TAG | Is Open Source software safe and secure? |
|
| by David M Williams | |
| Thursday, 22 May 2008 | |
|
Page 4 of 4 Other languages like Java seek to avoid exposing programmers to pointers. On the one hand, this removes the risk of harm but on the other hand can teach new coders sloppy habits which they then carry on to non pointer-safe languages.The second most common type of code defect across all projects are resource leaks. These accounted for 26% of problems. This type of flaw means that memory is seized for use but then not later released. According to Coverity, this happens most frequently when an application requires a resource like a disk file but finds, after allocating the memory, that the resource is unavailable. This is the type of programming problem which explains why a computer might slow down after days of continued operation; even though there are no, or few, programs apparently running, the computer’s memory is being choked up with allocations that aren’t subsequently freed. This is invariably the need for a computer to have routine reboots. Other faults uncovered by Coverity are the dreaded buffer overflows, which are a hacker’s delight, being a direct means of altering program code and being able to execute malicious instructions, as well as more pointer and memory problems – like memory being accessed after it has been freed. For any software developer interested in producing excellent code, the Open Source Report is interesting reading. What’s terrific, however, is how it statistically proves that open source software is on the whole a mature industry with skilled coders and that Scan has independently verified the safety of significant works like PHP and Samba. The open source projects on rung 0 ought to be encouraged by the community to participate but for the most part the news is good that software quality in open source is improving and users can have confidence in the applications they use.
|
Tags
| < Next story in category | Previous story in the category > |
|---|
