So I was riding with a friend the other day and I was commenting about how the bike always pulled to the right. We initially thought it might be the rear brake cable tugging on the handle bars. We stopped and inspected the bike and found that the front tire was actually sitting slightly off center to the frame. It turned out that the fork was bent to the side. So we removed the wheel and leveraged the fork and cold set it as best we could. For eyeballing, we got it pretty damn straight...straight enough I could ride without my hands on the handlebars (which was impossible to do before). But still, this seemed like a sloppy repair method. Is there a better way to straighten a fork so that it is 100% accurate?

I also recently used Sheldon Brown's string method to widen the spread between the dropouts to accommodate a wider wheel. The alignment seems very good, but I'm curious if there's a better way. Are there machines that can do this stuff with precision accuracy, and if so, where can I find them? Thanks.

Certainly.
https://www.jastein.com/Html/Tools_for_Frames_Forks.htm

Very nice. Is calling the only way to get prices?

This is probably the cheapest you will find it.
https://biketoolsetc.com/index.cgi?id..._id=SN-FCG/SET
I notice that it is out of stock. It's also out of stock at UBS, so maybe Mr Stein hasn't done a production run on them for a while.
You could e-mail Jim Stein and ask. He responds to enquiries very promptly in my experience.

Whoa. That's more expensive than a really nice new frame. I guess it's only good for someone that's doing that for a living.

Yeah, not really hobbyist stuff. If it's a one time repair you're looking for, look for your nearest frame builder with the equipment.

....or just use the string method and some care. You'd be amazed at how well basic techniques can be when used with a bit of planning and care taken to eliminate possible errors.

When doing the string method the big problem is holding it still. Sticks of wood clamped to the frame can be used to pinch down the string while measuring so that the settings don't move. A slip of paper pinched between the string and frame and then moving the string only just far enough to let the paper slip out can be a guide for achieving the same string to tube spacing on each side. By thinking about such stuff you can get your frame aligned with string, ruler and a few bits of stuff to a degree that is more than adequite. An error of one string's width means less than nothing. It's the 1/8 inches at a time that you need to correct. And the string method with some tricks can easily best that requirement if done right.

I use the Park FT-4, but it might not be available anymore -- I can't find it on the Park web site.