One solution I have found for this is to write my own logging application. I write mainly web-based software and AGI stuff, and all my software makes use of the same logging application. Basically what happens is that when I want to write to the log (which is stored in a MySQL database so analysis and viewing the logs is enhanced), the software simply tells the OS to execute the logging utility which can be niced or deprioritised on the system. This way, the software doesn't have to worry about I/O to the log, it doesn't have to wait for MySQL and it can carry on without the hassles involved in rewritting bits of logging code for each project. I have found this works really well for me, and having all the logs for all my code in a central database makes it easy for me to write some nice little AJAX/PHP app to apply filters to the logs or to analyse the logs, whatever I want basically. However, I know this method is not practical for many situations.
Never before has the term 'web' better described the nature of the internet. In the above article, it is questioned as to whether the fact that content cannot be directly navigated to suggests that the term 'web' is not as applicable. I would like to disagree. In the past, yes, there was a web-like network structure where content was neatly separate and distinct, thus it was possible to directly navigate to any content. However, due to the highly integrated nature of web 2.0 content achieved through such technologies as AJAX, XML, API's etc... information is shared and exchanged with such freedom and flexibility, that the content itself resembles a web-like structure, thus mimicking it's underlying hardware architecture. Thus, I feel that today's Web is even more of a 'web' than the web of yester-year.
If there is to be a new term used at some point in the future, perhaps the "Intermesh" would be an appropriate replacement.
Slashdot Mirror
One solution I have found for this is to write my own logging application. I write mainly web-based software and AGI stuff, and all my software makes use of the same logging application. Basically what happens is that when I want to write to the log (which is stored in a MySQL database so analysis and viewing the logs is enhanced), the software simply tells the OS to execute the logging utility which can be niced or deprioritised on the system. This way, the software doesn't have to worry about I/O to the log, it doesn't have to wait for MySQL and it can carry on without the hassles involved in rewritting bits of logging code for each project. I have found this works really well for me, and having all the logs for all my code in a central database makes it easy for me to write some nice little AJAX/PHP app to apply filters to the logs or to analyse the logs, whatever I want basically. However, I know this method is not practical for many situations.
Never before has the term 'web' better described the nature of the internet. In the above article, it is questioned as to whether the fact that content cannot be directly navigated to suggests that the term 'web' is not as applicable. I would like to disagree. In the past, yes, there was a web-like network structure where content was neatly separate and distinct, thus it was possible to directly navigate to any content. However, due to the highly integrated nature of web 2.0 content achieved through such technologies as AJAX, XML, API's etc... information is shared and exchanged with such freedom and flexibility, that the content itself resembles a web-like structure, thus mimicking it's underlying hardware architecture. Thus, I feel that today's Web is even more of a 'web' than the web of yester-year. If there is to be a new term used at some point in the future, perhaps the "Intermesh" would be an appropriate replacement.