Re: quitting netsurf

On 14 May 2013, Daniel Silverstone <dsilvers@netsurf-browser.org> wrote:
> On Mon, May 13, 2013 at 05:36:29PM +0000, Tony Moore wrote:

> > For each result below, NS was quit, the Log file and Cache deleted,
> > NS was loaded, and then quit, without displaying any web pages
>
> It would be helpful if, in each case, the resultant log file was named
> appropriately and put in a folder.
>
> > For each result below, NS was quit, the Log file and Cache deleted,
> > NS was loaded, used to display three web pages.
> > http://news.bbc.co.uk/ , http://www.theregister.co.uk/ ,
> > http://www.iconbar.com/ , and then quit
>
> Ditto.
>
> And then, that whole folder could be zipped up and popped as an
> attachment to a bug report on our Sourceforge project, detailing your
> notes about timings and reproduction techniques.
>
> It would also help if you ran with nothing which would differ from a
> totally fresh install. i.e. move your choices, hotlist, urldb etc out
> of the way.

After hiding choices and hotlist (in addition to the other files noted)
the problem goes away. For the record:

For each result below, NS was quit, the Log, Cache, Cookies, Hotlist and
URL files were deleted, NS was loaded, and then quit, without displaying
any web pages

version load quit log
sec sec KB

#417 9 1 14
v3.0 9 1 15
#1125 json 9 1 14

For each result below, NS was quit, the Log, Cache, Cookies, Hotlist and
URL files were deleted, NS was loaded, used to display three web pages.
http://news.bbc.co.uk/ , http://www.theregister.co.uk/ ,
http://www.iconbar.com/ , and then quit

version load quit log
sec sec KB

#417 9 1 173
v3.0 9 1 421
#1125 json 9 1 484

I have the relevant logs but, since they record apparently normal
behaviour, there seems little point in posting them to Sourceforge.

Reinstating my choices file makes little difference to the results,
above, but reinstating the hotlist file, causes #1125 quitting to be
greatly delayed, as noted in a previous post.

The hotlist file size is 93KB, which I would not have thought to be
extraordinarily large, and, anyway, #417 manages to write it in 3
seconds, when quitting, but #1125 takes an order of magnitude longer.
Would logs relating to this different behaviour be useful?

Tony