New SIDE3 Loader Testing and Development Thread

[TGB1718]

46 minutes ago, flashjazzcat said:

You run CMD.XEX from the loader, not from the SpartaDOS X prompt (or any other DOS, for that matter). That's why the 'Compatible DOS not found!' message is displayed.

 

Sorry, I just forgot, spent so much time getting things right with Windows 11, what a painless upgrade that was 👎🏾

 

Just tried and it works fine.

 

thanks

[flashjazzcat]

Loader update:

 

https://atari8.co.uk/firmware/side3-beta-loader/

 

Includes:

• Fix for CAR mounting bug
• Support for SIMCovox
• Various UI fixes
• DOS now maintains the last accessed date when opening files for read (loader doesn't yet, though)

This is an interim release until I get time to do yet another hefty re-write which should simplify the loader internally and improve maintainability (it currently carries a lot of code from older SIDE loaders which needs to be completely replaced). As usual, any howlers will be fixed as soon as I hear about them, however. This update should fix problems with cartridge loading in cases where issues aren't caused by bus timing/instability (which usually causes the loader to crash before you even get as far as mounting a cartridge).

 

DMA writes still aren't activated even though they work in Altirra, since they still don't work on real hardware. We'd be happy to hear from anyone expert in the SD card protocol, since everyone appears to be flat out of ideas why DMA writes proven to work algorithmically (via Altirra) nevertheless result in missing sectors and write errors on real hardware.

 

Thanks to @Beeblebrox and others for testing, suggestions, etc. SIMCovox was requested ages ago but I only just got the chance to test (and debug) it today.

 

Note the beta loader has now become the pre-installed default on SIDE 3.1 hardware, following agreement with the vendor. It wasn't my idea to include the beta loader on the new hardware, but since it found its way there anyway, I've removed the admonishment on the download page.

[+Stephen]

On 8/1/2022 at 6:14 PM, Stephen said:

How does it fail?  For me, it wouldn't read the SD card, and when I got to the config screen, there would be some spurious key presses and then it would just hang.  Even if I left it at the SD card screen upon booting, not pressing any keys would cause a hang.  The colours changed.

Well, I finally got around to testing the my SIDE3 on several other machines the other day and I'm sad to say, the issue is not with the 576, but with my SIDE3 cart.  I tried it on 6 different machines, both PAL and NTSC, some stock, some modified and it doesn't work with any of them.  Multiple different power supplies were tested.  I measured power at the input on 2 machines that were open.  Open circuit voltage was 5.01VDC and when powered on, 4.97VDC so the supplies are quite well regulated. The behaviour on every machine is the same.

 

75% of the time, the machine doesn't even act like a cart is plugged in (regardless of switch position, or SD card inserted or not).

On the rare occasion the machine sees the cart, it boots to a black screen and sometimes pressing reset will make the Loader load.

Just sitting and doing nothing, the Loader will crash.

 

The 6 machines I tested this on (listed below) are 100% rock solid with every single device I own (I have every flash cart ever released, SIC, SIDE2, AVG, etc.).  Also, I was able to re-flash the loader on a SIDE3.1 (not mine) and that cart was completely stable on 2 of the machines I tried it with.

 

130XE NTSC stock

130XE NTSC 320kB 32-in-1 OS

130XE PAL, U1MB

576NUC+ with cart port

1088XLD PAL (in fact this machine is running a 74F08 to fix a GTIA issue)

1088XEL PAL

 

Tomorrow I will be attempting a JED update.  I was able to use the cart long enough to install the 0.67 beta loader on it, but I couldn't test any of the new features.

[Frankie]

-I figured it out-

 

Edited November 21, 2022 by Frankie

[Beeblebrox]

10 hours ago, Stephen said:

Well, I finally got around to testing the my SIDE3 on several other machines the other day and I'm sad to say, the issue is not with the 576, but with my SIDE3 cart.  I tried it on 6 different machines, both PAL and NTSC, some stock, some modified and it doesn't work with any of them.  Multiple different power supplies were tested.  I measured power at the input on 2 machines that were open.  Open circuit voltage was 5.01VDC and when powered on, 4.97VDC so the supplies are quite well regulated. The behaviour on every machine is the same.

 

75% of the time, the machine doesn't even act like a cart is plugged in (regardless of switch position, or SD card inserted or not).

On the rare occasion the machine sees the cart, it boots to a black screen and sometimes pressing reset will make the Loader load.

Just sitting and doing nothing, the Loader will crash.

 

The 6 machines I tested this on (listed below) are 100% rock solid with every single device I own (I have every flash cart ever released, SIC, SIDE2, AVG, etc.).  Also, I was able to re-flash the loader on a SIDE3.1 (not mine) and that cart was completely stable on 2 of the machines I tried it with.

 

130XE NTSC stock

130XE NTSC 320kB 32-in-1 OS

130XE PAL, U1MB

576NUC+ with cart port

1088XLD PAL (in fact this machine is running a 74F08 to fix a GTIA issue)

1088XEL PAL

 

Tomorrow I will be attempting a JED update.  I was able to use the cart long enough to install the 0.67 beta loader on it, but I couldn't test any of the new features.

What loader firmware are you running on that particular SIDE3 again?  (Not that it should stop the cart being recognised by stock machines of course ) Just curious.

 

Wondering if either the cart has gotten corrupted at some stage or maybe it's edge connector is damaged/not making stable contact. Sounds like the former is reset sometimes allows the loader to boot.

[+Stephen]

3 hours ago, Beeblebrox said:

What loader firmware are you running on that particular SIDE3 again?  (Not that it should stop the cart being recognised by stock machines of course ) Just curious.

 

Wondering if either the cart has gotten corrupted at some stage or maybe it's edge connector is damaged/not making stable contact. Sounds like the former is reset sometimes allows the loader to boot.

Loader firmware is currently 0.67 beta.  It was I think .45 when I got the cart.  After work today I'll be trying different JEDs as well as the latest 0.68 loader.  I did even remove the cart PCB from the case thinking maybe the shell was preventing proper cart insertion, but that also did not make a difference.  I will report back this evening after re-flashing.

[flashjazzcat]

Another update:

 

https://atari8.co.uk/firmware/side3-beta-loader/

 

(No version number bump, but the revision date has changed).

 

This one implements support for the SD card write-protect tab, which I had forgotten to add after the loader gained write support (although the PBI BIOS always had it). This works in the loader and in the 'FATFMS' DOS. Note that if the card is write-protected, DOS can't maintain the last-accessed date for files opened for read (for obvious reasons).

 

I've also fixed crippled file copying speeds in yesterday's build. This was left over from tortous DMA-write tests from weeks or months ago; I had changed the copy buffer size to 1K. Nothing was accomplished by any of this, and I've changed it back to 8K now which makes file copying a lot quicker.

 

Lotharek asked me about support for TMC, CMC, etc, music playback. I know little about these other than there's a player/tracker for every file type, and many of these players probably work perfectly well when launched from the loader (and they can now write to the card, browse the FAT directory tree, etc). Some of the players are written for SDX, however, so may or may not work from the command-line DOS supplied with the loader. I suppose the vendor has some kind of super-player in mind whereby the loader can natively play back a wide variety of formats (and we still have to add video playback), so I invite suggestions and opinions on this.

 

The most requested feature is currently being able to browse SDFS and AtariDOS formatted ATRs and partitions in the loader, and perhaps be able to copy stuff into them. I'm invited to start porting the firmware to the new 'coming soon' devices as well, of course, so the to-do list continues to grow.

14 hours ago, Stephen said:

Tomorrow I will be attempting a JED update.

I've sent you the JED flasher and will be very interested to hear of your experiences, since we would like to deploy this tool more widely.

Edited November 21, 2022 by flashjazzcat

[+Stephen]

9 hours ago, Stephen said:

Loader firmware is currently 0.67 beta.  It was I think .45 when I got the cart.  After work today I'll be trying different JEDs as well as the latest 0.68 loader.  I did even remove the cart PCB from the case thinking maybe the shell was preventing proper cart insertion, but that also did not make a difference.  I will report back this evening after re-flashing.

Happy to say I had no problems flashing the JED 1.1 and 1.3 updates.  The cart seems perfectly happy now, and I have the loader 0.68 beta up and running.  Thanks so much for this.

[Beeblebrox]

10 minutes ago, Stephen said:

Happy to say I had no problems flashing the JED 1.1 and 1.3 updates.  The cart seems perfectly happy now, and I have the loader 0.68 beta up and running.  Thanks so much for this.

Great news.

[flashjazzcat]

On 11/21/2022 at 2:47 PM, flashjazzcat said:

since we would like to deploy this tool more widely

Scratch that. Since the JED update tool apparently can't be tested by anyone outside a core circle of 2-3 people (few of whom appear inclined to test said tool) without a bad flash necessitating return of the device to the vendor for repair (since the raw JED cannot be published), I'm not going to deploy the tool, since all it will do is generate more hardware-related complaints in my email inbox. I've just had to tell one user that he needs to send his cartridge back to the vendor after the JED update tool apparently bricked his device, and that's the last time I'm doing that.

[_The Doctor__]

there is always a chance of bricking stuff, that's nothing new... un bricking shouldn't be an issue, there's a tool for that too!

 

[flashjazzcat]

Read again carefully. You can have all the tools in the world, but without the JED, you're f***ed.

[_The Doctor__]

it would appear the individual tried with a jed update tool that must contain jed information to do the update... but since it's clear the individual doesn't have access to the stand alone jed file or possibly cables/blaster etc. they would probably have to have another individual do it for them.

In order to update you are pretty much stating they'd have to send it to the vendor (should they still be in business or not) to get it updated then, or un-bricked by the vendor or person whom may have provided whatever that may have led to the fail. I wouldn't sweat it, there was no harm in trying either way then... if it was a success then all that chance of loss or damage in shipping and money wouldn't have come into play. If it failed it would have needed to go through all of that for the update anyway.

 

Who knows maybe they will acquire the tools, knowledge, abilities, and will be able to do it themselves with jed file or extracted files and be a tester as well. Rather than just 2 or 3 testers you might have a larger pool to draw from and get a more thorough test base. All sounds like win win win eventually.

Edited November 23, 2022 by _The Doctor__

[flashjazzcat]

11 hours ago, _The Doctor__ said:

If it failed it would have needed to go through all of that for the update anyway.

No it wouldn't, since I believe the individual concerned (who lives in the United States) already has the USB cable which would have allowed him to un-brick the cartridge in five minutes had a climate of IP-theft and litigation not discouraged the author of the JED file from releasing it to the public in any capacity whatsoever.

 

The stand-alone flasher does not contain the JED file in any form usable by the platform cable, BTW.

11 hours ago, _The Doctor__ said:

Who knows maybe they will acquire the tools, knowledge, abilities, and will be able to do it themselves with jed file or extracted files and be a tester as well. Rather than just 2 or 3 testers you might have a larger pool to draw from and get a more thorough test base. All sounds like win win win eventually.

I would not condone or advocate the reverse-engineering of the stand-alone flashing tool even if it were possible to extract the JED file from it (which it is not). Since the flashing tool (which I wrote based on programming code kindly provided by Candle) was written three years ago and has not yet been adequately tested by 2-3 'trusted' individuals, let alone by a 'larger pool' (all of whom would require platform cables and copies of the JED to recover from mishaps, and presumably NDAs), I fail to see how the second private individual to try the flasher bricking his cart, having to send it overseas for repair, and opining that he will as a result think twice before purchasing anything from the vendor again can be regarded as a 'win win win' situation on any time-scale.

[_The Doctor__]

Not that any of it really matters but since it's approached in the way it it has been, it's hard to follow...

 

huh? was the device a knock off or something? was the vendor illicit?  All very strange, it's common for many of today's mainstream products, televisions, stereos and the like for even the casual consumer to receive updates/upgrades for not only standard flash portions of their products but also programmable arrays. Sure something can and will go wrong for some folks, but the entire industry isn't shattered by it.

 

that being said, many repair centers/repair service companies have folks who perform updates as well as un-bricking of many devices using all manor of such files and cables/programmers/blasters... almost none of them know anything about the actual information contained in such files, and certainly do not have NDA's for every device that passes through their hands.

 

Still amazes to this day how crazy folks get over the simple things.  It's appears nuts how a simple thing becomes a major incident to handle a mishap. I wonder what it will be like if a large number of the items are sold and all need an update to fix them... poor manufacturer receiving all of them for updates... we have seen how much the main suppliers of these goods enjoy that sort of thing on a smaller scale..

 

Maybe some sort of encrypted jed /key delivery system/tool would get the job done... unique and secure.

 

Best of luck to all involved. Interesting to see how it unfolds over time

Edited November 23, 2022 by _The Doctor__

[flashjazzcat]

1 hour ago, _The Doctor__ said:

huh? was the device a knock off or something? was the vendor illicit?

No, but you assume every vendor has the equipment and files with which to re-flash a bricked device. They may or they may not, but this still means that the end user - at the very least - has to send the cartridge back to the vendor in the post when - were it not for a breakdown of trust and confidence - the user could have otherwise fixed it himself in five minutes.

1 hour ago, _The Doctor__ said:

Sure something can and will go wrong for some folks, but the entire industry isn't shattered by it.

Where is it argued that 'the entire industry' is 'shattered' by it? I'm making the point that the breakdown of trust and confidence has inconvenienced the second user who tried the flashing tool, left a sour taste in his mouth because he couldn't recover from the bad flash in the way he expected, and that I feel as if I've been completely wasting my time promising to deploy the flashing tool once it had been subject the adequate testing which somehow never happened. There are indeed a huge number of SIDE3 cartridges out in the wild which might benefit greatly from JED updates (either for reasons of improved stability, as evidenced by Stephen's success in turning an ornament into a working device, or to provide new functionality), but unless the delivery method is proven to be 99 per cent reliable, and unless there's a way for the 1 per cent to recover without massive inconvenience, it's unlikely that in-field updates can happen.

1 hour ago, _The Doctor__ said:

Maybe some sort of encrypted jed /key delivery system/tool would get the job done... unique and secure.

Certainly, and this idea was floated about two years ago but it never went anywhere. It's rather a shame such things are necessary at all, though, bearing in mind I have vocally criticised those in the past who were completely pre-occupied with 'copy protection methods' on the basis that the entire community was supposed to be predisposed towards rampant piracy.

 

In any case: I simply intended to point out that a) I'm not releasing the update tool for the foreseeable future (or until circumstances change), and b) explaining the reasons behind this, which seem to me entirely logical.

 

End of discussion.

[Panther]

WTF?  I feel like I just stepped into the Twilight Zone.

 

[rensoup]

 Dear troll,

 

 Since you can't avoid posting garbage in the PoP thread, I'm just going to return it here.
 

Quote

You could say that, yes. A bit of skim-reading will satisfy any interested party of the solidly unarguable design decisions which resulted in that situation.

 

You'll be allowed to lecture me on design choices when your device is 100% compatible with all A8 software. 
 

[flashjazzcat]

11 minutes ago, rensoup said:

Since you can't avoid posting garbage in the PoP thread, I'm just going to return it here.

Oh dear. Symptoms of a tormented mind. You read too much into things.

[rensoup]

You eagerly keep posting nonsense in a thread you pretend to have put on ignore and given how crap your cart support was and the resulting accusations of "device protection" that followed, you're hardly the definition of sanity.

[flashjazzcat]

8 hours ago, rensoup said:

You eagerly keep posting nonsense in a thread you pretend to have put on ignore

I took it off ignore, since threads (and heads) generally cool off after a few months. Boy, was I wrong.

 

Regarding 'nonsense': Having pointed someone to a WUDSN update in another thread, I noticed in the PoP thread that @Beeblebrox was under the impression that the PoP cart images didn't work with SIDE3. So I (not 'eagerly', but helpfully) wrote (in my first post there since December 2021):

Even taking the 'L' by mentioning the bugs in the loader which had previously prevented said cartridge images from working, and thanking the PoP CAR files for being instrumental in said bugs being exposed. Beeblebrox was delighted by this news:

Nevertheless, Beeblebrox - being not the first and certainly not the last person who wonders why the ATR files don't work with Incognito (nor U1MB/SIDE, IDE Plus, etc) - seemed hopeful that an Incognito firmware update could make them work (which it couldn't), but conceded that the reasons for all this had been discussed earlier in the thread. Keen to prevent a re-hash of said discussion, I wrote:

Why a couple of people found my reverence towards the developer's solidly unarguable decision-making amusing, I do not know.😕

 

Finally, in response to some chatter about ATRs not working with 'RAM-based operating systems', etc, I point out (factually, paraphrasing the readme file you subsequently presented):

Which is to say:

1. The ATRs require an SIO device (so, no PBI-based SIO emulation, etc, for those who don't already know that)
2. The ATRs load at 1x SIO speed

8 hours ago, rensoup said:

given how crap your cart support was

Mature of you, in light of the fact I acknowledged and fixed the bugs in short order and publicly credit PoP with helping to facilitate those fixes.

8 hours ago, rensoup said:

accusations of "device protection" that followed

LOL. Well, PoP is certainly effectively protected from ever running on Incognito, etc. Don't blame me for the fact users STILL think PoP doesn't run on this or that device on which it actually runs perfectly well. I suppose that's an inevitable result of PoP ATRs having 'certain requirements' and the CAR images initially not working on SIDE3 because of some bugs that were later fixed, but simple clarification of the situation probably isn't going to cause offense to the majority of forum members.

9 hours ago, rensoup said:

You'll be allowed to lecture me on design choices when your device is 100% compatible with all A8 software. 

Although I no doubt lack your competence levels when it comes to the Atari OS, storage hardware and IO subsystem, I don't think I'll waste my time trying to make PBI-hosted ATR emulation (on hardware designed not by me, but by Candle O'Sin, whose solidly unarguable design decisions dictate the basic 'theory of operation' which my firmware must facilitate) support software whose readme file explicitly lists 'SD/DD disk drive' in the list of requirements.

 

In any case, I somehow think it's not so much what's written as who's writing it that's causing your issues, but feel free to keep going with this exhibition. It's educational.

Edited April 5, 2023 by flashjazzcat

[rensoup]

 

Amazing that you'd blame others for the very things that you keep doing. 

 

Talk about torment and maturity when you've been rebashing the same XBios nonsense for so many years in possibly every thread you join (Silkworm thread being the latest)

 

I don't care about your loader allergies but feel free to let the world know... but find new threads to spam. If you're going to claim again that it's a bad design decision, back it up with FACTS and point to a better solution otherwise you have shit in the mouth.
 

On 4/5/2023 at 10:59 AM, flashjazzcat said:

LOL. Well, PoP is certainly effectively protected from ever running on Incognito, etc. Don't blame me for the fact users STILL think PoP doesn't run on this or that device on which it actually runs perfectly well. I suppose that's an inevitable result of PoP ATRs having 'certain requirements' and the CAR images initially not working on SIDE3 because of some bugs that were later fixed, but simple clarification of the situation probably isn't going to cause offense to the majority of forum members.

 

All I know is that with an AVG, and the ATR or CAR files loaded on a CF, it just takes booting the machine, pressing ENTER on the selected files and PoP runs without an issue. Same as every other A8 software. User friendly or original hardware is a mad requirement for sure ...

 

On 4/5/2023 at 10:59 AM, flashjazzcat said:

I took it off ignore, since threads (and heads) generally cool off after a few months. Boy, was I wrong.

Just put it back, your contributions will not be missed.

[flashjazzcat]

I'm still waiting to hear why anything I wrote in response to Beeblebrox's queries in the PoP thread in any way justifies your ranting and raving or could in any way be classified as coordinated trolling or deliberate thread disruption. As said: I hadn't posted there in sixteen months and was simply clarifying the status of CAR and ATR files with regard to SIDE3 and Incognito. I would also point out that it wasn't me who smilingly brought up XBIOS in the Silkworm thread (I'm wondering: would a wordless upvote of a contentious post count as evidence of trolling as well?).

 

I have no desire to perpetuate these stupid quarrels (and this was the motivation for unblocking contentious threads, every single one of which aside from yours no longer requires one to walk on egg-shells), but nor can I disabuse myself of the knowledge that certain authors are either oblivious to or deliberately (even if for ostensibly practical reasons) unfriendly towards non-SIO loader and storage solutions. That's up to them, of course, but I and others have an opinion on the matter, just as you clearly have an opinion to the contrary, which you continue to voice loudly and colourfully in my threads at the slightest perceived provocation.

 

It's kind of a shame that we've apparently ended up at a 'works on SIO, so fuck you' situation here, and that differences of opinion I thought could now be relegated to light-hearted ribbing apparently remain burning, deeply personal hot-button issues about which you remain tremendously sensitive. Clearly you never got over the fact I held the door open for you when you flounced out of another of my threads, and therefore any comment I make - complementary or otherwise - immediately triggers you. Try chilling out, otherwise (since I manage my own thread ignore list, thanks, just as you clearly do) you're going to waste an awful lot of energy propagandising about my forum activity.

Edited April 7, 2023 by flashjazzcat

[Ricky Spanish]

Looks like I need to get a side3 for my 1200xl ! Nice work @flashjazzcat !!

 

EDIT;

Seems I can't find the SIDE3 anywhere

Edited April 7, 2023 by Ricky Spanish
added info

[flashjazzcat]

I have replied to your comment under my YouTube video.