Forum Moderators: open

Message Too Old, No Replies

Google stops understanding certain characters

Google is displaying the character » in the page title as?

         

mvl22

12:10 pm on Nov 25, 2002 (gmt 0)

10+ Year Member



I've noticed on the 20-30 sites that I run where I have used the character » in the page title, e.g.

Department of Foobar >> Contact details

(where the >> is actually a single character)

[this is what appears in the window heading in a webbrowser, or in google's main title in the listings]

it is now showing as the somewhat unprofessional-looking:

Department of Foobar? Contact details

(the question mark actually shows twice - but this forum doesn't seem to accept me typing this in..)

Has anyone else found this, and why has google changed its algorithm to break this? Surely » is legal HTML? (And all my pages are valid XHTML)

gstewart

12:18 pm on Nov 25, 2002 (gmt 0)

10+ Year Member



Type » into google and you get a blank page. Never seen one of those before.

msampson

2:55 pm on Nov 25, 2002 (gmt 0)

10+ Year Member



have you tried »?

Fiver

2:57 pm on Nov 25, 2002 (gmt 0)

10+ Year Member



I've noticed this from some computers and not others.

could be coincidence, but so far, the only computers it's messed up on are running IE 6.0+

all 5.5's I've tried it on display the chars the proper way.

Fiver

5:01 pm on Nov 25, 2002 (gmt 0)

10+ Year Member



Type » into google and you get a blank page. Never seen one of those before.

you get a blank page searching for any character that google sees as a space. including, a space.

But it's always been that way hasn't it?

I still think it's an IE thing, unless someone's seeing something I don't?

mvl22

5:19 pm on Nov 25, 2002 (gmt 0)

10+ Year Member



Looks like it's been fixed literally this afternoon. Perhaps googleguy has been watching this forum...

(NB - I had avoided » because this is less well supported by older browsers.)

Fiver

2:44 pm on Nov 26, 2002 (gmt 0)

10+ Year Member



not fixed from here :/

mvl22

2:47 pm on Nov 26, 2002 (gmt 0)

10+ Year Member



I agree, it's not fixed, in fact. For a short while yesterday I got a few correct results, but others since have not been correct.

I get the same effect on mozilla as IE, so I really don't think it's a browser issue, especially as it was fine before.

Fiver

3:43 pm on Nov 26, 2002 (gmt 0)

10+ Year Member



alright, so if it's not a browser issue, what are we dealing with here...

pretty open to possibilities I'd say.

what's strange is, FAST always had a problem with these characters (but fortunately replaced them with spaces instead of question marks) for the longest time. Just recently a site which uses double < on either side of the title shows the correct character trailing the title, but a space before it.

so while FAST has almost fixed their problem, google's gone from just fine to... all wrong.

mvl22

1:14 pm on Nov 27, 2002 (gmt 0)

10+ Year Member



I've noticed that google.co.uk doesn't have this problem, whereas google.com does.

Fiver

2:20 pm on Nov 27, 2002 (gmt 0)

10+ Year Member



you're right... alright, time to email searchquality I guess.

Fiver

3:17 pm on Nov 28, 2002 (gmt 0)

10+ Year Member



looks like the update is taking care of it (lets hope)

Fiver

4:16 pm on Dec 2, 2002 (gmt 0)

10+ Year Member



doesn't look like it has been fixed actually. This... is a problem. for some anyway.

emailed google. I'll keep the three of us in the world who are intersted, posted ;)

mvl22

10:25 pm on Dec 4, 2002 (gmt 0)

10+ Year Member



Well, this evening it seems to have been corrected, but who knows if this will stay?

Fiver

6:02 pm on Dec 5, 2002 (gmt 0)

10+ Year Member



still out for me mvl...

its a strange thing because during the update i saw the characters presented correctly occasionally on ww2... oh well.

no reply from search-quality@ ... but i was in a mood and called the example serps kinda ugly, maybe they got offended :)

Fiver

7:55 pm on Dec 6, 2002 (gmt 0)

10+ Year Member



hm, google's been uncommunicative so far on this, email and forum wise anyway. Think they just don't want people to make their titles stand out?

I already format the entire process of web site creation based around what they like... c'mon.

I hope this isn't the real reason for inaction..

mvl22

1:18 am on Dec 7, 2002 (gmt 0)

10+ Year Member



> I hope this isn't the real reason for inaction..

What do you mean?

They're probably just taking a while to look into it. There's probably a bug that's made its way into some of the servers, hence the inconsistencies.

It is a very annoying bug, though.

Fiver

3:26 pm on Dec 19, 2002 (gmt 0)

10+ Year Member



What I mean is I don't understand why they haven't addressed the issue. They haven't answered emails or forum posts about it.

They have question marks in place of characters flying all around their results pages... the pages that are google's main product... and it's not an addressable issue?

This bug hasn't resolved itself yet, and now I see it in the SERPs at bbc.co.uk, and aol.com

GoogleGuy

6:10 pm on Dec 19, 2002 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Question marks can be a sign that (for example) someone composed a page in windows and now a browser on Linux or another platform is using a different character encoding. I see people who use special Windows chars for '"' all the time and it can come out as a '?' in Linux and other browsers.

But that doesn't answer your question. I couldn't tell if people think it's an escaping problem with Google snippets, escaping on queries, or exactly what. Probably the best way is to give an exact search that fails. Also it's good to say what browser platform you see the problem on.

As far as doing a search for the single char '&#187;' many of those chars are treated as spaces. If someone can make a compelling case for why people need to search for that character, it might be worth changing.

mvl22

6:29 pm on Dec 19, 2002 (gmt 0)

10+ Year Member



Googleguy,

Glad you've picked up this thread.

To clarify - the issue is when the character appears in the <title> tag of the page, which therefore appears as the main highlighted entry in google's results. Instead of appearing as a >> it appears as?

I'm 99.9% positive it's a problem with Google for several reasons: it was working before, it works in other search engines, and the pages concerned are valid XHTML (the page titles themselves therefore appears correctly in all browsers). Furthermore, occasionally, google's results become correct, as if it was only affecting certain servers in your server farm, but this is only sporadic.

Examples: do a search for
[google.com...]
or for
[google.com...]

These give several examples of this problem. Searching within any of the sites concerned also shows how widespread it it.

I get the same bug when searching using IE6 or NN4.7 or Mozilla (all on Win2000 - I don't have a mac or linux shell to hand).

Feel free to sticky me if you would like more specific info. I'm wary of breaking the rules on revealing site info.

Martin

Fiver

8:27 pm on Dec 19, 2002 (gmt 0)

10+ Year Member



woop. I don't have much to add. Thanks for summing it up mvl22

and thanks for looking into GG.

mvl22

11:39 am on Dec 30, 2002 (gmt 0)

10+ Year Member



I believe this is a general problem with google reading this character: as well as showing up in the 'title' section, it also shows as a double question-mark for text from the page itself.

Sticky me if you want an example.

GoogleGuy, do you agree there is a bug here?

Fiver

7:03 pm on Jan 15, 2003 (gmt 0)

10+ Year Member



2pm EST

Google is reading the characters normally for me atm.

is the problem solved?

showing up right at the bbc.co.uk feed also... looking good!

mvl22

12:27 am on Jan 16, 2003 (gmt 0)

10+ Year Member



Yep, I just noticed this too. It's made my day, since I used the convention of the double >> on about 30 sites, each of which have looked amateur in the search results for two months now.

I'll check periodically to check it's not a server fluctuation reverting back to the double-? as before.

GoogleGuy

6:42 am on Jan 16, 2003 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Keep us honest by re-posting to this thread if you see future problems. :)

Fiver

5:35 pm on Jan 16, 2003 (gmt 0)

10+ Year Member



won't hesitate to GG, thanks ;)