WDCTools Limited C-Compilers

[BitWise]

WDC have today released free limited 65C02 and 65C816 C compilers for download here:

http://wdc65xx.com/wdctools-limited-c-c ... tter_2016)

I haven't tried them myself. I think the 250 line limit is a bit too restrictive. If you use any of the standard include files most of your line quota will be used up.

[HansO]

Indeed, severely limited. Crippled would better describe it. 2kb code size limit, 250 lines. No assembly listing.

Useless alas.

[BitWise]

HansO wrote:

Indeed, severely limited. Crippled would better describe it. 2kb code size limit, 250 lines. No assembly listing.

Useless alas.

Enough to make a LED flash but not much more.

[BigEd]

I feel like I can almost hear the discussion at WDC; they are torn between getting people interested in their tools and not wanting to give anything away. I'm sure they have not yet found the right balance. They should have confidence that support is what makes for a paying customer.

[GARTHWILSON]

I think free software tools is a major factor in what made PIC microcontrollers so successful in the 1990's. Part of this is the help it gives to hobbyists who can't afford much. They get cozy with the product, and later when they get into industry and they're now professionals, they bring their interest with them, and design your microprocessor or microcontroller into their products. Sometimes WDC seems to understand that, and sometimes not.

[Rob Finch]

I think free tools are a good way to build an audience for a toolset. Even for the non-free tools it looks like the license fee isn't too bad.
Do they have other tools besides 'C' / Assembler ?

I've an opinion that for the 6502 one may do almost as well as a compiler by running an interpreter if one wants to work in a high-level language. The focus of the 6502 isn't high speed and heavy duty apps. Is there an interpreter (BASIC ?) that can use more than 64k ?

[BigEd]

It may be that the "optimising" aspect of WDC's compiler is worth having - for anyone who finds they have time or space concerns. But one would need to try it to know that it is worthwhile.

John K's patch of Acorn's BBC Basic allows, I think, for 128k total RAM footprint on an '816: the interpreter and the user program live in bank 0 and the program's data lives in bank 1. It's a neat solution as it means the existing code dealing with 16 bit pointers is mostly untouched, and loading and saving can act as if bank 0 is all there is.

[Tor]

Well, this free compiler is frankly useless with those limitations. So I don't see the point of that release at all. On the other hand the full tools aren't very expensive IIRC (I should check again I guess). But let's look at the FPGA market for a moment: Altera (to take one) FPGAs are *everywhere* in the hobby arena. That overflows to the commercial sector too. This would not at all be the case if there was no free (as in beer) version of Quartus II. The deployment would be dramatically different.

[scotws]

I confess I don't understand the business logic behind not releasing the full version to the world for free either.

[teamtempest]

Quote:

Is there an interpreter (BASIC ?) that can use more than 64k ?

The C-128 uses a similar scheme as what I understand the BBC referenced above does. Program code lives in bank zero and variables are stored in bank one. This is all transparent to the user, as all the bank switching required is handled by the interpreter. However at the original 1MHz processor speed, BASIC programs are noticeably slower than on machines which do not bank switch.

[joe7]

I suspect they are providing "just enough" to do simple things with the SXB boards. For commercial purposes, the ($40?) registration fee isn't unreasonable, and the libraries support double-precision floating point, etc. But most people won't need that.

I'm going to try to run the trial version through it's paces in a few days.

[barrym95838]

teamtempest wrote:

... all the bank switching required is handled by the interpreter. However at the original 1MHz processor speed, BASIC programs are noticeably slower than on machines which do not bank switch.

Do you know if the unmodified string literals remain embedded in the code bank, or does the interpreter always copy them to the variable bank at run-time?

Mike B.

[teamtempest]

Quote:

Do you know if the unmodified string literals remain embedded in the code bank, or does the interpreter always copy them to the variable bank at run-time?

Good question. It seems to me the C128 interpreter would be simpler if it knew that every string manipulation was performed on a value in the variable bank (I'm thinking specifically of concatenation). Getting a string literal into the variable bank every time its value is requested would add considerable overhead time, but the alternative would seem to be to write every string manipulation routine to operate on values in either bank (which would take space).

[Gered]

joe7 wrote:

I suspect they are providing "just enough" to do simple things with the SXB boards. For commercial purposes, the ($40?) registration fee isn't unreasonable, and the libraries support double-precision floating point, etc. But most people won't need that.

I'm going to try to run the trial version through it's paces in a few days.

Is it even possible for an individual to buy a license? Their online store says "out of stock" and they don't seem very response to emails when I've tried. It seems kind of unbelievable that, as a customer, I have money I want to give them for their product but they will not accept it.

[joe7]

Gered wrote:

Is it even possible for an individual to buy a license? Their online store says "out of stock" and they don't seem very response to emails when I've tried. It seems kind of unbelievable that, as a customer, I have money I want to give them for their product but they will not accept it.

I tried to order it also, and wasn't able to add it to the cart either. Maybe give them a call, it could be a website bug or perhaps they had other reasons for pulling it.

http://wdc65xx.com/support/contact-us/