[Ur] Substantial memory leak in the JavaScript runtime

Adam Chlipala adamc at csail.mit.edu
Sun Aug 28 20:46:35 EDT 2016


Oh, I definitely understand the motivation for hunting down this kind of 
problem, and I appreciate your efforts.  I run a number of single-page 
Ur/Web apps in production.  The fact that no one is complaining about 
them provides some evidence that this isn't a deadly problem in practice.

On 08/28/2016 08:18 PM, Saulo Araujo wrote:
> Hi Adam,
>
> Memory leaks in the JavaScript runtime like these are not important in 
> classical web applications because the browser moves to a new page 
> very frequently, thus freeing all memory that the previous page has 
> allocated. However, In single-page applications 
> (https://en.wikipedia.org/wiki/Single-page_application), the browser 
> loads just one page, which makes the absence of memory leaks in the 
> JavaScript runtime crucial. Maybe most of the applications that are in 
> production are of the classical kind. This would explain why there are 
> no complaints about memory leaks in the JavaScript runtime. As I am 
> developing a single-page application 
> (https://github.com/saulo2/timesheet-ur), this is an important matter 
> to me. Therefore, I am willing to contribute by hunting those leaks 
> and suggesting fixes. Also, I believe that more and more single-page 
> applications will be developed with Ur/Web rather than classical ones.
>
> Sincerely,
> Saulo
>
>
> On Fri, Aug 26, 2016 at 1:12 PM, Adam Chlipala <adamc at csail.mit.edu 
> <mailto:adamc at csail.mit.edu>> wrote:
>
>     On 08/24/2016 03:02 PM, Saulo Araujo wrote:
>
>         I am happy to say that your patch also fixes the memory leak.
>
>
>     OK, great.
>
>         I believe there is another memory leak in the JavaScript
>         runtime (see the end of the previous message). I am gonna look
>         into it.
>
>
>     I'll appreciate any help finding more memory leaks, though in the
>     foreseeable future I probably won't be spending time tracking them
>     down.  There don't seem to have been any complaints yet about
>     memory leaks in connection to production applications.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.impredicative.com/pipermail/ur/attachments/20160828/9ee930b8/attachment.html>


More information about the Ur mailing list