Jump to content


  • Content Count

  • Joined

  • Last visited

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. jonpurdy

    Apple Adjustable Keyboard (M1242) keys repeating

    Just wanted to update this post in case anyone is following my path to resurrecting an Adjustable. I ended up finding problems with a bunch of other keys that couldn't be resolved by swapping out springs and such. So I ended up desoldering known good switches from my parts keyboard (A2) to my good keyboard (A1). To do this: Map out the known good keys on your parts keyboard. Make sure they work reliably every time the key is pressed, and make sure they feel correct relative to other keys. Desolder the good key switches. To desolder switches, you just need to take the top part of the keyboard off, no need to take the bottom apart. And watch out for this hook; it keeps the board in place. When removing the actual board where the keys are soldered to, remove the ribbon cable so that you don't accidentally bend or destroy it while desoldering. It's stuck in there pretty tight, but with equal pressure on both sides it can be removed (and reinserted). Each switch is connected by 4 contact points. Desolder these and the switch will come off. On your good keyboard, you'll need to desolder the broken switches. Once this is done, take a good switch and place it into the holes of the target key. Solder a couple of points. Then make sure that the switch is aligned correctly (it may not be all the way in). If correct, solder the rest of the points. Then just verify that the new switch(es) work once they're all soldered in. I will update this with a couple more photos and update it for clarity if there are any questions.
  2. I should mention that the day after I tried the CF in my PDQ, everything just started working. It's a 16GB Sandisk Extreme IV partitioned into two 8GB HFS+ partitions with the System Folder on the first partition. It's been working fantastically well for the past few weeks, which is the opposite of my previous experience with my 13" Wallstreet. Just wanted to mention this since it's counter to my previous experience that I posted before.
  3. That seems cool too. Perhaps I could just provide support for using macproxy and other services (like @TechEdison's nocss) without actually integrating an entire macproxy module?
  4. I gave this a try and registered a couple of test accounts, but I can't seem to log into either of them. Response is an "Incorrect login." page being returned. Any ideas?
  5. I've added @K Trueno's changes so that entries can redirect to loband. In the develop branch, I've also added a setting (among others) to enable or disable loband redirection since I was having trouble getting most loband links to actually work (like, the service itself seems to not work with most sites). I've also added a setting (in settings.py) for image width (I use 320 px wide images on older machines and smaller monitors, and 800 wide on my Wallstreet). Also number of entries per page, since my 9500 with iCab crashes with more than a few large entries (even with 50MB RAM allocated). Finally, image conversion is now threaded, so that images can get converted without waiting for others to finish. Most of these changes are in the develop branch since I wanted to test threading a bit more first before merging.
  6. jonpurdy

    Apple Adjustable Keyboard (M1242) keys repeating

    I was able to fix most of the bad keys on the good keyboard. Swapped a few springs out and one metal piece and the keys were back to normal. There's still one key that's giving me issues ("w") and I intend on desoldering the switch and replacing with a good one. I also gave each side of the keyboard a very detailed clean with, compressed air, q-tips, and alcohol. To protect the ribbon cable that connects each side from damage, I filed off a few plastic tabs. These were pressing down on the cable. I've attached a photo of the switches for future reference.
  7. Those images made my eyes hurt... The good news is that you can keep the near mint case and RAM and any other parts; you'll just need to find another II or IIfx board. But unfortunately that looks pretty bad; you'd need some serious electronics skills to repair the traces on that. If you're recapped stuff before you've got a fighting chance though.
  8. jonpurdy

    Apple Adjustable Keyboard (M1242) keys repeating

    Update: was motivated to look into this further this evening. Did some more testing and found that it was specific key switches that were the problem. Some of them (like the "u" key) are softer and activate with less pressure than others. I found that totally disabling key repeat in the Keyboard Control Panel did help. However, I took the chance and swapped the ribbon cable to my first M1242 (the one with the broken cable). Much better performance and the issue with the repeat keys basically went away. Of course, it turns out this keyboard has some issues of it's own: certain key caps require MORE pressure than others, so instead of repeat characters I get no characters at all I think the next plan is to figure out which key switches on each are good and bad, then remove or desolder the bad ones and replace them with good ones from the spare keyboard. Lots of links available for this keyboard for anyone else interested: http://www013.upp.so-net.ne.jp/cyberdog7/Adjustable.html https://deskthority.net/wiki/Apple_Adjustable_Keyboard https://applefan.tistory.com/41
  9. TL;DR: keys repeat sometimes, can it be fixed by cleaning switches or is it a design flaw? Full post: So I've always wanted this keyboard ever since seeing it in The Net (along with the 8100, PB 540c, etc.). Picked up a non-working but cosmetically great one a few months ago, and a working but yellowed one a month or so ago. The non-working one had a split ribbon cable (the one connecting both sides of the keyboard). Which is unsurprising since it's very tight and disassembling the keyboard seems to tug on the cable enough that it could split if one wasn't careful. The working one is mostly fine, but I've noticed that some keys have an issue where they repeat (I type "r", but end up with "rr" or "rrr"). Or if tying "the brown fox" I'll get "the ebrrown fooxo" (slightly exaggerated example, but shows that the extra presses still get inserted after I've typed the next character). This typically happens when typing at high speed and not as much at low speed. It's isolated to a few keys, but I've only tested it for a few minutes at a time so the issue could be more than those keys. If I was to wager a guess, it seems like the key switch is internally sticky, so electrical contact would be held for slightly longer than the split second that my finger hits the key, causing the repeat. (It's not a software setting, checked Key Repeat.) A second guess would be that the keys aren't debouncing correctly (google "keyboard debounce" for more info). Is this something that can be fixed? I was planning on getting the key caps off all the keys and cleaning the switches out individually to see if this fixes the issue. I could also pull the keyboard switch parts and circuit boards and plastics from the non working one and use the ribbon cable from the working one. Though I've heard that these keyboards are somewhat flawed from the factory and a lot of them had to be RMA'd when new.
  10. That's great! I created an issue for this and will merge the patch when I get a chance this week: https://github.com/jonpurdy/mosaicbin/issues/1 I made a lot of assumptions, so I presume there will be more issues like this. Especially cool to see the video of the SE/30!!
  11. That's great! Didn't even think of loband.org, but does it involve rewriting every <a href> to go through loband? If so, that's a great idea. Looking forward to the PR. What changes did you have to make to get it going? I've only tested this on a few feeds (simple and complex) but it still needs a lot of work. Figured I'd see if folks were interested first before putting a lot more into it. I'd like to restructure some of the code, plus have test coverage, plus handle errors properly rather than spit out debug info. And then eventually a bit more complex layout (could replicate the Feedbin UI with frames, lol).
  12. For those interested, I have released the source for "Mosaicbin" to Github: https://github.com/jonpurdy/mosaicbin Over the past week I've added pagination, style cleanup, and a lot of fixes. But I'd say the biggest feature is image conversion. When an entry is requested that has image links, Mosaicbin will download the image, convert it to fit within 640 pixels wide, save as a JPEG, and then serve that converted file locally to the client. (No more missing images from HTTPS URLs!) I've even emailed Icon Factory and licensed the icon from them (which I originally remember coming on a MacAddict CD around 1997), so Mosaicbin has a period-correct logo. If you're interested in trying it out in it's current state, I recommend doing so either in a Vagrant machine or a VPS, so you're not messing around with your local Python install. But it does work locally as well. Installation instructions are on the Github link above.
  13. So the PDQ I got is great: 300MHz, 512MB RAM, 8GB drive, and a DVD-ROM drive (unexpected surprise). Unfortunately, it too has the exact same issue as my 13.3" 250MHz. I didn't test the 44-pin adaptor, but I did test the CF card in a PCMCIA adaptor and it can read and even boot from the CF card, but as soon as data is written it hangs (clock stops, mouse can move, but need to hard reset). I tossed all my CF cards years ago (I think mostly 512MB and 1GB anyway), so I don't have any more to test with. I am considering getting an mSATA adaptor with a 32 or 64GB m.2 drive, but I'd like to wait a bit to see how much I actually use this machine (since I'm already in about $220CAD after buying two of them and the accessories). I may just copy the system folder and whatever I'm doing on it to a RAM disk and work off that (since I got the CF card for silence, not speed).
  14. I get the same issue when plugging the adaptor into my 250MHz Wallstreet, but it lasts for about 5 seconds then turns off. Then I can boot the machine normally. I've had the processor card out a few times due to other maintenance but it hasn't changed the result. I will try resetting the PMU this evening and report back.
  15. I'm actually having some trouble as well on my 250MHz Wallstreet (non-PDQ) with a Sandisk CF → 44-PIN IDE dual drive adaptor, as well as when the CF card is in a PCMCIA adaptor. The system hangs whenever I try to write data to it (mouse cursor can move, but clock and everything else freezes) using OS 8.6 (don't have X and wasn't planning on using X, but might do so for testing). I've actually got a PDQ waiting for me at home (shipped and delivered just today) so I'll update this post with my results after trying this machine out.