That looks like a similar problem I've experienced. For me it turned out to be bad drivers for a card that had reverted to old ones when I used a different pci slot.
The problem boiled down to a high number of interrupt requests from the device. I isolated it using Process explorer and windows device manager.
Use PE to monitor cpu usage across all running processes, and check to make sure there are no spikes on the interrupts coincidental with your spikes. You might spot a rogue process that's causing it otherwise. HTH
Edit:
Forgot to add. Process I/O is not displayed by default in PE, so will need turning on. This greatly helps identifying a regular spike like that.
Open PE
Right column bar - select columns
Process I/O tab - history
This will show a nice graph next to every running process. The same can be done for memory usage and CPU which is on by default.
The problem boiled down to a high number of interrupt requests from the device. I isolated it using Process explorer and windows device manager.
Use PE to monitor cpu usage across all running processes, and check to make sure there are no spikes on the interrupts coincidental with your spikes. You might spot a rogue process that's causing it otherwise. HTH
Edit:
Forgot to add. Process I/O is not displayed by default in PE, so will need turning on. This greatly helps identifying a regular spike like that.
Open PE
Right column bar - select columns
Process I/O tab - history
This will show a nice graph next to every running process. The same can be done for memory usage and CPU which is on by default.
Last edited by Squelch, .
Reason : added instructions about I/O
That is scary.
They are obviously talented and use sophisticated methods. My sense of Geek Honour is being severely tested when I think about it too much.:hide:
*Squelch dons tinfoil hat*
That is also possible in the editor, so not InSim specific. In a multiplayer editing session it can be both fun and annoying to place a barrier at another car position and watch them flip.
Related to this, I've tried but not been able to provoke a car reset that re spawns the car where there are already layout objects. If no checks are made, this could happen, but I can't confirm it does, and in fact some checking may take place anyway. If that's the case, we are halfway there when placing objects in the editor or via InSim. Resetting on a ramp does cause odd things to happen, either car ends up inside it, or flips immediately, but other objects seem to be ok in this regard.
Personally I don't consider this an important bug, and more like a mis-feature, although an intersection test with cars might be very useful.
I find it hard to support. Not on moral or ethical grounds necessarily, but on a more paranoid front.
We have to trust our governments and other official agencies with certain data about ourselves in the hope that this information is not abused by them, or employees of those organisations. Certain vetting procedures are undertaken, and on the whole the system has kept us relatively safe from harm. The same could be said of large companies who have a long term financial imperative. The problem with these "White Knight" organisations are they are by nature faceless and unknown. How sure can anyone be that the data they can access is not going to be used by a maverick individual within these so called organisations for short term gain? I'm not sure what the black market price for the personal details of Sony Playstation customers are, but I would bet it is not insignificant.
At first it was humorous, but now they are getting almost cocky and arrogant about it. If that is how they are starting to portray themselves, then it's not too much of a stretch to imagine certain individuals might not be playing as straight as they declare.
Flu jabs, and testing of security measures are very much a requirement in this day and age. Organisations like Anonymous do have a certain role to play, and it has been well documented that poacher turning game keeper is a good model when it comes to security. Lulsec don't seem to have any real motives, and the more powerful they become, the more dangerous their actions could ultimately be.
We have to trust our governments and other official agencies with certain data about ourselves in the hope that this information is not abused by them, or employees of those organisations. Certain vetting procedures are undertaken, and on the whole the system has kept us relatively safe from harm. The same could be said of large companies who have a long term financial imperative. The problem with these "White Knight" organisations are they are by nature faceless and unknown. How sure can anyone be that the data they can access is not going to be used by a maverick individual within these so called organisations for short term gain? I'm not sure what the black market price for the personal details of Sony Playstation customers are, but I would bet it is not insignificant.
At first it was humorous, but now they are getting almost cocky and arrogant about it. If that is how they are starting to portray themselves, then it's not too much of a stretch to imagine certain individuals might not be playing as straight as they declare.
Flu jabs, and testing of security measures are very much a requirement in this day and age. Organisations like Anonymous do have a certain role to play, and it has been well documented that poacher turning game keeper is a good model when it comes to security. Lulsec don't seem to have any real motives, and the more powerful they become, the more dangerous their actions could ultimately be.
The definition of "the red zone" isn't quite clear to me, but it is mentioned a lot in that article.
Does Caubet mean they have allocated the funding and are fully committed to 2013 (in the red financially) which if postponed until 2015 will break them? Or does he mean the internal board have been told that 2013 is the date, only to hear there will be a postponement and they are red faced now?
Anyone got a link to the exact proposed changes? I've tried to find something, but could only turn up speculation. Are four cylinders the only option? What about six? What about rotary? It seems like we will end up with a rule that makes only one engine layout available and would effectively ban innovation.
It all seems a bit messy right now. Indecision doesn't help anybody, and it seems the FIA are dithering over this as much as Bahrain. This particular issue could prove very costly for everyone involved.
I'm ambivalent about the new engines too, but concerned it is going to cause far too many problems if no clear decision is made soon. As a follower of F1 I'm thoroughly confused by the whole debate.
Does Caubet mean they have allocated the funding and are fully committed to 2013 (in the red financially) which if postponed until 2015 will break them? Or does he mean the internal board have been told that 2013 is the date, only to hear there will be a postponement and they are red faced now?
Anyone got a link to the exact proposed changes? I've tried to find something, but could only turn up speculation. Are four cylinders the only option? What about six? What about rotary? It seems like we will end up with a rule that makes only one engine layout available and would effectively ban innovation.
It all seems a bit messy right now. Indecision doesn't help anybody, and it seems the FIA are dithering over this as much as Bahrain. This particular issue could prove very costly for everyone involved.
I'm ambivalent about the new engines too, but concerned it is going to cause far too many problems if no clear decision is made soon. As a follower of F1 I'm thoroughly confused by the whole debate.
Are you running Win 7 or Vista?
Where is LFS installed?
Have you installed in "C:/Program Files"
Where is LFS installed?
Have you installed in "C:/Program Files"
Echo and light maps haven't been made for the [X,Y] configurations. This was mentioned when the open configs first appeared.
Black and Berg Cybersecurity home page still compromised.
That just about sums it up I suppose. Batman and Joker analogies aside, this one-upmanship with anonymous could get ugly.
That just about sums it up I suppose. Batman and Joker analogies aside, this one-upmanship with anonymous could get ugly.
Not so much a full library, but I came across it on GitHub. It's looking abandoned right now, but does seem functional with the limited tests I've made.
Ruby-Live-for-Speed GitHub
Source control doesn't have to be multideveloper or even distributed for that matter. I use Git for a whole load of non programming as well as programming revision control on an individual basis.
I was involved in a large project where we tried several of the distributed ones, and Git came out tops. I ended up being resident Git support.
Last edited by Squelch, .
Reason : Edit, change Ruby for speed link to the root
I'm just checking my history and looking for the LFS Ruby library I found.
Not forgetting SVN of course.
To expand the options, the web resources like GitHub, Codeplex, Sourceforge and even Google code could be included yes.
I'll nail my colours to the mast now and declare Git to be my favourite.
Edit:
You have the highest profile, so its why I made the suggestion rather than doing it myself.
Edit2:
The poll could be source control and issue tracking, seeing as there is also a close crossover where they are web based. Redmine is a very good issue tracker that supports all of the popular repositories.
Last edited by Squelch, .
I'm afraid not, and would have liked to have a video reference for his defence. If anyone finds one I'd be interested too
I'm surprised Ruby didn't make it onto the list alongside Java.
Like cargame.nl, I use netbeans for an ide because it runs well on my multiboot. Under Windows it is horribly slow at times though.
For most stuff I use Python. It's easy to scratch out a working script for trying an idea, and it lends itself to being easily ported to C if speed/efficiency is needed.
Under the "other" category I would list Ruby, Assembly, Lisp and SmallTalk and I think LUA might even count.
____
Any thoughts on making a repository/source management poll Dygear?
I see several people using Git and Mercurial here, but it would be interesting to see how its spread across them and the others available.
Like cargame.nl, I use netbeans for an ide because it runs well on my multiboot. Under Windows it is horribly slow at times though.
For most stuff I use Python. It's easy to scratch out a working script for trying an idea, and it lends itself to being easily ported to C if speed/efficiency is needed.
Under the "other" category I would list Ruby, Assembly, Lisp and SmallTalk and I think LUA might even count.
____
Any thoughts on making a repository/source management poll Dygear?
I see several people using Git and Mercurial here, but it would be interesting to see how its spread across them and the others available.
That's just it, different points of view, and I'm probably as happy with mine as you are with yours. If we all agreed on everything, the world would be a boring place.
Oh and edits seem to confuse things a bit - please see mine above
Not necessarily. the audi's were short fuelling too so they gained time in the pits too
The fastest lap time was matched. Yes Lotterer was demon behind the wheel, but he wasn't consistently setting records. This was during the golden hours of dawn when the cars are at their best.The lap times fluctuated between 3:27 and 3:30 for the most part, and for both teams. I think the Pugs were on medium, but I'm not certain on that. The Audi's won by spending the least time in the pits overall, not just outright pace on the track. That couple of kg's of fuel was the difference.
They were both on very similar pit strategies then, so yes it would be a concern. The extra pit stops that Audi required would have been a deciding factor in the last race. The SC let them off a bit, by nulling out an extra stop they would have had to make.
The Porsche curves were where the Audi was strongest I agree, but because they had higher downforce, the straights allowed the Pugs to pull away - significantly. The average lap times were close when fuel loads and tyres are taken into consideration. Individual sections of the track played to the individual strengths of the cars. I'm not offended by the BS accusation, but it really was much more balanced than you seem to be implying with terms like smoking and halt
Where it is so close, the lap times, tyres, and pit strategies do come into play. I stated earlier, that a handful of laps can very quickly get eaten up. I'll point out again, that #7 was on the lead lap and pushing #2 very hard until the off. Why would they stop pushing when they still have the same pace? Ant kept his line and nothing more. The race was still on, and he did not move to block an overtake. An incident like Wurz, a puncture, or any other mishap for the Audi team would mean the #7 would still be right at the sharp end again.
Some may argue that Gene was also keeping the line, and Lotterer made a lunge down the inside. Because they are laps down, does not obligate them to move off line to let others pass. A significant speed difference makes that harder to justify keeping your line, but lap times simply don't show this is true. I said I find Gene harder to defend and note you made a ninja edit, because we probably are in agreement here.
Edit:
I'll try to listen to it. It should be interesting to get "the professionals" view on it. My humble point of view is probably biased, misinformed, and completely at odds, but that is how I see it and try to keep balanced.
Last edited by Squelch, .
All commands are slowed when in slow motion for any replay. This happens in Z28 too, so is not a new thing. To speed things up I find pausing before issuing any new commands works. HTH
I noticed the alpha background overlapping in 0.6A1 but dismissed it. It seems to be more problematic with arrows, and only sometimes with the lines.
Yes it does exist, but I can't think of a situation where we would need multiple chalk markers all overlapping like that.
This falls into a category of "you can do it, but why do it?" just like closely spaced speed humps that create sampling problems. I find it hard to support as being a bug.
Edit: slightly OT FYI
On the subject of speedbump boosting, a note to layout editors:
A spacing of 1 metre between speedbumps seems to work without giving a seed boost for cars up to GTR.
GTR and above suffer severe suspension damage anyway, so their race is effectively over if they hit multiple humps although their speeds may well be enough to get a boost at 1 metre spacing.
Research Aliasing, Wave theory and Nyquist for a technical understanding into why the boost happens.
Last edited by Squelch, .
Reason : Added edit about speedhump boost.
They were hardly smoking them - slightly faster due to the pit strategies, but terms like smoking and halt are a bit strong
Exactly, there was no clear advantage in speed. If it was a clear air time trial, I'd support your view, but this is sportscar racing, and endurance at that. The Audi's were slower on the laps they encountered traffic not because they were necessarily held behind, but were unable to maintain the pace in dirty air.
This depends on your definition of blocking. I understand it to be a change of line that intentionally impedes the following driver who has made a move to overtake. Ant kept his line and didn't weave or change line so no blocking was involved. If on the other hand you define blocking as simply being on the race line and therefore in the way, then I'd agree, but please show me where this is written.
I wasn't talking about Gene, and agree that certain tactics are unsportsmanlike. Ant didn't cross that line in my opinion.
Blue flags are to warn of faster cars behind. There is a bit of leeway in the definition of faster, and a couple of seconds per lap is hardly significant - not insignificant granted, but certainly not a large closing speed in any one part of the track. #7 was hardly out of contention either. It was still lapping around the 3:30's - the lead pace - and was only lapped due to an incident. Races have been won in the past where the lead car was laps ahead, only to succumb to a problem in the closing stages. Over 24 hours, a couple of laps are relative.
Track limits is a completely different matter in this debate. They are there for safety. A broken down car, or marshal could very well be just off the track, and given this is a semi road course where run off areas are hard to make, this rule is enforced very tightly. The only relaxed place at Le Mans seems to be the Ford chicane where cars seem to cut it fine without too much attention. To leave the track to make a pass has been severely penalised in the past I believe, and to run the red lane along Mulsanne straight is a no no.
The Audi was on new tyres, and the Peugeot's tyres were on there fourth stint. The Audi had been slower on old tyres too, so again its relative, and we are talking about the very closing stages here. Have you forgotten the lead was swapping throughout the night? That hardly reflects the Audi romping away. In fact is was suggested the Audi short stint strategy may not have worked if there hadn't been a protracted 2 hour plus safety car. I think they played a blinder with what they had, and well done to them for the win of course.
I'll still defend Ant, but find it harder to do with Gene. It was borderline for me with regards to him. Did he break any written rules? I don't think so. Ask if he took some totally unnecessary risks and acted unsportsmanlike? I'd probably agree.
Last edited by Squelch, .
Reason : edited restructering mistake "was"
Davidson didn't actually make any defensive blocking moves that I saw. He did maintain his line throughout. He just made it hard for Lotterer who had the responsibility to make a clean pass. Had Lotterer been significantly faster than Davidson, then I believe there would have been more blue flags shown, fault could have been shown, and a possible infraction for ignoring them.
I'm probably wrong, but blue flags at Le Mans are an advisory only, and there's no such "pass ten blue's and you break the rules". They warn the driver that the car or cars behind are significantly faster. The only time an official inquiry happens is when they are ignored by significantly slower car which create a safety concern. Davidson was shown few if any blue flags, and was a match for Lotterers speed within a second or two. #7 was only a couple of laps down due to the off that Wurz had, and not because it had been consistently slower.
It could have been seen another way if the Audi was blocking the Peugeot who was trying to unlap. Say for arguments sake, the Pug was significantly faster after rejoining, then I would expect to see blue flags shown to the Audi even though it was laps ahead in the lead. A handful of laps lead is nothing in Le Mans, and can quite quickly, or easily be cancelled out by a minor incident - like Wurz running off line while on the lead lap and needing repairs.
No fault found with Davidson's behaviour from my point of view. He made Lotterer work hard to put him another lap down, and did relinquish the place once Lotterer started taking risks - like getting on the dirty line at the chicane.
I'm probably wrong, but blue flags at Le Mans are an advisory only, and there's no such "pass ten blue's and you break the rules". They warn the driver that the car or cars behind are significantly faster. The only time an official inquiry happens is when they are ignored by significantly slower car which create a safety concern. Davidson was shown few if any blue flags, and was a match for Lotterers speed within a second or two. #7 was only a couple of laps down due to the off that Wurz had, and not because it had been consistently slower.
It could have been seen another way if the Audi was blocking the Peugeot who was trying to unlap. Say for arguments sake, the Pug was significantly faster after rejoining, then I would expect to see blue flags shown to the Audi even though it was laps ahead in the lead. A handful of laps lead is nothing in Le Mans, and can quite quickly, or easily be cancelled out by a minor incident - like Wurz running off line while on the lead lap and needing repairs.
No fault found with Davidson's behaviour from my point of view. He made Lotterer work hard to put him another lap down, and did relinquish the place once Lotterer started taking risks - like getting on the dirty line at the chicane.
that is really beyond reasonable to expect though. Formula cars on rally tracks is pretty daft.
The server has setting to control ai guests already.
Edit:
This issue is way off from what I expected it to be, my apologies if this has become a diversion.
That was it - formation Pugs- 2 on the lead lap, and one a lap down.
I think that prompted a warning for off track excursions.
Regarding track limits - My understanding is that OCA are very strict on this because of marshal safety. That is, a warning is given for any and all track excursions that are not a result of an accident or avoiding one. Too many warnings results in stop/go. This is peculiar to this track I believe. I might be completely off the mark here however.
That particular incident could be interpreted two ways, over taking off track, or avoiding an accident, but I would hazard the former, and might even dare say Lotterer left the track trying to pass Davidson - you decide.
I think that prompted a warning for off track excursions.
Regarding track limits - My understanding is that OCA are very strict on this because of marshal safety. That is, a warning is given for any and all track excursions that are not a result of an accident or avoiding one. Too many warnings results in stop/go. This is peculiar to this track I believe. I might be completely off the mark here however.
That particular incident could be interpreted two ways, over taking off track, or avoiding an accident, but I would hazard the former, and might even dare say Lotterer left the track trying to pass Davidson - you decide.
I'm not sure of the positions at the time, but I posted this just after it happened if anyone has footage or lap records of that time.
I see the same in your replay - race line present before pitting, no race line after pitting.
If the replay is rewound, the race line is not present before your accident either. The replay must be reloaded to get it back prior to your pitting. I'm not sure if this has significance.
(I don't normally use race line, so cannot confirm that this happens live)
Something like skins download would work I expect.
There is one possible way to do it now. The local client issues an /axsave to create a local copy. I don't know if it's easy to check if layout name is already local though.
All of the fixes seem to work as expected, and the repeat based size/rotation feels much smoother than before, so is an improvement.
That one has already been reported and confirmed by Scawen.
With the collision detection changes that were made, the car does not end up flying like it used to when a reset was done on a ramp, but it does get stuck sometimes too.
FGED GREDG RDFGDR GSFDG