Prototype, periodically_call_remote & CPU overdrive


#1

Hello,

im playing around with “periodically_call_remote”
(frequency 1 sec) & even when i have it call an
empty “dummy” method it revs my cpu up to almost
100%!!?

i’m running Rails 0.13.1, Ruby 1.8, on a OS X 10.3.9 on a G5 w/
Firefox & i’ve alternated btw Prototype 1.3.0 &
1.4.0_rc2 with almost identical results.

Using the dummy call (frequency 1 sec) with Prototype
1.4.0_rc2 & checking JS errors at app startup & exit…
at startup i get a series of JS warnings from
prototype.

Then the “periodically_call_remote” call starts firing - without
(apparent)
incident - but the cpu revs way up into the 90’s!!

Then when i exit the app i get a whole slew of a particular error;

=======================================================

Error: [Exception… “Component returned failure code:
0x80040111 (NS_ERROR_NOT_AVAILABLE)
[nsIXMLHttpRequest.status]” nsresult: “0x80040111
(NS_ERROR_NOT_AVAILABLE)” location: “JS frame ::
http://localhost:3000/javascripts/prototype.js ::
anonymous :: line 678” data: no]
Source File:
http://localhost:3000/javascripts/prototype.js
Line: 678

======================================================

& then it can take up to 30 secs for cpu to rev back down after app
exit!!

Has anyone seen this error before or experienced such a problem?

Any callback to the server surely entails a CPU hit, but isn’t this
excessive?

And doesn’t the series of the same error on exit indicate something very
wrong?

Any wisdom would be greatly appreciated!

Thanks much!

Randy


#2

I see similiar behavior in Firefox on Windows…

Joshua Sierles


#3

jsierles wrote:

I see similiar behavior in Firefox on Windows…

Joshua Sierles

Hey Joshua,

Thanks for the response. So far, I haven’t been able to find any other
references to this problem…

I sent a detailed email to the author of Prototype, Sam S.
removed_email_address@domain.invalid, including the javascript (from the Firefox javascript
browser) warnings that poped up when I open the page with the
“periodically_call_remote” & the error message that pops up when I exit
the test app.

In my case, my cpu will continue to rev for 20 ~ 45 seconds after I exit
the app… seems to me that it must be caught in some kind of loop (?).

I’ll post again if I get an answer from Sam. In the meantime, however,
perhaps if you sent him an email too, he might be more likely to look
into it…

Randy