Doug_Park

I'm not sure of all the reasons why the 5K iMac is so slow. It sound's like they may have some OS level performance problems in addition to the the fact that the hardware is being asked to drive a massive number of pixels on low power mobile hardware. The number of recent files has near zero effect on a Mac. It is a Windows only issue.

Posting plans helps to get quicker more accurate answers. Often there are multiple possible causes so a plan may be necessary to give an accurate diagnosis. Posting a trimmed down plan that demonstrates your issue with the least amount of extra stuff in it is an even better thing and hopefully should address your intellectual property rights concerns as well.

The jump list performance is also tied to tab switching slowness in case you are seeing that. We are looking into a solution for this.

I agree that stairs, along with a long list of other things need to be improved. The problem that Perry has shown is something that we probably should take a look at fixing as it is difficult, but not impossible, to do correctly. As it turns out this is one particular problem that we have unfortunately not been as aware of as we probably should have been. Thanks for pointing this out to us Perry.

Crashing while zooming is not normal. I doubt that it is the mouse, although at one point in very distant past we had issues with a particular type of mouse that send scroll events to the program at a more rapid rate than normal which was a bug on our part. But that was fixed several versions ago. There also exists the possibility that the mouse driver is causing a problem. I've had a lot of trouble with drivers from Logitech and Microsoft in the past. Normally I don't install the mouse driver because they like to redefine what the buttons do in ways that don't follow standard practice. What kind of mouse are you using? More likely the cause is the video driver, especially if the crashes are occurring in elevations or 3D views. In that case I would recommend installing the latest video drivers from the vendor site (don't rely on Microsoft for them). If the problem persists please contact our support team. They can work through a variety of issues and hopefully get to the bottom of the problem much more quickly than we can do here.

The case that Perry has is still a mystery to us. I'm not sure what Ken is seeing. I looked at the video he refers to in the past. I'm pretty sure some of the slowness he reported in that video was fixed in the update. But because the video shows more than 1 issue it is hard to speculate as to what the performance is like now. The case Perry has seems likely to be precipitated by a setting. We were speculating yesterday that one possibility is the number of recent files. In X7 we populate a list on the task bar that wasn't there before. It gets updated when you save so it is possible that if it is very large that you might see a performance issue when you hit save. Setting it down to 0 should put the timing of that operation down to something closer to X6. That may be worth a try.

Chief 10 is not designed to run on anything newer than XP. There are probably half a dozen different issues that you might encounter when trying to run Chief 10 on newer operating systems. What you describe is one of them. In X1 we did preliminary support for the modern operating systems and completed that work in X2. We are now on X7. While you may be able to muddle through the issues of trying to get such an old version of Chief to sort of work on a new OS, you are kind of on your own there. Vista was released nearly 8 years ago now and my memory of those days and the problems that you might encounter is starting to fade. The first thing I would suggest is trying to run the program in XP compatibility mode. My recollection is that won't solve all of the issues but might solve the one you are describing. The best answer is to upgrade to the latest version of Chief which is designed to run on Windows 8.1.

Now that you pointed out the places where the problem exists I can see what you are talking about. Thanks. It appears that there is a somewhat transparent overlay on the layout. I assume this was put there on purpose. I don't know of any bugs where that would occur either in a printer driver or in Chief. Using the built-in Save As PDF there is enough information available to produce a Vector based PDF with transparency information intact. However, at least on Windows, printers don't support transparency so going through a print driver requires a conversion to a pixel based format where the transparency blending is done before sending to the printer. So ALL PDF printer drivers will suffer from this limitation. Chief, or rather the toolkit that Chief uses, does the blending before sending the output to the printer. If you need to use the printer driver to produce the PDF then you will need to go to a higher resolution (I suggest the DPI of the printer you are targeting for the paper copy). You will still be able to see the pixels in the viewer if you zoom in far enough, but if they are as small as the DPI of printer you are targeting there should be no problem with the printed output.

The auto exterior dimension shouldn't create a dimension through the middle of the plan as shown in the image if it is working right.

I am either dense and not seeing what you are talking about or it is not displaying as you describe on my system. I'm using the latest Adobe Reader XI (11.0.11). Also running Windows 8.1. I thought maybe my high resolution display may have been making it look better so I dropped it down to 1080p and still didn't see what you are talking about. Maybe someone who is seeing the problem can post a screen shot.

What you see when you choose the print preview option should be what you get when you print assuming that you printer can print in color. The feature to display and print images in color while everything else is gray scale or black and white has always seemed an odd feature. As I recall this was done because at one point in the past images were not converted to gray scale when printing and were coming out in color because of a bug. When we added the ability to display and print an image as gray scale we added the preference to keep the original bug. I don't recall whether that was because of customer complaints or just because it was different and we wanted to avoid complaints. In any case it sounds like that feature got broken in the latest release. Generally I would recommend choosing layer colors that are what you want and always printing in color. That way you can control everything the way you want.

The case you are describing sound's like a bug in the automatic dimension algorithm. It would probably be a good idea to send that plan to our support team so that they can get a bug logged if indeed that is the case.

I don't see an issue with your posted PDF so it is kind of hard to comment on what you are seeing.

I think the ability to group select a collection of items and add them to the library is a good solution to this. I think that all single or group selection of objects should be able to be put in the library.

The issue is that hardware accelerated OpenGL is disabled in Windows Remote Desktop software when you launch the program using Remote Desktop. If you can arrange to launch the program while Remote Desktop is not connected then it will work. This is a really odd limitation that Microsoft has put into the software that doesn't make any sense. At one point in the past I did this by writing a .bat file. I don't remember how I did that now. The other way to do this is to leave Chief running on the remote machine.

I would go for the cost effective solution. I don't actually have any favorites. I have one laptop with Intel graphics and one with an Intel/Quadro combination, a desktop with, I think an NVIDIA 750 and a Mac desktop with an ATI card (I forget the model). The 750 was a replacement for an ATI card that was giving me problems during the development of X6 that were subsequently corrected with some programming changes. That card is still in our stable of cards that we use of testing things. I would never spend big bucks on a video card if there were another cheaper option available. The Quadro I have was not what I would have picked, but came with the machine I wanted, which is a high end touch screen with high resolution display and really good CPU performance. But then I was looking more for a system to do development on. In fact for Chief the built-in Intel is almost as fast as the Quadro for most operations. The price on the Quadro cards is not justifiable to me for several reasons. 1) Benchmarks indicate that they aren't any faster with Chief, and even on other CAD programs where they are tested I've read a number of reports over the years that cheaper gaming cards were faster. 2) We have seen many cases where these cards don't work because they have buggy drivers. Those drivers are often not updated as quickly as the gaming cards. 3) The are are really expensive. As I understand the performance the Quadro cards are tuned to make old OpenGL technology fast. We have been moving away from this for years and will eventually have a fully modern OpenGL pipeline. They are also apparently quite fast in rendering wire frame models which we don't do. Although I can't imagine a wire frame model in Chief ever being anything but blindingly fast even on the slowest of gaming cards. Even when Chief was using the old OpenGL technology we never saw the performance benefits of the Quadro. Possibly because of the way we were using it. The benchmarks here are generally a good guide to performance of video cards within Chief. Actual performance will almost certainly not track exactly, but I feel confident that if you buy a card from this list with a good price performance rating you will be a lot happier than if you buy the most expensive one. http://www.videocardbenchmark.net/high_end_gpus.html Personally, I would probably never spend much more than $300 on a video card. It's just not worth it to me.

The easy part is creating a wall that has the desired shape using the same tools we use to create arbitrary 3D objects. That part of the programming would be straight forward once it is decided what to do. The hard part is figuring out what to do. Joe mentioned a way of doing a battered wall of relatively small angle. A larger angle would require some form of triangular framing. Perhaps using typical studs, perhaps using some other method. I just don't know how it is typically done or if there is typical. Also how is insulation and weather proofing done? As the angle increases there is a greater need for weather proofing such as what is required for a roof. As I said the ways that I could imagine doing this are varied and likely not an intersecting set with the way someone else might do this. While what Joe suggested seems pretty logical, that would only work for a limited set of possible designs. A retaining wall, which is possibly the more typical use of a battered wall is quite different in design that what Joe described. And highly variable. I'm not suggesting that we don't do this because it is hard, only pointing out that what we do here depends a lot on what is needed. It may be that there are so many different ways of doing this that the best thing to do would be to improve the manual tools for creating generic objects.

Go to page 0 and you can do this.

For ray tracing more cores are almost 100% utilized. 10 cores is almost twice as fast as 5 all other things being equal. This isn't typical for multiple threaded algorithms so in other areas while more cores can continue to improve performance it may not scale as well as for ray tracing. Whether the price difference between a 6 and 8 core is worthwhile depends on what you are doing. My read on the Titan vs the 980 is why spend all the extra money for marginally better performance. I don't have any reason to think that card will be appreciably faster than the benchmarks indicate for Chief. But I haven't bench marked it and am unlikely to do so given the price. I would not even consider a FirePro card based on really poor experience in the past with that brand, but I do know that the FirePro cards on the Mac Pro computers work OK with Chief, but are not as fast as their price would indicate.

The shape of the wall is the easy part. The questions you ask are the hard questions. I can imagine a lot of different ways of potentially doing this, but I don't know how they are actually constructed. Are there any common ways of doing this or does it require engineering? If an engineer gets involved I can imagine that the designs would be highly variable.

Xeons are a good choice. They perform better than i7s, plus if you do a dual CPU setup you can get even more cores working for you for ray tracing. https://www.cpubenchmark.net/high_end_cpus.html

Also don't forget that using angular dimensions you can force walls and lines to rotate to any specific angle you want without using the angle snap settings. Unless I was doing something with a lot of things at a particular odd angle I don't think I would even care about the extra angle feature or even the 7.5 degree option. Also make parallel can be used once a wall or line has been drawn at the correct angle. My point is only that there are a lot of ways of getting things on angle besides snaps.

The MS site is usually many versions behind what the vendor is with drivers. I'm not sure how they decide what to put up there. I generally don't trust that they have good drivers there. I've had many cases in the past where their drivers were bad but the hardware vendor had good drivers. But I've also seen it where the vendor driver was bad and the MS driver was good. So it is a bit of crap shoot when you update them. Video card drivers are unfortunately one of the most unreliable parts of a computer. It has been this way from the release of the very first accelerated graphics cards and while their quality has improved dramatically over the years, they remain at the bottom of the heap with regard to reliability of computer parts.

Reboot your computer. Check for updated video drivers. Make sure you have the latest Chief update installed.

In a case like this we would need the plan and steps to reproduce. I can't tell from the video what is going on. In theory dimensions can snap to things some distance away based on the reach distance. Certainly the dimension is locating something. Please work with support so that we can get an accurate reproduction of the bug entered in our database. Thanks.