-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathBUGS.html
260 lines (203 loc) · 6.94 KB
/
BUGS.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
<?xml version="1.0" encoding="US-ASCII"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<!-- $Id: BUGS.html,v 206.3 2003/09/29 12:46:38 kiniry Exp $ -->
<!--
OBJ3 2.06,2.08,2.09 Copyright (c) 2000-2003 Joseph Kiniry, Joseph Goguen
OBJ3 2.05 Copyright (c) 2000 Sula Ma, Joseph Kiniry, Joseph Goguen
OBJ3 2.04 Copyright (c) 1988,1991,1993 SRI International
TRIM Copyright (c) 1994,2001 Lutz Hamel
All Rights Reserved
-->
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<title>OBJ3 version 2.09 Bug List</title>
<link rel="stylesheet" type="text/css" href="docs/base.css" />
<link rev="made" href="mailto:kiniry@acm.org" />
<meta name="keywords" content="OBJ3" />
<meta name="description" content="" />
</head>
<body>
<h1 align="center">OBJ3 Bug List</h1>
<hr />
<!-- ============================================================ -->
<p> This document is a summary of all known bugs in this version of OBJ3.
To report a new bug, please send an email to <a
href="mailto:obj3-bugs@kindsoftware.com">obj3-bugs@kindsoftware.com</a>
</p>
<h2>OBJ3 2.09 [RedHat 8/glibc-2.3.2-4.80.6] (GCL+CMU)</h2>
<p> No known bugs. </p>
<h2>OBJ3 2.08 (TRIM)</h2>
<p> While not a bug per se, it is important to note that the TRIM compiler
has several limitations. In general, it cannot be used on arbitrary OBJ3
code because it does not support the following OBJ3 features: built-ins,
A/C rewriting, evaluation strategies, and non-left-linear equations. Even
so, the speed increase afforded by TRIM is quite welcome for evaluations of
high computational complexity. See the TRIM paper included with the
distribution for more information. </p>
<h2>OBJ3 2.04</h2>
<dl>
<dt> in(put) cannot accept filenames with a '_' in them. </dt>
<dd> noted: ?, fixed: 1993-04-07 by A. Stevens </dd>
<p>
No details available.
</p>
<dt> Module processing crash. </dt>
<dd> noted: 20/4/93, fixed: ?
<p>
Description: If an attempt is made to instantiate a
parameterized module with complete garbage, especially if
insufficient parameters are given, the system drops out with a
LISP exception. Missing error check someplace?
</p>
<p>
No fix currently available.
</p>
</dd>
<dt> Parser Limitations not commented on. </dt>
<dd> noted: 1993-04-20, fixed: ?
<p>
The parser is unable to handle operator syntax involving
brackets. For example,
</p>
<pre class="OBJ3">
op (.All _ ) _ : ...
</pre>
<p>
However, operator <em>declarations</em> of this form are
silently accepted.
</p>
<p>
No fix currently available.
</p>
</dd>
<dt> Problems with comm assoc matcher. </dt>
<dd> noted: 1993/03/18, fixed: ?
<p>
Summary. For example,
</p>
<pre class="OBJ3">
obj MATCH is
sorts S1 S2 .
subsort S2 < s1 .
op _+_ : S1 S1 -> S1 [assoc] .
op _+_ : S2 S2 -> S2 [assoc comm] .
op a : -> S2 .
op b : -> S2 .
op c : -> S1 .
op _*_ : S1 S1 -> S1 .
var x : S1 .
eq x * x = x .
endo
red (a + b)*(b + a). ***> should be: a + b
red ((a + b) + c)*((b + a)+ c). ***> should be: a + b + c
</pre>
</dd>
<dt> Problem with comm assoc id matcher. </dt>
<dd> noted: 1993/04/28, fixed: ?
<p>
If you open . a module instantiated with a module introducing an
operator needing matching acz matching, attempting to use the
operator causes a segmentation fault. Running within the
interpreter gives a more informative response.
</p>
</dd>
<dt> From original obj3/lisp/basic/counter.txt (unconfirmed claim): </dt>
<dd>
<pre>
Joseph,
I think I may have found a bit of a bug in the current lecture notes.
In section 5.5 you claim that if a Sigma term-rewriting system A with
no void sorts is confluent then so is A with fresh constant (i.e. A
treated as a Sigma U X term-rewriting system).
Here's what seems to be a counter-example to this claim.
Consider
obj A is
sort Nat .
vars X Y Z : Nat .
op 0 : -> Nat
op s : Nat -> Nat .
op _+_ : Nat Nat -> Nat .
op t : Nat Nat Nat -> Nat .
eq 0 + X = X .
eq s(X) + Y = s(X + Y) .
eq X+(Y+Z) = t(X,Y,Z) .
eq t(s(X),Y,Z) = s(X + (Y+Z)) .
eq t(0,Y,Z) = Y + Z .
endo
Termination is pretty trivial: e.g.
p(0) = 2
p(s(a)) = 1 + p(a)
p(a+b) = 4 + p(a)p(b)
p(t(a,b,c)) = 2 + p(a)(4 + p(b)p(c))
We now prove A is confluent. Let a,b,c,d,e etc denote A-terms
There are two critical pairs:
a + (b + (c + d)) 1=> a + t(b,c,d) (i)
1=> t(a,b, c + d)
and
a + (s(b) + c) 1=> t(a,s(b),c) (ii)
1=> a+s(b+c)
Now, it is easy to see that for any term t we either have
t *=> s(t') or a *=> 0
(Simply rewrite using an most-nested-first ``layered'' rewriting strategy.
I.e. alternately rewrite most-nested +'s, then most-nested t's)
Thus we can easiy show that a + t(b,c,d) and t(a,b, c + d) are
both rewriteable to one of
s(a' + s(b' + (c + d))) or s(b' + (c + d)) or s(a' + (c + d))
or (c + d)
depending on which of the two cases above hold for a and b .
Similarly, we can easily show t(a,s(b),c) and 1=> a+s(b+c) are
both rewriteable to one of
s(a',s(b+c)) or s(b + c)
depending on which of the two cases above holds for a.
Hence A is locally confluent and hence confluent.
Now let X = { a,b,c,d } and consider A as a Sigma^{A} U X
term-rewriting system.
Adding a,b,c,d as fresh constants we obtain
a + (b + (c + d))) 1=> t(a,b, c + d)
1=> a + t(b,c,d)
with both alternate rewritings normal forms. Hence A treated
as Sigma^{A} U X term-rewriting system is NOT confluent.
Andrew
</pre>
</dd>
</dl>
<hr />
<!-- ============================================================ -->
<!-- Toolbar at bottom of all major top-level pages -->
<p align="center">
[ <a href="index.html"> Index </a> ]
[ <a href="README.html"> Readme </a> ]
[ <a href="FAQ.html"> FAQ </a> ]
[ <a href="RELEASE_NOTES.html"> Release Notes </a> ]
[ <a href="TODO.html"> To-Do List </a> ]
[ Bug List ]
[ <a href="source/README.html"> Source </a> ]
[ <a href="docs/index.html"> Papers </a> ]
[ <a href="docs/obj_bib.html"> Bibliography </a> ]
[ <a href="LICENSE.html"> License </a> ] </p>
<hr />
<!-- ============================================================ -->
<p align="center">
<a href="http://www.anybrowser.org/campaign/">
<img src="docs/graphics/anybrowser.png"
height="31" width="88" border="0"
alt="Best Viewed With Any Browser." />
</a>
<a href="http://validator.w3.org/check/referer">
<img src="docs/graphics/vxhtml10.png"
height="31" width="88" border="0"
alt="XHTML 1.0 Checked!" />
</a>
<a href="http://jigsaw.w3.org/css-validator/check/referer">
<img src="docs/graphics/vcss.png"
height="31" width="88" border="0"
alt="CSS, Level 2 Checked!" />
</a>
</p>
<address>by Joseph R. Kiniry <kiniry@acm.org></address>
<!-- hhmts start -->
Last modified: Mon Sep 29 12:07:32 CEST 2003
<!-- hhmts end -->
</body>
</html>