On 11/09/2007, Ronan Klyne <[EMAIL PROTECTED]> wrote: > sebb wrote: > > On 10/09/2007, Ronan Klyne <[EMAIL PROTECTED]> wrote: > >> sebb wrote: > >>> By the way, BeanShell tends to leak more memory if the script is on > >>> the screen; using a file seems better. Best is probably to define the > >>> script as a method in a startup file and then invoke from a file or > >>> the screen. > >> Does this just apply to Samplers? I ask because there I see no option to > >> source pre/post-processors from files. In this situation, would it help > >> to keep leaks to a minimum if I had a single line of script on screen > >> calling a function? > > > > No - all the BeanShell test elements now have their own initial files > > - see jmeter.properties in 2.3RC4. > > Sorry, not quite what I meant to ask... > I saw a text field in the UI panel for a beanshell sampler which allowed > me to specify a file name. I couldn't see a similar option for > pre/post-processors, and was wondering if just putting a function call > in the script field would be as efficient?
It should be more efficient than including the code on the screen, but perhaps not as efficient as using a file. Some of the beanshell GUIs don't have file fields; they could/should be added in a later release. > > # r > > -- > Ronan Klyne > Business Collaborator Developer > Tel: +44 (0)870 163 2555 > [EMAIL PROTECTED] > www.groupbc.com > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
