Hacker News Reader

Comment by anonymousiam | Hacker News Reader

I've got a Model 102 which is pretty much the same computer, only manufactured later. I mostly used it as a terminal for UHF amateur packet radio (9600bps FSK) back before there were any turnkey solutions for that.

Rumor has it that the Model 100 contained the very last Microsoft product in which Bill Gates wrote any code.

View on HN

It's not a rumor!

> "Part of my nostalgia about this machine is this was the last machine where I wrote a very high percentage of the code in the product. I did all the design and debugging along with Jey." - billg in an interview with the National Museum of American History.

https://americanhistory.si.edu/comphist/gates.htm#tc35

Not to be all megalomanic.. but I wonder what will be the last machine I personally wrote code for. Code we write nowadays is mostly targeted at virtual machines, or even if it's low level at some idealised hardware (apparently based on the PDP-11?). Or maybe we'll go back to be coding for machine learning accelerators or whatever lies in the future that constrains us to specific optimised hardware.

You’ll end up like Firefly, where the trash collector was running Windows 98. Or was it ME?

The only issue with the 102 is that it uses surface mount components throughout, while in the 100 they are only in the display module, which makes 100's easier to repair, but YMMV.

Surface-mount components are *much* easier to repair than thru-hole components, as long as they aren't BGAs or TQFPs or other extremely high-density parts or parts with solder balls underneath. The 102 is so old it would likely only have 50-mil pitch chips, which are *far* easier to swap out than any DIP chip: all you do is blow some hot air on them and them come right off without any damage. DIP chips generally need to be destroyed to remove them without damage to the PCB.

It depends on the caps though - early surface mount ones can be terrible.

DIP is quite easy to desolder with a desoldering iron like this one https://www.ebay.com/p/2254498136 . You just have to avoid heating adjacent pin. For a DIP-8 like this:

I would desolder in that order 2, 5, 1, 6, 8, 3, 7, 4.

With an SOIC chip, I don't worry about "order": I just hold tweezers on the chip, blow some hot air, and it comes off in about 1 second. What you're describing is orders of magnitude more difficult and time-consuming.

Thank you for this thread, as it makes my fussing with soldering today to fix some shorts in a couple guitars (my own fault from poor soldering years ago) sound quite pedestrian so I should aim to do as best as possible, cuz y’all out here be doing advanced EE projects!!

Does this risk loosening up some of those chips' neighbors?

Yes, definitely, if they're really close together. But "loosening up" isn't a problem with SMT; as long as the legs are adhered to the pads with the proper amount of solder, it's correct. The risk is moving adjacent chips around so they're no longer properly centered on their footprints, so using the hot air gun does require some finesse: use enough heat and air velocity, combined with the distance of the tool to the board, to remove the chip you aim for, without affecting the adjacent chips.

Reply on news.ycombinator.com