| View previous topic :: View next topic |
| Author |
Message |
mteel
Joined: 30 Jun 2005
Posts: 435
Location: Collinsville, TX
|
 Posted: Sun Jan 15, 2006 10:31 am Post subject: wview 3.0.0 Released |
 |
|
URL: http://www.wviewweather.com
Disposition: Highly Recommended
3.0.0 01-15-2006
--------------------
radlib version required: 2.6.1 or newer
1) Added support for new HTML tags --stationDateMetric--, --dailyRainMM-- and
--baromTrend--. See parameterlist.txt for the exact tag expressions to use
in your templates.
2) Added AWEKAS support. Added html template files awekas_wl.htx-us and
awekas_wl.htx-metric to the examples/html directory in the distro and added
a new User Manual section for setting up AWEKAS with wview.
3) Totally reorganized and rewrote the station interface code to abstract the
station interface. This is done in preparation for Vaisala WXT510 support in
wview. The daemon directory is gone, replaced with a stations directory
which contains a common directory for all station-generic code and a
VantagePro directory for the Vantage Pro-specific interface code. Care was
taken to minimize the code required for station-specific interfaces to ease
the implementation of new stations in the future.
4) Added a new station interface API in stations.h. Clearly defines the
functions which must be implemented for specific station implementations.
5) Added a new storm rain API (stormRain.[ch]). wview now computes rain storm
events based on a 0.05"/hour rain rate start trigger and an idle (no rain)
condition of 12 hours to determine when the storm ends. Vantage Pro storm
values are no longer used so this is available for all station types. Added
a new HTML tag --stormStart-- which retrieves the storm start date/time.
6) Added a new sensor API (sensor.[ch]). This provides a robust API for storing
and retrieving sensor highs, lows and averages for any time frame (hour,
day, week, month, year).
7) Significantly enhanced the computed data API to use the sensor API for HILOW
data storage. HILOW computation and storage are now done generically by
wview - the Vantage Pro HILOW packet is no longer used. Sensor HILOW values
are exported to a binary file in the archive directory so they can be
restored when wview is restarted.
8- Added a new sun/civil/astronomical rise and set API. This includes a method
for "midday" and "day length". New HTML tags were added to access these new
values. The Vantage Pro sunrise and sunset values are no longer used so that
these values are available for all station types.
9) Renamed all HTML templates and references so that they do not contain the
words "Vantage Pro".
10) Enhanced the wviewconfig script so that it queries for station type and
does station-aware configuration.
11) Added new HTML tag --forecastIconFile-- which just inserts the name of the
icon file without any HTML img helper text.
12) Added native wview support for Rain Season Start Month. Added new parameter
to wview.conf and support in the wviewconfig script so it will create the
entry in wview.conf if it is not there. This only effects Yearly rain, ET
and rain rate calculation. Added a new HTML tag --rainSeasonStart-- so the
start month can be used in templates. Added it to all example home page
templates and the alamanac template.
13) Added sanity check logic and requirement to the station initialization so
that the station/user config setting for archive interval matches the
interval found in the archive files. Startup will not continue if they do
not match. This prevents accidental corruption of the archive files.
14) Added new HTML tags --intervalAvgWindSpeed-- and --intervalAvgWindChill--
along with the code to support them. These provide the most recent archive
interval average values for wind speed and wind chill.
15) Reorganized the User Manual and added content to the Building and
Troubleshooting sections. Added new content to the UPGRADE file.
16) Enhanced almanac, Current, Daily, Monthly and Yearly HTML templates to
include java buttons at the top of the page for site navigation. Created
alamanac_Plus.htx so it may be migrated in the same way the other files are
for extended sensor support. Updated the example html-templates.conf file.
Removed the davisticker applet. |
|
| Back to top |
|
 |
chrisale
Joined: 09 Nov 2005
Posts: 187
|
| Posted: Sun Jan 15, 2006 11:24 am Post subject: |
|
|
holy cow!
Quite the update! thank you very much for the awekas related metric stuff and that new storm functionality!
No doubt my wife will be sending daggers in your direction as I spend the rest of the day upgrading my station.
You're familiar with my rather... hacked... setup.  Any gotchas other than the regular changes to source files? |
|
| Back to top |
|
|
mteel
Joined: 30 Jun 2005
Posts: 435
Location: Collinsville, TX
|
| Posted: Sun Jan 15, 2006 11:43 am Post subject: |
|
|
Read the UPGRADE and ChangeLog files or view them online...
The configure options changed a bit, see the User Manual for those.
Other than what is documented there, no gotchas to my knowledge. |
|
| Back to top |
|
|
chrisale
Joined: 09 Nov 2005
Posts: 187
|
| Posted: Mon Jan 16, 2006 10:13 am Post subject: |
|
|
| the only thing I didn't catch right off the bat was the radlib 2.6.1 requirement. otherwise, looks like things are going... |
|
| Back to top |
|
|
NorwegianWould
Joined: 07 Jan 2006
Posts: 23
|
| Posted: Wed Jan 18, 2006 4:31 am Post subject: Re: wview 3.0.0 Released |
|
|
| mteel wrote: |
7) Significantly enhanced the computed data API to use the sensor API for HILOW
data storage. HILOW computation and storage are now done generically by
wview - the Vantage Pro HILOW packet is no longer used. Sensor HILOW values
are exported to a binary file in the archive directory so they can be
restored when wview is restarted.
|
I commend you on the new version. However, I do not like this (unless I have misunderstood). I want to be able to get at highs and lows that occur when my computer is off-line but my weather station is not. One of my main motivations for acquiring the Davis station was because of the datalogger as it allows me to do things that other weather stations cannot - namely run unattended - and retrieve historical data later. This will also mean that new wview users starting wview halfway through the year on a station that they have had for some time will only get HILOW data from when they first started wview? The abstracted version of wview seems to be heading towards a generic minimum denominator. My other weather station will run in a place with little communication and data will be retrieved infrequently by modem. I do not want to run a seperate computer at this remote location to do historical data logging. I also have a colleague with a need to set up a station for their kite surfing club, which will be similarly placed - so not just me that has this need.
wview may not be suited for this, but I was planning to talk to you about writing an executable based on wview that does a single data retrieval, and then exits when this is completed. If the HILOWS record is removed then I am not sure that this is going to be as useful as I thought.
Is it possible to reinstate HILOWS for the VP2 and populate the archive from that on startup?
As I said I could have misunderstood this since the almanac page is reporting year high/low values back to before I ran 3.0.0, and I don't know how!
I also see that the almanac is not correctly reporting monthly low temperature: http://neshamn.no/WeatherStation-VH87/almanac.htm (this is from an unaltered wview template rather than my own non-template site)
The date is munged up, and -17.8C has not occurred (yet). The monthly hi/low values should be the same as the yearly. Any ideas?
Jeremy |
|
| Back to top |
|
|
mteel
Joined: 30 Jun 2005
Posts: 435
Location: Collinsville, TX
|
| Posted: Wed Jan 18, 2006 7:47 am Post subject: |
|
|
I guess you don't get it...
The Vantage Pro stores archive records even when wview is not connected or running. These archive records contain high and low temperature as well as high wind for the archive interval. These records are used by wview to update the HILOW values when it starts up the next time.
So your only issue with this new approach would be if you don't synchronize with wview within the time allotted based on the data logger capacity (8.8 days for a five minute archive interval, longer for bigger archive intervals). Also note that high and low values taken from the archive records (and not observed real-time by wview) will have a time stamp of the end of the archive interval in which they occurred.
| Quote: | | Is it possible to reinstate HILOWS for the VP2 and populate the archive from that on startup? |
First of all, it would not be used to populate any archives, it was formerly used to display HILOW values for HTML tags and limits in buckets and dials. Even then, the VP HILOW packet is inadequate and many HILOW values were computed based on the archive records (as they all are now). So it isn't as big a change as you might think.
So of the majority of HILOW values for day, week, month and year, most were already being computed from the archive records. Now all are, and no, I'm not going back to using the few that the VP HILOW packet provides. It was always bothersome to me to use a few from that packet and the rest from the archive records. Now they all come from the archive records. wview now stores the HILOW data in a file named sensorstore.bin which it imports when restarted so that HILOW timestamps with better resolution than the archive interval that were observed real-time by wview persist across stops/starts of wview. Think of that file as the NVRAM storage formerly maintained on the VP console (but with much more info).
| Quote: | | As I said I could have misunderstood this since the almanac page is reporting year high/low values back to before I ran 3.0.0, and I don't know how! |
Because it uses archive records, as it always has for most values.
| Quote: | | The date is munged up, and -17.8C has not occurred (yet). The monthly hi/low values should be the same as the yearly. Any ideas? |
This could be a bug... Although four of us beta tested this and it did not show up, one of the beta testers experienced the same thing when he installed the release version - looks like low temp is getting messed up for one archive record reading...
-17.8 C equates to 0 F ... so maybe the low temp value is getting initialized to 0 errantly. I will look into this... Did that -17.8 occur at the time you started 3.0.0 for the first time?
Mark |
|
| Back to top |
|
|
NorwegianWould
Joined: 07 Jan 2006
Posts: 23
|
| Posted: Wed Jan 18, 2006 8:44 am Post subject: |
|
|
| mteel wrote: | I guess you don't get it...
|
Clearly...
| Quote: |
Because it uses archive records, as it always has for most values.
|
Ok, so I understand, and I saw entries in the log as these were downloaded. Even though it's no effective change since 2.0 I would still be concerned that users with an old VP2, starting to use wview for the first time, halfway through the year might not see consistency between their console/envoy and what wview is reporting.
| Quote: |
This could be a bug... Although four of us beta tested this and it did not show up, one of the beta testers experienced the same thing when he installed the release version - looks like low temp is getting messed up for one archive record reading...
-17.8 C equates to 0 F ... so maybe the low temp value is getting initialized to 0 errantly. I will look into this... Did that -17.8 occur at the time you started 3.0.0 for the first time?
|
Does seem to be a bug, and shows up with month lows for temperature, dewpoint and humidity.
I repeated this by stripping out all /etc/wview and /var/wview and starting from scratch. The problem is still there.
Regards,
Jeremy
[/quote] |
|
| Back to top |
|
|
mteel
Joined: 30 Jun 2005
Posts: 435
Location: Collinsville, TX
|
| Posted: Wed Jan 18, 2006 9:11 am Post subject: |
|
|
| Quote: | | Even though it's no effective change since 2.0 I would still be concerned that users with an old VP2, starting to use wview for the first time, halfway through the year might not see consistency between their console/envoy and what wview is reporting. |
Yep, well if they do not include their Weatherlink-generated archive files in /var/wview/archive before starting wview, some HILOWS may be different. Why wouldn't they include their archive files from before they used wview? If they do, there may be slight differences in a few values because the Weatherlink archive file record does not store many extreme values, mostly just the average over the archive interval. I would love to ditch that format for something more comprehensive, but that would leave VP users out in the cold. I think this is a reasonable compromise.
| Quote: | | Does seem to be a bug, and shows up with month lows for temperature, dewpoint and humidity. |
As I said, I will look into it. There are so very few bug reports with wview and I always address them promptly. Roughly 60% of the code was modified/gutted/added in the 3.0.0 release, so I am not surprised that something slipped through the cracks. Again, this did not show up in beta testing.
Mark |
|
| Back to top |
|
|
mteel
Joined: 30 Jun 2005
Posts: 435
Location: Collinsville, TX
|
| Posted: Wed Jan 18, 2006 10:02 am Post subject: |
|
|
I just noticed that your Up Time is 4 days and some change - 3.0.0 was not released 4 days ago!
This tells me you are not getting all wview daemons stopped with the wview start/stop script, you have some rogue daemons hanging around from starting them individually or when you had a failed initial start, etc.
Read the UM section on Troubleshooting to see how to confirm that all wview daemons are stopped and the IPC data is clear before restarting wview. Under normal operation, the wview start/stop script will stop all daemons properly and the IPC data will be properly destroyed.
So, your conclusions concerning "even after wiping ..." may be tainted by lingering daemons or IPC data... |
|
| Back to top |
|
|
NorwegianWould
Joined: 07 Jan 2006
Posts: 23
|
| Posted: Wed Jan 18, 2006 4:07 pm Post subject: |
|
|
| mteel wrote: | I just noticed that your Up Time is 4 days and some change - 3.0.0 was not released 4 days ago!
|
The uptime reported is exactly the same as the result from 'uptime' on the computer, and seems not to be anything to do with the length of time that wview has been up. My wview does start on boot, but there are no processes hanging around, they all started at 1600 today. How does wview calculate the uptime? Is there a file somewhere that it looks for?
Thinking about the -17C problem earlier - my VP2 has only been switched on since the end of December, so the archive records will not be filled up yet. maybe this is where the spurious value is coming from ?
Jeremy |
|
| Back to top |
|
|
mteel
Joined: 30 Jun 2005
Posts: 435
Location: Collinsville, TX
|
| Posted: Wed Jan 18, 2006 4:13 pm Post subject: |
|
|
radlib keeps track of when the first radlib process was started for a given radlib system (among many other things). It should definitely not be the same as your computer uptime (off by seconds for sure). If it is, you have changed how the HTML tag for systemUpTime is populated.
Is all your IPC data clear (see the UM) when you think wview has been stopped?
It is not possible for your uptime to be correct - again, 3.0.0 was not released when it says you started it... |
|
| Back to top |
|
|
NorwegianWould
Joined: 07 Jan 2006
Posts: 23
|
| Posted: Thu Jan 19, 2006 2:20 am Post subject: |
|
|
| mteel wrote: |
Is all your IPC data clear (see the UM) when you think wview has been stopped?
|
There were indeed some ipcs hanging around. Better now...
Still -17C... I'll try to do some debug output later...
Jeremy |
|
| Back to top |
|
|
NorwegianWould
Joined: 07 Jan 2006
Posts: 23
|
| Posted: Fri Jan 20, 2006 7:36 am Post subject: |
|
|
| NorwegianWould wrote: | | mteel wrote: |
Is all your IPC data clear (see the UM) when you think wview has been stopped?
|
There were indeed some ipcs hanging around. Better now...
Still -17C... I'll try to do some debug output later...
Jeremy |
Dear all,
This problem seems to be resolved in the next release (3.1.0). I have tested it on my site. Infact just running 3.1.0p2 once seems to have fixed it for 3.0.0 too!
Jeremy |
|
| Back to top |
|
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
