I enjoyed the Steve Gibson episode, and since he encourages us to get fluent (or at least convers...

[G+_Larry Weiss]

I enjoyed the Steve Gibson episode, and since he encourages us to get fluent (or at least conversational) with assembly language as a foundation, I'd encourage Coding 101 to run a series of shows on assembly coding, ideally with Steve Gibson as the Coding Warrior.

[G+_Fr. Robert Ballecer, SJ]

Hmmm... working with Steve Gibson on assembly.... 

 

/NoBrainer

 

:)

[G+_Nate Follmer]

Assembly causes brain bleeds... I'm game!

[G+_Lee Crocker]

I liked his comment that 8 bits wasn't enough to do much useful. I cut my teeth on the 6502 years before those newfangled 16 bit chips came out. And the project I'm working on now is an Atmel chip with 1k -- yes, that's 1024 bytes -- of RAM.

[G+_mike klaene]

The thing to keep in mind is that there is no 'one assembler' language. My first exposure to assembler was on an IBM/360 in 1969. In late '74 I learned assembler on a Singer 1500 w/4K of ram - just to move 8 bits you had to load the byte to the 'accumulator' first and then store it to the target location. This was on of the first machines with a 'stack' and the whole thing was implemented in TTL (no microprocessor). Next was 8086 on a Convergent Technologies system.

 

Assemblers assume that the programmer knows what he/she is doing unlike most of the newer languages that will try to protect you.

[G+_Larry Weiss]

I always found that floating point arithmetic was very hard to do with assembly.   I liked the languages like the original Microsoft C that allowed you to drop in some assembly in the middle of your C code so that you could get the best of both levels of language.

[G+_Lee Crocker]

Well, yeah, if you're using assembly, then you probably don't want floating point, and vice versa. Two of the most common mistakes I see novice programmers make is using floating point and strings for everything instead of really understanding the nature of their data and how to represent it efficiently.

[G+_Nate Follmer]

Lee Crocker Back when I first started in C++ I had a lab partner that would use doubles for EVERYTHING. "You never know how big you'll need your numbers to be" he'd say... Ugh! 

 

That same class, we did about 3 or 4 weeks of assembly. I made a calculator that could add, subtract, multiply and divide... Boy did I think I was awesome then. :) 

[G+_Larry Weiss]

Javascript only has one number type - double precision floating point numbers, following the international IEEE 754 standard  - http://www.w3schools.com/js/js_numbers.asp

[G+_Nate Follmer]

Larry Weiss That's true, but it's also very wasteful. There is trade offs with every language. It's nice in JS to not worry much about your number types, but at the same time it's hard to cut back on memory usage when you don't have many choices in terms of allocation.