Would be useful if, in chapter 1, section "Lisp is a huge language", the author mentioned approx how many pages of CLtL is devoted specifically to the core of the Language. Comparison to the size of Scheme would be helpful as well (one regularly hears that the size of the entire Scheme R5RS spec is smaller than the index of the CL book).
It takes 100 lines of bloated Common Lisp to do the same thing as 100 pages of tiny, elegant Scheme.
The standardized parts of Scheme are 45 pages. The parts of Scheme necessary to get work-done is the size of Common Lisp; without that pesky document from ANSI and ISO, and the support of 5+ vendors, or the large body of existent code that only comes with 30 years of cooperative work by a community with industry, government and academia.
I happily code in Scheme as well as Common Lisp; both languages have their places, and both need to be learned for the unique pleasure each has. Btw, Scheme SRFIs are some of the most enjoyable pieces of literature one has the fortune to read on Software Engineering.
No. I would use Scheme where I need to have absolute control over the application along with the implementation. That includes any type of hardware hacking, full-stack networked distributed apps, mobile apps, gaming platforms, experimenting on new OSes, etc.
Common Lisp is better on maturer platforms, while Scheme is like a little wedge that you can slip into cracks to open it wider, before you sneak in more and more Lisp. See how Kawa does Java applets while Common Lisp does the web server, etc.
Common Lisp is the 1-wood/driver, Scheme is the putter. To use a golf analogy. (and sadly, many of the people that advocate scheme-everything don't have messy problems; their ball is already on the green, so to speak.)
This discussion seems to focus on the plain spartan Scheme standard, when -- comparing to CL -- it should focus on an implementation that contains standard-ish libraries as well.
Although CL might be the whole well-stocked golf bag (1-wood, 2-wood, irons, wedges, putters, ball-cleaner, radio...), wouldn't, say, Chicken be a smaller but still fairly well-rounded set?
"It's not quite true that no mass market product uses Lisp. Microsoft's "Bob" environment for naive computer users was developed (and delivered) in Lisp."
7 comments
[ 6.1 ms ] story [ 28.2 ms ] threadThe standardized parts of Scheme are 45 pages. The parts of Scheme necessary to get work-done is the size of Common Lisp; without that pesky document from ANSI and ISO, and the support of 5+ vendors, or the large body of existent code that only comes with 30 years of cooperative work by a community with industry, government and academia.
I happily code in Scheme as well as Common Lisp; both languages have their places, and both need to be learned for the unique pleasure each has. Btw, Scheme SRFIs are some of the most enjoyable pieces of literature one has the fortune to read on Software Engineering.
What places? Common Lisp for production and Scheme for learning? Thanks.
Common Lisp is better on maturer platforms, while Scheme is like a little wedge that you can slip into cracks to open it wider, before you sneak in more and more Lisp. See how Kawa does Java applets while Common Lisp does the web server, etc.
Common Lisp is the 1-wood/driver, Scheme is the putter. To use a golf analogy. (and sadly, many of the people that advocate scheme-everything don't have messy problems; their ball is already on the green, so to speak.)
Although CL might be the whole well-stocked golf bag (1-wood, 2-wood, irons, wedges, putters, ball-cleaner, radio...), wouldn't, say, Chicken be a smaller but still fairly well-rounded set?
What exactly are arguing about then?
...