such a system, similar to the Geneve, you can have a TI-99/4A emulation mode, or possibly several ones.

[+adamantyr]

Maybe a "TI-99/4a+" is what we really want.

 

An F18A, SAMS (32k in console natively), speech synthesizer in console, a built-in USB/card reader system for floppy/hard disk emulation, and the entire internal architecture running on 16-bit rather than 8-bit multiplexing, with an optional "slow mode" for compatibility.

Basically, a TI console you can just hook up to your monitor and you don't need any external equipment at all.

[Arnuphis]

I know the TI UK User group was working on a Geneve II card but progress on that seems to have stalled judging by the lack of updates.

[matthew180]

...

 

I think a big issue with the TI though is it's much more than hardware. You are kind of trapped; If you want backwards comparability, you have to support GPL and other limiting technologies. But if you completely redesign the OS it's not really a TI anymore.

...

 

It is not that a new hardware implementation *would* redesign the OS (i.e. code in ROM / GROM), it is simply that users *could* change that code. Since ROM is in the low memory map we cannot use the vector table. With an FPGA-based system, the ROM could be implemented with RAM and would give hobbyists a ton of options to explore the computer.

[RickyDean]

The 99xxx processors offer all features of the original TI-990, and their instruction set is a superset of the TMS9900. So indeed, it can execute the 99/4A programs. But beyond that, it offers a kernel mode, so you get a true operating system support with privileged commands, system calls, and hence true virtualization, multi-user support and so on ... So this, plus a good video controller (F18A, or a v9938-compatible one), some system memory with a reasonable mapper, and there you go. I can imagine that in such a system, similar to the Geneve, you can have a TI-99/4A emulation mode, or possibly several ones.

I like the idea of several TI CPU's emulated, I would go for a TI99/4/4A, Geneve, TI99/8, + few others, hybrid. At the flip of a hardware/software switch to go from one to the other, sharing peripherals(included on board of course, for the most part), especially storage CF,SD, USB, emulated as SCSI, MFM, IDE and so forth. A guy can dream, can't he ?

[Sinphaltimus]

If I didn't have so many other projects on the docket, I I'd go for this just because...

 

http://makezine.com/2014/08/07/the-ti-pi-a-raspberry-pi-powered-ti-994a/

[Omega-TI]

I like the idea of several TI CPU's emulated, I would go for a TI99/4/4A, Geneve, TI99/8, + few others, hybrid. At the flip of a hardware/software switch to go from one to the other, sharing peripherals(included on board of course, for the most part), especially storage CF,SD, USB, emulated as SCSI, MFM, IDE and so forth. A guy can dream, can't he ?

 

Nothing wrong with dreaming, as that it how all good things eventually come to be.

 

If I remember correctly, back in the day the TRS-80 Model IV's default was Model III mode, the user needed to boot the computer with the Model IV DOS to gain access to the extended features. The same sort of thing could be accomplished with the TI. In this way, people would have whatever limitations, or lack thereof that they want and all parties could be satisfied. Considering the limited user base we have, this approach would appeal to a greater audience.

 

I agree with you, having access to currently available peripherals is a must, BUT having a modern emulated SSHD internal would be nice. It's fun to talk about this stuff even though the probability of coming to fruition is rather low.

 

All things considered, I'm still patiently waiting on the TIPI.

[RickyDean]

 

Nothing wrong with dreaming, as that it how all good things eventually come to be.

 

If I remember correctly, back in the day the TRS-80 Model IV's default was Model III mode, the user needed to boot the computer with the Model IV DOS to gain access to the extended features. The same sort of thing could be accomplished with the TI. In this way, people would have whatever limitations, or lack thereof that they want and all parties could be satisfied. Considering the limited user base we have, this approach would appeal to a greater audience.

 

I agree with you, having access to currently available peripherals is a must, BUT having a modern emulated SSHD internal would be nice. It's fun to talk about this stuff even though the probability of coming to fruition is rather low.

 

All things considered, I'm still patiently waiting on the TIPI.

Well, there could be a start up screen similar to the TI99/4A, that gave a person the ability to choose the system that they wanted to work with, like selecting a cartridge. then you would still keep the spirit of the systems, even in the selection.

[x24b]

I voted "Hell Yes!".

With that said, I assumed a few things were implied in this NEW 99X Console...
<ENGAGE PIPE_DREAM MODE>… NOW!

=======
Please, keep the old TI Boot Screen.

HARDWARE: Use the original ROM for fast boot and operations. No changes, only enhancements to what already exists. I love the internal structure of our TI 99/4A’s memory. Even as dumb as I am, I was able to wrap my head around its internal boundary’s/functions.
All physical ports should be replicated with their original function. They will of course be enhanced with added functionality (Data/Control Line Pads/Through Holes) above and beyond their original use.
All of the I/O devices will be expected to change over time.
However, all old hardware should have the ability to plug-in and fully function as intended by the Texas Instruments Engineers of the day. Think PEB, MBX, joysticks (blech), etc..

HARDWARE ENHANCEMENTS: Built-in, or Add-in Console/PEB Mobo 32K RAM, F18A video 40/80 (breakouts for future changes), Speech Synthesizer, RS232 etc.. I would look for an additional path to current “hardware standards” for adding new Printers, keyboards, mice, USB, Storage (Every Floppy drive known to man, MFM, IDE, SCSI, SSD, SD card, to thumb drives), monitors and joysticks.
For instance, the ability to use a USB 3.0 device is absolute overkill, but it should be supported, or the ‘ability to be supported’ in future changes. Think access to data and control lines. Maybe GPIO HAT board connectors and the like.
CPU? Obviously it must be in the 9900 family with full backward compatibility and ‘Overclock’ switching as needed (Original MHz as standard unless CPU won’t). 16/32/64-Bit? ROM/GROM: Yes. Should not be touched, unless there are known bugs that could/should be euthanized. RAM, GRAM? Yes.

=======
SOFTWARE: Backward compatibility with most, if not all, software and cartridges. It should use the native Assembly Language, Basic and Extended Basic we are all comfortable with.

SOFTWARE ENHANCEMENTS: Add as NATIVE, or plug-in upgradable GNU tools like GCC and Loki’s fabulous SDL (Simple Directmedia Layer) v2.0. In this way we can all dive into C, C++, Java Script, Python, and Objective-C.

=======
The 99X Console Mobo might be a PEB card (or cards) with many ports.
That same Mobo could be plugged into a small External-Console-Case with many ports, also connected via a modern version of the Firehose connector to the PEB.
So, my version of the new 99X would be a Mobo with really well thought out card edge connectors, GPIO HAT connectors in a nice 3D printed console case. Old Keyboards could be added as a peripheral or all the way up to modern PC style WiFi keyboards and mice.
New replaceable/reproduceable open source printed circuit Mobo with pluggable chips might be nice.

 

I'm thinkin' embrace, love and keep the old, but extend it to the moon.

 

Just sayin’…
<DISENGAGE PIPE_DREAM MODE>… now.

[atrax27407]

The problem is that even now there is NOT backwards compatibility with the projects that are coming out. F18A programs will not, in most cases, run on a 9938/9958 system. Which nano-PEB works with what existing gear? There appear to be significant differences depending on the version. Until some basic compatibility ground rules are established, a situation will exist where nothing works seamlessly with existing software. You are left with a fragmented community.

[x24b]

The problem is that even now there is NOT backwards compatibility with the projects that are coming out. F18A programs will not, in most cases, run on a 9938/9958 system. Which nano-PEB works with what existing gear? There appear to be significant differences depending on the version. Until some basic compatibility ground rules are established, a situation will exist where nothing works seamlessly with existing software. You are left with a fragmented community.

Can a list of these major differences be catalogued so we can all ruminate? A view from 10,000 feet so to speak.

[x24b]

 

On that path, I'd rather go for a 99105-based system. *That* would surely be an interesting thing with new challenges and potential.

I'm not enamored with a full FPGA-based system. Old Iron, er... Silicon is like a shiny old Harley. So... I'd prefer:

 

NICE!

 

http://www.cpu-world.com/CPUs/TMS99105/

 

"Texas Instruments TMS 99105 is the third generation of 16-bit microprocessors. The family includes all instructions from the two previous generations - TMS9900 and TMS9995, and fully object-code compatible with them. New instructions, included in the TMS 99105, are 32-bit arithmetic and logic, bit test, signed multiply and divide, and stack-related instructions. In addition to these new instructions, it is possible to re-define unused processor opcodes as new instructions. This can be achieved either with the help of an attached processor (i.e. co-processor), or by using a Macrostore feature. The Macrostore allows system designers to map unused microprocessor opcodes to custom functions located in a memory address space, which is separate from main memory. This allows the CPU to access up to 128 KB memory (64 KB of main memory + 64 KB of Macrostore memory). It is also possible to split main memory into two memory segments - 64 KB instruction segment and 64 KB data segment. In this case maximum addressable memory size is increased to 192 KB.

The TMS99105 family uses memory-to-memory architecture. The main advantage of this architecture is that the set of CPU registers (called "Workspace") can be located anywhere in memory. This architecture makes saving and restoring of the contents of all CPU registers as simple as switching the base address of the workspace. The disadvantage of this architecture is that the processor speed is highly dependent on memory speed. The 99105 CPUs have external clock frequency 24 MHz, which is divided by four internally. It takes three or more machine cycles for the CPU to execute any instruction when no wait states are required to access the memory. This translates to maximum execution speed 2,000,000 instructions per second or less. Execution speed drops significantly when the processor is used with slow memory - instructions may execute up to two times slower when memory access requires 1 wait-state, or up to 3 times slower when 2 wait-states are required.

The TMS99105 is almost identical to TMS99110 16-bit processor, except that the TMS99110 has 1 KB on-chip Macrostore memory with pre-programmed Floating Point instructions. Texas Instruments also planned to release TMS99120 microprocessor which was supposed to include commonly used utilities for PASCAL language as Macrostore instructions."

[Omega-TI]

Compatibility is an interesting term, which sometimes seems to have a variable definition, usually dependant on the opinion of one who owns a piece of rare or hard to obtain hardware. When I use the term, it's usually only in relation to what Texas Instruments released, not what a third party outfit released decades ago with marginal sales figures. If I remember correctly, back in the day even Myarc & CorComp could not even get on the same page when it came to format compatibility on their FDC's, but then again, why should they? They were in direct competition, why should they be forced to comply with some set standard, limit their options and be no different than their competition?

 

I do like backward compatibility, but I also believe in creating new and modern schemes or methods to accomplish a task, especially when it improves our systems and gives us all new and exciting capabilities. In the end, the market will decide with their PayPal account. The more choices the better. One of the reasons the TI is so hot right now is because of innovation. I seriously doubt people would find the TI very exciting if we did not have F18A's, FG99's or the promise of new toys like the TIPI.

[x24b]

To me, the ZX Spectrum Next should be the model for all classic computer recreations to follow. Basically, it provides close to 100% backwards compatibility with the additions of key modern conveniences. The catch is, not every classic gaming community could fund such an effort for their platform of choice. I'm skeptical if the TI community could.

Well, Linus Torvalds just plugged ahead because he wanted a computer all his own. He did not have Maker communities, the Internet and CAD. We could design a 100% open source platform, making a 1:1 version of our beloved 99/4A, with logical places to extend it so others can add on to it. All we need is a really well thought out Mobo design and populate it with the exact same chips off of a 99/4A.

 

The magic would be in a truly modular design. GPIO pin breakouts and pluggable areas. If we are really clever a PEB board design Mobo could be acheived that doubles as a new and modern console Mobo with room for experimenters HAT's. Leave the power supply, keyboard, mouse and all of that up to the community.

 

With a more modern (laughs) CPU like the TMS-99105 real computing power could be had with full backward compatibility. Added data lines, clock, for a new (extended) wider bus would have to be added for a larger pin CPU.

 

Am I WAY off? A configurable, modular Mobo is all I'm talking about. No need to reinvent the wheel.

[majestyx]

To me, the ZX Spectrum Next should be the model for all classic computer recreations to follow. Basically, it provides close to 100% backwards compatibility with the additions of key modern conveniences. The catch is, not every classic gaming community could fund such an effort for their platform of choice. I'm skeptical if the TI community could.

 

I would have to say that's the cold hard truth. The ZX Spectrum Next reached its funding goal in a day and had gone to more than double what they original wanted, with add-ons for each additional level beyond the goal, not to mention over 2400 people having made contributions. While the TI forum on here may be the most active, it's a very small base of users, at least based on how many post.

 

As far as whether or not a new machine would actually be a TI-99/4A - well, the PC is still with us after being introduced in 1981 but due to its evolution over time, it's nothing like it was 36 years ago. In fact, you need emulation to run software from back then. That being said, I suppose it's a matter of what people are looking for in a new machine - as realistic to the original, some modifications, advances on the technology... some combination of all of these?

 

And as for the TI-99/8 - while I think we'd all love to see that could have been, once it's made, what do you do with it? What software is available that you can use from day one? How many people are going to develop for it once it's released? And how many people would actually buy it and at what price? From what I've seen, it's an emulated system in MAME but how many people actually even use that FREE program to do anything 99/8 related?

 

I voted YES mainly because the machines that are still available are still over 30+ years old and will eventually wear out. I'd love a dedicated solution along the lines of an RPi that could act as a TI-99/4A with add-ons built-in such as expanded memory, F18A capability, speech, access to cartridges/disc/cassette images via SD card, an acceleration option - pretty much what Classic99 can do on a PC. But that's only MY vision of what I'd personally prefer. I prefer less clutter now that I'm getting older and not having to insert cartridges, worrying about diskettes getting damaged, drives dying, and all the other headaches of add-on hardware. Having everything (or almost everything) in one little box would be ideal for me... but I know that others probably have a completely different vision than my own.

[x24b]

I just jumped on the ZX Spectrum Next kickstarter as backer 2590.

https://www.kickstarter.com/projects/1835143999/zx-spectrum-next?ref=nav_search

 

THAT is a good way to go.

[Omega-TI]

 

But why should that restrict folks who's hobby is more in line with bringing that old tech into the 21st century??...those self same folks who's tinkering has brought us things like nanoPEBs, 32K cards, USB keyboards and F18A's!

 

Regardless of whether I see a point to creating a 4X, there would likely be ancillary benefits derived from it - even for old stuck-up authenticity snobs like me!

 

 

Well said! The tent is big enough for everyone... or us poor slobs with multiple personalities!

I love upgrading my main TI with new tech, but I'm also making a totally modern small footprint system, and I also keep an unmodified stock system as well.

[Diecrusher]

Nope, not in the slightest.

 

The reason I go to the TI99 is for my nostalgic kick and memories of using it as a teenager. For me if it morphs that far I'll stick with completely modern hardware. It's like putting a 100 watt stereo speaker setup on a Speak & Spell. Why?