|
a/website/BUGS.html |
|
b/website/BUGS.html |
|
... |
|
... |
| 19 |
|
19 |
|
| 20 |
<body>
|
20 |
<body>
|
| 21 |
|
21 |
|
| 22 |
<div class="rightlinks">
|
22 |
<div class="rightlinks">
|
| 23 |
<ul>
|
23 |
<ul>
|
| 24 |
<li><a href="index.html">Home</a></li>
|
24 |
<li><a href="index.html">Home</a></li>
|
| 25 |
<li><a href="download.html">Downloads</a></li>
|
25 |
<li><a href="download.html">Downloads</a></li>
|
| 26 |
<li><a href="doc.html">Documentation</a></li>
|
26 |
<li><a href="doc.html">Documentation</a></li>
|
| 27 |
</ul>
|
27 |
</ul>
|
| 28 |
</div>
|
28 |
</div>
|
| 29 |
|
29 |
|
| 30 |
<div class="content">
|
30 |
<div class="content">
|
| 31 |
|
31 |
|
| 32 |
<h1>Known bugs in current and older versions</h1>
|
32 |
<h1>Known bugs in current and older versions</h1>
|
| 33 |
|
33 |
|
| 34 |
<p><i>Bugs that are listed in an older version section are
|
34 |
<p><i>Bugs that are listed in an older version section are
|
| 35 |
supposedly fixed in later versions. Bugs listed in the
|
35 |
supposedly fixed in later versions. Bugs listed in the
|
| 36 |
topmost section may also exist in older versions.</i></p>
|
36 |
topmost section may also exist in older versions.</i></p>
|
| 37 |
|
37 |
|
| 38 |
<h2><a name="b_latest">Latest (recoll 1.12.0 + xapian 1.0.10)</a></h2>
|
38 |
<h2><a name="b_latest">Latest (recoll 1.12.1 + xapian 1.0.15)</a></h2>
|
|
|
39 |
<ul>
|
|
|
40 |
|
|
|
41 |
<li>Performing a full index with release 1.11 or later, over a
|
|
|
42 |
version created with a much older recoll release may
|
|
|
43 |
sometimes end with an error saying "backend doesn't
|
|
|
44 |
implement metadata". If this happens, you need to delete
|
|
|
45 |
the index directory (typically <em>~/.recoll/xapiandb/</em>)
|
|
|
46 |
and restart indexing. For big indexes, to avoid losing
|
|
|
47 |
time, removing the directory preventively may be preferable .</li>
|
|
|
48 |
|
|
|
49 |
<li> When Recoll is built with qt 4.4.0, the icons in the
|
|
|
50 |
result list are all displayed at the top of the page and
|
|
|
51 |
garbled. This appears to be a qt bug, fixed in 4.4.1. Use
|
|
|
52 |
either qt 4.3.x or 4.4.1</li>
|
|
|
53 |
|
|
|
54 |
<li> If the user-chosen result list entry format results in
|
|
|
55 |
several paragraphs (in the qt textedit sense), right clicks
|
|
|
56 |
will only work inside the first one for each entry.</li>
|
|
|
57 |
|
|
|
58 |
<li> When a mime type has an external viewer defined, but the
|
|
|
59 |
actual file is compressed (ie: xxx.txt.gz), recoll will try
|
|
|
60 |
to start the external viewer on the compressed file, which
|
|
|
61 |
will not work in most cases.</li>
|
|
|
62 |
|
|
|
63 |
<li>It seems that the recoll program sometimes segfaults when
|
|
|
64 |
exiting after the first execution ?</li>
|
|
|
65 |
|
|
|
66 |
<li> If you are seeing a delay of a few seconds before the
|
|
|
67 |
result list displays for the first query of a recoll
|
|
|
68 |
instance, try changing the result list font in the query
|
|
|
69 |
preferences. This is not a recoll problem, I don't know the
|
|
|
70 |
exact cause (I've seen it happen with "Sans Serif" and go
|
|
|
71 |
away with Helvetica or Arial).</li>
|
|
|
72 |
|
|
|
73 |
<li> Under some versions of KDE (ie: Fedora FC5 KDE
|
|
|
74 |
3.5.4-0.5.fc5), there is a problem with the window stacking
|
|
|
75 |
order. Opening the "browse" file selection dialog from the
|
|
|
76 |
advanced search dialog will stack the latter under the main
|
|
|
77 |
window, possibly making it invisible. This is quite probably
|
|
|
78 |
a Kwin bug, possibly related to
|
|
|
79 |
http://bugs.kde.org/show_bug.cgi?id=79183 or a correction
|
|
|
80 |
thereof.</li>
|
|
|
81 |
|
|
|
82 |
<li>Under Solaris, it is necessary to perform initial
|
|
|
83 |
indexing with the recollindex program (the recoll index
|
|
|
84 |
thread doesn't work for creating the database). Don't know
|
|
|
85 |
the reason. Only idea I have is problem with exception
|
|
|
86 |
handling (recoll catches an exception while trying the yet
|
|
|
87 |
inexistant db).</li>
|
|
|
88 |
|
|
|
89 |
<li>The default filter for files in microsoft word format
|
|
|
90 |
(application/msword, .doc), antiword, has trouble with some
|
|
|
91 |
relatively rare files with a very small text, resulting in the
|
|
|
92 |
following error message:
|
|
|
93 |
<blockquote>
|
|
|
94 |
I'm afraid the text stream of this file is too small to
|
|
|
95 |
handle.
|
|
|
96 |
</blockquote>
|
|
|
97 |
Only small files produced by Microsoft Word on a Mac, or by
|
|
|
98 |
OpenOffice will trigger this message. As a workaround, install
|
|
|
99 |
wvWare and modify mimeconf to use the rcldoc filter, which
|
|
|
100 |
will use vwWare if it is available. This will result in
|
|
|
101 |
slower indexing for doc files.</li>
|
| 39 |
<ul>
|
102 |
</ul>
|
| 40 |
|
103 |
|
|
|
104 |
<h2><a name="b_1_12_1">1.12.1</a></h2>
|
|
|
105 |
<li>Uncatched Xapian exceptions can crash the GUI when a query
|
|
|
106 |
is run while the index is being updated.</li>
|
|
|
107 |
<li>The result list right-click pop up menu does not appear
|
|
|
108 |
when the cursor is inside a table.</li>
|
|
|
109 |
<li>Multithreaded access to Xlib can crash the real-time indexer.</li>
|
|
|
110 |
<li>A looping filter (ie: rclps trying to index loop.ps) can
|
|
|
111 |
keep on running forever and stop the indexing while eating cpu.</li>
|
|
|
112 |
<li>Filter subprocesses can sometimes be left around after
|
|
|
113 |
indexing is interrupted. Two signals are sometimes necessary
|
|
|
114 |
to get recollindex to exit.</li>
|
|
|
115 |
<li>Signals SIGUSR1 and SIGUSR2 are not blocked.</li>
|
|
|
116 |
<li>Sort does not work on queries started from the command line.</li>
|
|
|
117 |
|
|
|
118 |
<h2><a name="b_1_12_0">1.12.0</a></h2>
|
|
|
119 |
<ul>
|
| 41 |
<li>To compile the Python interface for recoll 1.12, you need
|
120 |
<li>To compile the Python interface for recoll 1.12, you need
|
| 42 |
to edit setup.py and replace "rcldb/pathhash.cpp" with
|
121 |
to edit setup.py and replace "rcldb/pathhash.cpp" with
|
| 43 |
"utils/fileudi.cpp".</li>
|
122 |
"utils/fileudi.cpp".</li>
|
| 44 |
<li>Performing a full index with release 1.11 or later, over a
|
123 |
<li>rclman outputs control characters, causing problems with
|
| 45 |
version created with a much older recoll release may
|
124 |
preview and phrase searches in manual pages.</li>
|
| 46 |
sometimes end with an error saying "backend doesn't
|
125 |
<li>rcllyx has trouble with 8bit characters in file names.</li>
|
| 47 |
implement metadata". If this happens, you need to delete
|
126 |
<li>"recoll -q ..." processes incorrectly second and further
|
| 48 |
the index directory (typically <em>~/.recoll/xapiandb/</em>)
|
127 |
command line arguments.</li>
|
| 49 |
and restart indexing. For big indexes, removing the
|
128 |
<li><a name="XapianNearPatch">The</a>
|
| 50 |
directory preventively may be preferable to avoid losing
|
129 |
following problem was corrected by Xapian 1.0.11 or
|
| 51 |
time.</li>
|
130 |
1.0.12, and I can see no reason to use older versions and/or the
|
| 52 |
|
131 |
patches below. However, they're kept around in case someone
|
| 53 |
<li> When Recoll is built with qt 4.4.0, the icons in the
|
132 |
needs them.<br>
|
| 54 |
result list are all displayed at the top of the page and
|
|
|
| 55 |
garbled. This appears to be a qt bug, fixed in 4.4.1. Use
|
|
|
| 56 |
either qt 4.3.x or 4.4.1</li>
|
|
|
| 57 |
|
|
|
| 58 |
<li> If the user-chosen result list entry format results in
|
|
|
| 59 |
several paragraphs (in the qt textedit sense), right clicks
|
|
|
| 60 |
will only work inside the first one for each entry.</li>
|
|
|
| 61 |
|
|
|
| 62 |
<li> When a mime type has an external viewer defined, but the
|
|
|
| 63 |
actual file is compressed (ie: xxx.txt.gz), recoll will try
|
|
|
| 64 |
to start the external viewer on the compressed file, which
|
|
|
| 65 |
will not work in most cases.</li>
|
|
|
| 66 |
|
|
|
| 67 |
<li>NEAR expansion errors: recoll performs stemming expansion
|
133 |
NEAR expansion errors: recoll performs stemming expansion inside
|
| 68 |
inside NEAR clauses (except if prevented by a capitalized
|
134 |
NEAR clauses (except if prevented by a capitalized
|
| 69 |
entry). Because of a Xapian bug (at least up to 1.0.10),
|
135 |
entry). Because of a Xapian bug (up to 1.0.12 (or 11?)), NEAR
|
| 70 |
NEAR does not support multiple OR subclauses. This manifests
|
136 |
does not support multiple OR subclauses. This manifests itself
|
| 71 |
itself by a 'not implemented' Xapian exception or an
|
137 |
by a 'not implemented' Xapian exception or an explicit error
|
| 72 |
explicit error message. Workarounds:
|
138 |
message. Workarounds:
|
| 73 |
<ul>
|
139 |
<ul>
|
| 74 |
<li>Prevent expansion of NEAR terms (possibly except one) by
|
140 |
<li>Prevent expansion of NEAR terms (possibly except one) by
|
| 75 |
capitalizing them.
|
141 |
capitalizing them.
|
| 76 |
|
142 |
|
| 77 |
<li>Or apply the following patch to xapian, inside the
|
143 |
<li>Or apply the following patch to xapian, inside the
|
| 78 |
"api/" directory:<br>
|
144 |
"api/" directory:<br>
|
| 79 |
0.x versions:
|
145 |
0.x versions:
|
| 80 |
<a href="xapian/xapNearDistrib-0.x.patch">
|
146 |
<a href="xapian/xapNearDistrib-0.x.patch">
|
| 81 |
xapian/xapNearDistrib-0.x.patch</a>
|
147 |
xapian/xapNearDistrib-0.x.patch</a>
|
| 82 |
<br>
|
148 |
<br>
|
| 83 |
1.0.[0-9]:
|
149 |
1.0.[0-9]:
|
| 84 |
<a href="xapian/xapNearDistrib-1.0.0_9.patch">
|
150 |
<a href="xapian/xapNearDistrib-1.0.0_9.patch">
|
| 85 |
xapian/xapNearDistrib-1.0.0_9.patch</a>
|
151 |
xapian/xapNearDistrib-1.0.0_9.patch</a>
|
| 86 |
<br>
|
152 |
<br>
|
| 87 |
1.0.10:
|
153 |
1.0.10:
|
| 88 |
<a href="xapian/xapNearDistrib-1.0.10.patch">
|
154 |
<a href="xapian/xapNearDistrib-1.0.10.patch">
|
| 89 |
xapian/xapNearDistrib-1.0.10.patch</a>
|
155 |
xapian/xapNearDistrib-1.0.10.patch</a>
|
| 90 |
<br>
|
156 |
<br>
|
| 91 |
|
157 |
|
| 92 |
or fetch the already patched source from
|
158 |
or fetch the already patched source from
|
| 93 |
<a href="xapian/">xapian/</a>
|
159 |
<a href="xapian/">the local xapian/ directory</a>
|
| 94 |
then recompile, and install.
|
160 |
then recompile, and install.
|
|
|
161 |
</li>
|
|
|
162 |
</ul>
|
| 95 |
</li>
|
163 |
</li>
|
| 96 |
</ul>
|
|
|
| 97 |
|
164 |
|
| 98 |
I hope that an equivalent fix will make it into xapian at
|
|
|
| 99 |
some point (the current fix is not completely correct but
|
|
|
| 100 |
still handles most useful cases).</li>
|
|
|
| 101 |
|
|
|
| 102 |
<li>It seems that the recoll program sometimes segfaults when
|
|
|
| 103 |
exiting after the first execution ?</li>
|
|
|
| 104 |
|
|
|
| 105 |
<li> If you are seeing a delay of a few seconds before the
|
|
|
| 106 |
result list displays for the first query of a recoll
|
|
|
| 107 |
instance, try changing the result list font in the query
|
|
|
| 108 |
preferences. This is not a recoll problem, I don't know the
|
|
|
| 109 |
exact cause (I've seen it happen with "Sans Serif" and go
|
|
|
| 110 |
away with Helvetica or Arial).</li>
|
|
|
| 111 |
|
|
|
| 112 |
<li> Under some versions of KDE (ie: Fedora FC5 KDE
|
|
|
| 113 |
3.5.4-0.5.fc5), there is a problem with the window stacking
|
|
|
| 114 |
order. Opening the "browse" file selection dialog from the
|
|
|
| 115 |
advanced search dialog will stack the latter under the main
|
|
|
| 116 |
window, possibly making it invisible. This is quite probably
|
|
|
| 117 |
a Kwin bug, possibly related to
|
|
|
| 118 |
http://bugs.kde.org/show_bug.cgi?id=79183 or a correction
|
|
|
| 119 |
thereof.</li>
|
|
|
| 120 |
|
|
|
| 121 |
<li> Under Solaris, it is necessary to perform initial indexing with the
|
|
|
| 122 |
recollindex program (the recoll index thread doesn't work for creating
|
|
|
| 123 |
the database). Don't know the reason. Only idea I have is problem with
|
|
|
| 124 |
exception handling (recoll catches an exception while trying the
|
|
|
| 125 |
yet inexistant db).</li>
|
|
|
| 126 |
|
|
|
| 127 |
<li>The default filter for files in microsoft word format
|
|
|
| 128 |
(application/msword, .doc), antiword, has trouble with some
|
|
|
| 129 |
relatively rare files with a very small text, resulting in the
|
|
|
| 130 |
following error message:
|
|
|
| 131 |
<blockquote>
|
|
|
| 132 |
I'm afraid the text stream of this file is too small to
|
|
|
| 133 |
handle.
|
|
|
| 134 |
</blockquote>
|
|
|
| 135 |
Only small files produced by Microsoft Word on a Mac, or by
|
|
|
| 136 |
OpenOffice will trigger this message. As a workaround, install
|
|
|
| 137 |
wvWare and modify mimeconf to use the rcldoc filter, which
|
|
|
| 138 |
will use vwWare if it is available. This will result in
|
|
|
| 139 |
slower indexing for doc files.</li>
|
|
|
| 140 |
</ul>
|
165 |
</ul>
|
| 141 |
|
166 |
|
| 142 |
<h2><a name="b_1_11_4">1.11.4</a></h2>
|
167 |
<h2><a name="b_1_11_4">1.11.4</a></h2>
|
| 143 |
<ul>
|
168 |
<ul>
|
| 144 |
<li>Possibly harmful bug in strerror_r usage (GNU case).</li>
|
169 |
<li>Possibly harmful bug in strerror_r usage (GNU case).</li>
|
| 145 |
<li>Incorrect handling of "accents" inside Japanese katakana
|
170 |
<li>Incorrect handling of "accents" inside Japanese katakana
|
| 146 |
text.</li>
|
171 |
text.</li>
|
| 147 |
<li>Using the "Erase history" command on an empty history
|
172 |
<li>Using the "Erase history" command on an empty history
|
| 148 |
would cause recoll to crash.</li>
|
173 |
would cause recoll to crash.</li>
|
| 149 |
</ul>
|
174 |
</ul>
|
| 150 |
|
175 |
|
| 151 |
<h2><a name="b_1_11_1">1.11.1</a></h2>
|
176 |
<h2><a name="b_1_11_1">1.11.1</a></h2>
|
| 152 |
<ul>
|
177 |
<ul>
|
| 153 |
<li>Unicode space characters like
|
178 |
<li>Unicode space characters like
|
| 154 |
<em>0x3000, Ideographic space</em>
|
179 |
<em>0x3000, Ideographic space</em>
|
| 155 |
where not detected inside user entries like the main
|
180 |
where not detected inside user entries like the main
|
| 156 |
interface search entry. Badly parsed searches would retrieve no
|
181 |
interface search entry. Badly parsed searches would retrieve no
|
| 157 |
results, when the same search entered with ascii space characters
|
182 |
results, when the same search entered with ascii space characters
|
| 158 |
would have succeeded.</li>
|
183 |
would have succeeded.</li>
|
| 159 |
<li>Spaces were inserted inside CJK strings when building
|
184 |
<li>Spaces were inserted inside CJK strings when building
|
| 160 |
abstracts for the result list.</li>
|
185 |
abstracts for the result list.</li>
|
| 161 |
<li>Accent removal should not be performed for Japanese.</li>
|
186 |
<li>Accent removal should not be performed for Japanese.</li>
|
| 162 |
<li>When using the query language, an OR part with more than
|
187 |
<li>When using the query language, an OR part with more than
|
| 163 |
two terms will swallow preceding AND terms, one for each
|
188 |
two terms will swallow preceding AND terms, one for each
|
| 164 |
additional OR. Ex: (champagne ext:odt OR ext:sxw OR ext:lyx)
|
189 |
additional OR. Ex: (champagne ext:odt OR ext:sxw OR ext:lyx)
|
| 165 |
will be interpreted as
|
190 |
will be interpreted as
|
| 166 |
"champagne OR ext:odt OR ext:sxw OR ext:lyx"
|
191 |
"champagne OR ext:odt OR ext:sxw OR ext:lyx"
|
| 167 |
instead of the correct
|
192 |
instead of the correct
|
| 168 |
"champagne AND (ext:odt OR ext:sxw OR ext:lyx)"
|
193 |
"champagne AND (ext:odt OR ext:sxw OR ext:lyx)"
|
| 169 |
Workaround until the fix is issued: add non-existing terms
|
194 |
Workaround until the fix is issued: add non-existing terms
|
| 170 |
before the OR part and check the resulting query:
|
195 |
before the OR part and check the resulting query:
|
| 171 |
"champagne bogusxyztv ext:odt OR ext:sxw OR ext:lyx"
|
196 |
"champagne bogusxyztv ext:odt OR ext:sxw OR ext:lyx"
|
| 172 |
</li>
|
197 |
</li>
|
| 173 |
<li>The "Copy file name" and "Copy URL" entries of the
|
198 |
<li>The "Copy file name" and "Copy URL" entries of the
|
| 174 |
right-click menus only copy the data to the X11 primary
|
199 |
right-click menus only copy the data to the X11 primary
|
| 175 |
selection (use middle-button click to paste). This is
|
200 |
selection (use middle-button click to paste). This is
|
| 176 |
probably a mistake, the data should be copied to the
|
201 |
probably a mistake, the data should be copied to the
|
| 177 |
clipboard too (permitting the use of the "Paste" edit menu
|
202 |
clipboard too (permitting the use of the "Paste" edit menu
|
| 178 |
entry or Ctrl+V in the target).</li>
|
203 |
entry or Ctrl+V in the target).</li>
|
| 179 |
|
204 |
|
| 180 |
<li>Possibly harmful bug in strerror_r usage (GNU case).</li>
|
205 |
<li>Possibly harmful bug in strerror_r usage (GNU case).</li>
|
| 181 |
</ul>
|
206 |
</ul>
|
| 182 |
|
207 |
|
| 183 |
<h2>1.10.6</h2>
|
208 |
<h2>1.10.6</h2>
|
| 184 |
<ul>
|
209 |
<ul>
|
| 185 |
<li> If the locale is not utf-8, non-ascii command line
|
210 |
<li> If the locale is not utf-8, non-ascii command line
|
| 186 |
arguments to recoll and recollq are not converted to utf-8,
|
211 |
arguments to recoll and recollq are not converted to utf-8,
|
| 187 |
which may prevent, for example, the kde applet from
|
212 |
which may prevent, for example, the kde applet from
|
| 188 |
working. The workaround is to apply the following one-line
|
213 |
working. The workaround is to apply the following one-line
|
| 189 |
fix to qtgui/main.cpp, recompile and install recoll:
|
214 |
fix to qtgui/main.cpp, recompile and install recoll:
|
| 190 |
<pre>
|
215 |
<pre>
|
| 191 |
386c386
|
216 |
386c386
|
| 192 |
< sSearch->setSearchString(QString::fromUtf8(qstring.c_str()));
|
217 |
< sSearch->setSearchString(QString::fromUtf8(qstring.c_str()));
|
| 193 |
---
|
218 |
---
|
| 194 |
> sSearch->setSearchString(QString::fromLocal8Bit(qstring.c_str()));
|
219 |
> sSearch->setSearchString(QString::fromLocal8Bit(qstring.c_str()));
|
| 195 |
</pre>
|
220 |
</pre>
|
| 196 |
</li>
|
221 |
</li>
|
| 197 |
</ul>
|
222 |
</ul>
|
| 198 |
|
223 |
|
| 199 |
<h2>1.10.1</h2>
|
224 |
<h2>1.10.1</h2>
|
| 200 |
|
225 |
|
| 201 |
<ul>
|
226 |
<ul>
|
| 202 |
<li> A relatively simple error case can cause the indexer to
|
227 |
<li> A relatively simple error case can cause the indexer to
|
| 203 |
stop processing an mbox file (forgetting all subsequent
|
228 |
stop processing an mbox file (forgetting all subsequent
|
| 204 |
messages). More specifically, this happens when encountering
|
229 |
messages). More specifically, this happens when encountering
|
| 205 |
more than than a few dozen errors while handling
|
230 |
more than than a few dozen errors while handling
|
| 206 |
attachments. This is relatively common: for exemple if an
|
231 |
attachments. This is relatively common: for exemple if an
|
| 207 |
external helper application is missing and multiple
|
232 |
external helper application is missing and multiple
|
| 208 |
attachments of the affected type are found (ie: multiple
|
233 |
attachments of the affected type are found (ie: multiple
|
| 209 |
images and no exiftool). Workaround: install the helper
|
234 |
images and no exiftool). Workaround: install the helper
|
| 210 |
application.
|
235 |
application.
|
| 211 |
<li> The decoding of base-64 data in emails fails in a relatively uncommon
|
236 |
<li> The decoding of base-64 data in emails fails in a relatively uncommon
|
| 212 |
but sometimes encountered case.
|
237 |
but sometimes encountered case.
|
| 213 |
<li> In a preview window, when walking the search term hits with the
|
238 |
<li> In a preview window, when walking the search term hits with the
|
| 214 |
Previous/Next buttons, 'Previous' actually acts as 'Next' (it does work
|
239 |
Previous/Next buttons, 'Previous' actually acts as 'Next' (it does work
|
| 215 |
normally for the local search).
|
240 |
normally for the local search).
|
| 216 |
<li> Problems in detecting message separators inside Thunderbird mailboxes
|
241 |
<li> Problems in detecting message separators inside Thunderbird mailboxes
|
| 217 |
(quite probably mainly for messages imported from outlook?). Can lead to
|
242 |
(quite probably mainly for messages imported from outlook?). Can lead to
|
| 218 |
unindexed messages, and even apparently indexer crashes in some cases.
|
243 |
unindexed messages, and even apparently indexer crashes in some cases.
|
| 219 |
<li> File names indexed as terms can sometimes overflow the maximum term
|
244 |
<li> File names indexed as terms can sometimes overflow the maximum term
|
| 220 |
size, halting the indexing.
|
245 |
size, halting the indexing.
|
| 221 |
<li> For Phrase/Near searches, only the first term group is highlighted in
|
246 |
<li> For Phrase/Near searches, only the first term group is highlighted in
|
| 222 |
preview.
|
247 |
preview.
|
| 223 |
</ul>
|
248 |
</ul>
|
| 224 |
|
249 |
|
| 225 |
<h2>1.10.0</h2>
|
250 |
<h2>1.10.0</h2>
|
| 226 |
<ul>
|
251 |
<ul>
|
| 227 |
|
252 |
|
| 228 |
<li> If a filter fails while trying to extract the data from a file, the file
|
253 |
<li> If a filter fails while trying to extract the data from a file, the file
|
| 229 |
will not be indexed at all (not even the file name). The file
|
254 |
will not be indexed at all (not even the file name). The file
|
| 230 |
name should be indexed in this case. This happens in particular in the
|
255 |
name should be indexed in this case. This happens in particular in the
|
| 231 |
very common case where the helper application is not installed (ie:
|
256 |
very common case where the helper application is not installed (ie:
|
| 232 |
missing Exiftool -> no *.jpg names in the index).
|
257 |
missing Exiftool -> no *.jpg names in the index).
|
| 233 |
|
258 |
|
| 234 |
<li> If several query language "ext:" qualifiers are specified, they will be
|
259 |
<li> If several query language "ext:" qualifiers are specified, they will be
|
| 235 |
joined by an AND instead of OR, resulting in no results. Using an
|
260 |
joined by an AND instead of OR, resulting in no results. Using an
|
| 236 |
explicit OR doesn't work (actually OR + field names is generally
|
261 |
explicit OR doesn't work (actually OR + field names is generally
|
| 237 |
broken). In some cases, you can use a "type:" qualifier as a workaround.
|
262 |
broken). In some cases, you can use a "type:" qualifier as a workaround.
|
| 238 |
|
263 |
|
| 239 |
|
264 |
|
| 240 |
</ul>
|
265 |
</ul>
|
| 241 |
<h2>1.9.x</h2>
|
266 |
<h2>1.9.x</h2>
|
| 242 |
<ul>
|
267 |
<ul>
|
| 243 |
<li> Problems have been reported indexing big mailstores (several hundreds of
|
268 |
<li> Problems have been reported indexing big mailstores (several hundreds of
|
| 244 |
thousands of messages): resulting in a very big database and even
|
269 |
thousands of messages): resulting in a very big database and even
|
| 245 |
crashes.
|
270 |
crashes.
|
| 246 |
|
271 |
|
| 247 |
</ul>
|
272 |
</ul>
|
| 248 |
<h2>1.8.2</h2>
|
273 |
<h2>1.8.2</h2>
|
| 249 |
<ul>
|
274 |
<ul>
|
| 250 |
<li> Under ubuntu (at least, maybe debian too), the default awk interpreter
|
275 |
<li> Under ubuntu (at least, maybe debian too), the default awk interpreter
|
| 251 |
(mawk) is ancient, and the recoll pdf input filter does not
|
276 |
(mawk) is ancient, and the recoll pdf input filter does not
|
| 252 |
work (removes all space characters). This can be solved by installing the
|
277 |
work (removes all space characters). This can be solved by installing the
|
| 253 |
gawk package.
|
278 |
gawk package.
|
| 254 |
$ apt-get install gawk
|
279 |
$ apt-get install gawk
|
| 255 |
$ update-alternatives --set awk /usr/bin/gawk
|
280 |
$ update-alternatives --set awk /usr/bin/gawk
|
| 256 |
|
281 |
|
| 257 |
<li> There are sometimes problems with document deletions: the index can
|
282 |
<li> There are sometimes problems with document deletions: the index can
|
| 258 |
get in a state where deleted or moved documents are not purged from the
|
283 |
get in a state where deleted or moved documents are not purged from the
|
| 259 |
index (the log file says that the doc are deleted, but they aren't
|
284 |
index (the log file says that the doc are deleted, but they aren't
|
| 260 |
actually). When this happens, the only solution currently is to reindex
|
285 |
actually). When this happens, the only solution currently is to reindex
|
| 261 |
from scratch (recollindex -z). This is due to a xapian bug, which is
|
286 |
from scratch (recollindex -z). This is due to a xapian bug, which is
|
| 262 |
fixed in xapian 1.0.2, or you can apply the following patch to xapian
|
287 |
fixed in xapian 1.0.2, or you can apply the following patch to xapian
|
| 263 |
1.0.1 to fix it:
|
288 |
1.0.1 to fix it:
|
| 264 |
http://www.lesbonscomptes.com/recoll/xapian/xapian-delete-document.patch
|
289 |
http://www.lesbonscomptes.com/recoll/xapian/xapian-delete-document.patch
|
| 265 |
|
290 |
|
| 266 |
<li> The dates shown for email attachments in a result list are the email
|
291 |
<li> The dates shown for email attachments in a result list are the email
|
| 267 |
folder modification date. This should be inherited from the parent
|
292 |
folder modification date. This should be inherited from the parent
|
| 268 |
message instead.
|
293 |
message instead.
|
| 269 |
|
294 |
|
| 270 |
<li> There are a few problems in the qt4 version of recoll:
|
295 |
<li> There are a few problems in the qt4 version of recoll:
|
| 271 |
<li> Some accelerators (esc-spc, ctl-arrow) do not work, neither do
|
296 |
<li> Some accelerators (esc-spc, ctl-arrow) do not work, neither do
|
| 272 |
copy/paste between the result list and preview windows and x11
|
297 |
copy/paste between the result list and preview windows and x11
|
| 273 |
applications.
|
298 |
applications.
|
| 274 |
<li> The qt4 q3textedit::find() method is extremely slow, so that
|
299 |
<li> The qt4 q3textedit::find() method is extremely slow, so that
|
| 275 |
positionning to first search term in Recoll preview has been disabled,
|
300 |
positionning to first search term in Recoll preview has been disabled,
|
| 276 |
and the application will sometimes appear to be looping when using the
|
301 |
and the application will sometimes appear to be looping when using the
|
| 277 |
find feature in the preview window (it's not looping, it's searching...)
|
302 |
find feature in the preview window (it's not looping, it's searching...)
|
| 278 |
|
303 |
|
| 279 |
</ul>
|
304 |
</ul>
|
| 280 |
<h2>1.8.1</h2>
|
305 |
<h2>1.8.1</h2>
|
| 281 |
<ul>
|
306 |
<ul>
|
| 282 |
<li> This is not really a bug but .beagle really should be included in
|
307 |
<li> This is not really a bug but .beagle really should be included in
|
| 283 |
"skippedNames", or you end up indexing the beagle text cache, which is
|
308 |
"skippedNames", or you end up indexing the beagle text cache, which is
|
| 284 |
not really desirable.
|
309 |
not really desirable.
|
| 285 |
<li> Doc bug: the manual states that the query language supports a "mime:"
|
310 |
<li> Doc bug: the manual states that the query language supports a "mime:"
|
| 286 |
switch to filter mime types. There is currently no such thing.
|
311 |
switch to filter mime types. There is currently no such thing.
|
| 287 |
|
312 |
|
| 288 |
|
313 |
|
| 289 |
</ul>
|
314 |
</ul>
|
| 290 |
<h2>1.7.5</h2>
|
315 |
<h2>1.7.5</h2>
|
| 291 |
<ul>
|
316 |
<ul>
|
| 292 |
<li> Debian and Ubuntu: the rclsoff Openoffice filter doesn't work,
|
317 |
<li> Debian and Ubuntu: the rclsoff Openoffice filter doesn't work,
|
| 293 |
because of an incorrect shell syntax (understood by bash but not sh). To
|
318 |
because of an incorrect shell syntax (understood by bash but not sh). To
|
| 294 |
fix, you edit /usr[/local]/share/recoll/filters/rclsoff and can change
|
319 |
fix, you edit /usr[/local]/share/recoll/filters/rclsoff and can change
|
| 295 |
the line:
|
320 |
the line:
|
| 296 |
trap cleanup EXIT SIGHUP SIGQUIT SIGINT SIGTERM
|
321 |
trap cleanup EXIT SIGHUP SIGQUIT SIGINT SIGTERM
|
| 297 |
into:
|
322 |
into:
|
| 298 |
trap cleanup EXIT HUP QUIT INT TERM
|
323 |
trap cleanup EXIT HUP QUIT INT TERM
|
| 299 |
or download the updated filter from the filters page:
|
324 |
or download the updated filter from the filters page:
|
| 300 |
http://www.recoll.org/filters/filters.html
|
325 |
http://www.recoll.org/filters/filters.html
|
| 301 |
|
326 |
|
| 302 |
</ul>
|
327 |
</ul>
|
| 303 |
<h2>1.7.3</h2>
|
328 |
<h2>1.7.3</h2>
|
| 304 |
<ul>
|
329 |
<ul>
|
| 305 |
<li> Processing will stop on first error while indexing an mbox file. This
|
330 |
<li> Processing will stop on first error while indexing an mbox file. This
|
| 306 |
could happen just because an attachment could not be decoded, and can
|
331 |
could happen just because an attachment could not be decoded, and can
|
| 307 |
cause non-indexing of many messages. The most probable cause of error is
|
332 |
cause non-indexing of many messages. The most probable cause of error is
|
| 308 |
a missing filter (ie for ms-word files), so the temporary workaround
|
333 |
a missing filter (ie for ms-word files), so the temporary workaround
|
| 309 |
would be to install the missing filters. This bug is specific to 1.7 and
|
334 |
would be to install the missing filters. This bug is specific to 1.7 and
|
| 310 |
1.6 users need not worry. A correction will be issued very soon.
|
335 |
1.6 users need not worry. A correction will be issued very soon.
|
| 311 |
<li> Messages of type multipart/signed are not indexed.
|
336 |
<li> Messages of type multipart/signed are not indexed.
|
| 312 |
|
337 |
|
| 313 |
</ul>
|
338 |
</ul>
|
| 314 |
<h2>1.6.2</h2>
|
339 |
<h2>1.6.2</h2>
|
| 315 |
<ul>
|
340 |
<ul>
|
| 316 |
<li> Relatively unfrequent issue with message boundary detection in mbox
|
341 |
<li> Relatively unfrequent issue with message boundary detection in mbox
|
| 317 |
files, could cause miscellaneous problems.
|
342 |
files, could cause miscellaneous problems.
|
| 318 |
<li> Executing an external viewer for a file with single-quotes in the name
|
343 |
<li> Executing an external viewer for a file with single-quotes in the name
|
| 319 |
would not work.
|
344 |
would not work.
|
| 320 |
|
345 |
|
| 321 |
</ul>
|
346 |
</ul>
|
| 322 |
<h2>1.5.10</h2>
|
347 |
<h2>1.5.10</h2>
|
| 323 |
<ul>
|
348 |
<ul>
|
| 324 |
<li> If a defaultcharset was set in the configuration file for a subdirectory,
|
349 |
<li> If a defaultcharset was set in the configuration file for a subdirectory,
|
| 325 |
it would stay in effect for all subsequent files/directories (except if
|
350 |
it would stay in effect for all subsequent files/directories (except if
|
| 326 |
explicitely overridden), potentially causing many transcoding errors.
|
351 |
explicitely overridden), potentially causing many transcoding errors.
|
| 327 |
|
352 |
|
| 328 |
</ul>
|
353 |
</ul>
|
| 329 |
<h2>1.5.[1-7]</h2>
|
354 |
<h2>1.5.[1-7]</h2>
|
| 330 |
<ul>
|
355 |
<ul>
|
| 331 |
<li> Dates in result list come from the file's ctimes, which may be confusing
|
356 |
<li> Dates in result list come from the file's ctimes, which may be confusing
|
| 332 |
<li> Some rare MIME messages with null boundaries can crash the indexer.
|
357 |
<li> Some rare MIME messages with null boundaries can crash the indexer.
|
| 333 |
|
358 |
|
| 334 |
</ul>
|
359 |
</ul>
|
| 335 |
<h2>1.5.0</h2>
|
360 |
<h2>1.5.0</h2>
|
| 336 |
<ul>
|
361 |
<ul>
|
| 337 |
<li> Under some conditions, recoll startup and exit could be very slow: the
|
362 |
<li> Under some conditions, recoll startup and exit could be very slow: the
|
| 338 |
simple search history list had serious problems with non-ascii strings,
|
363 |
simple search history list had serious problems with non-ascii strings,
|
| 339 |
whose size sometimes doubled at each program startup/stop.
|
364 |
whose size sometimes doubled at each program startup/stop.
|
| 340 |
|
365 |
|
| 341 |
</ul>
|
366 |
</ul>
|
| 342 |
<h2>1.3.3</h2>
|
367 |
<h2>1.3.3</h2>
|
| 343 |
<ul>
|
368 |
<ul>
|
| 344 |
|
369 |
|
| 345 |
<li> Several of the external filters did not handle path names with embedded
|
370 |
<li> Several of the external filters did not handle path names with embedded
|
| 346 |
spaces (rcluncomp rclsoff rclps rclmedia rcldjvu). This is fixed in 1.4.
|
371 |
spaces (rcluncomp rclsoff rclps rclmedia rcldjvu). This is fixed in 1.4.
|
| 347 |
|
372 |
|
| 348 |
<li> If your QT installation is built with the QT_NO_STL flag, Recoll will not
|
373 |
<li> If your QT installation is built with the QT_NO_STL flag, Recoll will not
|
| 349 |
compile. I have a patch for this (will be fixed in the next release),
|
374 |
compile. I have a patch for this (will be fixed in the next release),
|
| 350 |
contact me if you get the problem. Typical error message:
|
375 |
contact me if you get the problem. Typical error message:
|
| 351 |
main.cpp:160: error: no match for 'operator+=' in 'msg += reason'
|
376 |
main.cpp:160: error: no match for 'operator+=' in 'msg += reason'
|
| 352 |
|
377 |
|
| 353 |
<li> The 'None of these words' field in the complex search does not work if
|
378 |
<li> The 'None of these words' field in the complex search does not work if
|
| 354 |
there are no other filled fields (it transforms into an ordinary
|
379 |
there are no other filled fields (it transforms into an ordinary
|
| 355 |
search). Workaround: enter very common term(s) in the 'any of these
|
380 |
search). Workaround: enter very common term(s) in the 'any of these
|
| 356 |
words' field.
|
381 |
words' field.
|
| 357 |
|
382 |
|
| 358 |
<li> Indexing cannot currently be conveniently and cleanly
|
383 |
<li> Indexing cannot currently be conveniently and cleanly
|
| 359 |
stopped when it's started. You can kill the process, and
|
384 |
stopped when it's started. You can kill the process, and
|
| 360 |
keyboard interrupt might work, but this may leave the
|
385 |
keyboard interrupt might work, but this may leave the
|
| 361 |
database in a bad state. This is fixed in the upcoming
|
386 |
database in a bad state. This is fixed in the upcoming
|
| 362 |
release, there is no current workaround.
|
387 |
release, there is no current workaround.
|
| 363 |
</ul>
|
388 |
</ul>
|
| 364 |
|
389 |
|
| 365 |
<h2>1.2.2</h2>
|
390 |
<h2>1.2.2</h2>
|
| 366 |
<ul>
|
391 |
<ul>
|
| 367 |
<li> The preview window is supposed to scroll after loading the document so
|
392 |
<li> The preview window is supposed to scroll after loading the document so
|
| 368 |
that the first search term is visible. This does not work in many cases.
|
393 |
that the first search term is visible. This does not work in many cases.
|
| 369 |
<li> The result list title is not shown for sorted lists
|
394 |
<li> The result list title is not shown for sorted lists
|
| 370 |
|
395 |
|
| 371 |
Notes on older versions:
|
396 |
Notes on older versions:
|
| 372 |
<li> Trouble compiling on some linux systems (Gentoo and Slackware?). There
|
397 |
<li> Trouble compiling on some linux systems (Gentoo and Slackware?). There
|
| 373 |
existed a quite common issue where the Recoll link will fail trying to
|
398 |
existed a quite common issue where the Recoll link will fail trying to
|
| 374 |
use a libstdc++.la file. This was due to a problem with the xapian-config
|
399 |
use a libstdc++.la file. This was due to a problem with the xapian-config
|
| 375 |
program. A workaround has been included in the configure script for
|
400 |
program. A workaround has been included in the configure script for
|
| 376 |
recoll 1.2.2, and the problem should not occur any more.
|
401 |
recoll 1.2.2, and the problem should not occur any more.
|
| 377 |
|
402 |
|
| 378 |
<li> Case-insensitive search should now work in most cases
|
403 |
<li> Case-insensitive search should now work in most cases
|
| 379 |
(used to not work except for accented ascii).
|
404 |
(used to not work except for accented ascii).
|
| 380 |
|
405 |
|
| 381 |
<li> All directories and files with names beginning with a dot were ignored
|
406 |
<li> All directories and files with names beginning with a dot were ignored
|
| 382 |
by the skippedNames directive in the default recoll.conf file from
|
407 |
by the skippedNames directive in the default recoll.conf file from
|
| 383 |
older versions (no indexation of mozilla or thunderbird email !). An
|
408 |
older versions (no indexation of mozilla or thunderbird email !). An
|
| 384 |
upgrade will not fix this (it will not modify an existing
|
409 |
upgrade will not fix this (it will not modify an existing
|
| 385 |
configuration). You need to edit recoll.conf by hand and remove the .*
|
410 |
configuration). You need to edit recoll.conf by hand and remove the .*
|
| 386 |
from skippedNames.</li>
|
411 |
from skippedNames.</li>
|
| 387 |
|
412 |
|
| 388 |
</ul>
|
413 |
</ul>
|
| 389 |
|
414 |
|
| 390 |
</div>
|
415 |
</div>
|
| 391 |
</body>
|
416 |
</body>
|