Re: Filtering webpages
Gavin Wraith wrote: > I would also like to be able to discriminate content by source URL and to > give permissions for which should be blocked or which allowed through. At the domain level, you could add entries to your hosts file for sources you want blocked, along the lines of: 127.0.0.1google-analytics.com This will effectively stop websites you visit on that computer *with any browser* from loading anything from that domain. If you can do something similar in your router, you will achieve the same result for any computer on your network. -- Vince M Hudd Soft Rock Software Don't forget to vote in the 2015 RISC OS Awards: www.riscosawards.co.uk/vote2015.html
Re: Very slow page rendering
Am Donnerstag, den 28.01.2016, 21:04 + schrieb Peter Slegg: > > https://www.royalmail.com/track-your-item > > Another page that took ages to display and looked like the css had > failed as well. > Early versions of the atari port crashed because of stack size issues (caused within mintlib regex module, called by NetSurf CSS parser, while it was processing exorbitant strings (3k, which expanded to several MegaBytes within mintlib regex module)...). I prevented the crash by compiling mintlib with the +DEFS=-DREGEX_MALLOC define. This is also applied to the CI builds. Just a guess: there is some kind of slowdown when doing excessive malloc operations with MiNT. Greets, Ole
Re: Very slow page rendering
> Date: Sun, 10 Jan 2016 16:43:40 +0100 > From: Jean-Fran?ois Lemaire > Subject: Re: Very slow page rendering > > On Saturday 09 January 2016 19:43:53 Peter Slegg wrote: > > >Date: Sat, 09 Jan 2016 16:11:40 GMT > > > >> Peter Slegg wrote: > > >>>http://git.netsurf-browser.org/netsurf.git/tree/atari/gemtk/guiwin.c > > > >>>This page takes abut 20mins to download and render, Highwire browser > > >>>takes about 6sec. > > > No criticism, I am hoping this might help the devs find some speed > > improvements. > > I have an Atari 2.9 version lying around and with that build it takes 100 secs > to render that page. Still very slow but much less so than with the 3.* > builds. > > Cheers, > JFL https://www.royalmail.com/track-your-item Another page that took ages to display and looked like the css had failed as well. Peter
Re: Big push on testing needed
In message <5548c5c513cvj...@waitrose.com> Chris Newman wrote: > In article <3bc72c4755.davem...@my.inbox.com>, >Dave Higton wrote: >> Big news... >> Current test (CI) builds are now release candidates. Yes, a new >> release of NetSurf is imminent. >> Please, everybody, download the latest test build (which will, >> of course, change as bugs are found and fixed), give it a good >> thrashing, and get your bug reports in. >> Please also note that, since it's now close to release time, the >> Javascript setting in Choices->Content is obeyed (and has been >> for a couple of days or so now). > Greetings from sunny Australia (gloat, gloat), > The 38 Degrees petition page at > https://secure.38degrees.org.uk/Scotland-stop-CETA > Is a bit of a pigs dinner. It takes ages to load, frames overlap & the > signing link doesn't work. > Dev CI #3315 on Virtual Acorn Adjust 4.39 > Same effects with JS on or off. > Works OK using Maxthon browser on the Windows side. > Does anyone see the same effects? 2.6s JS off, 5.4s JS on, CI #3312, 5.21 (RC14), Pi 2 @ 900MHz. Page display is substantially different compared to Otter 0.9.09 on RISC OS. The signing link doesn't work in Netsurf with JS on or off, but Otter seems fully functional with JS enabled. I didn't make a precise count of page loading time in Otter - it's considerably slower than NS, about 20-30 secs JS off/on. -- George
