Contents of /xml/htdocs/proj/en/glep/glep-0044.html

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1.5 - (show annotations) (download) (as text)
Fri Feb 10 23:30:46 2006 UTC (13 years ago) by genone
Branch: MAIN
Changes since 1.4: +5 -5 lines
File MIME type: text/html
change example filename to get past the QA checker

1 <?xml version="1.0" encoding="utf-8" ?>
2 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
3 <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
4 <!--
5 This HTML is auto-generated. DO NOT EDIT THIS FILE! If you are writing a new
6 PEP, see http://www.python.org/peps/pep-0001.html for instructions and links
8 -->
9 <head>
10 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
11 <meta name="generator" content="Docutils 0.3.9: http://docutils.sourceforge.net/" />
12 <title>GLEP 44 -- Manifest2 format</title>
13 <link rel="stylesheet" href="tools/glep.css" type="text/css" />
14 </head>
15 <body bgcolor="white">
16 <table class="navigation" cellpadding="0" cellspacing="0"
17 width="100%" border="0">
18 <tr><td class="navicon" width="150" height="35">
19 <a href="http://www.gentoo.org/" title="Gentoo Linux Home Page">
20 <img src="http://www.gentoo.org/images/gentoo-new.gif" alt="[Gentoo]"
21 border="0" width="150" height="35" /></a></td>
22 <td class="textlinks" align="left">
23 [<b><a href="http://www.gentoo.org/">Gentoo Linux Home</a></b>]
24 [<b><a href="http://www.gentoo.org/proj/en/glep">GLEP Index</a></b>]
25 [<b><a href="./glep-0044.txt">GLEP Source</a></b>]
26 </td></tr></table>
27 <table class="rfc2822 docutils field-list" frame="void" rules="none">
28 <col class="field-name" />
29 <col class="field-body" />
30 <tbody valign="top">
31 <tr class="field"><th class="field-name">GLEP:</th><td class="field-body">44</td>
32 </tr>
33 <tr class="field"><th class="field-name">Title:</th><td class="field-body">Manifest2 format</td>
34 </tr>
35 <tr class="field"><th class="field-name">Version:</th><td class="field-body">1.4</td>
36 </tr>
37 <tr class="field"><th class="field-name">Last-Modified:</th><td class="field-body"><a class="reference" href="http://www.gentoo.org/cgi-bin/viewcvs/xml/htdocs/proj/en/glep/glep-0044.txt?cvsroot=gentoo">2006/02/10 23:26:55</a></td>
38 </tr>
39 <tr class="field"><th class="field-name">Author:</th><td class="field-body">Marius Mauch &lt;genone&#32;&#97;t&#32;gentoo.org&gt;,</td>
40 </tr>
41 <tr class="field"><th class="field-name">Status:</th><td class="field-body">Draft</td>
42 </tr>
43 <tr class="field"><th class="field-name">Type:</th><td class="field-body">Standards Track</td>
44 </tr>
45 <tr class="field"><th class="field-name">Content-Type:</th><td class="field-body"><a class="reference" href="http://www.python.org/peps/glep-0012.html">text/x-rst</a></td>
46 </tr>
47 <tr class="field"><th class="field-name">Created:</th><td class="field-body">04-Dec-2005</td>
48 </tr>
49 <tr class="field"><th class="field-name">Post-History:</th><td class="field-body">06-Dec-2005, 23-Jan-2006</td>
50 </tr>
51 </tbody>
52 </table>
53 <hr />
54 <div class="contents topic" id="contents">
55 <p class="topic-title first"><a name="contents">Contents</a></p>
56 <ul class="simple">
57 <li><a class="reference" href="#abstract" id="id9" name="id9">Abstract</a></li>
58 <li><a class="reference" href="#motivation" id="id10" name="id10">Motivation</a></li>
59 <li><a class="reference" href="#specification" id="id11" name="id11">Specification</a><ul>
60 <li><a class="reference" href="#compability-entries" id="id12" name="id12">Compability Entries</a></li>
61 <li><a class="reference" href="#scope" id="id13" name="id13">Scope</a></li>
62 <li><a class="reference" href="#number-of-hashes" id="id14" name="id14">Number of hashes</a></li>
63 </ul>
64 </li>
65 <li><a class="reference" href="#rationale" id="id15" name="id15">Rationale</a><ul>
66 <li><a class="reference" href="#removal-of-digest-files" id="id16" name="id16">Removal of digest files</a></li>
67 <li><a class="reference" href="#reducing-redundancy" id="id17" name="id17">Reducing redundancy</a></li>
68 <li><a class="reference" href="#removal-of-checksum-collisions" id="id18" name="id18">Removal of checksum collisions</a></li>
69 <li><a class="reference" href="#flexible-verification-system" id="id19" name="id19">Flexible verification system</a></li>
70 </ul>
71 </li>
72 <li><a class="reference" href="#backwards-compatibility" id="id20" name="id20">Backwards Compatibility</a></li>
73 <li><a class="reference" href="#other-problems" id="id21" name="id21">Other problems</a><ul>
74 <li><a class="reference" href="#impacts-on-infrastructure" id="id22" name="id22">Impacts on infrastructure</a></li>
75 </ul>
76 </li>
77 <li><a class="reference" href="#reference-implementation" id="id23" name="id23">Reference Implementation</a></li>
78 <li><a class="reference" href="#options" id="id24" name="id24">Options</a></li>
79 <li><a class="reference" href="#credits" id="id25" name="id25">Credits</a></li>
80 <li><a class="reference" href="#references" id="id26" name="id26">References</a></li>
81 <li><a class="reference" href="#copyright" id="id27" name="id27">Copyright</a></li>
82 </ul>
83 </div>
84 <div class="section" id="abstract">
85 <h1><a class="toc-backref" href="#id9" name="abstract">Abstract</a></h1>
86 <p>This GLEP proposes a new format for the Portage Manifest and digest file system
87 by unifying both filetypes into one to improve functional and non-functional
88 aspects of the Portage Tree.</p>
89 </div>
90 <div class="section" id="motivation">
91 <h1><a class="toc-backref" href="#id10" name="motivation">Motivation</a></h1>
92 <p>Please see <a class="footnote-reference" href="#reorg-thread" id="id1" name="id1">[1]</a> for a general overview.
93 The main long term goals of this proposal are to:</p>
94 <ul class="simple">
95 <li>Remove the tiny digest files from the tree. They are a major annoyance as on a
96 typical configuration they waste a lot of disk space and the simple transmission
97 of the names for all digest files during a <tt class="docutils literal"><span class="pre">emerge</span> <span class="pre">--sync</span></tt> needs a substantial
98 amount of bandwidth.</li>
99 <li>Reduce redundancy when multiple hash functions are used</li>
100 <li>Remove potential for checksum collisions if a file is recorded in more than one
101 digest file</li>
102 <li>Difference between filetypes for a more flexible verification system</li>
103 </ul>
104 </div>
105 <div class="section" id="specification">
106 <h1><a class="toc-backref" href="#id11" name="specification">Specification</a></h1>
107 <p>The new Manifest format would change the existing format in the following ways:</p>
108 <ul>
109 <li><p class="first">Addition of a filetype specifier, currently planned are</p>
110 <ul class="simple">
111 <li><tt class="docutils literal"><span class="pre">AUX</span></tt> for files directly used by ebuilds (e.g. patches or initscripts),
112 located in the <tt class="docutils literal"><span class="pre">files/</span></tt> subdirectory</li>
113 <li><tt class="docutils literal"><span class="pre">EBUILD</span></tt> for all ebuilds</li>
114 <li><tt class="docutils literal"><span class="pre">MISC</span></tt> for files not directly used by ebuilds like <tt class="docutils literal"><span class="pre">ChangeLog</span></tt> or
115 <tt class="docutils literal"><span class="pre">metadata.xml</span></tt> files</li>
116 <li><tt class="docutils literal"><span class="pre">DIST</span></tt> for release tarballs recorded in the <tt class="docutils literal"><span class="pre">SRC_URI</span></tt> variable of an ebuild,
117 these were previously recorded in the digest files</li>
118 </ul>
119 <p>Future portage improvements might extend this list (for example with types
120 relevant for eclasses or profiles)</p>
121 </li>
122 <li><p class="first">Only have one line per file listing all information instead of one line per
123 file and checksum type</p>
124 </li>
125 <li><p class="first">Remove the separated digest-* files in the <tt class="docutils literal"><span class="pre">files/</span></tt> subdirectory</p>
126 </li>
127 </ul>
128 <p>Each line in the new format has the following format:</p>
129 <pre class="literal-block">
130 &lt;filetype&gt; &lt;filename&gt; &lt;filesize&gt; &lt;chksumtype1&gt; &lt;chksum1&gt; ... &lt;chksumtypen&gt; &lt;chksumn&gt;
131 </pre>
132 <p>However theses entries will be stored in the existing Manifest files.</p>
133 <p>An <a class="reference" href="glep-0044-extras/manifest2-example.txt">actual example</a> <a class="footnote-reference" href="#id7" id="id8" name="id8">[6]</a> for a (pure) Manifest2 file..</p>
134 <div class="section" id="compability-entries">
135 <h2><a class="toc-backref" href="#id12" name="compability-entries">Compability Entries</a></h2>
136 <p>To maintain compability with existing portage versions a transition period after
137 is the introduction of the Manifest2 format is required during which portage
138 will not only have to be capable of using existing Manifest and digest files but
139 also generate them in addition to the new entries.
140 Fortunately this can be accomplished by simply mixing old and new style entries
141 in one file for the Manifest files, existing portage versions will simply ignore
142 the new style entries. For the digest files there are no new entries to care
143 about.</p>
144 </div>
145 <div class="section" id="scope">
146 <h2><a class="toc-backref" href="#id13" name="scope">Scope</a></h2>
147 <p>It is important to note that this proposal only deals with a change of the
148 format of the digest and Manifest system.</p>
149 <p>It does not expand the scope of it to cover eclasses, profiles or anything
150 else not already covered by the Manifest system, it also doesn't affect
151 the Manifest signing efforts in any way (though the implementations of both
152 might be coupled).</p>
153 <p>Also while multiple hash functions will become standard with the proposed
154 implementation they are not a specific feature of this format <a class="footnote-reference" href="#multi-hash-thread" id="id3" name="id3">[2]</a>.</p>
155 </div>
156 <div class="section" id="number-of-hashes">
157 <h2><a class="toc-backref" href="#id14" name="number-of-hashes">Number of hashes</a></h2>
158 <p>While using multiple hashes for each file is a major feature of this proposal
159 we have to make sure that the number of hashes listed is limited to avoid
160 an explosion of the Manifest size that would revert the main benefit of this proposal
161 (reduzing tree size). Therefore the number of hashes that will be generated
162 will be limited to three different hash functions. For compability though we
163 have to rely on at least one hash function to always be present, this proposal
164 suggest to use SHA1 for this purpose (as it is supposed to be more secure than MD5
165 and currently only SHA1 and MD5 are directly available in python, also MD5 doesn't
166 have any benefit in terms of compability).</p>
167 </div>
168 </div>
169 <div class="section" id="rationale">
170 <h1><a class="toc-backref" href="#id15" name="rationale">Rationale</a></h1>
171 <p>The main goals of the proposal have been listed in the <a class="reference" href="#motivation">Motivation</a>, here now
172 the explanation why they are improvements and how the proposed format will
173 accomplish them.</p>
174 <div class="section" id="removal-of-digest-files">
175 <h2><a class="toc-backref" href="#id16" name="removal-of-digest-files">Removal of digest files</a></h2>
176 <p>Normal users that don't use a &quot;tuned&quot; filesystem for the portage tree are
177 wasting several dozen to a few hundred megabytes of disk space with the current
178 system, largely caused by the digest files.
179 This is due to the filesystem overhead present in most filesystem that
180 have a standard blocksize of four kilobytes while most digest files are under
181 one kilobyte in size, so this results in approximately a waste of three kilobytes
182 per digest file (likely even more). At the time of this writing the tree contains
183 roughly 22.000 digest files, so the overall waste caused by digest files is
184 estimated at about 70-100 megabytes.
185 Furthermore it is assumed that this will also reduce the disk space wasted by
186 the Manifest files as they now contain more content, but this hasn't been
187 verified yet.</p>
188 <p>By unifying the digest files with the Manifest these tiny files are eliminated
189 (in the long run), reducing the apparent tree size by about 20%, benefitting
190 both users and the Gentoo infrastructure.</p>
191 </div>
192 <div class="section" id="reducing-redundancy">
193 <h2><a class="toc-backref" href="#id17" name="reducing-redundancy">Reducing redundancy</a></h2>
194 <p>When multiple hashes are used with the current system
195 both the filename and filesize are repeated for every checksum type used as each
196 checksum is standalone. However this doesn't add any functionality and is
197 therefore useless, so the new format removes this redundancy.
198 This is a theoretical improvement at this moment as only one hash function is in
199 use, but expected to change soon (see <a class="footnote-reference" href="#multi-hash-thread" id="id4" name="id4">[2]</a>).</p>
200 </div>
201 <div class="section" id="removal-of-checksum-collisions">
202 <h2><a class="toc-backref" href="#id18" name="removal-of-checksum-collisions">Removal of checksum collisions</a></h2>
203 <p>The current system theoretically allows for a <tt class="docutils literal"><span class="pre">DIST</span></tt> type file to be recorded
204 in multiple digest files with different sizes and/or checksums. In such a case
205 one version of a package would report a checksum violation while another one
206 would not. This could create confusion and uncertainity among users.
207 So far this case hasn't been observed, but it can't be ruled out with the
208 existing system.
209 As the new format lists each file exactly once this would be no longer possible.</p>
210 </div>
211 <div class="section" id="flexible-verification-system">
212 <h2><a class="toc-backref" href="#id19" name="flexible-verification-system">Flexible verification system</a></h2>
213 <p>Right now portage verifies the checksum of every file listed in the Manifest
214 before using any file of the package and all <tt class="docutils literal"><span class="pre">DIST</span></tt> files of an ebuild
215 before using that ebuild. This is unnecessary in many cases:</p>
216 <ul class="simple">
217 <li>During the &quot;depend&quot; phase (when the ebuild metadata is generated) only
218 files of type <tt class="docutils literal"><span class="pre">EBUILD</span></tt> are used, so verifying the other types isn't
219 necessary. Theoretically it is possible for an ebuild to include other
220 files like those of type <tt class="docutils literal"><span class="pre">AUX</span></tt> at this phase, but that would be a
221 major QA violation and should never occur, so it can be ignored here.
222 It is also not a security concern as the ebuild is verified before parsing
223 it, so each manipulation would show up.</li>
224 <li>Generally files of type <tt class="docutils literal"><span class="pre">MISC</span></tt> don't need to be verified as they are
225 only used in very specific situations, aren't executed (just parsed at most)
226 and don't affect the package build process.</li>
227 <li>Files of type <tt class="docutils literal"><span class="pre">DIST</span></tt> only need to be verified directly after fetching and
228 before unpacking them (which often will be one step), not every time their
229 associated ebuild is used.</li>
230 </ul>
231 </div>
232 </div>
233 <div class="section" id="backwards-compatibility">
234 <h1><a class="toc-backref" href="#id20" name="backwards-compatibility">Backwards Compatibility</a></h1>
235 <p>Switching the Manifest system is a task that will need a long transition period
236 like most changes affecting both portage and the tree. In this case the
237 implementation will be rolled out in several phases:</p>
238 <ol class="arabic simple">
239 <li>Add support for verification of Manifest2 entries in portage</li>
240 <li>Enable generation of Manifest2 entries in addition to the current system</li>
241 <li>Ignore digests during <tt class="docutils literal"><span class="pre">emerge</span> <span class="pre">--sync</span></tt> to get the size-benefit clientside.
242 This step may be ommitted if the following steps are expected to follow soon.</li>
243 <li>Disable generation of entries for the current system</li>
244 <li>Remove all traces of the current system from the tree (serverside)</li>
245 </ol>
246 <p>Each step has its own issues. While 1) and 2) can be implemented without any
247 compability problems all later steps have a major impact:</p>
248 <ul class="simple">
249 <li>Step 3) can only be implemented when the whole tree is Manifest2 ready
250 (ideally speaking, practically the requirement will be more like 95% coverage
251 with the expectation that for the remaining 5% either bugs will be filed after
252 step 3) is completed or they'll be updated at step 5).</li>
253 <li>Steps 4) and 5) will render all portage versions without Manifest2 support
254 basically useless (users would have to regenerate the digest and Manifest
255 for each package before being able to merge it), so this requires a almost
256 100% coverage of the userbase with Manifest2 capabale portage versions
257 (with step 1) completely implemented).</li>
258 </ul>
259 <p>Another problem is that some steps affect different targets:</p>
260 <ul class="simple">
261 <li>Steps 1) and 3) target portage versions used by users</li>
262 <li>Steps 2) and 4) target portage versions used by devs</li>
263 <li>Step 5) targets the portage tree on the cvs server</li>
264 </ul>
265 <p>While it is relatively easy to get all devs to use a new portage version this is
266 practically impossible with users as some don't update their systems regulary.
267 While six months are probably sufficient to reach a 95% coverage one year is
268 estimated to reach an almost-complete coverage. All times are relative to the
269 stable-marking of a compatible portage version.</p>
270 <p>No timeframe for implementation is presented here as it is highly dependent
271 on the completion of each step.</p>
272 <p>In summary it can be said that while a full conversion will take over a year
273 to be completed due to compability issues mentioned above some benefits of the
274 system can selectively be used as soon as step 2) is completed.</p>
275 </div>
276 <div class="section" id="other-problems">
277 <h1><a class="toc-backref" href="#id21" name="other-problems">Other problems</a></h1>
278 <div class="section" id="impacts-on-infrastructure">
279 <h2><a class="toc-backref" href="#id22" name="impacts-on-infrastructure">Impacts on infrastructure</a></h2>
280 <p>While one long term goal of this proposal is to reduce the size of the tree
281 and therefore make life for the Gentoo Infrastructure easier this will only
282 take effect once the implementation is rolled out completely. In the meantime
283 however it will increase the tree size due to keeping checksums in both formats.
284 It's not possible to give a usable estimate on the degree of the increase as
285 it depends on many variables such as the exact implementation timeframe,
286 propagation of Manifest2 capable portage versions among devs or the update
287 rate of the tree. It has been suggested that Manifest files that are not gpg
288 signed could be mass converted in one step, this could certainly help but only
289 to some degree (according to a recent research <a class="footnote-reference" href="#gpg-numbers" id="id5" name="id5">[3]</a> about 40% of
290 all Manifests in the tree are signed, but this number hasn't been verified).</p>
291 </div>
292 </div>
293 <div class="section" id="reference-implementation">
294 <h1><a class="toc-backref" href="#id23" name="reference-implementation">Reference Implementation</a></h1>
295 <p>A patch for a prototype implementation of Manifest2 verification and partial
296 generation has been posted at <a class="footnote-reference" href="#manifest2-patch" id="id6" name="id6">[4]</a>, it will be reworked before
297 being considered for inclusion in portage. However it shows that adding support
298 for verification is quite simple, but generation is a bit tricky and will
299 therefore be implemented later.</p>
300 </div>
301 <div class="section" id="options">
302 <h1><a class="toc-backref" href="#id24" name="options">Options</a></h1>
303 <p>Some things have been considered for this GLEP but aren't part of the proposal
304 yet for various reasons:</p>
305 <ul class="simple">
306 <li>timestamp field: the author has considered adding a timestamp field for
307 each entry to list the time the entry was created. However so far no practical
308 use for such a feature has been found.</li>
309 <li>convert size field into checksum: Another idea was to treat the size field
310 like any other checksum. But so far no real benefit (other than a slightly
311 more modular implementation) for this has been seen while it has several
312 drawbacks: For once, unlike checksums, the size field is definitely required
313 for all <tt class="docutils literal"><span class="pre">DIST</span></tt> files, also it would slightly increase the length of
314 each entry by adding a <tt class="docutils literal"><span class="pre">SIZE</span></tt> keyword.</li>
315 <li>removal of the <tt class="docutils literal"><span class="pre">MISC</span></tt> type: It has been suggested to completely drop
316 entries of type <tt class="docutils literal"><span class="pre">MISC</span></tt>. This would result in a minor space reduction
317 (its rather unlikely to free any blocks) but completely remove the ability
318 to check these files for integrity. While they don't influence portage
319 or packages directly they can contain viable information for users, so
320 the author has the opinion that at least the option for integrity checks
321 should be kept.</li>
322 </ul>
323 </div>
324 <div class="section" id="credits">
325 <h1><a class="toc-backref" href="#id25" name="credits">Credits</a></h1>
326 <p>Thanks to the following persons for their input on or related to this GLEP
327 (even though they might not have known it):
328 Ned Ludd (solar), Brian Harring (ferringb), Jason Stubbs (jstubbs),
329 Robin H. Johnson (robbat2), Aron Griffis (agriffis)</p>
330 <p>Also thanks to Nicholas Jones (carpaski) to make the current Manifest system
331 resistent enough to be able to handle this change without too many transition
332 problems.</p>
333 </div>
334 <div class="section" id="references">
335 <h1><a class="toc-backref" href="#id26" name="references">References</a></h1>
336 <table class="docutils footnote" frame="void" id="reorg-thread" rules="none">
337 <colgroup><col class="label" /><col /></colgroup>
338 <tbody valign="top">
339 <tr><td class="label"><a class="fn-backref" href="#id1" name="reorg-thread">[1]</a></td><td><a class="reference" href="http://thread.gmane.org/gmane.linux.gentoo.devel/21920">http://thread.gmane.org/gmane.linux.gentoo.devel/21920</a></td></tr>
340 </tbody>
341 </table>
342 <table class="docutils footnote" frame="void" id="multi-hash-thread" rules="none">
343 <colgroup><col class="label" /><col /></colgroup>
344 <tbody valign="top">
345 <tr><td class="label"><a name="multi-hash-thread">[2]</a></td><td><em>(<a class="fn-backref" href="#id3">1</a>, <a class="fn-backref" href="#id4">2</a>)</em> <a class="reference" href="http://thread.gmane.org/gmane.linux.gentoo.devel/33434">http://thread.gmane.org/gmane.linux.gentoo.devel/33434</a></td></tr>
346 </tbody>
347 </table>
348 <table class="docutils footnote" frame="void" id="gpg-numbers" rules="none">
349 <colgroup><col class="label" /><col /></colgroup>
350 <tbody valign="top">
351 <tr><td class="label"><a class="fn-backref" href="#id5" name="gpg-numbers">[3]</a></td><td>gentoo-core mailing list, topic &quot;Gentoo key signing practices
352 and official Gentoo keyring&quot;, Message-ID &lt;<a class="reference" href="mailto:20051117075838.GB15734&#64;curie-int.vc.shawcable.net">20051117075838.GB15734&#64;curie-int.vc.shawcable.net</a>&gt;</td></tr>
353 </tbody>
354 </table>
355 <table class="docutils footnote" frame="void" id="manifest2-patch" rules="none">
356 <colgroup><col class="label" /><col /></colgroup>
357 <tbody valign="top">
358 <tr><td class="label"><a class="fn-backref" href="#id6" name="manifest2-patch">[4]</a></td><td><a class="reference" href="http://thread.gmane.org/gmane.linux.gentoo.portage.devel/1374">http://thread.gmane.org/gmane.linux.gentoo.portage.devel/1374</a></td></tr>
359 </tbody>
360 </table>
361 <table class="docutils footnote" frame="void" id="manifest2-example" rules="none">
362 <colgroup><col class="label" /><col /></colgroup>
363 <tbody valign="top">
364 <tr><td class="label"><a name="manifest2-example">[5]</a></td><td><a class="reference" href="http://www.gentoo.org/proj/en/glep/glep-0044-extras/manifest2-example">http://www.gentoo.org/proj/en/glep/glep-0044-extras/manifest2-example</a></td></tr>
365 </tbody>
366 </table>
367 <table class="docutils footnote" frame="void" id="id7" rules="none">
368 <colgroup><col class="label" /><col /></colgroup>
369 <tbody valign="top">
370 <tr><td class="label"><a class="fn-backref" href="#id8" name="id7">[6]</a></td><td><a class="reference" href="glep-0044-extras/manifest2-example.txt">glep-0044-extras/manifest2-example.txt</a></td></tr>
371 </tbody>
372 </table>
373 </div>
374 <div class="section" id="copyright">
375 <h1><a class="toc-backref" href="#id27" name="copyright">Copyright</a></h1>
376 <p>This document has been placed in the public domain.</p>
377 </div>
379 </div>
380 <div class="footer">
381 <hr class="footer" />
382 <a class="reference" href="glep-0044.txt">View document source</a>.
383 Generated on: 2006-02-10 23:30 UTC.
384 Generated by <a class="reference" href="http://docutils.sourceforge.net/">Docutils</a> from <a class="reference" href="http://docutils.sourceforge.net/rst.html">reStructuredText</a> source.
386 </div>
387 </body>
388 </html>

  ViewVC Help
Powered by ViewVC 1.1.20