Re: [Gnu-arch-users] Re: Tla spork

[Zenaan Harkness]

On Fri, Aug 27, 2004 at 07:20:33PM -0400, Miles Bader wrote:
> On Fri, Aug 27, 2004 at 07:09:32PM -0400, James Blackwell wrote:
> > For the sake of argument, lets assume your point is true. In fact, I'll
> > go you one step better and assume for the purpose of this argument that
> > prefix is *better*. We still loose, because the perception is there.
> 
> This is not clear.  I'd guess that a majority of people actually fall into
> two camps: those who have never heard of sexps (and thus have no existing
> prejudice against them), and those who both know of and can deal with them
> (after all it's hardly the case that what Tom is doing is from the esoteric
> fringe, it comes from a long and well respected tradition in CS).

I think you missed James' point - those in the first camp, who've never
heard of sexps (as this assumption goes), have a reaction when they
first see prefix-notated mathematical expressions. This reaction is not
what I'd call positive :)

I believe I had this reaction rather strongly - I perceived it as
difficult, challenging to grasp, requiring more than a non-trivial
effort to 'get my head around'.

So, I as hypothetical new programmer attacking tla, have decided I'll
 - put in the time to learn arch commands
 - read the tut/ browse the lists/ lurk
 - start using it for a project/ get some experience
 - ask for help when I get stumped
 - persist with the learning curve of tla

and then I come across a new language for configuration. I'm required
to learn prefix notation, and _my perception_ is that the learning
curve suddenly went up by [big amount/ double/ exponential].

That's the common perception.

As I said of myself, I tried to launch myself into the guile tutorial at
least twice over the last few years, for at least an hour on each of
those occasions.

I also read a bit about different scheme dialects, haskell, and did
some general browsing on the topic.

Before the "transformation introductions" given on this list today,
I never got to the point of feeling it might not be too hard to
'get it'.

So, yes there are two categories, and I think most people fall into
the first (have a lifetime of algebraic notation experience), _and_
that the mojority of these will find a new "grammatical" syntax
to be one too many things to learn.

Unless we find a way to gently introduce people like me into the
new syntax, and minimize gratuitous parens etc as Andrew Suffield
has been showing is possible.

Of course we can agree to disagree too...

cheers
zen