Even if the old mouse is the right choice, I was thinking that one kinda has to know that to know how much time/money one is gonna put into this. I mean, if the problem is "I want a cheap mouse and am not too particular about the specifics", you can get a new wireless optical mouse for $5. It's just not worth putting much time or money into getting an old one working; that kind of price seriously constrains the solutions you might be willing to do.

If one specifically strongly prefers a particular piece of hardware, they might be willing to put a lot of money and effort into getting the thing working. People have built new controller boards for buckling-spring IBM Model M keyboards. People have spent hundreds of dollars to get a Model M (or a modern remake) going in a modern environment, even though one can get a rubber dome keyboard for $6. They aren't doing it because they're unwilling to pay for a new keyboard, but rather because they very much want a piece of hardware that works the way it used to. For people like that, a new $5 mouse is not a reasonable option.

I've been on both sides of that equation.

I personally profoundly dislike "virtual buttons" on trackpads, despite the market generally shifting in that direction and them being somewhat cheaper. I am willing to limit myself to getting Lenovo Thinkpads, because they're one of the very few laptops out there that – on some models – have trackpads with (1) physical buttons and (2) the three physical buttons that Linux prefers. If I had to pay $100 to get a trackpad with three physical buttons on a given laptop, I would, though that's far more than what a "virtual button" trackpad would cost. Hell, I'd probably do $300.

On the other hand, I don't care much about mice. As long as it's got five buttons, a scroll wheel, and a USB connection, every mouse I've ever had has worked pretty much fine. I tend to just get whatever's cheapest on offer when I get a replacement mouse (my most recent mouse, a Razor, being an exception; my monitor refresh rate is now above the typical USB mouse polling rate of 125 Hz, and I wanted a mouse that would report position at something above my monitor refresh rate).

And I know that it'd be pretty hard to convince me otherwise on either point. Like, I've used expensive mice and cheap mice and they just haven't made much difference. On the other hand, I have used and very much dislike virtual trackpad buttons. So I wouldn't want to tell someone to go either route across-the-board without knowing where they're coming from on this.

Okay.

Personally, I would just get a new mouse, because you can definitely get a new ambidextrous wireless mouse with side buttons for a lot less than I'd consider the work involved with this being worth it to me.

If you really strongly want to do this, two possible routes:

• You don't actually need the PS/2 plug, just the chips and electrical connections. Dissassemble the PS/2 dongle and a PS/2 to USB adapter and rewire them into a smaller form factor that you're comfortable carrying around with your laptop.

• Rip out some or all of the internals of the mouse and replace them with internals from a newer, Bluetooth-compatible mouse that uses the same encoders that are encoding the ball's movements, assuming such a thing is out there. That won't give you the same mouse, but it'll give you something that has the same external feel. The shape will be the same, it'll click the same way, and it'll have the same rolling feel when moving.

Finally, a partial solution: it sounds like your concern with the dongle plus adapter is mostly that you don't want to carry more components with your laptop.

If you regularly carry other USB peripherals, you can just bundle them up with a USB hub and the dongle and then just have one box that you plug into your laptop. That might mitigate the issue of carrying the dongle, or might not. Depends on what it is that's problematic for you – a lot of parts or the bulk of extra components.

At one point in the past, I had an optical trackball with a translucent pink ball that I wanted to use in a dark room. The red LED was really disruptive, and while an opaque ball would have mitigated it a bit, it still would have shown around the edges. The fact that it went to sleep when being inactive and then woke up when moved, changing in brightness, made it annoying as hell. Those optical sensors can also see into the near-infrared, so I took a gamble and ripped out the red LED and put in an infrared LED. It mostly worked, except the sensitivity was reduced enough that when it was in sleep mode, any movement wasn't enough to reliably wake it up. I dug up the datasheet for the chip, found the pin that sent it into sleep mode, and wired it to the high voltage pin so that it never slept. That worked fine. All that being said, while it was satisfying to have it work, the only reason I was willing to to it was because I couldn't just buy an optical, off-the-shelf trackball at the time that didn't shine a red light all over. It simply would have made no financial sense to put the time and money into doing the modifications otherwise. In this situation, if there's an off-the-shelf product that does what you want, I'd probably drop the mouse off with Goodwill or similar if you don't want to toss it in the trash – maybe someone out there does want a mint condition PS/2 wireless mouse – and just buy the new mouse.