John Archambault:
Digitalist
|
|
|
|||
| Previous Entry << Blog Home |
ARCHIVE |
Subscribe:
 |
 |
 |
|
||
|
|
|||
| July 23, 2008 |
 Digg
|  del.icio.us
|  Facebook
|  newsvine
|  Reddit
|
||
| Plugin Blues | |||
|
A few months ago I began organizing the Boston JAVA Meetup Group, a sociable group
of programmers who meet for lectures about and discussions of the JAVA programming language. Recently, while discussing
Eclipse, several of us began to lament the performance (and diagnostic) issues that can
arise from having too many plugins. One group member said he has gone so far as to have three or four instances of Eclipse
installed, so as to assist with the diagnosis of plugin-related performance issues.
I left that meetup thinking about the problems I've noticed while developing for Firefox; a plugin I had installed was inhibiting the proper display of PDF files within the browser, and I only have about four plugins. I've grown accustomed to the fact that any Firefox-specific bugs I encounter may be plugin-related, but then I have to determine which plugin is creating the problem. This is complicated even more by issues that arise from software updates. What can be done to make dealing with plugin issues easier? It seems tedious to sift through developer forums for the answers to such problems. Should I really need to invest that much time? When I look at the Firefox Ad-On Page, I see very little that can help me. Is there anything to be found on the Eclipse site? I see nothing. Plugin-related problems are likely to grow exponentially as plugin usage becomes more prevalent, so a more practical solution needs to be arrived at to enhance the open-source experience. Perhaps a first step would be to develop some sort of web database tool to assist with tracking known issues that arise due to plugin incompatibilities - a place where users can search for entries by plugin name. But it would be really great if someone created a diagnostic tool that could examine what plugins are installed for a particular program, and then alert the user to known or potential performance issues. Ideally, it would be best to find out about potential problems the moment a user begins to install a plugin. Maybe we'll have a plugin to diagnose plugins ;) I have confidence that the open-source community will respond to this emerging issue appropriately. More than likely, a team will be formed, consisting of developers who have been motivated by their own bad experiences. Hopefully, this will happen soon - doing so will only accelerate the adoption of the open-source philosophy. |
|||