Zorro 2.53 is now available on https://zorro-project.com/download.php under "New versions".

List of new features: https://manual.zorro-project.com/new.htm

This release candidate will become the official release when no bugs are found in the next time. Please test everything and report any issues here! As usual, if you're the first one to find a new and serious bug in the release candidate, you can get a free Zorro S subscription or support extension.

Please see attached image (and t6 file for verification -> remove the .csv extension, forum doesn't let me attach a .t6 file so had to append .csv).

I see 2 problems with the chart's X axis:
1. inconsistent width in ticks: sometimes 1 division is 1 day, sometimes 2
2. invalid ticks: a candle is reported on Nov 6th but that was on a Sunday (this is a recent candlestick chart for AMZN)

The x axis on a financial chart is not determined by the Zorro version, but by the timestamps in your historical data.

Please see attached timestamps (after converting the .t6 file I've attached in the previous message to .csv). There is no entry for Nov. 6th.

Yes, there is. You can see in your list that you got a long bar from Nov 4 to Nov 7. That's a weekend bar. And since you have probably plotted your chart with BR_WEEKEND activated, you've suppressed that bar until Sunday 23:00 UTC when the new week begins. That's likely your Nov 6 bar. It has nothing to do with version 2.53.

https://zorro-project.com/manual/en/barmode.htm

Either change the default weekend to the real AMZN weekend, or simply resf(BarMode,BR_WEEKEND). I'll put a remark in the manual because other users might once encounter the same issue.

I actually didn't write any script. I simply set t6 as a file extension to be opened with Zorro. So when I double click the .t6 file Zorro launches automatically and displays the chart.

I've checked my process list and this is what gets executed on the double click:



From your reply I understand that the timestamps that it sees in .t6 are UTC and when displaying it converts it to my timezone which is UTC+1. That would explain the differences between chart and table. Maybe don't make any conversions by default?

Timestamps are converted only when you set up a time zone in a script. Not by default.

They need be UTC when you use multiple time zones but otherwise the time zone does not matter to Zorro. Still, for not mixing up data with different time zones better have them all in UTC.

From the Zorro UI looks like Chart.c is selected when the candlestick chart is displayed (I just double click the .t6 file to launch Zorro). When looking in Chart.c (one of the default scripts in Strategy/ folder, I don't see any timezone variable being set).

I've added the following line inside the if(is(INITRUN)) block and got this output:

The download link I received in the email is downloading version 2.50.2. Is there a different link for this specific version?

Never mind. I downloaded it from https://opserver.de/down/Zorro_253.exe

First bug reported, by 2 users at the same time: PRELOAD didn't work properly when historical data was too recent.

This is fixed in version 2.53.3, same URL.

New bug report with duplicate IDs in the recent TWS API. It happens under some circumstances when closing positions.

Zorro 2.53.6 contains a workaround. The new IB plugin can be separately downloaded from https://opserver.de/down/ib.zip.

Running Z10 "Test" in Offline mode,
I got this one:


Z10 .x ..
Z10.16: PH H 4 B 0 V 1
Loading 12 assets..
Error 111: Crash in run
Error 047: No bars to plot


Switched accounts and brokers, same thing.


I installed 2.53.6 on top of 2.50.2

Hopefully i'm the only one.

Tried installing 2.53.6 to a new install path.

new path | original path
------|---------------|----------------
Z10 | error 111 | error 111
Z9 | error 111 | error 111
Z12 | ok | ok

The Z systems are not yet ready for 2.53. They will be ready when it becomes the final release.

Happy New Year to @jcl & all the Zorro crew!

Running 2.5.3.6 on Windows Server 2012 VPS:

Possible Bug:
assetHistory Remarks states "Price history is only downloaded in the INITRUN or in the main function, and only when existing price history is not already up to date". In my tests downloading EOD in a main function, if the Asset file does not exist, assetHistory incorrectly returns a file-not-found error. I believe this is a bug, since the non-existent Asset price history is clearly not up to date! (For others who run into this in the interim, a workaround is to add the OVERRIDE flag...)

And... Possible Manual Bugs and related (which is why they're in this post) Enhancement Requests:

Possible Manual Bugs:
- Help -> Search -> "history": does not give a "History" result. Top result is "assetHistory", which does contain a link(s) to "History"; although only taking to the top of the "Predefined and user-supplied strings" entry. There should be a "History" result.

- History entry states "If not set, defaults to .t6 and the HistoryFolder". From prior tests, this is not correct. In my tests the correct statement is more like "If HistoryFolder is the default value of "History", History is not initialized. If HistoryFolder has been set to a different value, History is initialized to that value. When searching for Assets files, the path defaults to HistoryFolder, the suffix to nothing, and the extension to .t6. If History is set by the user, ...".

Related Enhancement Requests:
- Always initialize History to the value of HistoryFolder. (The user can still change it if desired.) This provides a script with the value of HistoryFolder without unnecessary gyrations. This is related to the next item:

- On the Performance Report, include the value of History so the user looking at it later knows what data was input. This would also allow a given script to be easily run against different data, with that difference clear.

- On the Performance Report, the System's details should also be shown. Esp. given the noted differences between Windows & Linux/Wine (& other?).

Note that while a user's script could do these, the user cannot do these with the Z Systems.

A file not found error by assethistory normally means that the file exists, but is not readable. Please check. Otherwise, please let me know if it only happens with your script, or also with download.c.

Thank you for the other suggestions.

Thanks @jcl.

The .t6 file was definitely not there. It was actually files - more than 1.

Thanks.

Hmm. Do you get the same message with the dowload script? Or only with yours? If so, can you post it?

I ran 2 tests:

Test 1:
- AAPL.t6 file exists.
- Trade Download
- Change to EOD
- Enter AAPL
- Click Download
Result: Download runs and AAPL.t6 modified date is updated

Test 2:
- Delete AAPL.t6
- Trade Download
- Change to EOD
- Enter AAPL
- Click Download
Result: "!Subscribe AAPL" and "AAPL history up to date" print in the message window. No AAPL.t6 file is created.

Clarification: My script tries to open the .t6 file, so that was the source of the file-not-found error. The actual Zorro problem is not creating the .t6 file...

Thanks.

The message means that the file exists and is up to date, i.e last updated less than a day ago. And the file name has normally a year number.

Thanks @jcl. Regarding each sentence:

- Yes, I know what the message means. But in this case it isn't accurate. The file was deleted before the test was run.

- This is EoD data, so no year in the file name.

Thanks.

New version 2.53.9. Some issues with BR_LOCAL have been addressed. This is probably the final version.

Something wrong. The https://opserver.de/down/Zorro_253.exe downloads a Zorro_235.exe and installs 2.3.5 version

The link that you used is for version 2.3.5, you can see it on the URL you posted.

If you are trying to download the latest version you can try https://opserver.de/down/Zorro_setup.exe