Early reliability assessment of UML based software models
WOSP '02 Proceedings of the 3rd international workshop on Software and performance
The Virtues of Assessing Software Reliability Early
IEEE Software
| Hi-index | 0.00 |
Can COTS software be tolerated in high assurance environments? Or is this hopelessly impossible? My position is that COTS software will exist in high assurance environments (in the near future) no matter what prudence or due diligence suggests. Prudence and due diligence would argue that it is foolish to expect dependable functionality from generic products that are mass produced, engineered for the typical user (who can tolerate failures because they are mere nuisances), suffer from shrunken development and testing schedules, and carry shrink wrap disclaimers. Prudence and due diligence would ask why we opt to use COTS software when we cannot even reach our high dependability goals via code that is written from scratch and according to standards that are known to improve dependability. After all, the COTS vendors do not follow these standards. Is it reasonable to expect software that is intended for the mass market to be highly dependable? Probably not.