On 9 Jan., 02:11, John J Barton <[email protected]> wrote: > On Jan 8, 2:51 pm, Sworddragon <[email protected]> wrote: > > > > So far I see no evidence of a memory leak. In your machine you have > > > lots of memory taken in the exotic case of tens of thousands of > > > reloads. Also in you see lots of memory taken for manual refreshes > > > over 7 days. Still the question remains as to the cause of your > > > problems. So far we have no way to know if the problem is a Firebug > > > memory leak. The fact that the problem disappears when you disable > > > Firebug does point to Firebug as being involved. But until we have a > > > test case that shows the problem on my machine I can't help you. > > > I don't know why you aren't willing to make just one longer test. You > > must only one time make my test with 20000 reloads (on my processor it > > needs ~1 hour) if the script panel is selected > > For two reasons: 20,000 reloads is not realistic and 1 hour is not > practical for fixing a memory bug. > > I really appreciate your efforts. But we need either a more realistic > test case or a faster one. I know you are convinced there is a > problem, but Firebug has lots of problems. So far the information > looks like this problem is not very significant. I'm sorry you have to > restart Firefox every few days, but really it's not such a big cost > for using Firebug. I think it makes more sense for me to work on > changing Firebug so users don't have to reload to debug. Then any > memory increase on reload won't matter.
I doesn't really have a problem with this memory leak it is just a little annoying. I'm reporting many bugs on many applications not just firebug. Not all reported bugs are beeing fixed and sometimes my reports are completely ignored. > If you want to do more experiments I will make more suggestions. > > First I would suggest you check that the script panel is truly related > to the problem. Check another panel, eg HTML and Console. If the > memory increase only happens when the Script is selected, then try the > script-limiter, the small menu on the tool bar that says "All". > > Another kind of change is to increase the HTML (by 10x) or the > scripts. At first I have updated firefox to version 3.5 RC1. I have made a lot of tests (with 1000 reloads) the last hours. I have opened the firebug main panel and closed them directly. After 20000 reloads there was no memory leak so the main panel must be opened to get the memory leak. I have called a debug function that does some things. Allocationg content to various types of variables and manipulating the DOM. Nothing of them increased the memory usage more as without this debug function. Even 30000 calls of this function (not in a loop, I have copied and pasted it) per reload doesn't change anything. Removing the CDATA from the javascript code doesn't change anything too. After this I have added more HTML elements. 15000 <div>Debug</div> and the memory usage was increasing after 1000 reloads to ~70 MB (the previous tests without this additional HTML content was using just ~55-60 MB. I have maded some tests with various things like nested elements and the memory usage was ~70 MB too. Using CDATA or a comment to the 15000 <div>Debug</div> needs just ~55-60 MB and the same happens with using only one div - container that contains 250000 zeros (0) as content. It seems that the memory leak appears on parsing HTML elements independent from the amount of content in the HTML elements. At next I'm doing 6 tests with 20000 reloads and 15000 <div>Debug</ div> with another panels selected. I have created 6 new profiles for this test and started them in seperate processes. The tests will need a long time (maybe a day or longer) and if they are finished I will post the results here.
-- You received this message because you are subscribed to the Google Groups "Firebug" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/firebug?hl=en.
