Discussions on scripts by thorbenhauer | Userscripts.org

Google update, replied by dob

dob — Sat, 24 Oct 2009 11:54:32 +0000

Due to a google update, there have been some CSS errors such as the page links being bordered.

This userstyle fixes that:
http://userstyles.org/styles/22066

Isn't working for me, replied by dwidmann

dwidmann — Fri, 09 Oct 2009 16:24:58 +0000

Yes.

Isn't working for me, replied by thorbenhauer

thorbenhauer — Fri, 09 Oct 2009 09:42:04 +0000

Does nothing mean that you still get this error on this page?

Isn't working for me, replied by dwidmann

dwidmann — Fri, 09 Oct 2009 07:51:50 +0000

Well that's just odd. I tried nuking my config dir and trying again. Nothing. Next I tried dowloading a neatly self contained tarball of the 386 version and trying it once more with that ... nothing. I'm perplexed.

Isn't working for me, replied by thorbenhauer

thorbenhauer — Thu, 08 Oct 2009 09:57:47 +0000

Using Opera 10 and a brand new profile this script works fine for me on your provided URL.

Isn't working for me, replied by dwidmann

dwidmann — Thu, 08 Oct 2009 07:43:51 +0000

Looks like a lovely script, but I get this error in Opera 10:

JavaScript - http://images.google.com/images?hl=en&um=1&sa=1...
User Javascript thread
Error:
name: TypeError
message: Statement on line 86: Cannot convert undefined or null to Object
stacktrace: Line 86 of User JS script
div.addEventListener('DOMNodeInserted', function (event) {
Line 51 of User JS script
(function () { // function wrapper for Opera

Request, replied by thorbenhauer

thorbenhauer — Wed, 16 Sep 2009 09:34:05 +0000

@Roccobot
As long as browsers don't provide a API for JavaScript to control link behavior, that's not possible for Greasemonkey. You have to configure your browser to fulfill you needs.

suggestion for "odd" urls, replied by thorbenhauer

thorbenhauer — Wed, 16 Sep 2009 09:26:20 +0000

You can use the same feature introduced here:
http://userscripts.org/topics/13339?page=1#post...
Just add /example.com/ to the list.

Long URLs break image layout, replied by thorbenhauer

thorbenhauer — Wed, 16 Sep 2009 08:32:37 +0000

Fixed in current version. Thanks for the hint!

script can break, replied by thorbenhauer

thorbenhauer — Wed, 16 Sep 2009 08:26:46 +0000

Please test the new version. Hope it fixes the issue.

script can break, replied by Joshatdot

Joshatdot — Wed, 16 Sep 2009 03:29:34 +0000

I have encountered this same error, whenever selecting Options, Larger then... relinker does not work anymore.

broken, replied by thorbenhauer

thorbenhauer — Thu, 03 Sep 2009 10:21:09 +0000

Please provide a test page on which this script fails.

broken, replied by sdfghjklkjhgfds sdfghjiuytr

sdfghjklkjhgfds sdfghjiuytr — Thu, 03 Sep 2009 02:00:21 +0000

it is not doing anything
not even generating an error in the error console
i am guessing elements.length is 0

Long URLs break image layout, replied by Wes M

Wes M — Tue, 18 Aug 2009 02:19:33 +0000

Having this script enabled can sometimes break the layout of the image results because the entire URL of the source page is displayed, pushing other images over to make room if the URL is too long. The problem is with the section of the code:
font.innerHTML = "<a href=\"" + contextLink + "\">" + domain + "</a>";

Change it to this and the links retain their original formatting with the ellipsis at the end for longer URLs:
font.innerHTML = "<a href=\"" + contextLink + "\">" + font.innerHTML + "</a>";

script can break, replied by ekbworldwide

ekbworldwide — Sun, 02 Aug 2009 03:40:57 +0000

Recently (the last 24-48 hours?) the google image layout changed.

Options are set up differently.

If you use "Larger than..." - the script fails - and an image opens in a frame.

request - block images by size, replied by ekbworldwide

ekbworldwide — Wed, 29 Jul 2009 02:48:22 +0000

How about if there was an option to block by minimum width and/or height?

If the image is not much bigger than the thumbnail - I doubt people are clicking on it.

I'm using AutoPager - so getting more results is easy. The problem is - as it often is now - I need/want a filter.

The "Show: [medium]" isn't good because I never want to see something that's about 250 x 250 or smaller. And I don't want larger images filtered out.

_______

AutoPager
https://addons.mozilla.org/en-US/firefox/addon/...

suggestion for "odd" urls, replied by ekbworldwide

ekbworldwide — Wed, 29 Jul 2009 02:39:23 +0000

>> A script could check for problematic... Only some unpredictable URLs are affected...

I should have said what I was thinking in my head.

How about having a bit of code to allow the user to enter problematic domains that need their images wrapped in image tags?

Like this:

// put problematic domains here
{
example.com
}

And then if the user can just enter the domains.

Doesn't work anymore with Fx 3.5.2pre, replied by myk04l

myk04l — Sun, 26 Jul 2009 20:48:05 +0000

thanks for update! works great now.

Doesn't work anymore with Fx 3.5.2pre, replied by Raloc

Raloc — Sat, 25 Jul 2009 15:09:01 +0000

I know that the code depend of the html of google image. But it worked for me with 3.0 and not anymore with 3.5 so I thought it was a TraceMonkey problem. And that's just weird it worked with 3.0.

But today it doesn't work anymore (with 3.0) since I upgrade to Relinker 0.5.6.

Thanks a lot for the fast update thorbenhauer. You're my new hero (:

Where is Extract, replied by thorbenhauer

thorbenhauer — Sat, 25 Jul 2009 14:04:38 +0000

This script is currently not maintained and I plan to remove it from userscripts.org. It was designed to work for the old eBay search pages. Therefore you can't see a menu entry under User Script Commands.

suggestion for "odd" urls, replied by thorbenhauer

thorbenhauer — Sat, 25 Jul 2009 13:44:18 +0000

The URL is not the problem here. Again it's the server that sends the wrong content type (application/octet-stream) for Firefox to display the image properly. Looking at your search for springfieldkarateandfitness.com one can see that this is not a generic case for this server. Only some unpredictable URLs are affected, unlike the cases we had before (blogspot, blogs.yahoo.co.jp). So adding a regular expression in the script code is some sort of an overkill and just based on trial-and-error.
A script could check for problematic HTTP header in advance, but this would bloat web traffic for each page load because this has to be done for every picture to be really sure. High costs for some rare cases.

Doesn't work anymore with Fx 3.5.2pre, replied by ekbworldwide

ekbworldwide — Sat, 25 Jul 2009 11:07:25 +0000

To anybody reading this - the code is dependent on the html of the google image page - if the html changes - the script might break. That's the bad news. The good news is that thorbenhauer tends to update the script very quickly.

If it stops working check the "Last update [when]" on the 'About' page.

Doesn't work anymore with Fx 3.5.2pre, replied by thorbenhauer

thorbenhauer — Sat, 25 Jul 2009 09:24:44 +0000

This had nothing to do with Fx 3.5.* or 3.0.*. It was just Google modifying their site - normal business that happens from time to time. Should be fixed with the latest update. Try it out!

Thanks for reporting!

Doesn't work anymore with Fx 3.5.2pre, replied by myk04l

myk04l — Sat, 25 Jul 2009 03:12:41 +0000

im using 3.0.12 and it just stopped working for me. i thought it was some google update thing.

symptoms experiencing is lack of links under images as well as image not redirecting.

Doesn't work anymore with Fx 3.5.2pre, replied by Raloc

Raloc — Sat, 25 Jul 2009 02:07:59 +0000

I got « font is null » on line 96 of google_image_relinker_mo.user.js file.

The line starting with « font.innerHTML = »

Maybe is it a bug that will be corrected with the official release of 3.5.2, maybe it's a new feature of TraceMonkey, maybe... something else I can't figure. I tested it on a new profile so I'm sure it's not due to a conflict with a another script or add-on.

I'm pretty sure it comes from Mozilla and I hope they will fix it. It would be painful that "old" javascript or DOM (I'm not a coder I don't really understand those things ^^) don't degrade gracefully with the new engine. If someone is involved with some firefox developer it would be great to tell them. I don't know who to contact nor what to say...

By the way, still works great with 3.0.12. And I must say that script is awesome. Can't live without it. So I'll return to 3.0 'til it works again (: