source: trunk/release_notes.v7.4 @ 5054

Revision 3975, 9.4 KB checked in by paulf, 8 years ago (diff)

release_notes set for 7.4

  • Property svn:eol-style set to native
  • Property svn:keywords set to Author Date Id Revision
Line 
1Release Notes:  Earthworm Version "working"
2Cleaned out after release of v7.3 on 12/23/2008
3
4These notes are for release v7.4
5
6NEW MODULES:
7***********
8ew2mseed from ISTI is now in the CVS repository. This was written
9by Ilya Dricker of ISTI. This only works on Solaris, Linux, and
10Mac OS X. It does compile with cygwin for WINDOZE.   Note there
11is no .desc file with this module since it is designed to operate
12as a daemon program external from an EW installation. It only
13reads from wave_serverV modules.
14Paul Friberg 3/10/2010
15
16debias:
17Computes a rolling-average for a specified time, subtracts from data to remove
18bias (hence the name).
19Scott Hunter 02/24/2010
20
21ewintegrate:
22Processes tracebuf packets in any non-empty subset of 3 stages: debias (see
23above), integrate, and high-pass filter.
24Scott Hunter 03/02/2010
25
26MODIFICATIONS/BUG FIXES TO EXISTING MODULES:
27********************************************
28rcv/vdl removed from data_exchange as these are no longer used by USGS NEIC
29or anybody else as far as we can tell.
30Paul Friberg 4/27/2010
31
32hyp2000_mgr
33Added a _GFORTRAN compile directive since g77 is now replaced by gfortran on
34some platforms (newer Linux installs). To get hyp2000_mgr to compile, be
35sure to set FC=gfortran and to add -D_GFORTRAN compile time flag to GLOBALFLAGS
36Paul Friberg 4/27/2010
37
38tankplayer
39Allowed tankplayer to play back data faster with new setting:
40InterMessageDelayMillisecs
41Paul Friberg 3/13/2009
42
43tankplayer
44Had a bad check for size of message based on nsamps. This let through bad
45tracebuf packets if there was an error in writing the tank. This is now
46limited to tracebuf max size.
47Paul Friberg 3/19/2010
48
49wave_serverV
50Fixed a ReadBlockData() error for end of tanks read, thanks to Dave Kragness.
51I just integrated the code and compile tested. Earthw users tested!
52Paul Friberg 3/18/2010
53
54sgram
55Was never made to automatically compile for a windows build. It is now.
56Thanks to Jeff Marso for  pointing this one out.
57Paul Friberg 3/18/2010
58
59sac2tb (in tankplayer_tools):
60Allow (and require, for Windows) output file to be specified as a command line
61argument (instead of writing output to stdout); also added a switch to allow
62for appending to output file.
63Scott Hunter 3/17/2010
64
65localmag:
66Added in Matteo Quintiliani feature to have %v for event version
67in the directory naming of where ML messages are written.
68See the new option:  MlmsgOutDir in the cmd html file.
69I just did the cvs merge, Matteo did all the work.
70Paul Friberg 3/15/2010
71
72sniffwave:
73Now has the ability to show latency of data (Matteo's addition) and
74also run for n seconds and show the amount of wave data in a ring. (EW class)
75Paul Friberg 6/12/2009
76Made it now able to do: sniffwave ringname
77which translatest to "wild wild wild wild n". As of
78today, "sniffwave ringname y/s" is also possible.
79These improvements come after being sick and tired of
80typing wild wild wild wild n too many times :)
81Paul Friberg 3/15/2010
82If WaveMsg2MakeLocal fails, tracebuf is rejected
83Scott Hunter 3/19/2010
84Allowed SCN and SCNL searches for the same command line as it was before.
85Paul Friberg 3/25/2010
86
87copystatus:
88Now copies TYPE_STOP and TYPE_RESTART messages since statmgr needed these for
89stopmodule actions (especially if statmgr was not monitoring the first ring
90in the startstop listing).
91Paul Friberg 3/8/2010
92
93statmgr:
94When specifying the maximum number of pager or email messages, a negative
95value is interpreted as infinity (i.e. no limit) -- USE WITH CAUTION!
96Scott Hunter 12/13/2009
97Modified the statmgr to log any TYPE_STOP messages received.
98Paul Friberg 3/8/2010
99
100startstop:
1011) Stop and restart requests are now run in independent threads, so startstop
102will still be responsive to further requests no matter how long the stopping
103and/or restarting takes
104Scott Hunter 12/13/2009
1052) If there is a problem encountered when reconfiguring, instead of aborting,
106startstop now proceeds as if no such request had been made.
107Scott Hunter 12/13/2009
1083) Removed corruption of PATH environment variable & fixed display of paths in
109status
110Scott Hunter 01/11/2010
1114) Eliminated potential hang of status by having it read process info using
112non-blocking reads
113Scott Hunter 02/12/2010
1145) Should now reliably create its logfile
115Scott Hunter 03/01/2010
1166) Smarter display of arguments in Unix Status display (i.e. dropping trailing
117.d if that is all that makes line too long; truncating last argument from front,
118replacing dropped chars with <; use trailing > to indicate 1 or more trailing
119arguments not displayed)  Also added optional .d command, maxStatusLineLen.
120Scott Hunter 03/04/2010
1217) Consolidated unix code for stopping modules; will first try setting a flag
122for the module to self-terminate, then try sending a terminate signal (which
123modules can be written to ignore); finally, if new optional .d command
124HardKillDelay is non-0, will use a kill signal which CAN'T be ignored.
125Scott Hunter 03/10/2010
1268) More consistent Status display between Windows & Unix; in particular,
127showing arguments to the right.  Also reports modules that did not properly
128start in both as NoExec
129Scott Hunter 03/25/2010
1309) Better enforcement of column widths in status; allow process name to be
131truncated as necessary
132Scott Hunter 04/05/2010
133
134nmxptool - updated version 1.2.4
135Matteo Quintiliani 2009-12-22
136
137waveman2disk: Add support for consolidating multiple stations/channels in one
138wfdisc file
1392010/01/11 Tim Zander
140waveman2disk: allowed no logging by running it interactively like this:
141
142        waveman2disk waveman2disk.d n
143This effectively allows a user to run waveman2disk outside of an EW environment.
144Paul Friberg 2010/03/25
145
146carlsubtrig
147Allowed a way to point to the trig_id.d file instead of having it in a stock
148location: TrigIdFilename is the new directive. If this is used, the line
149@trig_id.d should be commented out from the .d file
150Paul Friberg 2009/08/28
151
152CHANGES TO CONFIGURATION FILES and DESCRIPTOR FILES:
153****************************************************
154
155startstop:
156Added optional HardKillDelay (# seconds before using method of last resort when
157killing a module in Unix) and maxStatusLineLen (maximum of individual module
158status lines)
159
160KNOWN BUGS:
161***********
162
163liss2ew:
164liss2ew has been observed intermittantly producing malformed
165TRACE_BUF messages.  Currently the conditions for causing this problem
166are unknown.  Due to this, liss2ew should be treated as suspect.  Use at
167your own risk.  JMP 6-18-2003
168
169wave_serverV:
170appears not to reply to requests for a single sample of data. I noticed
171when testing wave_viewer, that if the start time and end time of a request were
172equal (in which case there should be one sample of data) then wave_serverV did
173not reply to the request (ASCII request) at all. No Data, No Flags, No Reply,
174No Nothing.  It needs to issue a reply to every request.
175DavidK 09/25/01
176
177Automatic restarts of adsend (using the "restartMe" line in the descriptor
178file) can cause an NT system to hang. Therefore, you should never
179use the autorestart feature with adsend, but you should bring down
180the entire Earthworm system if adsend needs to be restarted.
181LDD 5/31/2000 Comments added to adsend.desc, but leave this warning here!
182
183libsrc/utils/site.c: The strings used for station, channel and network are
184required to be fixed length with trailing spaces added to short names. If
185the strings given to site_index do not have these trailing blanks, SCN names
186will not match. This is not documented anywhere.  PNL 10/15/00
187
188
189KNOWN DEFICIENCIES:
190*******************
191ringtocoax:
192In Windows NT, the time resolution of sleep_ew() is about 16 msec (one clock
193tick).  On Solaris, the resolution is about 10 msec.  This is a problem for
194ringtocoax, since packet delays need to be set to a few milliseconds.
195
196threads functions: The KillThread function on WindowsNT and Solaris
197terminate the thread without ensuring that no mutexes are held. If a thread
198holds a mutex when it dies, no other thread can get that mutex. PNL 1/12/2000
199
200carlsubtrig:
201The system time must be set to GMT and ew_nt.cmd must have
202TZ=GMT for carlsubtrig to work.  Comments in ew_nt.cmd done 5/25/00. Barbara
203       
204localmag:
205needs GMT set on the system
206
207ew2seisvole:
208on NT, exits with horrible crash when system is stopped.
209
210NUMBER OF RINGS LIMITED ON SOLARIS:
211Under Solaris 2.6 (and probably other versions as well), the maximum number
212of shared memory segments is six. This means that on an out-of-the-box machine
213you can only configure six rings. If you try to configure more than that, you
214will see a cryptic message from tport_create about too many open files.  The
215fix to this problem is to add the following lines to the /etc/system
216file, and then reboot the system.
217
218 set shmsys:shminfo_shmmax = 4294967295
219 set shmsys:shminfo_shmmin = 1
220 set shmsys:shminfo_shmmni = 100
221 set shmsys:shminfo_shmseg = 20
222 set semsys:seminfo_semmns = 200
223 set semsys:seminfo_semmni = 70
224
225This allows for 20 rings.
226
227     Lucky Vidmar (7/6/2000)
228
229
230TRACEBUF messages.
231The definition of `endtime' of the TRACEBUF message is not documented.
232Some programmers are taking it as the "expected start time of the next
233TRACEBUF packet (if the sample interval is uniform.)" The more accepted
234practice is that `endtime' is the time of the last sample of the current
235TRACEBUF packet; that is, one sample interval less than the expected
236start time of the next TRACEBUF messsage. Using this last definition, if a
237TRACEBUF packet has exactly one sample, then its starttime and endtime are
238the identical. Clearly this distinction needs to be documented. The file
239waveform_format (in the /home/earthworm/DOC directory) gives no specifics
240about start or end times.  PNL 1/24/01
241
Note: See TracBrowser for help on using the repository browser.