WXSIM 2019, Build 1.0 Released!

Sorry about these errors! Which happens first, error 70 or error 6? (I would think 70 might happen AFTER 6, but I’m not sure). I don’t know if it’s something about the new version, or if there’s been a data problem lately. If you see it again, please send me custinit.txt, rgcode.txt, localdat.txt, localcal.txt, wxsmini.txt, wdata.txt, cty.fdt, fdata.txt, and cor.txt, as those are probably all the files that could be involved. Also, if you revert to another version and can definitely show that the new version has the problem and the old version does’t, that would be helpful.

Thanks!

Tom

Tom the problem is intermittent and as I was not watching it I don’t know for sure which error came first, as a guess it was the 70. Has not happened for acouple of days now but am wondering if it could be related to when lite runs the analysis which is not every night. I’ll watch it over the next few days and will send the files if it does happen again.

Stuart

Well this morning WXSim managed to crash my PC - it hung up with an error 480 from WXSim and I needed to power off my PC to recover as nothing worked. I have reverted to the previous version for now… I have gathered the files mentioned previously but with this error I dont know if they are valid.

Stuart

Have been running this build for 10 days - no problems so far.

Hi Everyone,

Build 1.0 running well here.

Thanks Tom.

John

Hi Tom,
As you know I’m in a rather rough situation at present with past weeks of living in thick smoke and fires threatening us (all occurring at the wrong time as we are a third way through preparing our property for sale due to my health declining, which will render the weather site offline eventually = as we have no idea where we will be after we sell). So at last yesterday I had a breather from the smoke as it cleared slightly here and from the exhaustion of 24hr monitoring (though still on Watch and Act). I found this thread and decided to update wxsim even though I will shut down in due course and although had a few teething issues, got it running OK.
However with ref to your comments quoted below, I found that my personal mysql script failed due to the inclusion of a new field in latest.csv this new field is the UTC date/timeUTC stamp though does not have a header.
I have rectified my own personal scripts (not used by others) so all good.
If you add a Header it will not impact mine. Though surprised I could not find any mention of this on the forum aside from my original post and Wim’s reply at http://discourse.weather-watch.com/p/520860 which was all prior to this upgrade.
I also found a few other things in other .csv files which I’d like to question one day (when all this fire/move/etc is over, though no capacity left to go into detail on those.

[quote]
Version 2018 Build 1.3, (September 19, 2018) had the following change relative to Version 2018, Build 1.2 (described below):

(1) In the latestshort.csv file, the time will now be consistent with the Daylight Saving Time setting. Previously, during DST, the date would change according to DST, but the time wouldn

Regards Tony

These information have left me worried.
Above all, I wish you better in your health.
I also hope that your problems end with the risk of fire, smoke, sale of the house …

I’m sorry that you have to close the page and, I’m sure you’ll be back soon.
I am an enthusiast of your scripts for WxSim.

Effectively, with the WxSim update of October 2018, your scripts stopped working on my page.
In January 2019, I managed to solve the problem as indicated in:

http://discourse.weather-watch.com/p/523652

I pray for the solution of all your current problems.

Greetings from San Sebastian, Spain

Javier

Hi Javier,
As always I thank you for your kind words.
Though at this stage I feel I may not have the capacity to continue after our eventual move, only time will tell.
Sadly I thought all was good after recent post, however I am now getting error 6, all was brilliant before the upgrade and also at a time of need while during ongoing fire situation I feel now I should have left well enough alone.
I am in a position where I am not able to monitor wxsim/website as I used to due to current situation so when I check it now and then I now find errors. such a pity, if it continues I’ll just shut it down as that is not my standard displaying incorrect old not updated data.
Will see what happens, kind regards,

Tony I had issues with the 2019 release and since going back have not (touch wood) had any.

Stuart

Hi Stuart,
Thanks matey, will try that, if it fails I’m done.
Tom as per your request on error 6 which I get on some runs, wonder if it is Cumulus or your wxsim source data or maybe 9 day vs say 4 day forecast.
zipped files attached.
Kind regards,


wxsimissues19.zip (380 KB)

Hi reinstated old version and did manual run, all good.
Off to bed now, will see if auto runs are fault free when I awake.
One thing I did note is that in wxsimate, I was running :
Metar : proxy/wxsim (in old and new version)
Synop : proxy/wxsim (in old and new version)
RAOB : rap.ucar.edu (in old but (proxy/wxsim) in new version). did not realise till revert so not sure if it was an underling issue.

kind regards,

SYNOP is the only one I see in old version now with proxy value in. RAOB is

www.rap.ucar.edu

Stuart

Tony, thanks for the files from when you got the error. I’m working on it! :slight_smile:

Tom

Thanks Tom,
Old version forecast’s ran ok overnight, though I had to remove my fix for the additional UTC stamp in latest.csv (I forgot to last night).
Only issue I have now in old is that autolearn will not run, it crashes with error 62 input past end of file.
Smoke back thick as again today. :frowning:
Kind Regards,

Hi Tony,

I DID manage to reproduce the error, though I haven’t quite found the variable that’s overflowing, because I have lots of error trapping, that sends the error to the handler instead of leaving it where it occurred. However, I’m almost sure it is associated with the fact that you are running a long forecast, with 60 iterations per interval, which has the counter up to 15,900 when the error occurs. I think some related variable must be about twice that, and is an integer exceeding the bounds of such things (32,768, I think). I easily solved this by cutting back to 20 iterations per interval. In fact, I see no reason to use more than 6. This is under parameters.

I’m confused about why there would be any difference between the older and new versions here, though. I can’t think of anything I changed related to this. I also don’t know whether Stuart was seeing the same thing. In any case, I’d be interested to hear how the new version does with 6 (or as much as 20) iterations per interval. If that solves the problem, I can prevent this somehow.

With autolearn, does it crash immediately? Can you send me a screen shot? Can you send me alrninit.txt?

Thanks!

Tom

Hi Tom,
Greatly appreciated, just reinstalled upgraded version again and reduced iterations to 6, will see how it goes next auto run.
Tried autolearn again but same issue. yes crashes as soon as i click run immediately. have attached as requested.
Kind Regards,


alrninit.txt (94 Bytes)

Hi Tom,
I have found the reason for autolearn error 62, though can only assume there must have been a save issue somewhere during the upgrade and next run stage.
I found that the correc.txt file was not line format structured as it normally is. eg seemed to be centred vs left aligned with line breaks.
Any way lucky I had previous day backup, so added recent day line to old file and all ran perfect.
So now running new version no issues.
Strange about the iteration issue as I have been running at 60 for months now (just to see if it would differ or break) I like doing that.
Kind Regards,

Hi Tom,
Overnight runs went fine no errors, however this mornings 05:00 run (which is gfs 12z) failed after the 7th day with error 6 in wxsim.
I have run 9 days without error for years, so unsure what is causing it.
For now I have cut back to 7 days.
Kind regards,

Hi Tom,
another issue, my 11:00 (gfs 18z) run failed this time wxsimate unable to download (not sure if it was Metar (proxy.wxsim.net…) or RAOB (proxy.wxsim.net…) sorry did not capture the windows exception error.
So I changed source as per image below and all ran ok again, will wait and see what happens next run.
Kind regards,


Tom I’m wondering if you have managed to get to the issues which several of us have had with this release?

Stuart