<?xml version="1.0" encoding="utf-8"?>
<feed xml:lang="en" xmlns="http://www.w3.org/2005/Atom"><title>Recent changes to feature-requests</title><link href="https://sourceforge.net/p/utool/feature-requests/" rel="alternate"/><link href="https://sourceforge.net/p/utool/feature-requests/feed.atom" rel="self"/><id>https://sourceforge.net/p/utool/feature-requests/</id><updated>2005-04-26T13:19:46Z</updated><subtitle>Recent changes to feature-requests</subtitle><entry><title>short vs. long labels "graphical" codecs</title><link href="https://sourceforge.net/p/utool/feature-requests/11/" rel="alternate"/><published>2005-04-26T13:19:46Z</published><updated>2005-04-26T13:19:46Z</updated><author><name>Stefan Thater</name><uri>https://sourceforge.net/u/thater/</uri></author><id>https://sourceforge.net0d6c6eadf31b9f95be5719e83ee49df23d2e4df3</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;The current implementation of the MRS codec uses feature &lt;br /&gt;
structures (encoded as strings) as labels, which leads to a &lt;br /&gt;
complex layout if the MRS structure is displayed as a graph with &lt;br /&gt;
Leonardo or daVinci.&lt;/p&gt;
&lt;p&gt;The layout could be improved if the graphical output-codecs &lt;br /&gt;
would use "short" labels.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>USR descriptions</title><link href="https://sourceforge.net/p/utool/feature-requests/10/" rel="alternate"/><published>2005-02-23T16:42:01Z</published><updated>2005-02-23T16:42:01Z</updated><author><name>Alexander Koller</name><uri>https://sourceforge.net/u/akoller/</uri></author><id>https://sourceforge.net5b5ee0269d299575b9202363d75c48d9e906000e</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;It would be useful to put some information into the&lt;br /&gt;
output that says where an underspecified description or&lt;br /&gt;
solution comes from. For an underspecified description,&lt;br /&gt;
this could be a sentence ID; for a solution, it could&lt;br /&gt;
be sentence ID + solution number. This would be quite&lt;br /&gt;
useful for keeping track of information flow.&lt;/p&gt;
&lt;p&gt;There are different places where this information&lt;br /&gt;
should be put for different codec formats. In GXL there&lt;br /&gt;
is a graph/id attribute that is specifically for this&lt;br /&gt;
purpose. In other formats, the information could be put&lt;br /&gt;
into a comment.&lt;/p&gt;
&lt;p&gt;If the input file for a convert or solve operation&lt;br /&gt;
already contains a description, the default should be&lt;br /&gt;
that it is passed on into the output file. It should&lt;br /&gt;
also be possible to specify the (base) description on&lt;br /&gt;
the command line; solution descriptions could still add&lt;br /&gt;
the solution number to this description string. When&lt;br /&gt;
both the input file and the command line specify a&lt;br /&gt;
description, the command line should win.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Clarify compact/weakly normal constraints</title><link href="https://sourceforge.net/p/utool/feature-requests/9/" rel="alternate"/><published>2004-11-05T11:34:59Z</published><updated>2004-11-05T11:34:59Z</updated><author><name>Alexander Koller</name><uri>https://sourceforge.net/u/akoller/</uri></author><id>https://sourceforge.netd5fb4f4a2f1ec7d77f00713dbcc64094557195d4</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;At the moment, there is some confusion in the code as&lt;br /&gt;
to what constitutes a weakly normal dominance constraints. &lt;/p&gt;
&lt;p&gt;According to the papers, weakly normal dominance&lt;br /&gt;
constraints must be compact; but this is a technical&lt;br /&gt;
restriction that may be violated by real constraints&lt;br /&gt;
from grammars, and the solver doesn't depend on it.&lt;br /&gt;
Also, normal constraints according to the classical&lt;br /&gt;
definition are not weakly normal under this definition.&lt;/p&gt;
&lt;p&gt;I propose the following redefinition: A constraint is&lt;br /&gt;
weakly normal if it is mostly normal, but dominances&lt;br /&gt;
may also go out of  roots (not internal nodes!). Then&lt;br /&gt;
the SM-solver should be able to run directly on WN&lt;br /&gt;
constraints.&lt;/p&gt;
&lt;p&gt;A side effect is that WN constraints can always (but&lt;br /&gt;
never have to) be compactified. Modify or remove&lt;br /&gt;
compactify.cc accordingly.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Solvability command</title><link href="https://sourceforge.net/p/utool/feature-requests/8/" rel="alternate"/><published>2004-11-05T11:30:34Z</published><updated>2004-11-05T11:30:34Z</updated><author><name>Alexander Koller</name><uri>https://sourceforge.net/u/akoller/</uri></author><id>https://sourceforge.netb9a5b0babee2bddd5be174f1886b221cfd2a81e4</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Add a "solvable" command to the command-line interface&lt;br /&gt;
that checks whether a constraint is solvable, but&lt;br /&gt;
doesn't compute all solved forms.&lt;/p&gt;
&lt;p&gt;The only nontrivial aspect of this is to decide how the&lt;br /&gt;
program will return the true/false output to a caller.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Command-line type checking</title><link href="https://sourceforge.net/p/utool/feature-requests/7/" rel="alternate"/><published>2004-11-05T11:29:07Z</published><updated>2004-11-05T11:29:07Z</updated><author><name>Alexander Koller</name><uri>https://sourceforge.net/u/akoller/</uri></author><id>https://sourceforge.netfc4545c025c541f2a980f439094afb5c00332304</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Find a way in which the command-line arguments can be&lt;br /&gt;
checked for consistency in a clever way. For example,&lt;br /&gt;
the --port option only makes sense for the "server"&lt;br /&gt;
command. The --output option and the --no-output option&lt;br /&gt;
contradict each other and shouldn't be useable in the&lt;br /&gt;
same call to utool. &lt;/p&gt;
&lt;p&gt;At the moment, contradictory command-line options are&lt;br /&gt;
ignored (more or less gracefully).&lt;/p&gt;
&lt;p&gt;The consistency check should not be much longer than&lt;br /&gt;
the current code for parsing the command-line&lt;br /&gt;
arguments. In particular, no code in the switch&lt;br /&gt;
statements must be duplicated, because this would be&lt;br /&gt;
insanely error-prone.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Net test</title><link href="https://sourceforge.net/p/utool/feature-requests/6/" rel="alternate"/><published>2004-11-05T11:25:32Z</published><updated>2004-11-05T11:25:32Z</updated><author><name>Alexander Koller</name><uri>https://sourceforge.net/u/akoller/</uri></author><id>https://sourceforge.netff8a8f4f6b2c64cc34e9facf03fb1d9cec28b70e</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Implement Stefan's test for hypernormal connectedness.&lt;br /&gt;
This is a prerequisite for the Hole Semantics and MRS&lt;br /&gt;
codecs, because these implement a conversion that is&lt;br /&gt;
only correct on nets.&lt;/p&gt;
&lt;p&gt;Also make the net test available as a new command that&lt;br /&gt;
can be called from the command line.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Hole Semantics</title><link href="https://sourceforge.net/p/utool/feature-requests/5/" rel="alternate"/><published>2004-11-05T11:24:23Z</published><updated>2004-11-05T11:24:23Z</updated><author><name>Alexander Koller</name><uri>https://sourceforge.net/u/akoller/</uri></author><id>https://sourceforge.neta53b1e890efbb0aaf81ef145120c26709593b10d</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Create an input/output codec for Hole Semantics.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Domcon-XML</title><link href="https://sourceforge.net/p/utool/feature-requests/4/" rel="alternate"/><published>2004-11-05T11:23:56Z</published><updated>2004-11-05T11:23:56Z</updated><author><name>Alexander Koller</name><uri>https://sourceforge.net/u/akoller/</uri></author><id>https://sourceforge.net20e43e8c383e3e5290c00a5b55ea5995d90ac942</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Create an input/output codec for dominance constraints&lt;br /&gt;
in XML syntax. This should be compatible with the&lt;br /&gt;
dominance graph syntax that Leonardo expects.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>MRS-XML codec</title><link href="https://sourceforge.net/p/utool/feature-requests/3/" rel="alternate"/><published>2004-11-05T11:23:08Z</published><updated>2004-11-05T11:23:08Z</updated><author><name>Alexander Koller</name><uri>https://sourceforge.net/u/akoller/</uri></author><id>https://sourceforge.net883fefc431a777fdb30014c69c4a888e35ca6d71</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Create an input codec for MRS in XML syntax.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Server mode</title><link href="https://sourceforge.net/p/utool/feature-requests/2/" rel="alternate"/><published>2004-11-05T11:22:02Z</published><updated>2004-11-05T11:22:02Z</updated><author><name>Alexander Koller</name><uri>https://sourceforge.net/u/akoller/</uri></author><id>https://sourceforge.net575ddc47e09bccb1da556d5edac7536bb74c34f0</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;If started in server mode, utool should wait for socket&lt;br /&gt;
connections on a user-specified port. Once a connection&lt;br /&gt;
is established, it reads one-line messages from the&lt;br /&gt;
socket. Each line is expected to be an XML element that&lt;br /&gt;
specifies the command. Then the result is computed and&lt;br /&gt;
sent back as an XML element.&amp;lt;p&amp;gt;&lt;/p&gt;
&lt;p&gt;An example for an XML message to the server could be&lt;/p&gt;
&lt;p&gt;&amp;lt;code&amp;gt;&lt;br /&gt;
&amp;amp;lt;solve input-codec="domcon-oz"&lt;br /&gt;
output-codec="domcon-oz"&amp;amp;gt; blah blah blah (constraint&lt;br /&gt;
in domcon-oz format will be CDATA) &amp;amp;lt;/solve&amp;amp;gt;&lt;br /&gt;
&amp;lt;/code&amp;gt;&lt;/p&gt;
&lt;p&gt;The server-mode utool should support all commands and&lt;br /&gt;
codecs that the command-line utool accepts as commands&lt;br /&gt;
on the command line.&lt;/p&gt;&lt;/div&gt;</summary></entry></feed>