[9fans] abaco and google links

[9fans] abaco and google links

[Steve Simon]

Abaco can nolonger follow googles links as they now redirect through
googles own servers so they can "help" you.

anyone fixed abaco to follow such links?

-Steve

Re: [9fans] abaco and google links

[cinap_lenrek]

works just fine on 9front

clicking:

http://www.google.de/url?q=http://de.wikipedia.org/wiki/Plan_9_(Betriebssystem)&sa=U&ei=7k99UtrDBceItQaenoDgBA&ved=0CCgQFjAB&sig2=qgPElOPSaFdndrFNpQgT-g&usg=AFQjCNHQJbuSIfAqYc3tErY7YT2PnH-qzw

gets me to:

http://de.wikipedia.org/wiki/Plan_9_(Betriebssystem)

--
cinap

Re: [9fans] abaco and google links

[Steve Simon]

> works just fine on 9front

I sould have guessed :-)

I will look at the diffs.

-Steve

Re: [9fans] abaco and google links

[erik quanstrom]

On Fri Nov  8 16:21:55 EST 2013, steve@quintile.net wrote:
> > works just fine on 9front
>
> I sould have guessed :-)
>
> I will look at the diffs.

also works in 9atom.

- erik

Re: [9fans] abaco and google links

[erik quanstrom]

iirc, the issue is very minor—a problem interpreting urls by webfs.

- erik

[9fans] abaco and google links

[Jonas]

[-- Attachment #1: Type: text/plain, Size: 205 bytes --]

for my native plan9 (on ibm t42) abaco's google links are without one "/", so i have add one manualy in address bar. (google link shows for example http://www.google.com/url?q=http:/9fans.net/archive/..)

[-- Attachment #2: Type: text/html, Size: 580 bytes --]

Re: [9fans] abaco and google links

[erik quanstrom]

On Thu Nov 14 04:06:16 EST 2013, puta2001-goo@yahoo.com wrote:

> for my native plan9 (on ibm t42) abaco's google links are without one
> "/", so i have add one manualy in address bar.  (google link shows for
> example http://www.google.com/url?q=http:/9fans.net/archive/..)

the problem is urlcombine and urlcannon.
try the versions in /n/atom/plan9/sys/src/cmd/abaco/urls.c

- erik

Re: [9fans] abaco and google links

[hiro]

somewhat related:

s./url?q=..g
and while we're at it
	/<div id="tads"/,/^<\/div>/d;
	/<script type/,/<\/script>/d;
	/<div id="gbar">/,/<div class="gbh" style="right:0">/d;
	/<td class="sfbgg" valign="top">/,/<\/td>/d;
	s/<div class="lsbb"><button/<div class="lsbb"><input/;
	/<div id="rhs_block"/,/^<\/div>/d;
	/<h1><a href="\/webhp?hl=en"/d;
	/<div class="flc"/,$d;
	/^<h3 class="r"><a href="\/images?q=/,/<li class="g">/d'

Still experimenting what's the best place to put such code. I'm
definitely not going to use java scripts in the browser for this, so
right now I have my own /search script on my home server that wgets
from google and eventually runs this fix on the output.
Oh, don't forget to format nicely (I'm using tidy on linux for
html-tidying, there might also be utf-8 issues, so ie=utf8&oe=utf8 in
the http query to google might help)

I'm not presenting a finished product, because I'm on linux and my
script sucks. I didn't yet manage to make a better, reusable
"framework" for this. Although I've been playing with the thought of
using a trusted, SSL-capable http proxy for the purpose. But the SSL
part of it made me stale in disgust.