“In Ed’s previous column, he focused on socket programming and performance within a single system. In a future column he will pick up where he left off, but his topic this month is management of threads and processes in Linux and Windows systems. He walks through the differences between processes and threads, shows how to create and destroy them, and writes a program you can use to study thread management on your systems.” Read the intersting aricle at IBM DeveloperWorks.
Managing Processes and Threads
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
http://java.sun.com/docs/books/vmspec/html/Threads.doc.html
Also by IBM:
http://publibn.boulder.ibm.com/doc_link/en_US/a_doc_lib/aixprggd/ge…
Sorry the url is so long
While overall the article points outs the high overheads of Windows, and I certainly would trust the figures, one issue gets under my skin.
When he ran his test longer his results got worse on Windows than on Linux. He later points out he had a bug in the program causing a memory leak, but it just doesn’t seem professional to publish graphs that show degrading performance when the test program had bugs in it which he willingly admitted. Why bother at all, and why didn’t he publish the corrected graphs. A reader who skimmed the article and didn’t understand the finer points of the bug would conclude that windows was worse under that scenario.
I would say that was bordering on manipulative journalism. What do others think?
P