<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">On 12/24/2014 11:41 AM, Stefan Scott
Alexander wrote:<br>
</div>
<blockquote
cite="mid:CAFwK6aud6SgVHn20W-MQcv7o+udFmYhN6Xhta5Rt_A2ZML0LYQ@mail.gmail.com"
type="cite">
<div dir="ltr">
<div>I was trying to learn more about the Grid1 demo, because it
seems like a very important example, but it's not always
straightforward to figure out what a lot of the code is doing
- so I had been hoping that even though UrWeb uses the MLton
whole-program optimizing compiler and there is no UrWeb REPL,
I might be able to use SMLNJ to interactively experiment with
individual functions to get a better understanding of what
each one of them does. But I guess this won't be possible.<br>
</div>
</div>
</blockquote>
<br>
Sorry, no interactive execution of Ur/Web code. Your strategy of
writing web apps to test features is the right one, IMO.<br>
<br>
Although, the type system here is so rich that you don't necessarily
need to <i>run</i> code! You can stop at type checking by passing
the '-tc' flag to 'urweb'.<br>
<br>
<blockquote
cite="mid:CAFwK6aud6SgVHn20W-MQcv7o+udFmYhN6Xhta5Rt_A2ZML0LYQ@mail.gmail.com"
type="cite">
<div dir="ltr">
<div>I thought I had a pretty good understanding of ML - but now
I see that all I have learned was the usual "toy" examples
from reading textbooks. When confronted with a real
application such as the Grid1 demo (a non-trivial example
using metaprogramming in UrWeb to implement an Ajax-y CRUD
interface for a 3-tier web app), I felt a bit overwhelmed.<br>
</div>
</div>
</blockquote>
<br>
FWIW, probably most of the intricacy that you're running up against
is found in neither ML nor Haskell. It takes time to get used to
the ways of dependent type theory!<br>
</body>
</html>