Independent analysis of Timebomb Disposal Kit by VNU European Labs.

Abstract

The purpose of the investigation was to establish how effective Timebomb Disposal Kit (henceforth referred to as Timebomb) is at diagnosing and correcting PCs with the inability to handle dates beyond 1999. During our testing we found that Timebomb produced 100% accuracy in diagnosing which of the PCs had a problem and which were able to handle the rollover to the next century. Where an incompatibility was detected, the supplied fix was also 100% successful at overcoming the date handling failure. During the testing, observations were also made on the general stability of the product and on any other aspects of its use that were relevant to its viability in the marketplace. A separate investigation was made on the validity of an unrealized TSR fix and this again proved to be 100% successful in correcting date incompatibilities.

The Product

Timebomb employs a DOS interface and is supplied on a single copy protected floppy disk that contains both the diagnostic and fix components of the application. This is a hardware only tool and makes no claims to diagnose or correct problems associated with software applications or data entry. All diagnostic testing is performed from the bootable floppy disk and no modifications are made to the local hard drive during this phase. Where the application determines that the PC will not handle Year 2000 dates correctly then the fix is installed automatically as part of the testing cycle. Tests are performed on the ability to handle dates up to 2040, to check correct leap year identification up to 2040 and to test for correct century rollover of the system clock and the Real Time Clock. During testing a single manual reboot is required. Once completed, a report is produced showing whether the PC has passed or failed in each of the individual sections.

Test methodology

Tests were carried out on fourteen PCs, seven of which were established to be year 2000 compliant and seven of which could not handle year 2000 dates. These PCs comprised a mixture of models with processors ranging from 25MHz 486 SX versions through to 333MHz Pentium II varieties. The operating system installed also varied from DOS 5 through to DOS 7 and Windows 95 and NT. Exact details of the PCs under examination are detailed in conjunction with the test results in table 1. VNU European Labs used manual date rollover and low-level diagnostics tools that examined the RTC, BIOS and system clocks to establish which of the PCs were Year 2000 compliant. Timebomb was installed and run on each PC according to the manufacturer’s instructions, and the results of the testing were recorded.

The following areas were examined during the testing of Timebomb:

1) Examination of validity of system clock diagnostics beyond Year 2000.

2) Examination of validity of system clock diagnostics for leap years beyond Year 2000.

3) Examination of validity of system rollover tests.

4) Examination of validity of RTC rollover tests.

5) Examination of correct installation of Timebomb fix.

6) Examination of correct functioning of Timebomb fix.

Analysis of results

The following analysis is based on the test sequence outlined under the methodologies section and should be read in conjunction with table 1 which shows results for each individual test.

1. Examination of validity of system clock diagnostics beyond Year 2000. In all instances VNU found Timebomb to be 100% accurate in reporting that the installed system clock under test could correctly handle dates between the years 1999 and 2040. No failures were observed (or are ever likely to occur) so we were unable to report whether Timebomb would correctly identify such an occurrence.
2. Examination of validity of system clock diagnostics for leap years beyond Year 2000. Timebomb reported that all systems under test could correctly identify leap years between 2000 and 2040. VNU European Labs knows of no PCs that would fail this test so although the results are entirely accurate, this is considered to be one of the least significant results.
3. Examination of validity of system rollover tests. As with the previous test there was a 100% pass rate in this section which was in agreement with our own findings about how the PCs handled the century rollover.
4. Examination of validity of RTC rollover tests. This is the most important test included in the Timebomb diagnostic suite. We found Timebomb to be 100% accurate in its findings that seven PCs passed and seven PCs failed in this section.
5. Examinations of correct installation of Timebomb fix. The seven PCs that were reported as failing in the RTC rollover test (test 5) all had the fix installed automatically as a result of this reported failure. This was confirmed by examining the autoexec.bat file to see that the fix file had been added, and by a manual reboot to ensure that the executable activated on start up.
6. Examinations of correct functioning of Timebomb fix. We could find no instances in which the installation of the fix program did not correct the incorrect date rollover of the RTC. Correct functioning was established by performing a manual rollover and then using low level tools to ensure that the three system clocks could all handle dates beyond 2000.

General observations

In all the instances that Timebomb was used, we did not encounter any occasions where the application failed to complete the tests successfully or caused any system failures on the PC on which it was being used. There were also no instances of the fix causing system failure or interacting with another application. We did observe two occasions in which, having completed the diagnostics, we were unable to exit the application by selecting the appropriate choice from the menu of the user interface. However this did not affect the test results and was no more serious than requiring a manual reboot of the PC.

From a cosmetic point of view there were a few irregularities in the reporting of test results. Firstly the system rollover test is incorrectly labeled as being a real time clock rollover test. As such the report gives two separate RTC rollover results, which can be in contradiction, and therefore confusing to the user. The screen display of the results (not the printed version) also states that the product is a demo and that a fix is not installed even if one has been. Similarly the instructions supplied with the product state that if a failure is detected then the report merely suggests that a fix be installed. In reality the fix is installed automatically.

There are also a few observations that we feel would add to the usability of Timebomb. Firstly we would have preferred it to be possible to run the tests directly from the user interface – i.e. selecting option 2 starts the tests as opposed to exiting the program and telling the user to reboot. We would also like to have seen the option to retest once the fix had been installed. If it is considered necessary to include leap year tests, why not include non leap years to see that these are also handled correctly. Finally there was no warning if the fix was removed or disabled.

Conclusion

As a result of this investigation, VNU European Labs has found that Timebomb was 100% successful at identifying and correcting all the test PCs that could not handle Year 2000 dates. Equally no occurrences of failure was reported on those PCs that did not have a Year 2000 hardware problem. We also found the program to be extremely stable and found no occurrences where the diagnostics did not complete successfully or where the fix failed to install.

Analysis of Timebomb TSR fix

A separate study was made on an unreleased TSR version of the fix utility. We tested functionality of this TSR by using the seven PCs that were reported by the diagnostics tool as being RTC rollover failures. The fix was loaded into conventional memory and then the PC manually rolled over to see whether the fix would correct date anomalies. Additionally a rollover was performed with the power off to see whether the TSR could handle this eventuality. In both cases, and on all systems, we found that the TSR performed its job successfully and corrected date incompatibilities. Occupying just 384 bytes of memory, it was also one of the most compact TSR fixes we have come across.

Table1

Table of Results

	Year 2000 compliant  (y/n)	System clock diagnostics  (pass/fail)	System clock leap year diagnostics  (pass/fail)	System clock rollover  (pass/fail)	RTC rollover  (pass/fail)	Fix installed  (y/n)	Fix worked  (y/n)
Compaq 486 SX/25  (Compaq 15/10/92)	n	p	p	p	f	y	y
IBM PC 340  (IBM 30/5/96)	y	p	p	p	p	N/A	N/A
Compaq Prolinea 4/50 (Compaq 15/12/90)	n	p	p	p	f	y	y
IBM ValuePoint 433 (IBM 30/3/95)	n	p	p	p	f	y	y
CIC Ventura (SystemSoft 23/12/94)	n	p	p	p	f	y	y
Compaq Proliant (Compaq 20/9/96)	n	p	p	p	f	y	y
IBM 330 (IBM 7/7/95)	y	p	p	p	p	N/A	N/A
Apricot Xen Apricot 1.2r1.7)	n	p	p	p	f	y	y
Hurricane (Award 4.51PG)	y	p	p	p	p	N/A	N/A
Acer P133 (AcerBios 2.1)	y	p	p	p	p	N/A	N/A
Roldec AMD 486 DX4/100 (Award 4.50G)	n	p	p	p	f	y	y
Carrera PII 266 (AMI V1.3)	y	p	p	p	p	N/A	N/A
HP Pavillion (Phoenix 13/5/97)	y	p	p	p	p	N/A	N/A
Panrix Lightning P2 333 (Award 11/11/97	y	p	p	p	p	N/A	N/A

 

index1.htm Back to Options.

  Reports2.htm What Solice Consulting Services have to say.