The real fight Atari versus Commodore

[carlsson]

Beyond the links I posted above, some more:

49 simple games in the Cassette 50 Charity Competition.

THEC64 Winter Game Development Competition however only saw a handful C64 entries (but far more VIC-20 entries).

Of course all of you are aware of the BASIC 10Liners that started as an A8 event but has grown to support multiple platforms. So far this year there has been 4 entries for A8 and 23 for C64.

 

Sure, neither of those are major, seriously developed games but an indication that it is about more than scene demos.

 

[Mclaneinc]

As Anders said, CSDB is a good old site, yes it shows scene releases as well but unlike the PD you cannot download them for a year after their sales release. I saw a mention of a drop in C64 releases but I simply don't see that, ok some stuff is recracks or improved versions but there's still a decent trickle of new original software. Obviously its never going to go back to the levels of when the machines were new but I still see enough software for both the Atari and the C64 to be happy to have both. Even the Amiga see's a slow but sure set of releases..

 

Life in the old dogs still....

 

Thankfully neither side listen to Emkay, he's enough to dishearten some coders.. 

Edited March 25, 2021 by Mclaneinc

[emkay]

2 hours ago, youxia said:

What's unbelievable is that you seem to lack basic English comprehension and yet continue to spam these threads talking to yourself. My point was that I'd much rather read the off-topic stuff than yet another 20 page tirade about some perceived, minuscule hardware differences, which you are mostly confused about, and which don't change the basic truth that (at least back in the day) they were plenty of skilled coders capable of coding for different platforms.

 minuscule ?

 

Please, read some books... before I get a laughing fit. 

 

Did you even get the sense? 

 

[emkay]

4 minutes ago, Mclaneinc said:

 

Thankfully neither side listen to Emkay, he's enough to dishearten some coders.. 

Words from the guy who is only happy, if a thread is fully derailed.

 

 

[ivop]

1 hour ago, carlsson said:

Aha! So people are modding their computers with custom hardware, and expect that software will support it or at least not utilize features that the custom hardware doesn't support.

Yes, more or less. If it's not needed for speed, I agree, don't use undefined opcodes for no reason. But if it helps to get that extra vertex on screen, do it! It's the exact same hardware that was sold by the millions.

 

Quote

For that matter, doesn't the 65C02 implement new instructions that the old 6502 doesn't do? It means that software that embraces the custom hardware on the other hand wouldn't work with stock machines, which more commonly would be a problem IMHO.

AFAIK there's no, or not much, specific 65C02 code was written for the Atari. There is for the 65C816 though, but nobody expects that to run on stock hardware

Edited March 25, 2021 by ivop

[carlsson]

Undocumented opcodes may be used for space savings too, e.g. if you're making a game within 1024 bytes those may come handy. So both speed and code size.

[ivop]

16 minutes ago, carlsson said:

Undocumented opcodes may be used for space savings too, e.g. if you're making a game within 1024 bytes those may come handy. So both speed and code size.

Definitely. I used LAX #$00 in both of my C64 sizecoding experiments. The only LAX imm that is stable

Edited March 25, 2021 by ivop

[+gnusto]

48 minutes ago, Faicuai said:

Greatest 25 PCs of all time

 

Really surprised this list doesn't include the Tandy 1000, but does have several laptops that are fairly similar to each other. The 1000 was often looked down on by the business types who were buying IBM, but it was a DOS machine that came with the same processor/memory configurations as many of those IBM (XTs), but had much better graphics and sound, built in. For several years it was the best DOS gaming machine you could use, and had serviceable built-in productivity apps for home use, all while costing pretty much the same as an IBM.

 

[Goochman]

52 minutes ago, Faicuai said:

Greatest 25 PCs of all time:

 

https://www.pcworld.com/article/126692/greatest_pcs_of_all_time.html

 

By:

• Innovation: Did the PC do anything that was genuinely new? Did it incorporate the latest technology?
• Impact: Was it widely imitated? Did it become part of the cultural zeitgeist?
• Industrial design: Was it a looker? Did it have clever features that made using it a pleasure?
• Intangibles: Was there anything else about it that set it apart from the same ol' same ol'?

 

Enjoy!

 

Atari should be 2nd behind the Apple I/II

[Faicuai]

6 minutes ago, gnusto said:

Really surprised this list doesn't include the Tandy 1000,

LESSON:

 

When you put yourself next to the IBM/PC, you are not going to win office crowds with flashy graphics and features having a stronger recall in kiddie-gaming domains.

 

You see, by the time the IBM/PC brought discipline, reliability and service to the office environment (which ended up being the REAL propeller of the PC computer architecture, up to these days, because those were the customers that would PAY RECURRENTLY), graphical computing was not yet seen as productive means / tools, but rather gaming. That would begin to change (main-stream) with the Apple McIntosh, and later with its Windows/PC nemesis.

 

I do agree with your assessment, overall, of the Tandy 1000. And so it was the Victor 9000, which should be on top of the Tandy itself. I would add to the list the BBC-Micro, not for its industrial design, or its mass-consumer appeal, but more because of how well it brings together the 6502 architecture, as well as its expansion capabilities (limited with respect the Apple II, but still very well present, compared to other similar machines of the time, which had essentially NONE).

 

Notice the C64 did not even make it there... 8-)

 

[drpeter]

32 minutes ago, Faicuai said:

LESSON:

 

When you put yourself next to the IBM/PC, you are not going to win office crowds with flashy graphics and features having a stronger recall in kiddie-gaming domains.

 

You see, by the time the IBM/PC brought discipline, reliability and service to the office environment (which ended up being the REAL propeller of the PC computer architecture, up to these days, because those were the customers that would PAY RECURRENTLY), graphical computing was not yet seen as productive means / tools, but rather gaming. That would begin to change (main-stream) with the Apple McIntosh, and later with its Windows/PC nemesis.

 

I do agree with your assessment, overall, of the Tandy 1000. And so it was the Victor 9000, which should be on top of the Tandy itself. I would add to the list the BBC-Micro, not for its industrial design, or its mass-consumer appeal, but more because of how well it brings together the 6502 architecture, as well as its expansion capabilities (limited with respect the Apple II, but still very well present, compared to other similar machines of the time, which had essentially NONE).

 

Notice the C64 did not even make it there... 8-)

 

 

Or the Psion 5, probably the last general purpose computer ever to be built with the entire bundled software stack written from scratch... Psion 5 - The Last Computer.  Faintly ridiculous it's overlooked in favour of a number of less distinguished handhelds/notebooks.

[Tuxon86]

13 minutes ago, Faicuai said:

LESSON:

 

When you put yourself next to the IBM/PC, you are not going to win office crowds with flashy graphics and features having a stronger recall in kiddie-gaming domains.

 

You see, by the time the IBM/PC brought discipline, reliability and service to the office environment (which ended up being the REAL propeller of the PC computer architecture, up to these days, because those were the customers that would PAY RECURRENTLY), graphical computing was not yet seen as productive means / tools, but rather gaming. That would begin to change (main-stream) with the Apple McIntosh, and later with its Windows/PC nemesis.

 

I do agree with your assessment, overall, of the Tandy 1000. And so it was the Victor 9000, which should be on top of the Tandy itself. I would add to the list the BBC-Micro, not for its industrial design, or its mass-consumer appeal, but more because of how well it brings together the 6502 architecture, as well as its expansion capabilities (limited with respect the Apple II, but still very well present, compared to other similar machines of the time, which had essentially NONE).

 

Notice the C64 did not even make it there... 8-)

 

And without RS, Apple, Atari and Commodore to demonstrate to IBM that there was a new emerging market in the personnal computers, IBM would have skipped it. The 5150 was a rush job, brillant but reactionary. They goofed in their approach by going with off the shelf and open architecture that was easilly copied. We benefited by it since it gave rise to the cheap clone market, but they tried to go back on that when  they introduced the PS/2 line and MCA.

 

None of the four had the finances or brand recognition to compete with IBM either in the business world. 

 

 

 

 

[+gnusto]

1 hour ago, Faicuai said:

When you put yourself next to the IBM/PC, you are not going to win office crowds with flashy graphics and features having a stronger recall in kiddie-gaming domains.

 

Oh believe me, I was there for all of it and witnessed it, so I totally understand, at the time, how that happened.

 

But this article is written with a full retrospective, and all the learning that came since. I think in that light it's correct to identify the PC/XT for all the reasons you make; IBM brought the micro mainstream, capturing the high dollar customers who would keep coming back, pulling the PC as an appliance out of the hobby realm and into the day to day (whether that was a good thing or not is questionable, but IBM's role in doing so pretty much isn't).

 

Looking back now though, I would say the Tandy 1000 belongs in there not only as an interesting technology play, but also for recognizing that home users were still a real thing, including dare we say it, gaming. Other industry titans like IBM and Apple left huge dollars on the table and ensured the long term existence of console gaming by ignoring the entertainment audience (Steve Jobs was famously against being serious on gaming because he was sure it didn't move units; it took the iPhone to convince him otherwise, and then only grudgingly). So what I like about the Tandy, even though I didn't own one at the time (I had an Amiga, obviously!) was it recognized the home user as a direct customer and went to considerable effort to improve for them, at almost no cost to the business use potential. The suits at IBM could never do that, it was "below them".

 

Anyway, there's always one left out of a list like this somewhere. I actually enjoyed the article and read through the entire thing, they did put a lot of thought into the choices they made.

[Mazzspeed]

5 hours ago, zzip said:

I've been looking for C64 homebrew, and I found some, but nothing like the quantity I see for Atari.   Obvously AtariAge seems to be a hub for these developments, is there a similar hub in the C64 world?

There's tonnes of homebrew software and new games being released. As stated, CSDB is a great site for this.

 

Using the 1541 UII+ with Ethernet and UltiTerm 2.4 you can connect to Retrocampus and download software from CSDB to your C64 directly. You don't need to mod your C64 internally to have a great machine, all you need is a 1541 UII+, that gives you: Dual cycle accurate 1541's with the capability to load alternate roms via .bin files as well as capability for the addition of virtual RAMBoards (up to 40k) - Storage is provided via USB and you can plug a USB HDD or SSD into the UII+ as well as optical drives, DMA loading as well as a software IEC hard drive and print to PDF functionality (I believe Commodore actually had print to file before Atari), JiffyDOS 6.01 as well as the ability to load alternate kernel's via .bin files, an REU expansion up to 16MB to Commodore specifications, FTP support meaning you can transfer files directly to your C64 from your PC, Telnet support meaning you can control your C64 remotely, a paged RAM expansion to Georam specifications, TAP (tape) file loading, CRT (cartridge loading), the provision of a number of freezer/fastloader carts including the Final Cartridge III (all supplied with the device), dual SID support as well as Ultimate Audio with the provision to playback up to 8 channel MOD files 'perfectly' and Ethernet support with the networking stack implemented in hardware - You can run an alternate kernel (JiffyDOS), two cycle accurate 1541's with alternate roms (Say: S-JiffyDOS as it's the fastest), a freezer cart and a game based on a CRT file (so emulating a cartridge, right down to the fact that it's still there if you press reset) and a 16MB REU expansion as well as a 16MB ram drive all at once. Furthermore, drive speeds are comparable to the A8 running SIDE 3, Ultimate 1MB and SDX, as I've highlighted in other posts. What A8 users are getting with FujiNet, Commodore users had for years prior. Development on the Commodore platform has reached a plateau as all of this has been available for Commodore users since ~2008, there's little else that really needs to be achieved (although that FPGA VIC-II sounds tasty). However, software and new games are constantly being released, I'd have to say the C64 is probably still the most popular 8bit software and game development platform. It's the one retro machine that you can still build from brand new parts today (excluding the keyboard at this point in time) - I have a clear 64C case made using factory tooling in 2018 that one day I plan on fitting a brand new Ultimate 64 board to for 48Mhz 6510 goodness as well a Dolphin DOS and virtual parallel interface's from the virtual 1541's...

 

Yes, there are a number of demo's available under the Commodore platform, as the VIC-II as well as the SID have capabilities that allow for some very impressive demonstrations of what can be achieved when pushing the C64 chipset beyond anything it was even intended to do. Furthermore, this is a good thing as a number of demoscene discoveries actually filter down to games (VSP, sprite multiplexing techniques, SID advancements using little CPU overhead). Disappointingly, some of the comments I read here just highlight that certain A8 users are fairly clueless, what they believe they know about the C64 vs reality are two polar extremes.

 

The only reason for the success of PC clones was the fact they were essentially based on an open platform, there wasn't a lot about them at the time that was in any way innovative - But they were stupidly expensive and it took until the mid 90s until they even began to see widespread adoption by home users.

 

 

Edited March 25, 2021 by Mazzspeed

[nadir]

4 hours ago, gnusto said:

 

Really surprised this list doesn't include the Tandy 1000, but does have several laptops that are fairly similar to each other. The 1000 was often looked down on by the business types who were buying IBM, but it was a DOS machine that came with the same processor/memory configurations as many of those IBM (XTs), but had much better graphics and sound, built in. For several years it was the best DOS gaming machine you could use, and had serviceable built-in productivity apps for home use, all while costing pretty much the same as an IBM.

 

I had an XT clone and a 1000 for awhile.  Honestly, the performance of the 1000 was dismal in comparison at the same clock speed.  The XT clone had Plantronics video, so was theoretically capable of the same graphics modes, but besides Lotus 1-2-3's graphing module I never found anything that supported it.  So, it was CGA until upgraded later on (ultimately ended up with 1 MB VGA card!).  The 1000's stock audio was obviously way better, but I dropped an Adlib in the XT and then it was better.

 

So, yeah, the 1000 was technically better graphics/audiowise out of the box.  And it could have been upgraded to match the upgraded XT audio/video, but the performance was just sluggish it was hard to use.  It always made me think of the PCjr.  Lots of promise with better video and audio than the standard PC or XT, but hobbled by poor performance out of the system.

[Faicuai]

1 hour ago, Mazzspeed said:

and a 16MB REU expansion as well as a 16MB ram drive all at once. Furthermore, drive speeds are comparable to the A8 running SIDE 3, Ultimate 1MB and SDX, as I've highlighted in other posts

 

Keep in mind that some users in this community, especially those that have more time vested on the current "lineage" of upgrades closely designed around the Atari architecture, may not necessarily find the upgrades and functionality you describe as astonishingly novel (putting aside any solderless or tool-less approach, which is always desirable from a practical point of view).

 

By closely reading the above description, it sounds more like praising the integration of alien hardware in charge of an array of functionality that cannot be really carried out with the host's CPU or even its own logical or physical architecture. Sure, we can say that the Ultimate-II is hung from the system-bus, but... that's about it. It sounds as if the architectural formula for the C64  nowadays is to keep its own resources out of the equation as long as you can, while increasing reliance on external ones (pretty much alien) as much as possible. The Ultimate-II functional scope and cohesion bodes really well for its developer / designer, but not so well for the C64's inherent capabilities (and much less its "OS"), if we are going to pay tribute to the truth. 

 

And, as a practical example of the impact that architectural differences may have, we can re-visit I/O speeds in the context of procedural-control (6502+ storage controller) vs. bus-control (DMA + supporting controller). On the Atari, you can achieve raw-speeds of  up to 105 KB/sec, sustained, fetched from HD media, under the 6502 and OS / PBI command (and the storage controller, of course, which is essentially a traditional formula seen in the past). With DMA (as seen on SIDE3 own version), we have clear indication that raw speeds would jump up to 1.5MB/sec., approx.

 

It may sound counter-intuitive, but in reality I am not that interested on DMA speed, because, when I work on SDX (or even DOSXL, Atari DOS, etc), a LOT of the I/O is heavily transactional, in nature. That is, tons of small and large files, tracking sector reads, writes, file tables, attributes, sub-directories, etc. This is a REAL operating environment (instead of loading a kiddie PANG! +245 KBytes file, in 5.5 secs., from SIDE loader) . To this extent, just a simple "COPY /brsv {Dn:folder} {Dn:folder}" operation issued at SDX prompt (which I perform frequently, with hundreds of files involved), instantly becomes a total tour-de-force on the Atari, which I can't simply imagine how it would be conducted by going through DMA's negotiation overhead, and much less a slower CPU, for instance. I already know it is likely to be painful.

 

The above suggests, in turn, that there will be significant differences in effective throughput and I/O speeds under heavy, transactional loads, though. And if relief is sought by off-loading such transactional framework onto another alien piece of HW that will work in tandem, it would be much harder to praise the host HW itself for not doing any of the heavy-lifting, at all (!) This would also apply on the A8 domain, as well. No one questions, however, how much more "usable" the systems may be. 

 

There is a lot more than the eyes can meet, though. It all seems to boil down to the A8's system architecture. Extensible, driver-based OS, 120 Kbps SIO universal bus, on-CPU-bus banked ram (beginning with AXLON on $4000-$7FFF window), dual cart-ports (supporting read and write traffic), and system expansion slots are concepts already present and supported on my 1980 CTIA 800 (!) It goes from 48KB to 544Kb in under 1 min. No need for internal degutting or external alien hardware, except maybe a DOS+CF-based HD plugged on the cart. port?

 

For some of us, that's more like it. No wonder why the 800 made it to that 25-Best-Ever PCs list, though.

 

 

[Faicuai]

4 hours ago, Tuxon86 said:

And without RS, Apple, Atari and Commodore to demonstrate to IBM that there was a new emerging market in the personnal computers, IBM would have skipped it.

It was Apple who showed it, in reality. IBM paid special attention to it. It later considered the 800 as a possible OEM-like personal computer product, with its own design sketch.

 

And, if anyone here ever wondered how far the 1981 5150 could go with a meager CGA adapter and its color monitor (and essentially NOTHING else), you may be in for quite a wake up call:

 

 

[Mazzspeed]

33 minutes ago, Faicuai said:

 

Keep in mind that some users in this community, especially those that have more time vested on the current "lineage" of upgrades closely designed around the Atari architecture, may not necessarily find the upgrades and functionality you describe as astonishingly novel (putting aside any solderless or tool-less approach, which is always desirable from a practical point of view).

 

By closely reading the above description, it sounds more like praising the integration of alien hardware in charge of an array of functionality that cannot be really carried out with the host's CPU or even its own logical or physical architecture. Sure, we can say that the Ultimate-II is hung from the system-bus, but... that's about it. It sounds as if the architectural formula for the C64  nowadays is to keep its own resources out of the equation as long as you can, while increasing reliance on external ones (pretty much alien) as much as possible. The Ultimate-II functional scope and cohesion bodes really well for its developer / designer, but not so well for the C64's inherent capabilities (and much less its "OS"), if we are going to pay tribute to the truth. 

 

And, as a practical example of the impact that architectural differences may have, we can re-visit I/O speeds in the context of procedural-control (6502+ storage controller) vs. bus-control (DMA + supporting controller). On the Atari, you can achieve raw-speeds of  up to 105 KB/sec, sustained, fetched from HD media, under the 6502 and OS / PBI command (and the storage controller, of course, which is essentially a traditional formula seen in the past). With DMA (as seen on SIDE3 own version), we have clear indication that raw speeds would jump up to 1.5MB/sec., approx.

 

It may sound counter-intuitive, but in reality I am not that interested on DMA speed, because, when I work on SDX (or even DOSXL, Atari DOS, etc), a LOT of the I/O is heavily transactional, in nature. That is, tons of small and large files, tracking sector reads, writes, file tables, attributes, sub-directories, etc. This is a REAL operating environment (instead of loading a kiddie PANG! +245 KBytes file, in 5.5 secs., from SIDE loader) . To this extent, just a simple "COPY /brsv {Dn:folder} {Dn:folder}" operation issued at SDX prompt (which I perform frequently, with hundreds of files involved), instantly becomes a total tour-de-force on the Atari, which I can't simply imagine how it would be conducted by going through DMA's negotiation overhead, and much less a slower CPU, for instance. I already know it is likely to be painful.

 

The above suggests, in turn, that there will be significant differences in effective throughput and I/O speeds under heavy, transactional loads, though. And if relief is sought by off-loading such transactional framework onto another alien piece of HW that will work in tandem, it would be much harder to praise the host HW itself for not doing any of the heavy-lifting, at all (!) This would also apply on the A8 domain, as well. No one questions, however, how much more "usable" the systems may be. 

 

There is a lot more than the eyes can meet, though. It all seems to boil down to the A8's system architecture. Extensible, driver-based OS, 120 Kbps SIO universal bus, on-CPU-bus banked ram (beginning with AXLON on $4000-$7FFF window), dual cart-ports (supporting read and write traffic), and system expansion slots are concepts already present and supported on my 1980 CTIA 800 (!) It goes from 48KB to 544Kb in under 1 min. No need for internal degutting or external alien hardware, except maybe a DOS+CF-based HD plugged on the cart. port?

 

For some of us, that's more like it. No wonder why the 800 made it to that 25-Best-Ever PCs list, though.

 

 

Almost any time we add to the architecture of any machine, we do so via the system bus. Unlike the A8 which has limited resources available to the cartridge port with the whole system bus available to the PBI port, the C64 has both ports accessible via the cartridge port and more (one very important edge connector on the cartridge port is the DMA pin). Furthermore, the Commodore 64's cartridge port allows for Ultimax mode, which is where much of the 1541's extensive functionality comes from - As I stated, essentially, dual cartridge support isn't limited to the A8, as using Ultimax mode the C64 can do the same (hence the reason we can have a kernel loaded, a cartridge loaded and a memory expansion loaded all at the same time). Furthermore, the I/O speeds you mention are achievable by the C64 without screen blanking (bearing in mind you always quote maximum theoretical speeds, forgetting overheads), you know I have proven this.

 

Versatile DMA is a game changer in the 8bit world, the CPU is a limitation and best avoided. I'll be grabbing a SIDE 3 for this feature as well as the cartridge support I'm used to, as the support of these features is game changing.

 

The 1541 Ultimate II+ is not simply using the C64 as a host like the Turbo Chameleon. The 1541 UII+ is leveraging unique and interesting systems put in place by Commodore engineers when the system was designed - Your belief that the C64's resources are kept out of the equation couldn't be more incorrect. All cycle accurate 1541 emulation is done via the regular Commodore IEC port. The C64 can also bank switch memory via the cartridge port, really no different to the 800XL, users choose DMA mode as it's faster and more versatile.

 

I understand the A8's architecture, I went out of my way to understand and appreciate certain aspects of it. It appears you know little about the architecture of the C64, which appears to be a common theme here. I also know all about SIDE 3 and DMA mode, it appears you know little about the 1541 UII+ (which has had DMA storage transfer for about five years or more now).

 

 

Edited March 26, 2021 by Mazzspeed

[Faicuai]

23 minutes ago, Mazzspeed said:

It appears you know little about the architecture of the C64,

Let's put it to the test:

 

RUN the command "COPY /brsv {Dn:folder} {Dn:folder}" (or equivalent) with a folder containing 100 or more files and subfolders (of diverse size) and let us know the timing of such operation. This will uncover pretty much anything (I promise). 

 

Now, if you feel braver, fire up TWO hard-drives (not partitions) on the C64 and run a sector-level copy from one partition in one drive to another partition, in another drive (I do this to maintain only an active portion of Incognito in sync with my SIDE2 partitions, for use on my 800XLs). Consider approx. 20MB worth of data. It takes about 8mins on the Atari. This will also be a (real) eye opener.

 

Take your time.

Edited March 26, 2021 by Faicuai

[Mazzspeed]

29 minutes ago, Faicuai said:

Let's put it to the test:

 

RUN the command "COPY /brsv {Dn:folder} {Dn:folder}" (or equivalent) with a folder containing 100 or more files and subfolders (of diverse size) and let us know the timing of such operation. This will uncover pretty much anything (I promise). 

 

Now, if you feel braver, fire up TWO hard-drives (not partitions) on the C64 and run a sector-level copy from one partition in one drive to another partition, in another drive (I do this to maintain only an active portion of Incognito in sync with my SIDE2 partitions, for use on my 800XLs). Consider approx. 20MB worth of data. It takes about 8mins on the Atari. This will also be a (real) eye opener.

 

Take your time.

Faicuai,

 

Due to the fact that the 1541 Ultimate II+ emulates multiple 1541's and all 'disks' as well as drives are mounted within the 1541 UII+ and accessible via the C64, I can [Ctrl] & C to copy files and [Crtl] & P to paste files within the 1541 UII+ itself, the process takes seconds from the keyboard of my C64.

 

I'm sure you consider that cheating, even though it achieves the exact same end result. IMO, such an approach is far more elegant than relying on a limiting 8bit CPU running at under 2Mhz. As stated, the CPU is a limitation best avoided.

 

However. The 1541 UII+ and it's internal storage system aside, all cycle accurate emulated 1541 data transfer is performed via the Commodore IEC bus with loading times comparable to the A8 using cycle accurate emulation and high speed SIO (basically, the argument the 1541 is slow has been dead for around 20 years now). Naturally, DMA storage transfers are done via the cartridge port using the exact system intended by Commodore engineers, technically I see little reason why the A8 and C64 wouldn't achieve identical speeds when using DMA transfer and I have shown video's in the past demonstrating the streaming very high quality video in 16 colours - If that's not sustained transfer I don't know what is! Using DMA off mass storage, speeds of 90,000 bytes per second are achievable considering all overheads and standard FAT partitioning. Using the ram disk via DMA, speeds of 113,000 bytes per second are achievable considering overheads with absolutely no screen blanking whatsoever in both cases.

 

Streaming video on the A8 involves dumping what is essentially raw data onto the CF card, which makes the CF card essentially no different to a raw ram disk. You can hardly sustain those speeds once correct formatting and partitioning is applied considering overheads.

 

Anyway, I'm done. I've got a monitor cable to make and monitor port mods to perform on my 600XL.

 

TLDR; I just copy/paste via the keyboard of my C64 within the 1541 UII+ itself and the process takes seconds. There's no need to involve an ancient CPU. I transfer data from my PC to the 1541 UII+ using FTP from my PC, takes seconds.

Edited March 26, 2021 by Mazzspeed

[Faicuai]

31 minutes ago, Mazzspeed said:

I'm sure you consider that cheating,

Absolutely.

 

But whether cheating or not, it seems to invalidate your own opinion about the functional role of the 1541 Ultimate-II.

 

There were only two (2) possible results out of this test: slow-to-a-crawl or much faster. If the latter were the outcome, that would clearly suggest the heavy-lifting (file search, sector-location & transfer, folders-creation, attributes updating, etc.) is being done by something else other than the C64 own resources. Again, no one here is questions the resulting "usability" of the systems, as I stated before.

 

And, to many us, that is where the actual beauty lies in all of this: the core system executing the show. In fact, what you otherwise call not elegant or archaic, is in fact SDX's core design philosophy, as well as Incognito, SIDE and even the low-level BIOS menu, which are all built under this concept... The Atari host machine runs the show and calls the shots, single handedly (essentially). Why? Because IT CAN. Period. If not, I simply pull out my CF from my host machines, go to my PC of choice, and access the content there anytime, or even run Altirra at Full speed with a sector image of the cards !!! 8-))

 

The video streaming I have seen on the C64 is NOT real, 60fps video-playback. It can't address the required linear physical storage (already dead from the get-go), and there is no source-content audio track (in the samples I have seen so far). While resolved spatially higher (not chroma-wise), it gives the impression as if the C64  lacks the necessary muscle or raw power needed to timely route and render 60 fps of video transport stream, as well as sequencing and routing the audio, simultaneously. But I would not mind at all being proven wrong, I assure you.

 

The real bottom-line of all this is simple: being realistic and considering apples-to-apples is probably the healthiest approach.

 

 

Edited March 26, 2021 by Faicuai

[Mazzspeed]

40 minutes ago, Faicuai said:

But whether cheating or not, it seems to invalidate your own opinion about the functional role of the 1541 Ultimate-II.

No it doesn't. It's no different a situation to connecting your A8 to a PC via SIO2USB and transferring files from one folder to another, or from one virtual drive to another using the A8 (or even on the PC itself) - Assuming that's possible (as yet I haven't used the system, hence the reason I'm actually getting an A8 and exploring everything it has to offer as opposed to making pointless claims out of what appears to be some form of bitter resentment).

 

Why involve an ancient CPU when you don't have to? Most A8 SIO storage devices have their own display allowing for a number of features, perhaps they also allow for file copy between virtual drives on the device itself? The difference is that on the C64 you can do it from the C64 itself seamlessly. I can jump in and out of the Ultimate II+ menu without interrupting my workflow at all, my system state is simply paused - And all because of Ultimax mode, something built into the C64 by Commodore engineers. I can jump into the Ultimate II+ menu in the middle of a game and the game isn't affected at all, because of Ultimax mode.

 

The C64 is not simply IO and display for the UII+ as you're implying, the reality is far more complex and useful.

 

40 minutes ago, Faicuai said:

And, to many us, that is where the actual beauty lies in all of this: the core system executing the show. In fact, what you otherwise call not elegant or archaic, is in fact SDX's core design philosophy, as well as Incognito, SIDE and even the low-level BIOS menu, which are all built under this concept... The Atari host machine runs the show and calls the shots, single handedly (essentially). Why? Because IT CAN. Period. If not, I simply pull out my CF from my host machines, go to my PC of choice, and access the content there anytime, or even run Altirra at Full speed with a sector image of the cards !!! 8-))

Once again, this idea you have that the C64 is no more than IO and display for the UII+ is incorrect beyond the manipulation of files on the storage sub system itself (as outlined above, I believe the A8 has similar devices, some not as well integrated with the host machine as others). Mass storage on the C64 runs either FAT16, FAT32 or ExFat, so if you want to remove the SD card, USB stick or in my case 500GB HDD and plug it into a PC to manipulate files you can - However I don't, as I don't need to due to the ability to simply FTP files across to storage. About the only time I actually remove my SD card and plug it into my PC is to transfer files onto the card I use in my SD2IEC HDD.

 

Most of the time I download files from CSDB directly from my C64 via Ethernet straight onto storage.

 

Both machines are capable of doing what you describe, I'm sure the A8 holds a slight advantage due to outright clock speed unless you consider the Ultimate 64 which allows for 6510 speeds from 2Mhz right through to 48Mhz and virtual cycle accurate parallel 1541 connection using Dolphin DOS where the 64 will definitely have an advantage. Furthermore JiffyDOS with CMD-HD commands is very capable, and from my perspective 100% compatible with all games and existing software using 1541 cycle accurate emulation. I haven't come across a single game or item of software that won't run under JiffyDOS (in my case I run S-JiffyDOS roms on the emulated 1541's with the JiffyDOS kernel on the C64 for the highest speeds possible).

 

40 minutes ago, Faicuai said:

The video streaming I have seen on the C64 is NOT real, 60fps video-playback. It can't address the required linear physical storage (already dead from the get-go), and there is no source-content audio track (in the samples I have seen so far). While resolved spatially higher (not chroma-wise), it gives the impression as if the C64  lacks the necessary muscle or raw power needed to timely route and render 60 fps of video transport stream, as well as sequencing and routing the audio, simultaneously. But I would not mind at all being proven wrong, I assure you.

The video streaming I've seen on the A8 is of poorer quality with a very interlaced look with artifacting as the storage swaps banks. As stated, when it comes to outright storage transfer speeds, you're splitting hairs between the C64 and the A8 - Furthermore, that CF card on the A8 is not formatted, it's streaming raw data to provide the needed transfer speed - Therefore as far as hard drives are concerned, technically in such a scenario the CF card is in no way acting as a hard drive. Basically, it's similar in concept to a ram disk.

 

Look. This is all bullshit. They're both great machines, they both trade blows and a lot of what the A8 is getting only now the C64 has had for years. I'm not interested in an argument, I'm not getting drawn into an argument, I'm simply correcting the misconceptions I see regarding the C64. I own both machines, one fully expanded, the other soon to be fully expanded and I love them both equally. Go yell at the wall. Better yet, buy a C64 and get a 1541 Ultimate II+.

 

 

Edited March 26, 2021 by Mazzspeed

[_The Doctor__]

so like a black box on an Atari you can jump in and out, or a MIO for than matter, same thing, BB has high speed built in for SIO drives that support it, but that's an 1980's device use case, and of course you could use it's floppy board for parallel drive speed and use. Though most folks just used them for ram disks and hard drive... all of this was super fast stuff bitd

 

Edited March 26, 2021 by _The Doctor__

[Mazzspeed]

1 minute ago, _The Doctor__ said:

so like a black box on an Atari you can jump in and out, or a MIO for than matter, same thing, BB has high speed built in for drives that support it.

I assume so?

 

The reason I got my 600XL is because I'm so keen to try all of these system's out on real metal. What is this Doc? Can you still buy it?

[_The Doctor__]

oh I don't think any of this is in production anymore...

replicas of the MIO were made...

but the BB hasn't been... closest thing to these today are U1M plus SIDE2/B, plus the freezer... etc.