Official Fixes Needed In Next Update/ Problems From This Update

[Alex Folds]

So I thought it might be a good idea to make an offical thread so that DJI can address the issues that we have been having in one cohesive thought.

So could people specifically lay out the problems they have been having since the most recent round of updates that we need fixed in the next update?

I know VPS has been an issue, and some people have been having IMU and compass problems. I feel like I have had issues with my gimbal roll being off. Anyone care to enlighten DJI with specific details so they can get us a solution worked out ASAP.

Also, Could anyone enlighten me how these things happen? How does something go from working in a previous firmware version, to being broken in the next? I don't understand the process very well, I assume it has something to do with the complexities of coding. Are there situations where they try to fix one thing (like lag and app shutdowns on androids) and something else they didn't touch gets messed up (like VPS)?

 

[FredSheldon]

As a programmer myself I can assure you that a misplaced period can cause all kinds of heck in a program. Anytime you add or modify code in a program you can cause issues elsewhere in the code, areas that were once working. Happens to me all the time Since you didn't intend to cause issues with once working areas of the program, you don't necessary test those areas.

 

[bbfpv]

As a programmer myself I can assure you that a misplaced period can cause all kinds of heck in a program. Anytime you add or modify code in a program you can cause issues elsewhere in the code, areas that were once working. Happens to me all the time Since you didn't intend to cause issues with once working areas of the program, you don't necessary test those areas.

But as a programmer, you'd generally have unit tests set up to regression test your code after any update IMO, they clearly dont.

 

[Oso]

There never seems to be enough time for regression testing during an implementation, but then you always need to find time to fix a new issue when the end user finds it immediately after go-live.

It seems that a small amount of testing would have at least uncovered the VPS issue before going live with the FW.

Yes of course the P3 is very complicated and an amazing machine, but that doesn't mean testing is not critical. As one person said earlier, the P3 is a flying lawn mower.

 

[bbfpv]

There is never enough time for regression testing during an implementation, but then you always need to find time to fix a new issue when the end user finds it immediately after go-live.

I code for a large financial institution, and generally any issues an end user finds after go-live means a pink slip come review time.

 

[Kingfish28]

It would be good for you to specify which exact update version number you are referencing for having issue, since there have been two come out rather rapid fire.

 

[Retired Pops]

If I skip this FW update can I use the next one or do I have to do this new one that has caused some problems