Changeset 11949


Ignore:
Timestamp:
05/24/09 12:09:53 (14 years ago)
Author:
ehuelsmann
Message:

Move more from the front page to the FAQ.
Update the headers.

Location:
public_html/staging
Files:
3 edited

Legend:

Unmodified
Added
Removed

  • public_html/staging/contributing.shtml

    r11937 r11949  
    77  <link rel="stylesheet" type="text/css" href="style.css"/>
    88  <meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1"/>
    9 <style type="text/css">
    10 td { font-size: 90%; padding: 0 5px 0 5px }
    11 th { font-weight: bold; align: center; padding: 0 5px 0 5px }
    12 dt { font-weight: bold }
    13 dd dt { font-weight: bold; font-style: italic }
    14 </style>
    159</head>
    1610
    … …  
    2317
    2418<div style="margin-left:auto;margin-right:auto;width:20cm">
     19
     20
    2521</div>
    2622

  • public_html/staging/faq.shtml

    r11948 r11949  
    44<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
    55<head>
    6   <title>FAQ: Armed Bear Common Lisp (ABCL) - Common Lisp on the JVM</title>
     6  <title>FAQ: ABCL - Common Lisp on the JVM</title>
    77  <link rel="stylesheet" type="text/css" href="style.css" />
    88  <meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1" />
    9 <style type="text/css">
    10 td { font-size: 90%; padding: 0 5px 0 5px }
    11 th { font-weight: bold; align: center; padding: 0 5px 0 5px }
    12 dt { font-weight: bold }
    13 dd dt { font-weight: bold; font-style: italic }
    14 </style>
    159</head>
    1610
    1711<body>
    1812 <div class="header">
    19    <h1>FAQ: Armed Bear Common Lisp (ABCL) - Common Lisp on the JVM</h1>
     13   <h1>FAQ: ABCL - Common Lisp on the JVM</h1>
    2014 </div>
    2115
    … …  
    2721  <li><a href="#general">General</a>
    2822  <ol>
     23    <li><a href="#what">What is ABCL?</a></li>
    2924    <li><a href="#license">What license is used for ABCL?</a></li>
    3025    <li><a href="#bug-reporting">How/Where should I report bugs?</a></li>
     26    <li><a href="#speed">Is ABCL faster or slower than implementation XYZ?</a></li>
     27    <li><a href="#qa">What is the quality of the implementation? How can you tell?</a></li>
    3128  </ol>
    3229  </li> <!-- general -->
    3330</ol>
    34 </div>
    3531
    3632
    3733<div class="h2" title="general" id="general">
    3834<h2>General</h2>
     35
     36<div class="h3" title="what" id="what">
     37<h3>What is ABCL?</h3>
     38<p>ABCL is an implementation of the full Common Lisp specification, with
     39the exception of the implementation of the long form of
     40DEFINE-METHOD-COMBINATION.</p>
     41
     42<p>Unfortunately, the CLOS implementation is not accessible through
     43a MOP (MetaObject Protocol). Any contributions in this area would
     44be greatly appreciated, ofcourse.</p>
     45
     46<p>One thing which is considered almost standard lisp - because all
     47implementations deliver it - is "Gray streams". Unfortunately ABCLs
     48version is broken [as per 05/2009]. It should be noted this is by no
     49means the final state of affairs, though, merely a warning that one
     50can't depend on this feature at the moment.</p>
     51
     52</div>
    3953
    4054
    … …  
    6175
    6276<p>If you found a bug which is not on the list, or you want to stress
    63 the importance of one that is, please mail our <a href="https://lists.sourceforge.net/lists/listinfo/armedbear-j-devel">mailing list</a> about it.</p>
     77the importance of one that is, please mail our <a
     78href="https://lists.sourceforge.net/lists/listinfo/armedbear-j-devel">mailing list</a> about it.</p>
    6479</div>
     80
     81
     82<div class="h3" title="speed" id="speed">
     83<h3>Is ABCL faster or slower than implementation XYZ?</h3>
     84
     85<p>General comparisons are hard to make, the relative speeds depend
     86on a lot of factors. For example timing outcomes of specific bits
     87of Java may have different timings depending on the settings of
     88the HotSpot JIT compiler (if the tests are run on Sun).</p>
     89
     90<p>Some statements can be made in general though. Due to the fact that
     91ABCL has been implemented in Java, it inherits some of the aspects of
     92Java itself as well as the fact that it can't directly manipulate
     93CPU content. Implementations such as SBCL and Closure CL can do that and
     94take that to their advantage: for example in SBCL a boxed fixnum is a
     95direct register value, while in ABCL it's encapsulated in an object.</p>
     96
     97<p>On the other hand, ABCL - like SBCL - supports unboxed fixnums. ABCL's
     98fixnums support the full 32 bit range of integer values, while SBCL due
     99to its boxing strategy can only use 29 bit integers (on 32bit platforms).</p>
     100
     101<p>Given ABCL's age - a young project by Lisp standards - there is also
     102plenty of room for improvement in the area of execution speed and optimization.
     103The project welcomes initiatives to improve performance.</p>
    65104
    66105</div>
    67106
    68107
     108<div class="h3" title="qa" id="qa">
     109<h3>What is the quality of the implementation? How can you tell?</h3>
     110
     111<p>The project recognises there are several dimensions to quality:</p>
     112<ol>
     113<li> The level of compliance to the standard </li>
     114<li> The level of 'useability': whether (or not) the application is able
     115  to run existing Lisp code</li>
     116</ol>
     117
     118<p>The plan is to add to the list above software from Edi Weitz, who
     119 wrote some great libraries.</p>
     120
     121<p>The first item is being measured by running the ANSI test suite compliance
     122  tests. The second item is measured by compiling and running the test suite
     123  in the <a href="http://maxima.sourceforge.net/">Maxima application</a>.
     124  Additionally, compilation of <a href="http://ap5.com/">AP5</a> is used to
     125  improve this measure too.</p>
     126
     127<p>ABCL 0.15.0 fails 34 out of 21702 tests in the ANSI test suite
     128  in interpreted and compiled modes, coming from ca 44 in the last
     129  release.</p>
     130<p>As a measure of 'improvement achieved', the development team refers
     131  to the number of failing tests in the Maxima test suite too.
     132  ABCL 0.15.0 is able to run the test suite with 'only' ca 75 failing
     133  tests, coming from ca 1400 failures around October 2008.</p>
     134
     135</div>
     136
     137</div>
    69138
    70139<div style="float:left;width:100%">
    … …  
    77146<div style="float:right;font-size:10px;font-family:monospace">$Id$</div>
    78147</div>
     148</div>
    79149</body>
    80150</html>

  • public_html/staging/index.shtml

    r11947 r11949  
    143143        the testimonials page for their own words.
    144144        <br /><br />
    145         ABCL 0.15.0 fails 34 out of 21702 tests in the ANSI test suite
    146         in interpreted and compiled modes, coming from ca 44 in the last
    147         release.
    148         As a measure of 'improvement achieved', the development team refers
    149         to the number of failing tests in the Maxima test suite too.
    150         ABCL 0.15.0 is able to run the test suite with 'only' ca 75 failing
    151         tests, coming from ca 1400 failures around october 2008.
    152         <br /><br />
    153145        ABCL's CLOS is slow and does not handle on-the-fly
    154146        redefinition of classes correctly. There is no support for the long
    … …  
    156148        features are also missing. Enough CLOS is there to run ASDF and
    157149        CL-PPCRE, if you're in no hurry. There's no MOP worth mentioning.
    158         <br /><br />
    159         Due to the age of the source code (when compared to several other
    160         implementations) you're more likely to find bugs in ABCL.  However,
    161         we're committed to fixing any bugs you find.  Patches (bugfixes as
    162         well as features) are most welcome.
    163150        <br /><br />
    164151        Please report problems to the <a href="https://lists.sourceforge.net/lists/listinfo/armedbear-j-devel">j development mailing list</a>
Note: See TracChangeset for help on using the changeset viewer.