> The physical work itself is then completely different to software engineering.
It’s funny, but I often think how similar they are after doing a project. Maybe I haven’t written enough software for it to feel routine, but I often enjoy the unique and unexpected challenges in each that require creative thinking to solve.
One interesting woodworking tip I carried over was “Know your tolerance”, i.e. does it need to be accurate to within an inch, an eighth, etc. I often apply that when working on software now (i.e. how “perfect” and bug free does this component really need to be?)
People often say that “code is malleable” and you can change anything, whereas working with wood and tools is unforgiving. I actually think they’re comparable again, in that your time is what costs the most, and you can fix anything with minor materials cost compared to the time cost. Throwing away a piece of wood I worked on for a while and messed up isn’t much different than deleting a working branch that didn’t pan out. (Working with a contractor makes this more apparent. They’ll happily change almost anything after they’ve done it if you’re willing to pay for it).
Oh, and another similarity: I start lots of projects that interest me, but the ones that get finished are the ones my boss (the wife in this case) keeps hassling me about ;-)
Wood is malleable as well. Building a swing set for my children I learned the trick of taking the bow out of a warped 2x4 by applying leverage and screws as I assembled sections.
Better glue and screw that or it will surprise you at some point (unless you used screw overkill :) ). Better still: ripsaw, reverse one part then glue back together. (a lot more work, but that way the internal stresses in the wood will cancel out).
Would you use screws over internal dowels/biscuits for such a project? curious about best practice for something that is going to be outdoors 24x7. i agree that you need something to counteract the shear forces from the wood trying to bend back to its original warp. Not arguing, just intellectually curious.
In something like a swingset my first consideration would be safety, make everything from stuff that I'm 100% sure isn't going to surprise me one way or another, and given that it is a dynamically loaded structure with kids doing 'creative' stuff with it I'd aim for overkill rather than getting it 'just so' or working for something that looks good but that may not be strong enough.
If it's going to be outdoors in the weather over a very long period of time I would not count on glue (unless you have very good process control and know exactly what the long term properties will be) but I'd use steel to connect the various wooden bits, typically hole through (large diameter), then epoxy the hole and then drill again (smaller diameter). That stops the wood around the hole from rotting away. You also want to take care in what kind of wood you use in the first place a nice wood for stuff like that is Tamarack, which doesn't rot.
another principle applicable I think is knowing what may or may not change in the future. being able to switch from legs to wheels is useful, so maybe put an interface in between. being able to change a table into a bed - not useful and will sink your project.
It’s funny, but I often think how similar they are after doing a project. Maybe I haven’t written enough software for it to feel routine, but I often enjoy the unique and unexpected challenges in each that require creative thinking to solve.
One interesting woodworking tip I carried over was “Know your tolerance”, i.e. does it need to be accurate to within an inch, an eighth, etc. I often apply that when working on software now (i.e. how “perfect” and bug free does this component really need to be?)
People often say that “code is malleable” and you can change anything, whereas working with wood and tools is unforgiving. I actually think they’re comparable again, in that your time is what costs the most, and you can fix anything with minor materials cost compared to the time cost. Throwing away a piece of wood I worked on for a while and messed up isn’t much different than deleting a working branch that didn’t pan out. (Working with a contractor makes this more apparent. They’ll happily change almost anything after they’ve done it if you’re willing to pay for it).