Hello all,
I creates here a separate thread regarding Smartport behaviour.
Smartport is not exaclty working the same between the IIGs, IIc & IIe
On the IIc & IIe via Smartport SP card, issuing PR #5 starts the smartport with the command 0x85 to init devices and start reading first boot device (id the lowest)
On the IIGS it is a bit different, at boot time it, it start checking the SELECT pin LOW and init device.
If he find not drive or no boot drive, then it start looking at C600 for boot disk
It means that if the IIGS start with an emulator on DISKII emulation, after switching to SMARTPORT emulation there is no way for PR #5 to boot on smartport (like for IIe & IIc)
It will put on the screen no connected devices...
Is there a smartway to restart the device detection on IIGS ?
My Goal is Start a program on DiskII emulation and switch to smartport emulation on IIGS
Thanks
Vincent
In wish I could help you as I'm very interested in your Floppy Emu project, but, alas, I have no Apple IIgs.
But I do have a hunch that the behaviour of the IIgs (as far as SmartPort is concerned) may be different because the IIgs might probe for smart and non-smart floppy drives (or HDDs which always would be smart) and if your floppy emu chokes on these probing procedures, the best way to avoid the problem is to make your side of the protocol (in the Floppy Emu) ignore everything which is not Apple II+ (Liron card) or Apple IIc - ish in terms of protocol and then just wait for the next attempt from the Apple II+ / IIc / IIgs. This is based on the conjecture that at some point even the IIgs should use a request for a classic SmartPort device which is similar to what the II+ / Liron or IIc does.
It is very unfortunate that SmartPort is such a mess. I wrote about this elsewhere here on Applefritter. I think the problem is rooted in Apple's haphazard handling of 'SmartPort' bug fixes and extensions and adding new devices (i.e. Floppy disk drives) without properly documenting what they did and how it works and telling the users / third part developers outside Apple by issuing "Technical Notes". So all what is left for us is a riddle which can only be solved by reverse engineering, hacking and being patient.
Keep up the good work, but my advice is: don't waste too much time on the IIgs issues. If you get stuck there, put it on the backburner and spend your time on further development of non-IIgs features. This advice is based on my own prospective venturing into the IIgs realm, and the result of this prospector's work is that the IIgs has much uncharted terrain (aka lack of documentation available on the internet) that I don't want to go there. It just would bog down my work on the Replica 2e which still lacks some IIc capabilities. It will never support a IIgs mode. Here is another way to probe this topic:
ask "Google" the question "Was there an Apple IIgs clone" and their "AI" spits out the following "AI overview":
"Yes, there were Apple IIgs clones, though they are less common than other Apple II clones like the Apple IIe or II+. While many companies produced clones of earlier Apple II models, the IIgs proved more difficult to clone due to its advanced features. Some companies focused on creating "cloning" boards that could be installed into an existing Apple IIe case to upgrade it to a IIgs, rather than producing a complete clone system.
Apple IIgs Clones: Some companies, particularly those in Asia, attempted to create their own versions of the Apple IIgs, but the complex nature of the machine made it challenging. These clones were often less compatible with software and hardware than their genuine counterparts.
Upgrades and Mods: A more common approach was to create upgrade boards that could be installed in an existing Apple IIe case, bringing the functionality of the IIgs to the older hardware. This was a popular method because it allowed users to get the power of the IIgs without having to buy a new machine.
Limited Availability: Due to the technical challenges of cloning the IIgs, fewer such clones were produced compared to other Apple II models. Most were created for the Asian market and were not widely available in the United States."
This tells us that trying to clone an Apple IIgs always failed to produce a perfect clone with 100% compatibility. Which for me means I will no try it for my Replica IIe project. But the AI overview also implies that the Apple IIgs documentation available in the user space was inadequate for cloning attempts (back in the day, mind you) and unless there were substantial "leaks" until today, nobody is any the wiser. So what are your chances to solve the IIgs riddle in a useful amount of your time invested ?
- Uncle Bernie
That's the problem with so-called "AI": it spits out a ream of authoritative-sounding verbiage, without any details that would make it straightforward to verify. You should regard this output as 100% lies until proven otherwise, for each specific instance.
Harry Frankfurt described bullshit as something which a speaker:
[On Bullshit, 2005. Princeton University Press.]
It's not accidental that this description is also an accurate summary of the LLM technique.
As for clones of the IIgs, I very much doubt it. "Asian" cloners, like VTech, reached the apex of their reverse-engineering capabilities with boards like the Laser 128EX and the Universal Disk Controller. A IIgs clone would require re-engineering multiple custom ASICs, something they had never done before. And there was never going to be much money in IIgs clones, since by the 1990s the whole Apple II architecture was being regarded as passé, even in Asia.
In post #3, "robespierre" wrote:
" That's the problem with so-called "AI": it spits out a ream of authoritative-sounding verbiage, without any details that would make it straightforward to verify. You should regard this output as 100% lies until proven otherwise, for each specific instance."
Uncle Bernie comments:
True that, "AI" as we have it today is very problematic because it "hallucinates" and makes up "facts" which are simply not true, but I would not call it "lies" as a machine is unable to lie (so far). A "lie" always carries evil intent, this is why scripture calls the Devil the "lord of lies" or similar. So to be more precise, and back to the scientific foundation, it's "hallucinations" because within the the LLM, an input triggers certain statistics based pathways which are similar, but not the same, as the real facts from which the LLM was trained. And then it makes up nonsense. The trouble is, this cannot be fixed on the LLM level and IMHO has been the bane of "Neurocomputing" since the first baby steps (the "Perceptron"). I've wasted years of my life trying to design and develop a true sentient AI and I even studied Neurophysiology to get insight how the human brain works ... which was a disappointment as I found out that even the professors who held the lectures had no real understanding how the brain works. All they knew was which parts do what. But they were unable to explain the "how". This was in the 1980s, so some progress may have been made in the 40 years which have passed but IMHO "AI" is a fraud and will go bust, and investors will lose their money. But this is just my personal opinion based on my few interactions with AI chatbots. I recently tested an "AI" card for the Apple-1 which gives the user the (false but stunning) impression that 'ChatGPT' sits inside this 50 year old computer. I took screenshots of the whole chat with all the "lies" aka "hallucinations" but did not have the time yet to transcribe it. The results were sobering. But it's not the card as such, it works fine. (If anyone reading this is interested, comment and I'll put something about this card into the Apple-1 forum).
But where the Google "AI" summary is great is that it quickly scans Google's whole database on topics and turns it into a summary.
Which however requires human review and scrutiny to sort out the "hallucinations". These mostly would be "false positives", that means, the "AI" claims that something exists which doesn't. For instance, when I asked it about my YAAK keyboard it went like this:
Q: Who designed the YAAK keyboard ?
ChatGPT: The YAAK keyboard was designed by a team of engineers and designers at YAAK TECHNOLOGIES.
Sure ! I had a good laugh ! The company exists (I did not now that) but it does not make keyboards. This is a typical example for "hallucinations" of so-called "AI".
But so far I did not get any "false negatives", that it claims that something does not exist but it does (can be found on the internet). With the exception that the "AI" once claimed that the CV-19 "vaccines" killed only about 55 people worldwide ... this makes me wonder which fraudulent data sets they use to train their LLMs on such "hot potato" topics.
Now, to the specific "AI" summary I posted in post #2, I of course checked it for being truthful enough for the purpose and found no fault. The Apple IIgs indeed appears to be extremely difficult to clone perfectly, and if a perfect clone could be found anywhere on the internet, the "AI" would have pointed to it by a link to that website.
Hope this sums it up how to deal with "AI". It requires some disbelief and scrutiny by the human user.
- Uncle Bernie
P.S.: as some more off-topic remarks, despite I think that "AI" as we have it today is a largely a fraud and will not deliver what the perpetrators of the fraud promise, IMHO it will have a place for such internet searches, and it might also find applications in "lawfare" where lies, scams, fraud and made up nonsense reigns. So it might replace some human lawyers who earn their living by such vile and disgusting conduct. The idea here is to use "AI" to produce some 100000 pages of legalese motions and discovery etc. to overwhelm the counterparty with this worthless stack of paper that was cheap to write (by "AI"). Now the lawyers of the counterparty must spend gazillions of US$ to read and analyze that morass of "facts" by human lawyers or, use their own "AI" to disprove it and write rebuttals (again, 100000's of pages). Take the "AI" aspect out and you have the way how large corporations conduct lawfare (and this is why some lawsuits, like who invented the microprocessor, went on for decades). With the use of "AI" this method of producing millions of pages of "lawfare ammunition" will go into overdrive. And if you don't believe me, watch the Y2019 movie "Dark Waters" and pay attention to how the evil corporation swamps the lawyer (Robert Billot) with truckloads of "discovery" materials. But he took the time to go through it with a fine comb (others would have given up). The movie is based on true events and hence, worth watching. You will be smarter afterwards.
The comments about AI hallucinations are correct. There were no commercial IIgs clones and for that matter the number of //e and //c clones (or hybrids like the Laser 128 family) were actually fairly few compared to ][+ clones. The reason for that is the custom chips (MMU and IOU) in the //e and //c were harder to clone than the ][+ which was built pretty much entirely with off the shelf parts plus by the time the //e came out Apple was startingt o try to use legal pressure to keep clones off the market. They eventually successfully got Franklin out of the clone business (they went to just making handhelds) but the Laser family they were less successful because of how Central Point, VTech's partner on that was able to use "clean room" techniques and license Microsoft BASIC to make a ROM that was hard to challenge in court. Most of the direct //e clones were either made and sold in countries like Brazil where Apple didn't have the same reach or they were imported to the US in parts and then things like the ROMs installed by small clone shops. Also by the time the //e and //c were the norm, a lot of cloning had moved on to doing PC clones beause that market was booming.
thanks for your answer and comments,
I have tried many things to restart the Smarport detection after initial booting phase but it is not working.
the way I need to manage it is:
- Start the Smartdisk emulation smartport and provide init phase for 4 virtual image, and switch to 5.25 emulation...
For the moment, I am triggering a soft reset of the IIGS, (I know it is dirty ;)
You can tests this in the last release of the test version of the Smartdisk in the testUF2 directiory on github.
to be continued,
By the way, Yellowstone card is not working with Smartidsk smartport emulation, I will have to investigate and fix it as well
Vincent
No, it is your smartdisk emulator that is not working with Yellowstone controller. The Yellowstone firmware is derived from Laser UDC firmware, so one can assume that your emulator is not working with Laser 128 computer and UDC cards, too.
Yes this what I meant of course
and I will fix the smartdisk soon
Vincent
Does anybody know how's the bank switching of the emulated Yellowstone ROM controlled and how to exctract the Apple2 ROM part of the Yellowstone?
I have a UDC card so I could test it to see if it works or not.
You are a talent of meaningless messages. Do you expect us to beg you to test it? Are you boasting you have a UDC card? I also have a UDC card. This does not make me smarter.
You are an arrogant jerk who is unnecessarily abrasive for no reason.
There isn't much point in me testing the UDC card with VIBR's device if someone has already done it. That is why I ask. You did not mention whether you had a UDC card that you had already tested or verified whether it works or not.
Based on your comments you obviously believe that you are smarter than everyone else, but your communication skills stink. Not because you don't know how to write, but because you are ajust a nasty person.
So, if someone else has ever connected this smartport emulator to a UDC controller there is no need for you to put such enormous effort in testing that combination? It is easier to write, of course.
This is completley inappropriate and unwarranted.