Dr Memory

Yes, well, I was and am one of those having issues. I'm seeing it on Firefox and Chrome on a Windows 10 machine, Chromium on a Raspberry PI, Safari on a Mac, and Chrome on a Samsung Phone through an entirely separate path (not even the same ISP). So you are not correct in asserting that this has anything to do with people who have messed around deep in their browser security settings. The Raspios install is ultra-fresh (like < 24 hours old). I haven't touched the Mac in years - it sits quietly there serving music and has no updates beyond whatever Apple pushes out. I've certainly not touched the phone settings - those are pretty well locked down by AT&T. Whatever is going on is not due to user error nor caused by user configuration changes. I just rechecked and I'm still getting the problem on all the platforms described, even after rebooting. I think you are optimistic in your assertion that "It was never *NOT* working for most people" as very few responded to the other thread, where I first reported this. We have no evidence for what percentage of people were/are experiencing the problem and not enough of a geographic sample to know if it's a regional issue. Also, the specific error I'm getting is not triggering the "Advanced" option - just "Learn more..." and "Try Again". I am very familiar with the Advanced dialog and allowing exceptions and this is not a situation where that is available. Sorry, but the evidence doesn't back up your assertions. Just because not everyone is seeing a problem, that doesn't mean there is no problem. There is a real problem on Lotharek's end and it isn't fixed yet. Could it be worked around by configuring an exception? Maybe, I'd gladly try that but every browser I've tried on every platform I've tried has failed to offer me a path to doing so. For those who didn't read the original thread, I can reach his site with ping and traceroute and the like, but all of my attempts at connecting to his web site are rejected with security errors. Also, I just dug into the security settings on one of the browsers on one of my PCs and I can see that I'm set up to query OCSP servers (one of the normal ways to validate certificates). Nothing looks wierd or is disabled. Further, I'm able to access everything else I've tried, including other sites in Europe and specifically Poland. This is a configuration issue, on his end, guys. I imagine he's working on it but have no way to verify that. Oh, I did one last test - I logged in to work using my corporate laptop using VPN. Once I've done that, I'm effectively accessing the internet from a different state (specifically, Washington). Got the same symptoms there with Firefox, IE 11, and Edge. All of those are locked down by corporate and most assuredly have their security, root certificates, and CRLs or OCSP settings up to date. I really shouldn't use work resources for this but I'm really frustrated (and it was a fast, simple test).

This is sort of a coda to the U1MB installation, the original purpose of the thread, from way back in the days before we got sidetracked into figuring out if lotharek.pl is really down or what. I decided it was time to put the modified machine back in its case. It now has MUCH better video than previously, auto-switches between composite and S-video, and has Ultimate 1MB installed and working. Yay! Unfortunately, this caused it to gain substantial weight. I can certainly sympathize! I was able to replace the bottom shield, no problem, but not the top. A moment's thought and I saw this should have been obvious. The U1MB sticks up far too far for the shield to go back on. No problem. I searched on here and found that others had come to the same conclusion. Some cut away some of the shield, most went without. Where it got bad was when I tried to put it in the case. Ugh. It is... tight, to say the least. I think the case is actually a little warped now. Even with the cables folded over and layered differently to make it as low profile as I could manage, I still had to pretty much force it. I've got it installed in the same spot as FJC's video. My ribbon cables are longer because I didn't modify them, but I was able to fold them out of the way pretty good. It's still just a lot thicker. I think the thing that stuck up the most was actually the four-pin CPU connector. If I ever open up this machine again (and I dearly hope to avoid that) I'll try and replace it with something lower profile, with the wires coming out the side instead of the top. Anyway, it's a darned tight fit. I don't remember seeing comments on that. Did I do something wrong with my install or is that normal? On the bright side, it's working. Now I need to learn how to use it - how it interacts with SIDE 3, that sort of thing. But that's not a problem, that's an adventure.

LOL! Ya maybe. I'm just concerned because I've been on his site in the past without issue. I can still access Mathy's site, which has a .nl domain, and atari8.info, which I believe to be in Poland. So it's not like I'm cut off from Europe or anything. I'd be less concerned if I wasn't blocked on so many different platforms. Wierd.

No, I'm in southern Indiana, so it's clearly not a U.S. too Lotharek issue. Ok, that's odd. I tried to get there from a raspberry pi, just to try an entirely different platform and browser in case something is hosed on my PC. I'm able to look it up. nslookup lotharek.pl Server: 208.67.222.222 Address: 208.67.222.222#53 Non-authoritative answer: Name: lotharek.pl Address: 85.128.198.36 and traceroute to it. 3 agg60.igwdcaew02h.socal.rr.com (24.30.172.221) 17.404 ms 18.329 ms 18.236 ms 4 agg30.lsaicaev02r.socal.rr.com (72.129.17.152) 21.180 ms 21.350 ms 21.260 ms 5 agg26.tustcaft01r.socal.rr.com (72.129.17.2) 22.217 ms 21.852 ms 24.327 ms 6 ae-5-0.cr0.chi10.tbone.rr.com (66.109.6.202) 25.001 ms 22.514 ms bu-ether16.tustca4200w-bcr00.tbone.rr.com (66.109.6.64) 23.186 ms 7 66.109.5.247 (66.109.5.247) 19.932 ms 16.482 ms 22.379 ms 8 las-b24-link.ip.twelve99.net (62.115.156.224) 28.989 ms 29.750 ms 24.857 ms 9 * cogent-ic352552-las-b24.ip.twelve99-cust.net (80.239.160.65) 22.834 ms * 10 be3271.ccr41.lax01.atlas.cogentco.com (154.54.42.101) 23.253 ms be3360.ccr42.lax01.atlas.cogentco.com (154.54.25.149) 26.156 ms be3271.ccr41.lax01.atlas.cogentco.com (154.54.42.101) 23.937 ms 11 be3176.ccr21.sjc01.atlas.cogentco.com (154.54.31.190) 36.201 ms 35.898 ms 35.794 ms 12 be3178.ccr21.sfo01.atlas.cogentco.com (154.54.43.69) 33.697 ms be3179.ccr22.sfo01.atlas.cogentco.com (154.54.43.149) 36.444 ms 28.318 ms 13 be3716.ccr22.sea02.atlas.cogentco.com (154.54.86.106) 52.396 ms 52.276 ms be3717.ccr21.sea02.atlas.cogentco.com (154.54.86.210) 50.788 ms 14 be2074.rcr21.b052488-1.sea02.atlas.cogentco.com (154.54.47.110) 54.432 ms 53.355 ms 54.208 ms 15 38.104.127.26 (38.104.127.26) 51.574 ms 58.989 ms 57.917 ms 16 sea1-fw1.net.nazwa.pl (85.128.133.150) 56.621 ms 56.312 ms 55.851 ms 17 sea1-a2.net.nazwa.pl (85.128.132.187) 56.460 ms sea1-a1.net.nazwa.pl (85.128.132.186) 56.183 ms 55.866 ms 18 shared-amp36.rev.nazwa.pl (85.128.198.36) 49.227 ms 54.726 ms 54.631 ms (Personal addresses redacted) I can even ping it! 64 bytes from shared-amp36.rev.nazwa.pl (85.128.198.36): icmp_seq=1 ttl=50 time= 49.3 ms 64 bytes from shared-amp36.rev.nazwa.pl (85.128.198.36): icmp_seq=2 ttl=50 time= 49.7 ms 64 bytes from shared-amp36.rev.nazwa.pl (85.128.198.36): icmp_seq=3 ttl=50 time= 49.1 ms 64 bytes from shared-amp36.rev.nazwa.pl (85.128.198.36): icmp_seq=4 ttl=50 time= 57.4 ms etc. but when I try to access it using Chromium, I get the same ERR_SSL_PROTOCOL_ERROR result. So from here, it appears to be a config issue on their end, but if others can access it, then I dunno. I even tried my lonely Mac Mini and my phone, which is not attached using wifi at this time. Same results with slightly different error message text in all cases. Safari says "Safari can't open the page "https://lotharek.pl" because Safari can't establish a secure connection to the server "lotharek.pl"". Phone, which is currently using 4G and not wifi, says "This site can't provide a secure connection" and "lotharek.pl sent an invalid response ERR_SSL_PROTOCOL_ERROR". So whatever this is, it isn't just my home ISP or my router.

Grats on getting a payment from YouTube finally! You deserve it! I agree with everything you said there. I write a lot in my job, which used to be architecting, designing and sometimes developing non-trivial software but has evolved into more leadership, bureaucracy, and communication. You can't satisfy everyone at once. People seem to really like the stuff I write, but unfortunately that has caused them to demand more of that and less of the things I find interesting and fun. The management and financial parts are just... not fun. This born-again Atari hobby is often the most fun thing I get to do on any given day. I'm trying to help pay that back by contributing when I think I can. Maybe I should just start a blog instead of posting these long-form write-ups.

As I had guessed, there is good info out there on how to mod ribbon cables. For example: https://startingelectronics.org/articles/IDC-ribbon-cable/ According to the author of that page, these are a form of IDC connector, which is an Insulation-displacement Connector. https://en.wikipedia.org/wiki/Insulation-displacement_connector There are many types - I've encountered other types of IDC cables in the past and didn't realize this was one of those. It makes sense that SCSI and floppy cables would be in the same family, but I wouldn't have guessed DE-9s were. Those are the 9-pin connectors most people erroneously call DB-9. I only learned that about a year ago, I called them that for years also! Anyway, the first link explains how to make ribbon cables using the appropriate IDC connectors, and the video shows how exactly to line things up and such, from multiple views. You can even see how the connector grabs the wires, which I was curious about.

No drilling needed (of any kind, whatsoever) to install U1MB and make it fully reversible... (at least on Rev.C MoBo, Hong-Kong, fully socketed) Now, in the case that U1MB is not going out in the future (like my case), the shields will require precise-cutting (non-reversible) to open space for U1MB, as they should be installed back in place, for normal operation. That, of course, will push the work clearly longer than an hour... Another option is to partially dismantle the shields, store them again, but they will likely corrode in contact with your hands, and unfriendly storage conditions. Interesting approach! I never even thought of cutting the shield. Unfortunately, my board is not fully socketed, so I'm not sure how well this would have gone for me, but I'll certainly consider it in the future. It looks a lot cleaner than what I did. I don't think I have any way to make precision cuts in metal like that though. Thanks for sharing that. I'm not concerned about it taking longer than an hour - I was just poking at the "5 minutes" thing in the original procedure.

I agree fully. I was just pointing out that for many of us, additional info is needed. Really, this is the biggest difference between a video showing an example installation and a written procedure - it's great as far as it goes but does not and can not cover all situations. Also, and this is the part that bugs me about the modern tendency to make a video and call it a day, I can't do a text search on a video nor print it for later study. It's a lot harder to do research on points of confusion - can't just copy a sentence that puzzled me and put it in google, for example. Still agree that it isn't your problem, not like you're being paid to do this.

Tried that, no joy. So maybe US -> lotharek routing issue.

I wish I had thought to check for something like that. Thanks for the info. If I end up loving the U1MB and get another, I'll try it with that next time.

Now that's interesting. I still can't get there. Maybe some sort of routing or ISP issue between here and there? I normally can. Usually I get a different error if it's routing or connectivity... The error I'm getting implies that the site is up but has certificate issues. I'm tempted to blame Firefox updates, except it happens using Chrome as well, like so. Can anyone in the Los Angeles area reach lotharek.pl? I can't do anything about it either way, just curious.

Sorry dude, but not all of us do well with video. You are correct that you show the vias in your video, briefly, but I do a lot better with something I can read and print and have on the workbench next to the system being worked on. Or study while sitting in a comfy chair, not in the computer room. Or maybe even make notes on. I'm not sure if this is a personal preference thing or an old school thing - I just do far better with written material with pictures. Like if I want to look again, or more closely, at one of your inset pictures, I have to scroll through the video to find the few seconds where it is on-screen, pause, and stare at it. I suppose I could at that point grab the screen, save it as an image, crop it, and print it, but that just gets me to where I started with the picture I had. So I still have to run around in circles, just different circles. So despite the cool video, which is indeed helpful because it shows how to do the things that would be hard to explain in print, I still would be far happier to use that as an optional supplement to a written procedure with pictures. I've learned a lot from your videos, such as soldering and de-soldering techniques, but I'm sure the lessons I take away aren't always what you're trying to teach. Clearly, that is NOT YOUR PROBLEM. It's just my preference and I know I work best that way. As for the ribbon cable thing, this is an example of the problem with video. Even when you explain it, you describe the things one would need that you don't assume they will already have laying around or know. I don't know what the connectors are called that need to be pressed onto the cable. I don't know how to align them properly. I'm not really sure how an electrical connection is made by the connector - maybe it has little needle pins inside? A diagram and BOM would help me immensely here. Again, not your problem, just trying to explain why I don't find video only tutorials totally satisfying. I'm sure I can do a quick search and pull up the rest of the info I lack. I'm guessing that you already have a stockpile of the needed connectors and have installed them many times, so those mundane details seem obvious to you and not in need of explanation. Agree that if you follow Lotharek's perfunctory instructions you get what you deserve in the end. Still, starting with a checklist like that helps, it's just not sufficient.

If it isn't copy protected, you could use any number of programs to image the disk to another disk or a file. Every DOS I can think of has that capability. If that fails, you try something like Diskey, which would give you more info and let you poke around.

I watched FJC's video. I tried to get the installation procedure from lotharek.pl, and immediately hit a serious problem. That problem is that lotharek.pl appears to be down or at least inaccessible. When I try, I get this: It's been that way all day, possibly longer - I believe I first tried to get this info last night. I tried a few things to get there, including different web browsers and a different computer, but no joy. So I went hunting for other sources of these instructions. I actually bought the thing from TBA so I went there. They had a link to FJC's "extensive user manual", which worked. This document contains a lot of good info, but alas, no actual installation instructions for the board. They also had a link to an installation manual, which led to, you guessed it, the lotharek page. I assume this is a short-term problem and will be fixed soon, but that doesn't help me today. So, off to the Internet Archives. The various deep links to the product page I found in articles on AtariAge didn't work for me there, but going directly to an archived copy of the top-level lotharek.pl site let me navigate to a product page for the version with the FJC firmware, which is what I have. One problem with this still is that the embedded links don't really work so well, so I had to do some additional gymnastics with the IA to get readable copies of the pictures. To be clear, I don't think any of this is the fault of the video nor of the product - it's just a problem that added some challenges to the install. Anyway, I saw several mentions of using a CPU diagram (that I couldn't access) to see what CPU pins you need to solder the 4-pin connector to. I managed to dredge up a copy, again using the IA and a couple of different links to that picture that people had posted previously. Once I found a working version of that, things sped up. So that helped immensely. It'll become easy to get again once lotharek.pl comes back online, but can't hurt to have here. Somewhere along the way, I found this http://atariprojects.org/2020/12/30/purchase-and-install-an-ultimate-1mb-board-for-your-atari-800xl-1-3-hours/ Which was helpful, but led me to try to figure out what pins, exactly, were being pointed at. I'm a big fan of pin numbers, or at least knowing what the mods I'm installing are actually doing in terms of interfacing with the computer. So I used a multimeter to try and figure out what pins these are and to verify that they actually match the CPU table above. Three of them did. One did not. Wire 1 did not have continuity with pin 39 of Sally. This one was pretty funny, in a way - I traced out the circuits in the schematics, and actually went back to the table I made the other day when I looked at all the Phi signals, and found to my surprise that the pin shown (pointed at) only gets Phi 2 when the board is powered! This can be verified by checking it against POKEY pin 7 or PIA pin 25, which have the same feature. Phi 2 from Sally is ANDed with +5V in U18D (pins 12 & 13 -> 11) and then passed along to the other chips and to the mystery hole. The other three wires were fine as shown in the picture and showed continuity even w/o power. I guess I could have ignored that pic and just soldered directly to Sally but that made me uncomfortable so I wanted to do it the other way. Which worked. I don't have any better way of describing which spots to use than is shown in that pic, and that's the highest rez I could find. Drilling was a bit of an adventure, just because I've never tried to drill holes in a motherboard before, and one that I care about at that. Dremel did the job just fine. On the bright side, the location of the holes doesn't have to be precise, just the spacing, so long as you don't drill into a trace of course. FJC's video showing hole drilling was helpful here. For me, being ready to back off if the bit starts to skitter was pretty important. Once I got the holes drilled and the 4 pins soldered on, the rest was easy. Needless to say, my install isn't as pretty as FJC's - I don't have any of the things needed to shorten the ribbon cables, and I'm not quite sure what he did there when he put on the new connectors anyway, so I've got full length cables. Meh. I'll fold em over and make it work when I'm ready to put it back in the case. Amusingly, the author of the lotharek procedure says "Installation of this upgrade takes me approx. 5 minutes and even rookie can handle it at home...". Um. Maybe it's a true statement if you have done it before and have ready access to the info, but each challenge you encounter can add quite a lot of time. Like, socketing the OS ROM and the MMU takes time, and puzzling out the 4 pin wire diagram does too, if you aren't 100% confident you are interpreting the arrows properly like me. Drilling the holes COULD go bad, but did not in my case. If you screw up the drilling you could kill your board or at least make it need expert repair attention. I agree with the others who have mentioned they thought this product would benefit from a clear, step-by-step installation procedure. This is the entire procedure from the archived lotharek.pl page: It's not lying, that's what you have to do, but it reads more like an elevator speech you'd give someone when you were in a tearing hurry to not be there than an actual procedure. I can't see getting my time down to 5 min any time soon - maybe a half hour, on the same system type? In the end, it wasn't THAT bad, but I'm not sure I'd claim it was the easiest mod ever. Oh, after all that, it powered up and worked first try. Go team Atari!

_The Doctor_, thank you so much for that info! I definitely will use it when I get to the "debug CRT" step in my master plan.

It was mainly the noise of the drives seeking, but it was being caused by the thumbnail making daemon. Which I disabled, first through the GUI, then by killing off the startup script, and eventually by deleting the program (last two via shell). It always came back and would run, continuously, even if I could see it was explicitly disabled. Very frustrating. I tried every fix I saw suggested... It wasn't just me - there are many threads about this. Also, it was definitely that - I could see it running in top whenever I got annoyed enough to check. Disappointing. The only thing I never tried was removing all images and such from the unit, to see if it would cut it out if there was nothing to thumbnail-ify. I suppose I could try that if I get really bored someday - unit is still around. It either never finished or immediately started a new pass when it was done, dunno which. I suspect it might not have been such an issue if I had a Synology with more RAM. Adding more disk space certainly didn't help! Yours are much higher end. That could well be the difference.

Nice! That's quite a nice cluster of Synologys there also, Albert. I ran a Synology DS209 (+?) for many years but switched to a QNAP last time I ran out of space. No regrets. The QNAP does real-time transcoding and doesn't continuously make noise due to a persistent, aggressive thumbnail daemon, unlike SOME NASes I could mention. They are a little bit nicer about apps and licensing too I think. No real regrets about the Synology though, it served me well for quite some time.

So when the OP says that, and tries this 20 OPEN #1,4,0,"N:TNFS://127.0.0.1/MISC/TEST.TXT" Does that mean that the file he is trying to open is on a TNFS: server running on the same PC he's running FujiNet on? I guess I got confused when he mentioned he was communicating via SIO2PC/USB. It seemed like one end or the other should be on an Atari, but maybe he's looping back to the PC? Not sure I am visualizing this properly. I only posted because I find IP/routing/protocol stuff pretty easy to debug usually, but then I usually have a better idea of how things are configured and what is being attempted. I'll shut up now.

It's an assortment of systems I've accumulated over the years. Some where bought new in store! Some from ebay, some from other places. I think only one 800 is totally dead at the moment but a couple of systems are impaired. I try fairly hard to fix things myself before reaching out for help like this. I'm sure nobody really cares, but the current roll call: 3 Atari 800 4 Atari 800XL 2 Atari 130XE 1 Atari XEGS Sadly, this is only part of my collection - there are also Atari STs, a nice Amiga 4000, and quite a collection of gaming systems. Some of the Ataris have various upgrades and/or problems. The one I'm working on now is the best-looking of the 800XLs. I want to do the U1MB upgrade before I start messing with possible electrocution er monitor repair. I got slowed down by lotharek being down so I couldn't get the installation instructions, but I got them from the Internet Archive project and am now trying to psych myself up to drill holes in my motherboard. (eek!) You can kind of see why I've been putting a lot of effort into learning how to repair stuff - everything is getting older and developing problems, including me. Hard to say who will get video upgrades and who will not, depends on which is getting the most use I suppose.

Dumb question. 127.0.0.1 is normally the loopback address, which tells a network stack to connect to the local host. Is that really what you want it to do? I would think this would be the Atari itself in this case, and that you'd want this to be the address of the PC running TNFS. But I could very well be entirely wrong, as I haven't tried TNFS yet and am basing this suggestion on decades of experience with IP networking. Maybe they did it that way on purpose, and there is configuration somewhere that tells Fujinet to route such connections to another IP address, I dunno. I'm keying off of the statement that you're trying to connect to a PC using a protocol running atop IP and the address used only. Sorry if this turns out to be a dumb comment.

Start about 2:30 into this to see what I'm worried about (regarding working on CRTs). Imagine his ad hoc probe were an unwary hand...

I did the LCD test. The problem did not manifest there, using either composite or s-video as the input to the up-convertor. So that's confirmed - this has something to do with CRTs, but not one particular CRT or brand. Also, it doesn't look nearly as nice on the LCD - the colors are washed out, the image is stretched, it isn't as bright... I tend to blame the cheap up-converter for all of that. The LCD video quality is beautiful with modern sources. But the blooming lines or however you want to describe this don't happen. I'm pretty hesitant to open up my working CRT monitors but perhaps I should at least look around for obviously blown caps. I've been doing a lot of semi-modern electronics work lately, obviously, but old CRTs have much higher voltages and larger capacitors (which can hold bigger charges for longer). Just because I know the risks doesn't mean I'm competent to deal with them.

That sounds correct. I plan to dig up my HDMI up-converter and test with that. It shouldn't be as sensitive to being over-driven. I'll read the SV documentation yet again also - maybe I missed some subtle change meant to address excessive video signal levels.

I've heard a lot of good things about the UAV. I can think of a good reason to go with a (basically free, handmade) upgrade like SV over the UAV - if you have something like 10 Atari 8-bit systems it gets freaking expensive! It adds up. Especially when some of those systems already have various upgrades installed, such as RAMBO or Omnimon. If I went with UAV for everything I'd have to deal with possible conflicts with other upgrades. Also, SV and similar video upgrades allow me to continue to use the already installed monitor jacks. Maybe I could wire a UAV to do that too? Not sure, haven't installed one yet, but I can see it has screw terminals for its outputs and I remember people talking about wiring up S-Video ports and the like. I am very happy that the SV thing works as well as it does. I did buy one UAV board, at the same time as I picked up the Ultimate 1MB upgrade I've been agonizing over, and have been trying to decide what system to put it in. I really didn't want to put them in the same system unless I couldn't get acceptable video without. I don't think it would be good for this thread to morph into a discussion of the pros and cons of UAV, especially before I finish reading the massive Rev D Upgrade Thread.

Honestly, I don't think it's a problem at all. I believe it's just a subtle effect I never noticed before, and only noticed now because of a bad setting on my Toshiba monitor. Like I said, I am NOT claiming this is a bug, and have verified that by reproducing it with an entirely different piece of software. I love your software, BTW. As for it being a CRT problem... maybe? If it is, it's happening on two entirely different monitors from different manufacturers, using different computers, different video cables, different power supplies... The only thing in common in my tests is the carts and again, it happens on more than one different kind of cart also. If I get bored maybe I'll try this out with a HDMI up-convertor. I have one, it just isn't set up right now. It might be interesting to see if it happens on a modern LCD.