Jump to content

Sign in to follow this  
CaptKornDog

The PMDG 777 VAS Data Collection Thread

Recommended Posts

Hello PMDG 777 Enthusiasts,

 

Background

Some of you know me from recently participating in the various VAS/OOM/leak threads involving the 777, as well as some of the lengthier ones dating back to late last year.  I had the pleasure of helping Ryan M. and the PMDG Team test the most-recent 777 release after a few of us (myself, pcubine, and others) identified what appeared to be a memory leak in a weird combination of circumstances.  Robert described this in pretty lengthy, accurate detail here: http://forum.avsim.net/topic/468259-07may15-detailed-discussion-of-resolved-777-vas-leak/ .  He used the term "pesky" to describe it, I certainly didn't choose such polite language trying to track it down :lol: .  (Robert, if you're reading this, and someone criticizes your team for "taking so long" on the 747v2, please feel free to blame me as the sole root cause :lol: ).

 

How we identified the source of the VAS leak was a very trying process.  Trying our collective patience, more than anything, I'd say.  By the time I started realizing I was having a problem, to the identification of the source here...was somewhere on the scale of 3-4 months maybe, if I recall correctly. 

 

One of the keys, as I've harped on over and over again (as has the PMDG Support Team), is isolation of variables and repeatability.  Just like a good old science fair project, keeping constants, changing one variable at a time, hypothesizing, and proving/disproving such a hypothesis in a conclusion.  This is absolutely vital to generate the valuable data developers need to identify potential problems and act upon them.  I come from an engineering/flight test background so I will likely sound like a broken record by the end of the thread, repeating those last few sentences.

 

With a few other users posting about VAS usage still and OOMs after the latest 777 patch, I thought I might be able to help rope in the scope and breadth of all of these posts across multiple threads.  Some contain some helpful information, others don't.  So, I think it may be helpful for the PMDG Team, if we can do our best, as community, to perform the same type of controlled tests, and hopefully summarize this information and give something discernible to the developers to decide if further action is required.  I'm not going to spend any time here describing how OOMs and VAS usage infinitely varies based on individual systems (PMDG manual introductions do a good job of that).  In fact, I'm a firm believer that *most* of the time, VAS usage and increases over time are a result of settings and setup.

 

OK, so you need some controlled data...how do you propose doing that?  How can I participate?

So, what I'd like to do is compile a set of scenarios for any interested users to take part in.  That includes both users who still suspect abnormal 777 VAS usage after the most recent build, and those who don't.  These scenarios (I'll call them "Test Cards") will take on a build-up approach, again attempting to isolate variables, starting with a VAS baseline and eventually expanding into specific flights.

 

The best way I can think of distributing these Test Cards is via DropBox as I create them.  I'll post links to updated Test Cards and users can download at their leisure.

 

These Test Cards specify individual tests and settings for users to configure their sims to and execute.  The first few tests will be baselines and ground tests, with middle-of-the-road, conservative settings.  I understand that most would like to run with some higher settings, based on the more powerful systems today.  It is important, though, to start in the middle of the envelope, baseline as much as we can, then start adding in the extras.  However, this testing must be focused on the 777's DIRECT effect on VAS (and it's increase over time in certain scenarios, if present).  Adding in 100 sceneries and ultra HD textures does not help the overall community cause.

 

The front of each Test Card contains a Test Data Sheet.  This form will be your way of submitting data back to me.  After the completion of a specific test, please fill out the information I ask for accordingly and either send me a link in this thread, a screenshot, or a private message.  I'll try my best to create a master archive of this data which we may be able to use to generate conclusions down the road. 

 

I'll be honest, this is going to be an experiment in itself to gauge what kind of feedback and useful data I can gather from the community.  Maybe it'll fail horribly.  Hopefully not.  Hopefully, I can see how things start off and adjust these tests as necessary to create some kind of beneficial data-reduced product.

 

Full Disclosures and Disclaimers

1) PMDG has not asked me to do this.  I'm also not an FSX aircraft programmer, nor have any inner-working knowledge of the PMDG 777.  Repaints here and there?  Yes.  Gauges, aircraft, scenery?  Nope.  Searching and reading the forums, and the vast amount of discussion on the subject, I see countless posts regarding 777 VAS.  I can only imagine my head would be spinning if I were a PMDG Moderator (though I'm sure they all have more message board patience than I could ever have).  If this data collection project is completely out of line, Moderators, please feel free to lock this post and smack me upside the head.

 

2) I do not own P3D or the 77W.  I only own FSX SP2 and the 77L.  And I do use DX10/Steve's Fixer, but not ReShare or SweetFX currently.  Unfortunately, with recent discussions on the 77W, I can't test these out myself.  For any settings I specify in Test Cards, if you own P3D, use your best judgment for similar settings. 

 

3) It is absolutely imperative to start with a FULL, fresh install of the 777.  No liveries, no previous cockpit states, nothing.  Even AIRACs must be taken into consideration, so we will be starting off with the default PMDG AIRAC in the 777 (should be 1309).  It would be even better if programs like REX, ActiveSky, FTX Global/Vector were uninstalled.  (I mean, even more ideally, would be a fresh install of FSX - but I'm a realistic guy and know the participation rate would be near zero if I asked everyone to do that :wink: ).  Many of the initial tests will ask for such add-ons and plug-ins to at least be disabled.

 

4) I do have a day job with a roller coaster of a schedule at times.  Compiling data or creating new Test Cards for users to try may sometimes be slow.  I love simming but there are just some days where I can't get on a computer to check forums or fly.

 

5) In general, this type of methodical testing can be a VERY tedious, time-consuming process.  Since the ugly VAS monster reared its head previously on 5-8+ hour flights, our virtual Jet A bill is probably going to be through the roof.  Time acceleration is worthless for these types of tests.  You can easily fly EGLL-KLAX at 16x and "outrun" possible VAS leaks.

 

6) Feel free to call me crazy, make suggestions, comments, or provide feedback as we go.  Also feel free to private message me.

 

 

 

 

 

With that said...

 

Here is the link to the first Test Cards/Test Data Sheets.  These MUST be read in their ENTIRETY before testing.

 

Test Card #B77L.1.1, #B77W.1.1, #B77L.1.1a VAS Baseline - http://bit.ly/1BaUc7e

(B77L.1.1 must be performed before the 77W is installed, B77W.1.1 and B77L.1.1a must be performed after the 77W is installed)

 

 

 

 

--------------------------------------

Thanks for reading and for your participation in advance, hopefully I can help a few users out with this project.

 

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

×
×
  • Create New...