Ok it was the small print and it was me! Sorry for the inconvenience. Here it is again without the space. Take off the doc. Should run faster and show the cyclone directions with arrows. Sean would like confirmation that this works without problems, so download, test and confirm. Thanks.
Also uploaded to the Opencpn Beta File Thingie [rguser, rgpass]
Look under Plugins-Win-Opencpn-3.3.1303
Testing with MSVC++ in debug mode, there does not appear to be any data. I have put the data under ..opencpn-git/build/plugins/climatology/data/ Should it be under opencpn-git/plugins/climatology/data?
...therefore I have no output for errors.
Testing in Release mode, the data does appear properly.
All data values appear to work properly except
Sea Depth with Isobars on. It hangs up building the isobar map halfway through for some reason, even when put on "very course" and 5000 feet.
When I hit "Cancel" on the Isobar Map there continues to be high cpu use for a very long time and in one instance it stopped successfully, in other instances it seems to keep running with high cpu and I have to cancel program in task manager. Also found when I minimize and then maximize program, the isobar mapping will stop (but this is after a long wait) and then Opencpn 3.3.1303 seems to have recovered, leaving Seadepth checked and in Config Isobars is found unchecked (which is correct) because I had to cancel isobar mapping. I am wondering why isobar mapping does not complete? I have succeeded in mapping the sea depth isobar with 8000 feet and normal with the 05 version, when I try fine it also maps successfully. When I try 8000 feet and very fine, it hangs up. I can hit "cancel" and then wait just a little, minimize Opencpn and then maximize Opencpn and the "cancel" of Isobar mapping stops, with the Climatology pop left intact. I can then try another setting....on 6000 feet and normal isobar mapping seems to hang up... go to Minimize Opencpn and Maximize and the Isobar Mapping window is gone, the checkbox for Sea Depth is checked, and in the Config window the Seadepth Isobar checkbox is checked and set at 6000 feet, so it must have completed ok. --There must be some loop that does not finish, or a screen rewrite that is not refreshed or something. Tried 6000 feet course, isobar map hung up, then minimize opencpn and maximaze and the isbar mapping has completed ok.
Cyclones with arrows work great and seem to be faster. I can turn all on and have all months checked and it works very well on this computer. Slower computers may have different results but it seems much improved.
I tried Climatology v05 in MSVC++ debug mode again, this time with new compile and also copying the data into ...opencpn-git/plugins/climatolgy/data/ directory.
There still is no data available, wonder why. The data type check boxes remain are greyed out. Am I missing something?
Data if found in Release mode. Maybe we need a way to browse and point to the data folder? Wouldn't that be better? But plugin is initialized to be the normal data directory until user changes it.
Here is a newer version of Climatology. Everything appears to work with the exception of Sea Depth, which the Isbar Map (when enabled) when set towards "fine" and smaller intervals like 2000 feet will require heavy calcs and the graphic freezes... workaround is to "minimize" Opencpn and then "maximize" opencpn and the display has then been rewritten.
Take the doc off and make sure you have the data in the correct data directory please. (Also watch out for white spaces in the name! I've done it too, sorry.)
Maybe the correct solution is that sea depth is the only data unit that has nothing to do with climatology and can be removed from this plugin. I think it's basically useless anyway.
Maybe the correct solution is that sea depth is the only data unit that has nothing to do with climatology and can be removed from this plugin. I think it's basically useless anyway.
Sean as author of the plugin, I can certainly understand your point. However this data does provide some useful information regarding the massive size and depth of the oceans which may correlate to climatology, it provides context.
Having helped you with acquisition and compressing this and other data, I would simply like to keep the data as it is presented. I seldom/never use the Isobar feature under Sea Depth and we could put some limits on the range of use [preferred], perhaps, or just grey it out for the time being.
I happen to think that your optimization of the isobar procedure is working very well, but there is something in the code where this procedure is called that is not returning properly after a certain period of time has elapsed, and it is probably a very small change in that area of code that would allow the isobar map to complete and rewrite the screen with the new isobars. By the way "Cancel" works after a long wait, so that is ok.
I can set Sea Depth Isobars to "250 feet, fine" and the Isobar mapping progress bar operates properly up to just before 1 second, then it freezes.
Then when I minimize Opencpn, then maximize, the isobar mapping progress bar shows 1 second, ...I wait
Then minimize..maximize Opencon the progress bar shows 24 seconds, but is frozen....wait.
Then minimize...maximize Opencpn, the Isobar progress bar shows 45 seconds and is frozen.
Then min...max opencpn, the Isobar progress bar has completed and the Isobars are draw on the chart at "250 feet, fine"
So here are a few suggestions in priority.
1. fix the isobar mapping pause/freeze in sea depth with isobar checked.
2. limit the selection of interval, units and coarseness to one second or so it works ok.
3. grey out isobars in sea depth until this can be fixed.
Thanks for bearing with me on this. I think the data is useful.
Serge, I do not understand your point, Climatology is essentially a 30 year fixed database about the oceans. There are potentially other useful safety data-types that may not fit "climatology" exactly which have been considered
1. Shipping routes database by month like cyclones
2. SART calls database by month
There are other items which could be useful.
Quote:
I do agree with you, each pluging must work for a specific fileld and not embrace several.
I'm working on the translation Climatology your plugin.
Enclosed two zip files. Each zip file contains a version (not final ) language file in French .
I use the plugin version downloaded from post # 230 of this thread. I use the data file CL -DATA- 9.7z is in github ( see screenshot )
With "OpenCPN - climatology_pi_FL1_Valide_A.po" French version of the file, the data loading is not done correctly. See screenshot.
With another file, "OpenCPN - climatology_pi_FL1_Valide_A_vent_courant.po" , loading the data quite well.
Similarly, the loading is done if I do not use the French language.
The difference between my two versions of the language file is in the translation of two items: wind and current (without upper).
If I translate, I replace "wind" by "vent" and I replace "current" by "courant". But then, instead of trying to load currentxx.gz and windxx.gz files, the plugin tries to load the files "courantxx.gz" and "ventxx.gz".
And, of course, these files do not exist.
The problem would be the same in all other languages.
The wind atlas display shows the percentage of time wind is in each of eight directions as a length of the arrow, or in the case it is 35% or more a number is shown.\n
\n
The barbs indicate wind speed, each barb being 5 knots.\n
\n
The numbers in the circle, give percentage of calm (less than 3 knots) in blue or storm (above 47 knots) in red.
When I try to translate it, it gives this:
Dans l'atlas affiché, la longueur des flèches dans chacune des huit directions montre le pourcentage de vent durant la période. Si supérieur à 35 pourcent, un nombre est affiché.\n
\n
Les barbules montrent la vitesse du vent. Chaque barbule vaut 5 noeuds.\n
\n
Le nombre, dans le cercle, donne le pourcentage de calme (moins de 3 noeuds) en bleu ou un orage (plus de 47 noeuds) en rouge.
But when I try to record the po file to make the mo file, Poedit told me there was a problem. "Number of format specification in 'msgid' and in 'msgstr' does not math.
Is it a problem in english version ? and so in in french version.
Second :
In the pot file there are two item (89 and 90), "Very course" and "Course". With Pyteas54 that I ask, we thing that it should be "Very coarse" and "Coarse".
If true, it is necessary to correct the pot file. If not, sorry to ask this.
This is a pretty good idea, had not thought of it:
Quote:
I use the plugin version downloaded from post # 230 of this thread. I use the data file CL -DATA- 9.7z is in github ( see screenshot )
Its also available on Opencpn Beta File Thingie [rguser, rgpass] which you know about. Look for Climate Data 5
In MSVC++ Debug mode the data is not found, probably because of the way the pathname is done (just guessing)
Quote:
In the pot file there are two item (89 and 90), "Very course" and "Course". With Pyteas54 that I ask, we thing that it should be "Very coarse" and "Coarse".
If true, it is necessary to correct the pot file. If not, sorry to ask this.
Yes, you are correct, he means "coarse" as in "rough" versus "fine"
I have recently added ability to interpolate between months to an individual day. This gives a much smoother display, and also improves integration with weather routing (where I am developing several new algorithms to greatly improve accuracy of using climatology there)
To render the overlay maps, the opengl combine extension is used, so it may not work with all drivers, but it should work with most (core since opengl 1.3 which is ancient) Otherwise it falls back to using a monthly display. The arrows numbers isobars and wind atlas works at the daily display on all systems.
There are also several other important changes since previous versions (lightning support cyclone map key, and bug fixes) so testing is appreciated.
Also, please post the latest translations. I have only an incomplete french one.
I can build linux (amd64, i386, armhf-armv6, armhf-armv7) as well as windows mingw, so basically any platform gcc can target, but for other platforms (ie: windows visual studio/ macosx clang-llvm) it is not possible to cross-compile from a linux system. I would like installers built which can be used with any released beta opencpn if possible.
Sean, How do I build an "installer". I can compile MSVC, but building an installer has me stumped. I did some searches on this but did not get anywhere... just a few hints to get me headed in right direction please. Then perhaps I can build the installers needed.
Linear Mode
Recent Discussions