This file contains answers to some commonly asked questions about Winstone 98. For detailed information about Winstone 98, see the benchmark's online help file.
Winstone 98 is a system-level, application-based benchmark that measures a PC's overall performance when running today's top-selling Windows-based 32-bit applications. It runs real 32-bit business applications through a series of scripted activities and uses the time a PC takes to complete those activities to produce its performance scores.
At the end of a test, Winstone 98 produces a measure of a PC's performance as it runs 32-bit applications under either Windows« 95 or Windows NT« 3.51 (or later). Winstone 98 uses no 16-bit applications for its tests.
You can compare a PC's score with the scores of other PCs — higher scores mean faster overall performance.
Back to Questions about Winstone 98
No. Winstone 98 uses only 32-bit applications for its tests. To fully test a computer's 32-bit performance on a 32-bit operating system, Winstone 98 uses leading 32-bit business applications and our research on how typical people use applications. The result is an accurate and repeatable benchmark you can use to determine a PC's overall performance when running today's top-selling Windows-based 32-bit applications.
Back to Questions about Winstone 98
The Business Winstone 98 tests are "market-centered" tests. Business applications are the popular applications employed by most users everyday. The Business applications and the categories in which the benchmark groups them are:
The High-End WinBench 98 disk and graphics tests address the needs of users who employ demanding styles of work or specialized applications, such as 3-D visualization or application development. The High-End tests aren't grouped into categories.
The Winstone 98 High-End applications are: MicroStation« 95, Adobe« Photoshop« 4.01, Adobe« Premiere«, AVS/Express« 3.1, PV-Wave« 6.1, and Microsoft« FrontPage« 97.
Back to Questions about Winstone 98
Although you can run the Business Winstone 98 tests on either Windows 95 or Windows NT systems, the High-End tests run only on Windows NT. The reason for this is that Winstone 98's high-end applications take advantage of NT's advanced system capabilities, making NT the environment most users would employ to run these applications.
Back to Questions about Winstone 98
Winstone 98 adds a new Business category that is not part of the overall Business score: Business Task Switching. The goal of this category is to show how well PCs perform when switching between multiple applications. One task-switching test alternates between Word and Excel, while the other switches between CorelDRAW and WordPerfect. The task switching script copies and pastes material from one application to another. For example, the Word/Excel task switching test copies charts from Excel to Word and text from Word to Excel. Winstone 98 groups the applications in these tests to reflect office suite packaging (grouping together only the Microsoft Office applications, for example), which is how a user would normally run these applications.
While the Task Switching applications are a part of Overall Business Winstone, they do not weigh into the overall Business Winstone 98 score. However, when Task Switching is excluded from the test, no Business Winstone 98 score is produced. If there's an error during the test run, and you click Ignore at the Abort/Retry/Ignore dialog box, an overall score is produced if the PC successfully runs the other Business Winstone scripts .
Back to Questions about Winstone 98
We wanted to included it, but ran into difficulties because IE 4.0 was not ready before we had to finish the benchmarks, requiring us to use IE 3.01, and Microsoft could not ensure that it would be possible to install IE 3.01 on a machine with IE 4.0 already installed. Given that many machines that will appear during the life of these benchmarks are likely to have IE 4.0 installed, we could not include IE in Winstone 98.
Back to Questions about Winstone 98
The Browser test behaves unlike real Web browsers in one key way: It runs totally locally and does not access the Web. We chose this design because there is no way we could guarantee consistent performance results and still access the Web.
Back to Questions about Winstone 98
Winstone 98 reports unitless scores, classing them as "Winstone 98 units." For all tests, higher scores mean better performance. The results indicate your system's relative performance as compared to the Winstone 98 base machine.
The base machine returns a score of 10.0 for the Business and High-End Winstone 98 suites, and a score of 1.0 for the category suites and the individual High-End applications. Some if your system returns a score of 20.0 on the Business Winstone 98 suite, then it's twice as fast at performing the Business Winstone 98 tasks as the base machine.
Back to Questions about Winstone 98
The following questions are the ones we receive the most often from our users. These questions cover a wide range of general technical support issues. For specific Winstone 98 technical issues, review the questions covered in Troubleshooting Winstone 98-specific problems.
Which button should I press: Abort, Retry, or Ignore?
If a test halts with an error, press OK at the error message dialog box. WinBench 98 will then display a dialog box with three buttons: Abort, Retry, and Ignore.
Back to Common technical support issues
Most any task you can perform in the benchmarks is described in one of two places.
You can access both files under the benchmark's Help menu.
Back to Common technical support issues
The current releases of the benchmarks all contain sample results. All of the benchmarks include sample results in the directory where you installed the benchmarks.
We don't, however, intend those results to be more than useful reference points. The best source of benchmark results continues to be the Ziff-Davis publications.
You may also find the most current published results online on the ZDNet web site at http://www.zdnet.com , and in the ZDNet Products channel, at http://www.zdnet.com/products/.
Ziff-Davis does not support older versions of the benchmarks. Each year we revise the benchmarks to reflect the way the current generation of applications use PCs. All the Ziff-Davis publications then move to the new benchmarks, and we encourage you to do likewise.
Back to Common technical support issues
When we have new products ready for release, we'll let you know via the usual channels (ZD publications and online announcements on ZDNet™ web sites, for example). ZDBOp developers are working constantly on new benchmarks and new releases of existing benchmarks.
Back to Common technical support issues
Historically we haven't performed this kind of work. Again, if we produce a new benchmark, or port an existing benchmark to a new operating system or hardware platform, we'll announce it.
Back to Common technical support issues
Probably not. ZDBOp develops, supports, and distributes only the core, publicly available ZD benchmarks. Core benchmarks are those that all ZD publications use for performance testing. Currently those benchmarks are the most recent releases of MacBench«, NetBench«, ServerBench«, WinBench«, Winstone«, and BatteryMark™.
In addition, however, individual publications also regularly create their own performance tests to cover areas, such as printer performance, that the core benchmarks don't address. The publications typically use those benchmarks for a specific review or for a limited number of reviews, so the publications don't invest the considerable time, effort, and expense it takes to turn those benchmarks into distributable products.
Back to Common technical support issues
If you want to know how ZD Labs or a ZD publication performed any kind of test, then you should send your questions to them in a letter or email. All ZD publications (as well as ZD Labs and ZDBOp) have their own web sites and are happy to field your questions there. (The main ZDNet web site is at http://www.zdnet.com.)
Back to Common technical support issues
Sure. See the help file topic on creating batch files (Running Tests->Selecting and running tests->Using batch files). It describes how to use the point-and-click options to create a self-running batch file that, upon double-clicking it, will start the benchmark, run tests, save results, and exit the benchmark, among many other options.
Note, however, that several of the .ZDR batch files included with the benchmarks are self-documented and contain many more capabilities (such as rebooting the PC, defragging the hard disk, and setting environment variables) than are available through the dialog box.
Back to Common technical support issues
Some of our benchmarks require the Windows OS have a requirement like "Service Pack 3 or higher." A service pack is extra software released by Microsoft to enhance the functionality of the OS software. To find out more about service packs, go to the Microsoft web site and go to the page dedicated to the OS you're using.
Back to Common technical support issues
To find out why the benchmark won't run, check that your computer meets the benchmark's basic hardware and software requirements. (Note also that the benchmark conducts a System Configuration review before you run a test; this review will highlight some hardware or software problems that may prevent the test from running.)
The next step is to review the benchmark's documentation to ensure that you're following the correct test procedures.
If the system still refuses to run, then you may need to strip your system down to its bare essentials. First, strip out background tasks, drivers, resident programs, add-ons, and switch your system's display to one of Microsoft's generic SVGA or VGA display drivers. If the benchmark runs with this stripped-down system, then add in your original setup components one piece at a time until you find the culprit that's preventing the benchmark from running.
Back to Common technical support issues
When you're comparing results for two PCs, you need to ensure the machines are configured and tested in the same way. You need to follow the testing procedure outlined in the magazine and ensure your machine's configuration matches the one used by the magazine. You need to do this because machine configuration affects real-world performance.
To find out about a PC's system information at the time you ran the benchmark's tests, from the Edit menu select System Info.
Back to Common technical support issues
The way to get the best and most repeatable results is to follow the same standard testing procedure every time you use the benchmark. Each time you run a test, you should defragment your hard disk, reboot the PC, and, if possible, run the test using a .ZDR batch file.
You need to make sure you're testing exactly the same combination of hardware and software starting from exactly the same initial state every time. If you do so, our research shows you won't see much variance in benchmark results. If you're positive you've done so, however, and you still see variance higher than a few percent, let us know.
Back to Common technical support issues
The current releases of the benchmarks all contain sample results. We don't, however, intend those results to be more than useful reference points. The best source of benchmark results continues to be the Ziff-Davis publications.
Back to Common technical support issues
The Configuration Information window will inform you of any applications that are running when you start a test. Ideally, the benchmark should be the only program that's running and there should be no programs running in the background.
Occasionally, a component from an already installed application (possibly an application also included with Winstone) can also conflict with the tests. Unless the application is running at the time you start a Winstone test, however, you shouldn't have problems.
However, if you can't finish running a test, try some of the suggestions for recovering from a timeout.
If you're testing Windows 95 systems, you can use the benchmark's Startup Manager program to disable the loading of some programs when the computer boots up. (Startup Manager also works on Windows NT, but presents an incomplete list of applications.)
Back to Troubleshooting Winstone 98-specific problems
Error messages that reference driver modules are usually due to bugs in the video driver. In addition, other driver bugs can cause some applications to crash or hang.
If you suspect video driver bugs are causing Winstone 98's problems, first try running Winstone 98 with a generic video driver for Windows. If you can run Winstone 98 successfully that way, then you need to consult the vendor who made the video driver card to get a more recent video driver.
If running Winstone 98 with a generic video driver for Windows doesn't fix the problem, please contact us.
Back to Troubleshooting Winstone 98-specific problems
During a test, you might see an error message with the window title "Winstone 98 Timeout." This usually happens when Winstone 98, which is running a timed script, is waiting for some onscreen object, such as a dialog box or menu, that never appears or doesn't receive focus. When the object doesn't appear or receive the focus, Winstone 98 stops the script and displays the timeout message.
To recover from a timeout error:
- Cancel cleans up the test run and reboots the machine
- Retry reruns the test that showed the timeout
- Ignore skips that test and moves on to the next selected test if there is one.
The usual timeout workarounds include:
Back to Troubleshooting Winstone 98-specific problems
If you have a problem with Winstone 98 and need to abort the test run, restart Windows. On restart, Winstone 98's cleanup program will automatically restore the PC's original system files to their pretest state. You can then run the test suite again.
Back to Troubleshooting Winstone 98-specific problems
If Winstone 98 has problems during installation, it will display an error message. The message that appears on the screen will explain what happened and what you should do to correct the problem. Problems that can generate a setup error message include a failed installation, an invalid disk drive or path name, or insufficient disk space.
Back to Troubleshooting Winstone 98-specific problems
If Winstone 98 attempts to find a file that is not accessible, it will display an error message.
To recover from this message:
Back to Troubleshooting Winstone 98-specific problems
If a path Winstone 98 uses is invalid, it will display an error message. The message that appears on the screen will tell you what happened and what you should do to correct the problem. Problems that can generate a path error message include an invalid base directory or working directory path. (You can set these paths in the test settings dialog box.)
Back to Troubleshooting Winstone 98-specific problems
If Winstone 98 encounters compatibility problems with a library or the operating system, it will issue an error message. In this case, you should:
Winstone 98 doesn't support Windows 3.x or OS/2«. If you try to run the software on these operating systems, you'll receive an error message.
Back to Troubleshooting Winstone 98-specific problems
If Winstone 98 has problems with the test PC's screen resolution and font size, it will display an error message. The message will tell you what happened and what you should do to correct the problem.
Back to Troubleshooting Winstone 98-specific problems
If Winstone 98's files become corrupt, the benchmark will issue an error message.
To recover from a corrupt file message:
If you still have problems running Winstone 98, contact us.
Back to Troubleshooting Winstone 98-specific problems
We've tested Winstone 98 on the following language versions of Windows 95, OSR2, and Windows NT. If Winstone 98 doesn't work and you're not using one of these versions, then that may be the problem. If possible, you should try to acquire one of the language versions we've tested on, and run the benchmark from there.
| Languages | Windows 95 |
Windows 95 OSR2 |
Windows NT 4.0 |
| English | X |
X |
X |
| French | X |
X |
X |
| German | X |
X |
X |
| Italian | X |
X |
X |
| Polish | X |
X |
|
| Spanish | X |
X |
Back to Troubleshooting Winstone 98-specific problems
Winstone messed up my hard drive. What's up?
When you exit Winstone 98, it automatically restores the PC's original system files. If anything happens and you need to interrupt Winstone 98 during the test run, or if Winstone 98 halts, stops, or hangs the PC and you must exit Winstone 98 in any way other than clicking the Exit button, Winstone 98 automatically cleans up the system when you restart Windows and will restore the PC's system files to their original state.
Don't try to fix the problem yourself. You should always let Winstone 98 clean up after a test run.
Back to Troubleshooting Winstone 98-specific problems
Winstone's demo mode keeps rebooting the system. What's up?
Running Winstone 98's demo over and over eventually exhausts the system's resources and may cause the machine to hang. Therefore, it's a good idea to stop the demo after a few runs and restart Windows to refresh the system's resources. You can specify the number of demo runs before a restart in Winstone 98's Test Settings dialog box.
Keep in mind that Winstone 98is a performance test, not a demonstration program and not a system burn-in tool. Winstone 98 wasn't put on this earth to run in demo mode.
See Winstone 98's online help for more information on demo mode.
Back to Troubleshooting Winstone 98-specific problems
You should first read the benchmark's online help to make sure the PC meets the minimum requirements to run the benchmark and you're running the benchmark correctly.
If you continue to have trouble, send a Problem Report Form to Winstone 98 Technical Support at ZDBOp. You can find a Problem Report Form on ZDNet/World Wide Web Edition (http://www.zdnet.com/zdbop/techfrm.html) and in the benchmark's online help.
You can send the Problem Report Form to us in a variety of ways, though we can answer your questions more quickly if you use email. Fill out the form and send it via:
When you have a problem, it's most helpful to us if you provide the following information:
The Problem Report Form provides space for you to answer these questions. If you don't want to send in a Problem Report Form, please include the above information in any faxes or communications you send to us.
Back to Contacting ZDBOp for technical support
The complex interactions of PC software, hardware, BIOS settings, and so on, can make locating any problem difficult. Although we can't guarantee we'll always be able to find an answer, having the system information focuses our troubleshooting efforts. You can find your PC's system information in the ZDBENCH\SYSINFO.INI file.
Back to Contacting ZDBOp for technical support
If you have a modem and communications software, you can download most of ZD's publicly available benchmarks from the following Ziff-Davis online services from ZDNetTM/World Wide Web Edition (http://www.zdbop.com).
The online versions of some of the benchmarks are limited, however.The online version of WinBench 98, for example, doesn't contain the graphics, CD-ROM, or video performance tests; those tests and the files they require are too large for downloading. Winstone 98, at over 100MB in size, is also not available for downloading.
But you can get a copy of the CD-ROM that contains complete versions of both WinBench and Winstone, as well as copies of other CD-ROMs containing other Ziff-Davis benchmarks, from the Ziff-Davis Benchmark Operation (ZDBOp).
You'll need to pay a shipping and handling fee of $5 (US), $6 (Canada), or $7 (International) for your first CD-ROM. Add $1 for each additional CD-ROM you request. Send credit card information (a VISA, MasterCard, or American Express credit card number, the expiration date, and the name as it appears on the card) via:
A CD-ROM takes four to six weeks to arrive. To receive it sooner, include your Federal Express account number and shipping instructions with your request. ZDBOp will then send the benchmark to you via Federal Express and charge the FedEx shipping cost to your account. (The FedEx charge is in addition to the ZDBOp shipping and handling fee.)
Back to Getting copies of ZD benchmarks
ZDBOp develops, supports, and distributes only the core, publicly available Ziff-Davis benchmarks. Core benchmarks are those that all Ziff-Davis publications use for performance testing. Currently those benchmarks are the most recent releases of MacBench«, NetBench«, ServerBench«, WinBench«, and Winstone«.
In addition, however, individual publications also regularly create their own performance tests to cover areas, such as printer performance, that the core benchmarks don't address. The publications typically use those benchmarks for a specific review or for a limited number of reviews, so the publication doesn't invest the considerable time, effort, and expense it takes to turn those benchmarks into distributable products.
Back to Getting copies of ZD benchmarks
No, the fax machine isn't broken—it stays busy. In fact, we have several fax machines, all of them pumping out requests from all over the world for our benchmarks. Lots of people are interested in our products. Please keep trying, or email your request to us.
Back to Getting copies of ZD benchmarks
Our benchmarks are licensed software. While the software is free, you'll need to pay your normal connection time and downloading charges (if any) if you download a benchmark from one of ZD's online services. (If you order a CD-ROM, you'll need to pay a shipping and handling fee.) You must license and register your copy of the software before you can use it. The benchmark displays its license agreement the first time you run it.
Back to Getting copies of ZD benchmarks
ZDBOp is a division of Ziff-Davis dedicated to the research and development of the core, publicly available benchmarks ZD publications worldwide use for performance testing.
Sorry, but we don't use external beta sites for our benchmarks. Although we welcome input from the user and vendor communities, we are first and foremost journalists and can't afford to let some—but not all—vendors see early releases of the benchmarks.
MacBench«, NetBench«, ServerBench«, WinBench«, and Winstone« are registered trademarks and 3D WinBenchTM, BatteryMarkTM, CPUmarkTM, WebBenchTM, WinMarkTM, ZDigitTM and ZDNetTM are trademarks of Ziff-Davis Inc.
1-2-3« and Lotus« are registered trademarks of Lotus Development Corporation.
Adobe«, Photoshop«, and Adobe Premiere« are registered trademarks of Adobe Systems Incorporated.
AVS/Express is a registered trademark and AVS is a trademark of Advanced Visual Systems Inc.
Corel«, Quattro«, and WordPerfect« are registered trademarks and CorelDRAW!™ and Corel GALLERY™ are trademarks of Corel Corporation.
Encarta«, Microsoft«, MS-DOS«, PowerPoint«, Visual C++«, Win32s«, Wingdings«, Windows«, and Windows NT« are registered trademarks and ActiveMovie™, DirectDraw™, Direct3D™ , and DirectX™ are trademarks of Microsoft Corporation.
MicroStation« is a registered trademark of Bentley Systems, Inc.
Netscape«, Netscape Navigator«, and the Netscape N logo are registered trademarks of Netscape Communications Corporation in the United States and other countries.
PV-WAVE« and Visual Numerics« are registered trademarks of Visual Numerics, Inc.
Copyright ⌐ 1997 Ziff-Davis Inc. All rights reserved.
Revised: 11/3/97