Patch Notes September Update - Patch 2

yelboc

yelboc
Zantor said:
Has anyone had a issue that I got several times last night.
Was doing a passanger mission to visit tourist beacons, but everytime I got a message prompting the new target, once I went into the nav panal to select the target, I lost all keyboard functions.

I had to task manager to kill elite, then reload to continue.

I don't really want to do another passanger run if this is going to keep happening?
Click to expand...

This is because of the chat box - you can escape without killing the game - enter head-look (middle mouse button? or whatever its bound to) and look at the comms panel - hit 'esc' then your keyboard is freed again..
 

Old Duck

Old Duck
Dutchman141 said:
Ugh, if this screen tearing becomes the EBL for the Xbox then I'm out.
Rediculous, we'll probably have to wait another week at least untill this and/or the mission boards get fixed.
Click to expand...
It's sad, considering it's probably a single line of code that needs changed. Frontier's inability to fix SIMPLE bugs in a timely manner has always baffled me. They screwed up our gyroscopic head look on PS4 a year or so ago, basically mixing X and Y axis. When asked why it was taking so long to fix (weeks and weeks), they gave some excuse about the need for an "investigation" to make sure bugs like that never happen again. Ha!!!!
 

MDHansen

M
Old Duck said:
It's sad, considering it's probably a single line of code that needs changed. Frontier's inability to fix SIMPLE bugs in a timely manner has always baffled me. They screwed up our gyroscopic head look on PS4 a year or so ago, basically mixing X and Y axis. When asked why it was taking so long to fix (weeks and weeks), they gave some excuse about the need for an "investigation" to make sure bugs like that never happen again. Ha!!!!
Click to expand...
Priceless. Investigation, makes it sound like they have a QA department:unsure::giggle:
 

Soron Silin

Soron Silin
Old Duck said:
It's sad, considering it's probably a single line of code that needs changed.
Click to expand...
A couple of years ago I wrote probably the largest and most intricate piece of code that I've ever written. It wouldn't be an exaggeration to say that it was on par with the whole planetary landing stuff, maybe add atmosphere to that, (although probably not the shader bits.) It was a tour de force and a nightmare all in one. It also had a bug. After running for around two hours the program it was part of would crash. I worked for months on that bug. Every time I got close to it the symptoms would change; it still crashed but now it crashed miles away from where I had tracked it to. It was like hunting a man-eating tiger in the forests of rural Nepal and finding its empty den just as the news reports sightings of it in Delhi. I was taking so long that the management put two other guru-level developers on the task along with me. After spending probably six man-months on this single bug, one of them found it.

The function to create a new thread takes the stack size of the new thread as a parameter. If you specify zero, then Windows uses a default size of 2Mbytes. This was plenty and I had had no idea how much I needed, so that was what I had done. But the environment I was using was not Windows, and it used a default of only 2Kbytes. Even that would have been enough, except for one single function. It wasn't called often or even all the time, but when it was called it used 6Kbytes. Even then it didn't crash all the time, the memory it corrupted was usually empty, but every now and then, maybe after about two hours, it hit something important.

Six man-months, and the fix was five characters long: delete "0", insert "16384". The particular digits aren't important; anything over about 8000 would have done the job.
 

Old Duck

Old Duck
Soron Silin said:
A couple of years ago I wrote probably the largest and most intricate piece of code that I've ever written. It wouldn't be an exaggeration to say that it was on par with the whole planetary landing stuff, maybe add atmosphere to that, (although probably not the shader bits.) It was a tour de force and a nightmare all in one. It also had a bug. After running for around two hours the program it was part of would crash. I worked for months on that bug. Every time I got close to it the symptoms would change; it still crashed but now it crashed miles away from where I had tracked it to. It was like hunting a man-eating tiger in the forests of rural Nepal and finding its empty den just as the news reports sightings of it in Delhi. I was taking so long that the management put two other guru-level developers on the task along with me. After spending probably six man-months on this single bug, one of them found it.

The function to create a new thread takes the stack size of the new thread as a parameter. If you specify zero, then Windows uses a default size of 2Mbytes. This was plenty and I had had no idea how much I needed, so that was what I had done. But the environment I was using was not Windows, and it used a default of only 2Kbytes. Even that would have been enough, except for one single function. It wasn't called often or even all the time, but when it was called it used 6Kbytes. Even then it didn't crash all the time, the memory it corrupted was usually empty, but every now and then, maybe after about two hours, it hit something important.

Six man-months, and the fix was five characters long: delete "0", insert "16384". The particular digits aren't important; anything over about 8000 would have done the job.
Click to expand...
While I appreciate the sentiment, I don't think the recent lack of vertical sync on XBox is as tricky to find and fix as your problem was.

ps - we're all / were software developers in this forum it seems, myself included.
 

Soron Silin

Soron Silin
Yorper said:
Yet FD has 500 people to cherry pick from and still can’t fix some bugs from 5yrs ago, let alone 2 weeks ago.
Click to expand...
Half of those are going to be admin, sales, marketing, customer support etc. leaving 250 working on the software.
About a fifth of those will be in the QA team, leaving about 200 developers.
They've got around three projects in full-time development, so 66 developers working on Elite.
Of those, most are working on 2.0, so call it 20 developers working on this release, maybe fewer.
There are many areas of specialisation in game development so, for every bug, you may have three to five developers that could either find the bug or work on new features.

Some bugs may take weeks, even months, to track down and only slightly inconvenience a few players. Out of the 3,000 bugs on the tracker, there's probably only around 300 that could do with fixing. The rest are reports of single events, are so minor they don't matter at all or have already gone away due to changes elsewhere. Even so that is around 300 person-weeks of effort, or the whole team working for three months producing no new content. In reality, a handful of those 300 would be difficult to find or require complex fixes, so it might be four or five months. Add a month for regression testing. Six months with no work on new content.

I think they should do that. But it wouldn't be free or easy.
 

AveCandido

AveCandido
Soron Silin said:
Half of those are going to be admin, sales, marketing, customer support etc. leaving 250 working on the software.
About a fifth of those will be in the QA team, leaving about 200 developers.
They've got around three projects in full-time development, so 66 developers working on Elite.
Of those, most are working on 2.0, so call it 20 developers working on this release, maybe fewer.
There are many areas of specialisation in game development so, for every bug, you may have three to five developers that could either find the bug or work on new features.

Some bugs may take weeks, even months, to track down and only slightly inconvenience a few players. Out of the 3,000 bugs on the tracker, there's probably only around 300 that could do with fixing. The rest are reports of single events, are so minor they don't matter at all or have already gone away due to changes elsewhere. Even so that is around 300 person-weeks of effort, or the whole team working for three months producing no new content. In reality, a handful of those 300 would be difficult to find or require complex fixes, so it might be four or five months. Add a month for regression testing. Six months with no work on new content.

I think they should do that. But it wouldn't be free or easy.
Click to expand...


Do not you think it is a bit unreal that statement of distribution of employees that you do?
I do not think that Frontier has published how many are dedicated and that not even that the distribution of developers is equitable as you propose?
 
You must log in or register to reply here.
Top Bottom