| Pages:
1
..
23
24
25
26
27
..
54 |
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
.82 then .83 was released. has fixes for mc ode that didnt compile.
.82 was false flagged as a virus by MSE, so we made .83 to have a different signature.
.83 has minor change in random data naming
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
some users have had the opt settings reset to default in workers. I dont understand how of why this happened.
This will change share keys back to public cloud. If this happens change them back to your private share key, and check your cache settings if you
have chosen to make them larger in the past
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Later note. I know how this happened. Unused file on workers only are now being cleaned up. There was a bug in this code.
Fixed in next build. Apologies for the issue.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
55.84 released. Fix for file cleanup bug
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
I am getting no connection between my workers and managers, and all my workers, public and my own are used by others right now and i get no workers
for my own development, my managers is searching for over an hour now for workers, been looking for the 55.84 update, but not seeing any 55.84
updates.. what to do?
I erased all sharekeys for my workers and then they seem to be connected, however do i need to do this on every new workers that starts up ?
i saved the app setting as default for the workers, every time i change i need to save it it seems
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by Daniel UK1  | I am getting no connection between my workers and managers, and all my workers, public and my own are used by others right now and i get no workers
for my own development, my managers is searching for over an hour now for workers, been looking for the 55.84 update, but not seeing any 55.84
updates.. what to do?
I erased all sharekeys for my workers and then they seem to be connected, however do i need to do this on every new workers that starts up ?
i saved the app setting as default for the workers, every time i change i need to save it it seems |
if you set the worker to use local network etc, and you an put in a share key(not needed if first step is done) then save the app settings. then kill
all workers, and open them via rm
Regardless you should have picked up some free gcloud workers if you have the free cloud share key as well. Worst ccase we can do teamviewer to look
at this.
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
Peter, When you say "set the worker to use local network" which setting control this?
If i pick one worker (started from RM or not ) and save the app setting as default with no keys used meaning key field empty, which as i understand
shall be the used if i understand your post about this issue a few days ago... will this mean all started workers going ahead will use this app
setting i just saved which i assume is correct? started from RM or not?
EDIT, SEEMS TO BE SOLVED WITH .84
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
coming in the next day or two, lots of bug fixes including some systems with indicator cross entry type not trading,
and improvement on miss-match issues.
I should be back in Australia < 24 hours, so will be given a bit higher service levels than the last week.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
.87 released to a few beta testers.
Bug fixes for no trades in ts, and tc <> gsb.This build is worse for MC - hence no public release.
https://www.dropbox.com/s/k76qcecjbrmn5h9/GSBWorker.55.870.z...
I am flying back to au today, so no time for me to test.You need to update workers and manager
Other fixs and mc fixes are being working on.
Peter
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
50.87 released.
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
Not receiving 50.87, has it been pushed out trough RM? Thanks
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
My mistake
released now
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
coming soon
55.90
release notes for .88 & .87
* 1.0.55.88 / 2019-11-07:
Fixed bars-ago reference of secondary-data indicators in GSB to match TS, ex. RSI[1] of Data2. Note: not tested with older bars or higher data
(ex. RSI[3] of Data5).
Fixed the three new cross-based entries (AnyIndicatorCrossed, AllIndicatorsCrossed, and NoConflictsCross), as a result of the bars-ago reference
fix. Note: AnyIndicatorCrossed only tested.
Fixed CrossAndCloseD entry type in GSB and TS.
Still to be done, Partially-implemented BarsAgo (applies to indicators in GSB only, not to TS, MC, or NT yet) in IndicatorsTester,
* 1.0.55.87 / 2019-11-05: Multiple fixes for the mismatch issues
Fixed TS/MC script to use inputs/vars in function calls. Note: this is a recent bug where GSB exported TS/MC script with hard-coded, static values
(instead of inputs or vars).
Hiding ExitModesAll from trial users.
Writing MC debug files to a .mc.txt file to differentiate from TS files (of .ts.txt extension).
Hiding indicators vars in the TS script (that uses inline-calls) to avoid users’ confusion. Note: MC still used vars-calls of
indicators/oscillators functions.
Moved the result and decision calculation outside the start-end date check in TS/MC script to match GSB’s HL2 and CS2 norm types. Note: HH3 and
CS3 are currently Alpha user only and disabled by default with plans to remove altogether in later builds.
A minor improvement in debug script readability/editability for better diagnosis.
A minor updates to IndicatorsTester strategy script.
Fixed the mismatch caused by the built-in SEOC (when enabled), and renamed it from “BiSEoC” to “EL-SEoC”
Thanks received (1):
+1 emsjoflo at 2019-11-28 08:04:46
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
55.90 released. Has one new ts function and implemented BarsAgo (applies to indicators in GSB only, not to TS, MC, or NT yet) in IndicatorsTester,
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
55.93 under testing. Has a few bug fixes and data folder trees are shrunk bu default. I may not release till more significant improvements made.
Current build 55.90 seems quite good.
Support ticket placed to MC about their bugs that cause mc <> gsb was done today. 
The ticket was
"We have recently discovered two serious bugs in indicators calculation that return incorrect values for many, and sometimes all, bars.
The first issue appears whenever we call an indicator of another (for example, HighestFc(AverageFC)) of a secondary data (such as Data2, Data3, etc.).
This issue is mainly related to filtered data (or in other words, missing bars in the secondary data), which is quite common.
Moreover, the second issue is that the vars-call (as in value1=AverageFc; value2 = HighestFc(Value1) gives different values compared to inline-calls (as in HighestFC(AverageFC)), which is very confusing. After multiple tests, we have found that the vars-call is specifically wrong for some bars
after Data2 date-gap (missing-bar compared to Data1), but the inline-call is nearly wrong in most bars, and even in Data1, unfortunately.
In the attached workspace, we have used a test data (CL as Data1 and HO as Data2) and a small indicator (MC_Test_Data2) to plot both var-call and
inline-call and log them to files for an easy compare. "
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
50.96 under testing. (its failed)
It will have worker than can do wf, and build systems - but not both at the same time.
I think thats the most efficient use of cpu and ram.
Other small tweaks too.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
50.98 released
bug fixes, and some minor improvements
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
51.01 released shortly.
bug fixes
|
|
|
Carl
Member
Posts: 342
Registered: 10-5-2017
Member Is Offline
Mood: No Mood
|
|
Peter, i think it is 56.01?
Thanks received (1):
+1 admin at 2019-11-22 01:51:28
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Your correct. Next we are finishing off Ninja Trader.
|
|
|
Daniel UK1
Member
Posts: 470
Registered: 4-6-2019
Member Is Offline
|
|
| Quote: Originally posted by admin | 55.93 under testing. Has a few bug fixes and data folder trees are shrunk bu default. I may not release till more significant improvements made.
Current build 55.90 seems quite good.
Support ticket placed to MC about their bugs that cause mc <> gsb was done today.
The ticket was
"We have recently discovered two serious bugs in indicators calculation that return incorrect values for many, and sometimes all, bars.
The first issue appears whenever we call an indicator of another (for example, HighestFc(AverageFC)) of a secondary data (such as Data2, Data3, etc.).
This issue is mainly related to filtered data (or in other words, missing bars in the secondary data), which is quite common.
Moreover, the second issue is that the vars-call (as in value1=AverageFc; value2 = HighestFc(Value1) gives different values compared to inline-calls (as in HighestFC(AverageFC)), which is very confusing. After multiple tests, we have found that the vars-call is specifically wrong for some bars
after Data2 date-gap (missing-bar compared to Data1), but the inline-call is nearly wrong in most bars, and even in Data1, unfortunately.
In the attached workspace, we have used a test data (CL as Data1 and HO as Data2) and a small indicator (MC_Test_Data2) to plot both var-call and
inline-call and log them to files for an easy compare. " |
Peter, have you heard back from MC support in regards to the bug? i propose that someone also post it on their forum, i would be happy to do it,
however i am not sure i can argue my case for the bug since i am not a developer with proper knowledge of EL.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
There has been zero response from mc. I will confirm when I hear anything. Yes, im open to numerous people complaining - as that might help.
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
Next build of GSB has all but two indicators matching Ninja Trader.
Newest video is just uploaded. This is critical for all user to see.
Significant changes in methodology, plus what works and doesn't work.
the video. Please hit like
https://www.youtube.com/watch?v=6bPAvyMXIwA&feature=youtu.be
Thanks received (3):
+1 JasonT at 2019-11-29 17:01:59 +1 Daniel UK1 at 2019-11-27 07:15:32 +1 emsjoflo at 2019-11-27 02:26:36
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
under testing
1.0.56.04 / 2019-11-28:
Implemented LoadExtraPriceFiles(), ex. "Price Data.xyz.txt". This is so we can add price files (use in video demo) without over writing users price
files)
Expiry extended to Mar 2020.
Updated Show Favorite C and D to be visible to all users.
defaults changed..
Disabled defaults of all SF indicators except Close/CloseD and CloseBpv.
Updated EntryLevelMode default from Exact to SignedPower.
Updated EntryLevelValue default from null to be 4.
Added VolumeLessPrevAverageFc, VolumeOverPrevAverageFc, and FractalDimension to all except trail users, enabled and SF disabled by default.
Removed CrossOrCloseD entry mode from trial users access.
Enabled CrossDualLevels (visible to all except trial users) by default.
Removed/Hided OptimizePriceData from trial users.
Updated "Entry Params." and "Entry Params. 2" defaults from 0:200:5 to -10:10:1 (and for WF params. as well).
|
|
|
admin
Super Administrator
Posts: 5060
Registered: 7-4-2017
Member Is Offline
Mood: No Mood
|
|
56.04 released. 30 nov 2019
|
|
|
| Pages:
1
..
23
24
25
26
27
..
54 |
![[*]](images/davis/default_icon.gif){kind=icon}
