Also running python on a microcontroller to do cryptography is fucking insane.
I feel like I should confess to snubbing this (and some other) hardware wallet projects for the fact it used micropython. I think it's hard to draw the line between systems programmer snobbery and good advice... and I feel a little like a cop guilty of stop-and-frisk on the basis of skin color.
Because in general the ideas necessary to produce reliable software aren't well understood or agreed on there is a risk of letting style preferences which are only correlated with good engineering but aren't causative of good engineering get mistaken-- and this can cause errors in both directions, both mistaking stuff as good because it uses the "right" tools, or mistaking something as bad because it doesn't.
Comments
I feel like I should confess to snubbing this (and some other) hardware wallet projects for the fact it used micropython. I think it's hard to draw the line between systems programmer snobbery and good advice... and I feel a little like a cop guilty of stop-and-frisk on the basis of skin color.
Because in general the ideas necessary to produce reliable software aren't well understood or agreed on there is a risk of letting style preferences which are only correlated with good engineering but aren't causative of good engineering get mistaken-- and this can cause errors in both directions, both mistaking stuff as good because it uses the "right" tools, or mistaking something as bad because it doesn't.
I think some things telegraph competence and some the opposite -- and our experiences over the years tend to shape what we see that way.