Re: [Flightgear-devel] GIT and FGDATA
Alan Teeder wrote Same here. But the slow-down has been quicker than the increase in data over time. I am to be yet to be convinced that there isn't something else going on besides the increase in the amount of data. I'm beginning to suspect that the whole concept is fundamentally flawed with larger repos. I hope we can devise a work around - and fast. It's taking me longer to push stuff than to write it in the first place. Work has become almost impossible here. Vivian Looking at other reports on the net it may be something to do with the number of changes, not just the size of the repo. I had reached the same conclusion: it does conform to the observed fact that the slow-down has been faster than the increase in data. The splitting of the repo might not overcome the problem if this is indeed the case. Meanwhile, Git-bash works here. No fancy graphics, but I can push/pull. I'm not always sure what I'm pushing, but hey ho, it's still hanging on in there. Vivian -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Color change
Hi there, is there a way to change colors of objects or submodels during flight? Perhaps with nasal? For example i can imagine an arrow outside an airplane that shows the resulting load factor and changes its color due to the normal load factor from e.g. green to red... Thx for help, Peter -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] GIT and FGDATA
-- From: Vivian Meazza vivian.mea...@lineone.net I had reached the same conclusion: it does conform to the observed fact that the slow-down has been faster than the increase in data. The splitting of the repo might not overcome the problem if this is indeed the case. Meanwhile, Git-bash works here. No fancy graphics, but I can push/pull. I'm not always sure what I'm pushing, but hey ho, it's still hanging on in there. Vivian Could something similar to Flightgear´s terasync option be implemented? By this I mean that a complete list of aircraft is always available and kept locally for use by both flightgear and fgrun. If an up-to -date copy of the selected aircraft is not available locally, then it is downloaded (unfortunately together with dependencies) as required. Just a suggestion. The Git symptoms I experience are a 100% cpu load and memory usage increasing above 100M. This goes on for several HOURS. It has got to this state over the last week, before then it was slow. Now it is dead. The only practical way to get a complete set of Fgdata is to start again and make a fresh git clone. Although this takes several hours on my connection it is much quicker than a git pull. Alan -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] GIT and FGDATA
I am not seeing the slowdown. I just ran a simple script that updates simgear, flightgear source, and fgdata. I would estimate that the git pull origin run in fgdata took about 3 minutes. This performance has been typical for me. I ran System Monitor (F12) during the fgdata pull and the data rate was between 96 and 100 KB/s and the memory usage was constant. I connect via Earthlink dsl. Dave P. On 08/06/2010 02:11 AM, Vivian Meazza wrote: Alan Teeder wrote Same here. But the slow-down has been quicker than the increase in data over time. I am to be yet to be convinced that there isn't something else going on besides the increase in the amount of data. I'm beginning to suspect that the whole concept is fundamentally flawed with larger repos. I hope we can devise a work around - and fast. It's taking me longer to push stuff than to write it in the first place. Work has become almost impossible here. Vivian Looking at other reports on the net it may be something to do with the number of changes, not just the size of the repo. I had reached the same conclusion: it does conform to the observed fact that the slow-down has been faster than the increase in data. The splitting of the repo might not overcome the problem if this is indeed the case. Meanwhile, Git-bash works here. No fancy graphics, but I can push/pull. I'm not always sure what I'm pushing, but hey ho, it's still hanging on in there. Vivian -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] Color change
Hi Peter, Absolutely. Look at the bluebird, the inside colors change with lighting and door-open-close, besides the usual livery choices. It is done with nasal, which changes properties that the model material tags point to. Stewart Peter Völk wrote: Hi there, is there a way to change colors of objects or submodels during flight? Perhaps with nasal? For example i can imagine an arrow outside an airplane that shows the resulting load factor and changes its color due to the normal load factor from e.g. green to red... -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] GIT and FGDATA
-- From: dave perry skida...@mindspring.com Sent: Friday, August 06, 2010 2:59 PM To: vivian.mea...@lineone.net; FlightGear developers discussions flightgear-devel@lists.sourceforge.net Subject: Re: [Flightgear-devel] GIT and FGDATA I am not seeing the slowdown. I just ran a simple script that updates simgear, flightgear source, and fgdata. I would estimate that the git pull origin run in fgdata took about 3 minutes. This performance has been typical for me. I ran System Monitor (F12) during the fgdata pull and the data rate was between 96 and 100 KB/s and the memory usage was constant. I connect via Earthlink dsl. Dave P. Dave From earlier posts I surmise that you have a 64bit Linux box. I have an old single processor XP box (limited to 1 GB memory by the motherboard) and a newer dual-core laptop. These run Xp and Vista respectively. Not in the same league as yours :-( My searching on the net suggests that the Windows version of Git is not as fast as the Linux original. The net result is that Git on Windows needs a higher spec processor than flightgear does when used with a large and frequently changing set of data such as fgdata. During pull origin Git´s CPU usage goes to 100% (or 50% on the dual core box) and Gits´s memory increases to over 100M, but there is negligible net usage. I have let it run for 3 hours with no sign of any completion. My guess is that Windows Git is having trouble sorting out all the changes and updating its internal databases. Alan -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] GIT and FGDATA
Alan Teeder wrote I am not seeing the slowdown. I just ran a simple script that updates simgear, flightgear source, and fgdata. I would estimate that the git pull origin run in fgdata took about 3 minutes. This performance has been typical for me. I ran System Monitor (F12) during the fgdata pull and the data rate was between 96 and 100 KB/s and the memory usage was constant. I connect via Earthlink dsl. Dave P. Dave From earlier posts I surmise that you have a 64bit Linux box. I have an old single processor XP box (limited to 1 GB memory by the motherboard) and a newer dual-core laptop. These run Xp and Vista respectively. Not in the same league as yours :-( My searching on the net suggests that the Windows version of Git is not as fast as the Linux original. The net result is that Git on Windows needs a higher spec processor than flightgear does when used with a large and frequently changing set of data such as fgdata. During pull origin Git´s CPU usage goes to 100% (or 50% on the dual core box) and Gits´s memory increases to over 100M, but there is negligible net usage. I have let it run for 3 hours with no sign of any completion. My guess is that Windows Git is having trouble sorting out all the changes and updating its internal databases. Yes, I'm seeing this on a Core2Quad - all cores go to 100% and then nothing. However, good old Git-bash shell will work from the command line. Not really 21st Century, but it will let you do most anything you want. This is not inconsistent with Dave P observations, because I guess his script uses shell commands stuff. I think we might have predicted something like this happening when we could only clone fgdata in the first place by downloading it in 2 sections. That's another fine mess we have gotten ourselves into! Never mind, we'll get it sorted (and perhaps shadows at the same time) :-) Vivian -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] Shaders and textures
Hi all, I see a lot of improvements with recent shaders. But there comes some renewals I do not how to work with. - Snow: Activating the landmass shaders gives me some snow. Some material is not covered i.e. glacier. And snow is starting by default at 2000 (2000 what?). Is this real i.e. for season 'summer'? And how can I change appearance of this soft-edge 'geometrical' snow fields? - Mystic gray blobs: When I activate landmass shader everywhere there are some 'random' gray blobs now. What is the goal of this blobs? Is it possible to add a texture to this blobs? - Crop shader: This is a very interesting shader of course. But when I activate it- only one crop type is covered by this shader. Is the crop shader thought to have multiple crop effects? How can I work with this shader and different textures to get things like irregular crop and many many other crop types like we have in recent landcover? Thanks for your answer, Yves -- This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel