Bugs? Especially international bugs..
Every few weeks I like to call for Google bugs. I’ve got a joint meeting tomorrow with several people to talk about potential bugs, so if you know of any, feel free to mention them in the comments. To remind people of what I’m looking for, here’s the guidelines:
Just to be clear, pruning will be ruthless for this post: I only want to see specific queries that seem to show bugs, and the more concisely you can explain something, the better. I’ll probably keep just the first example of what looks like a bug. I’ve got a meeting at noon tomorrow to talk about search bugs, so I’ll probably lock the comments after that.
Please don’t include stuff like “it’s a bug that you (do something I don’t like, don’t index/rank my site as much as I want).”
Instead, I’m looking for specific bug reports to pass on.
In addition, I’d be curious to hear about any international bugs that you know of. If we don’t handle the top-level domain of Serbia correctly, or if we don’t do a currency conversion correctly, or concrete things like that are welcome..
Related Posts:- Bugs?
I'm in a bug meeting tomorrow, so I thought I'd do a follow-up to this bug post from a while ago. The same rules of... - Re-opening bugs thread
I've got a meeting at noon to talk about potential bugs in search syntax or search results, so I'm re-opening this search bug report thread... - Smaller issues
Okay, you've got a fall "search weather" forecast, and it's mostly quiet. Most of the recent infrastructure changes are out at most/all datacenters. What now?...
Duncan Said,
October 23, 2007 @ 11:58 pm
Maybe not so much a bug, but a deficiency: Google Australia doesn’t have the same top nav bar layout that Google.com does. I’d use my local site more if it had this, it’s handy when switching to blog search etc…
Frank Said,
October 24, 2007 @ 12:07 am
I am not sure if this is relevant but try a search on google for “search engine” what I see is funny because I do not see google anywhere in the first page!!! I guess goofle does not have enough PR
Mikhail Said,
October 24, 2007 @ 12:24 am
On currency convertor:
1 RUR in USD (Russian rouble) gives 24.8805732 for today (10/24/2007).
Central bank of Russia gives us another 2 numbers: 24.8066 for 10/23/2007 and 24.9225 for 10/24/2007.
I found nothing at citibank.com about RUR currency rate.
Central Bank of Russia has a daily XML export for currency rates to Russian currency (I can help with this if you have interest), so that could be more correct to make calculations. Otherwise, the currency conversion is a bit useless for Russian google.ru, probably the same for other local Googles.
Maurice Said,
October 24, 2007 @ 12:36 am
It would be nice to be able in webmaster tools to set one or more target countries ie a .com hosted in the usa but targeting the uk
for one exampel I work with a local hotel and they dont work as well in google.co.uk as they do in .com and not much point people looking for hotels in bedford texas or ma geting my clients site when its Bedford Uk specific.
Hans Nordhaug Said,
October 24, 2007 @ 12:39 am
Not a bug, but a bad feature - searching for “ære” returns “aere” as top results. (It’s slightly better at google.no.) My point is: How likely is it that a English speaker (with a US keyboard) types “ære” in stead of “aere”? That “aere” returns results for “ære” makes more sense since there isn’t “æ” on the US keyboard. (Google is at least consistent - the same problem is present for “åre” and to some degree for “Øre”. Æ, Ø and Å is the extra characters in Norwegian outside US-ASCII.)
Jeremy Said,
October 24, 2007 @ 12:40 am
This may be a feature, but we consider it a bug…
http://www.google.com/search?hl=en&rlz=1B3GGGL_enUS220US220&q=thai+san+francisco+yelp&btnG=Search
Yelp Mobile results show up in the main index even though we’ve blocked Googlebot from crawling them (we’ve tested our robots.txt in the Webmaster console seems fine). Looks like you’re feeding mobile-Googlebot content into the main index which can cause dup content issues and potentially a bad user experience.
Greg Brown Said,
October 24, 2007 @ 12:42 am
Hi Matt,
This isn’t search-related, but I’ll try my luck anyway. I’m trying to log into my blogger account from Taiwan, and it has decided that my being here means I am obviously fluent and is showing me the page in traditional Chinese. Very clever for figuring out my location correctly, but VERY annoying especially given I can’t see a ‘view in English’ link anywhere.
Maybe my signup info just needs to take precedence over my physical location when the content negotiation wizardry takes place?
Greg
Diablos Said,
October 24, 2007 @ 12:51 am
I have recently been having a problem in Webmaster Tools.
When I choose the “Links” section then choose “external links” I am presented with m,y list of pages with their varying amount of links as usual, however, if I try to drill down and see the links for a specific page I get a server busy, try again later message.
Now I would simply believe this is true but I get it 80% of the time when examining my home page. Can we look into this?
-Diablos
mascix Said,
October 24, 2007 @ 12:59 am
I found a bug at google webmaster tools . actually I am not sure is it bug or international thing. but when I try to verify my site with “Upload an html file” it does not work and it always says there is temporarily system problem. but its been more then 6 month I mean come on
I have not verify any of my site with this property.. its not about search but a google bug anyway.
thanks
Ian Mansfield Said,
October 24, 2007 @ 1:09 am
This is for Google Analytics when accessed via Firefox browsers.
It is possible to crash Firefox under a specific situation - I see this happen maybe once a month.
Log into Analytics account
Click on “view reports” for a website account
Click on the date selector “reveal”
Click on yesterdays date
Click on Apply Range
Click on Page Views
Click on Hourly
Click again on the date selector “reveal”
Now, at times if then click on today’s date - the Firefox webbrowser will crash and close down. I then restart using the session restore.
As it is intermittent, it is probably going to be a right pain to track down - sorry!
adrian steele Said,
October 24, 2007 @ 1:12 am
Matt
UK Google mobile (beta) index bug:
Sites designed for mobile devices in XHTML that are in google’s main cache and listed with verified mobile sitemaps in the google webmaster console do NOT appear in the google mobile (beta) index when viewed from a mobile device, but are in the main index. See my URL for an example. I see no way to nominate sites for inclusion in the mobile (beta) index, and see this as a bug.
Howard Said,
October 24, 2007 @ 1:16 am
I’ve seen a small number of pages (from two separate sites) disappear from the Google index/cache. Previously in the Google index, some minor updates were made 1-3 months ago (e.g. changing a title or a few words of alt text for an image), was checking to see when the updates had been indexed. Was able to check date of cached versions until about 10 days ago (Google hadn’t seen the new content yet) - but since then, the pages have vanished.
Searching for the URL, doing a site search or searching for a unique phrase from the pages all give same result - nothing for the pages in question are returned.
No webmaster console warnings either (for the site I mainly work on).
Am hoping this is just a temporary blip, but I’ve been saying that for about 10 days so far…!
Howard
Stan Said,
October 24, 2007 @ 1:29 am
Hey, Matt.
I am quoting a guy who in his personal blog mentioned that google uses a couple of different words for “account” in Bulgarian. It is kind of confusing. I guess the translations come from different sources but using “сметка” for “account” is quite ridiculous. And using TWO different words when referring to the same thing is confusing.
Could you have a look at that?
Thank you.
http://www.optimiced.com/bg/2007/10/23/google-account-vs-google-smetka/
Sam I Am (aka Sam Daams :) ) Said,
October 24, 2007 @ 1:32 am
In most Google products I use the “My Account” link at top right always takes me back to my account which I believe is logical behaviour.
In Analytics, clicking on My Account takes me to a page where I can edit my name and language, I can only guess specific to Analytics not the rest of my G account. I was going to write “I’ve never figured out how to get back to my real My Account page from Analytics” but I just now noticed that that Analytics page has a small text link to take you back to the real My Account page.
Not sure if this is what you refer to as a bug, but it’s certainly a usability bug.
ScotchAL Said,
October 24, 2007 @ 1:32 am
Hi, I’ve got a few bugs I’ve encountered with Google Local Business;
- company names not being displayed properly, eg. McVies WS being displayed as Mcvies Ws (this led to the removal of the listing for the client’s brand’s sake after several attempts at requesting it be fixed).
- wrong URL’s being displayed against Google Local Business listings despite being correct in the actual listing (this has led me to remove the URL as I’d rather it not be displayed at all than be incorrect).
- Listings taking more than 6 weeks to upload.
- GLB click stats showing the same figures for the past x-months despite being “24hrs old”.
Please feel free to get in touch for specifics relating to the above.
Google does a great job as a SE but things like this, when reported and ignored, can really irk.
Cheers, AL.
rene Said,
October 24, 2007 @ 1:32 am
Hello Matt,
the currency conversion misses BAM and the former european currencies that got replaced by the Euro, (DEM, ESP, …)
[100 HRK in EUR] works,
[100 BAM in EUR] does not work
[100 ESP in EUR] does not work
René
Dave Starr --- ROI Guy Said,
October 24, 2007 @ 1:53 am
Bug or Feature, search and service such as Google Alerts should respect case (if enclosed in quotes or otherwise specifically requested). Search for “GPS” (as in Global Positioning System) and you get many results about British GPs (or GP’s) complaining about their work hours. This wastes my resources and yours.
Nothing earth shattering, but respecting case would go a long way toward improving the user experience. And thanks for asking.
Tyler Said,
October 24, 2007 @ 1:57 am
Another Google Analytics usability issue:
In most Google products, clicking the product icon in the upper left hand corner will take you back to the main page for that product. However, when I am using Analytics (ex. log-in and then click Filter Manager), clicking the product icon in the upper left-hand will actually log me out of Analytics and take me back to the log-in page (where I have to resubmit my password).
Rich Said,
October 24, 2007 @ 2:05 am
Matt
International news articles are being served to google.co.uk SERPS - they are totally irrelevant and are taking up a top 10 place in the organic listings.
When you search “pages from the uk” the articles don’t show, but I don’t think they should show regardless, unless it is a uk article.
Example http://www.google.co.uk/search?hl=en&q=mortgages&btnG=Google+Search&meta=
Rich
Errioxa Said,
October 24, 2007 @ 2:05 am
when we search from Google.es “spanish courses in San Sebastian” the first result we can get is a web page with a .dk domain (result of Google Maps) which does not correspond to our main web page but to one of our affiliated agents in Denmark.I think that happens because .dk is a local domain (even though it is not .es) so it comes first than a .com domain.
For instance if we search “company name + san sebastian” it appears .dk domain and .es domain but not the domain .com
is this a bug?
Erez Said,
October 24, 2007 @ 2:09 am
The problem still exist:
http://groups.google.com/group/FFToolbar-group-advanced/browse_thread/thread/dcc792cfd0930739
Dave (The Other One) Said,
October 24, 2007 @ 2:31 am
Google has bugs? Funny, I thought only cockroaches worked for Google??
as Stan Lee used to say….
’nuff said….
killerch0 Said,
October 24, 2007 @ 2:32 am
In Google Webmaster Tools while in Top search queries (with no search location filter) the links are not accurate all are linking to google.com instead of google.whats the top position for.
Dr. J Said,
October 24, 2007 @ 2:40 am
I don’t know if this issue should be posted here, but sometimes my Google Calendar, which should be in Spanish, appears in japanese characters, and it stays that way for a couple of days. Then, suddenly, it all comes back to normal.
This keeps happening now and then.
Dave (original) Said,
October 24, 2007 @ 2:45 am
I don’t eveny your job Matt, you ask for bugs and get all the idiots under the Sun replying with what THEY want.
Johan Said,
October 24, 2007 @ 2:55 am
Google UK: http://www.google.co.uk/search?hl=en&q=something&btnG=Search&meta=cr%3DcountryUK%7CcountryGB
When I scroll through the first ten results and end up at the bottom, I can use the blue-background search box there to refine my search query. Makes sense when you don’t like the first ten results. However, that one defaults to ‘worldwide’ i.e. forgets my CountryGB/UK preference. IMO it should carry on all GET parameters but q.
Google UK:
http://www.google.co.uk/search?hl=en&q=sleep+apnea&btnG=Search&meta=cr%3DcountryUK%7CcountryGB
Google’s language algo’s don’t take into account the official Oxford British spelling of words. “Sleep Apnea” yields suggested refinements but spelling it like “Sleep Apnoea” as is custom here doesn’t.
Google UK (FF/Netscape/IE):
http://www.google.co.uk/search?hl=en&q=zzz-pap&btnG=Search&meta=cr%3DcountryUK%7CcountryGB
Bizarrely, Google says “Results 1 - 10 of about 580,000,000 for zzz-pap. (0.24 seconds)” The results have nothing to do with the query and more results are reported than for “the” on “pages from UK only” Searching world wide for zzz-pap yields just 710 results, quite a contrast with 580 million in the UK! Very strange one and if you were to find the cause, I’d be happy to hear what it was.
Analytics:
When there are more than one website profile in an account, the drop-down at the top which allows you to jump between profiles has stopped working since roughly June this year. Whichever one you select, it always jumps to the top one. I use FF but remember it also happening in other browsers. Real PITA!
Lewis Said,
October 24, 2007 @ 2:58 am
A minor bug I’ve noticed, which can sometimes be frustrating.
Whilst signed in to my google account, the first time I got to http://www.google.com/analytics or …/webmasters or …/adwords etc, I’m always taken to the iGoogle homepage.
toby kesterton Said,
October 24, 2007 @ 3:05 am
In the new My Maps feature there is a directory of public maps. One of them is for the London underground (heres your international slant).
However if you click on a line it starts a tour along the map which you have little control over jumping from station to station. Can we have some proper controls please.
toby kesterton Said,
October 24, 2007 @ 3:15 am
When I open an Gmail email directly from the IGoogle homepage or from Google Desktop there is no option to achive the email once you have read it. You have to go to the inbox to achive it from there.
A little time saver here would be appreciated.
Kloubi Said,
October 24, 2007 @ 3:46 am
Matt,
The bug concerns French Google.
I think that Google is trapped by a webmaster when he puts a keyword in the URL.
For example a webmaster just includes bookmakers’ names in domainnames only to have a very good ranking in Google for these keywords.
http://www.bwin.foot-bookmaker.com
http://www.bwin.paris-sportifs.tv
http://www.betclic.foot-bookmaker.com
http://www.betway.paris-sportifs.tv
http://www.gamebookers.paris-sportifs.tv
http://www.interpari.paris-sportifs.tv
http://www.sportingbet.paris-sportifs.tv
http://www.winamax.paris-sportifs.tv
I’m sure that Google thinks that each website is owned by the company Bwin, Unibet, or Interpari etc etc
So when you write on French Google Unibet or Bwin just see what happen
http://www.google.fr/search?hl=fr&q=unibet&btnG=Recherche+Google&meta=
http://www.unibet.foot-bookmaker is number 3
http://www.google.fr/search?hl=fr&q=Bwin&btnG=Rechercher&meta=
http://www.bwin.foot-bookmaker is number 2 and has a better ranking than Bwin.com !!!!!
I think it could be great for Google to penalize websites which have a brand in their domainname and those whois shows that they are not owned by the company but by a ‘normal’ person.
I am honnest I have a website which is opponent of these websites for same keywords. But I can’t understand why my website is penalized (I don’t cheat, I don’t sell or buy backlinks, I don’t cheat with my domainname) and why such websites have a great ranking in Google.
Robert Lesic Said,
October 24, 2007 @ 3:55 am
Hi,
I don’t know exactly if it’s a Google or an Opera bug. But most times such behaviour is caused by the site owners code - as far as I know.
Browser: Opera 9.24 (identifying as Opera) and older versions, Linux
- Go to the Google start page and click “Language Tools”
- Opera crashes
- Start Opera again and hit the “Continue where Opera crashed” Button
- Click again “Language Tools”
- It works
- Close Opera and start from the beginning
- Same procedure…
It would be nice, if you could fix it.
greetz
Robert
Andy Beard Said,
October 24, 2007 @ 4:00 am
Can we assume that PR drops happening today are not bugs?
Problogger PR4
Copyblogger PR4
SearchEngineJournal PR4
It goes without saying that I feel there is a bug with your manual procedures on paid links. If I was using Google Checkout for payment for consulting services on my own site rather than a reputable escrow service I am sure there wouldn’t have been a problem.
Ramoney Said,
October 24, 2007 @ 4:03 am
Google somehow managed to index the virtual folder “http://domain.tld/defaultsite”
from my website. It’s implemented by my ISP by default, but I didn’t want it to be indexed.
So, I blocked it with my robots.txt (webmastertools analysis: Blocked by line 12: Disallow: “/defaultsite”)
I did a 301 redirect via .htaccess from “http://domain.tld/defaultsite” to “http://domain.tld/”
Additionally, I removed “/defaultsite/” from G’s index (via webmastertools) successfully (it wouldn’t let me remove “/defaultsite” without the trailing slash).
Although it did all this months ago and G crawled, spidered and indexed my website completely, “http://domain.tld/defaultsite” still shows up in the SERPs.
Richard Barber Said,
October 24, 2007 @ 4:13 am
Hi Matt, I haven’t comented for a while, but you’ve given me an opening…
Ads showing up in Google News - especially from yahoo.
I am showing the problem using one specific to the company i work for, but believe it must be a wider problem.
I searched for Martin Lewis in Google News http://news.google.co.uk/news?sourceid=navclient&ie=UTF-8&rlz=1T4GGIC_en-GBGB223GB223&scoring=n&q=%22martin+lewis%22
and found that a large number of news items appearing were not news items but yahoo ads that we have running syndicated on news sites, this is problematic as it makes searching for real news difficult and for real “non techie users” could mean that they regard the pages that they are going to as spam…
Sorry about using my site, but i am sure you can see why this is a problem.
Kind regards,
Richard
Tom Coppeto Said,
October 24, 2007 @ 4:28 am
Here’s one: if I look up specific pages with internal or external links, “Our system is currently busy. Please try again in a few minutes.”
It’s been doing this for the past few days.
Alex Said,
October 24, 2007 @ 4:29 am
Hi Matt, the bug in gmail, when sending letter, googletalk disconnecting for 10 seconds.
jeff Hall Said,
October 24, 2007 @ 4:40 am
Google Images keeps producing random blank pages. Its hard to show An example as it does fluctuate frequently. Currently for example if I google image search ’silverstall’ and then proceed to page 14 click on next and page 15 is blank. The images then continue on page 16. (Last week it happened on page 2-4).
I am pretty sure the problem is not specific to our site as I have seen the same issue for all manner of search terms. e.g. image search ‘matt cutts’ and page 2 disappears.
It might not be a bug as the user might just be catching it between updates however once confronted with a blank page I fear many users mistakenly think that is the end of the images for their search term.
Errioxa Said,
October 24, 2007 @ 5:09 am
In Google SiteMaps I can’t to see my external links.
“Our system is currently busy. Please try again in a few minutes.”
This often happens
Philippe Yergeau Said,
October 24, 2007 @ 5:10 am
The major bug I still have to deal with is in “Google Maps”, which the postal code for the Quebec region is not updated.
Last year (or nearly 2 years now), they were a lot of change in the Quebec rural system, changing street name, postal code, even annex small town to bigger town.
Anyway, it’s really hard to make some map localisation with theses non-updated information.
Thanks,
Alex Said,
October 24, 2007 @ 5:25 am
Not sure if this can be classified as a bug, but it is something that I have noticed a while back, and can cause some problems regarding IDNs ( internation domain names ) in the Greek language, and how the URL is displayed in the Google results.
In Greek, the letter “s” is Sigma “σ - lower case” and “Σ - upper case”.
However, when a word in Greek ends with the letter Sigma “σ” is actually uses another way to display the letter - this is “ς”.
So, for example, the word Skopelos in Greek, is actually spelt as Σκοπελος and not Σκοπελοσ.
I have developed several websites with Greek IDN domains, and when these sites ( when the word ends in the letter Sigma ) appear in Google, the URL listed always uses the “σ” at the end of the correct “ς”.
So, whereas the correct URL should be http://www.σκοπελος. , it is actually showing up on the results page as http://www.σκοπελοσ. which is not correct.
Are there any plans in the pipeline to fix this small matter ???
Again, sorry if this is not the type of “bug” you were asking about, but I do think that it is something that is worth addressing.
Thanks
Alex
Guido Said,
October 24, 2007 @ 5:58 am
Hello Matt,
I wote something, then got slapped in teh face by the spam protection field error, clicked back and my post was gone. I’ll keep it short.
http://www.digitaaltvkijken.nl ranks 1st for “digital television” in Dutch (digitale televisie).
Yahoo link search:
http://search.yahoo.com/search;_ylt=A0geu_U2Ph9HFDMApPxXNyoA?p=linkdomain%3Awww.digitaaltvkijken.nl+-site%3Awww.digitaaltvkijken.nl&y=Search&fr=yfp-t-471&fp_ip=NL&rd=r1&meta=vc%3Dnl
Yes, these people annoy me and I think it’s a bug that algorythms don’t recognize a blatant domain/link spam like this. They were on 93k links 3 weeks ago and today they have 101k.
Already reported to Google NL… hope they do something with it…
Don’t mean to sound cross, just annoyed by these people and the fact that your blog nuked my post.
Keep up the good work!
Kind regards,
Guido
Brent Said,
October 24, 2007 @ 6:17 am
Matt,
Here’s one. When I log in to my Adwords account, then click on the Analytics tab at the top I am correctly sent to a page showing my website profiles - ok good. Then I select one of my sites in the drop down list near the top and that works. But, after this point, if I am done viewing the sites stats and want to look at another - and use the same drop down list to pick another site, it doesn’t work. I am always sent to the first one in the alaphabetically sort list. So I have to click another tab like Reports or My Account then back to the Analytics tab to select the site I want next. I haven’t tried testing this from the Anaytics login, only from Adwords.
-Brent
Philipp Lenssen Said,
October 24, 2007 @ 6:22 am
Hard to tell whose fault this is, but I’m having some sites with e.g. Chinese content and high-ranked search keywords for those are often not showing correctly in Google Analytics, even though I can display Chinese fine in other contexts on this PC (WinXP/ Firefox). Sometimes all I get in Analytics is stuff like “ƴͼ” or “×öãõ”. At other times e.g. Japanese and Chinese get through fine as keywords in Analytics, so it’s not a general issue.
Not an international issue, but there was also another bug recently in Google Calendar… I wanted to change the time for an event. This was an existing ongoing event so I just changed the time for once, and got that usual “do you want to change this for every event in this series” dialog. No, I only wanted to change for this single even. But I tried several times, and it always ignored my new time setting, even though there were no visible error messages being sent by Calendar. I finally gave up and the event stayed the same. Again, not a general issue, at other times it works fine so…
Google Maps as usual often don’t load, or take minutes to load from Germany. Sometimes tiles show “no data” even though at other times the same tiles contain data. Not sure if this is a local problem. (I’m always accessing the international Google Maps though.)
corey Said,
October 24, 2007 @ 6:34 am
I am unable to change sites in Google Analytics. In IE6, changing the dropdown combo box to a different site just takes me to the dashboard of my first website.
To get to a different one, I have to click Analytics Settings and then click the link to the site I want to view. The drop down has been funky for about a month.
Thomas Said,
October 24, 2007 @ 6:55 am
Hi Matt,
I am in Thailand and search for German results with http://www.google.de, for results from USA with http://www.google.de/ncr and for results from Thailand with http://www.google.co.th. But sometimes I become at http://www.google.co.th german results at first, when I search the next shop it’s this result not very helpful or some people here become results in thai with http://www.google.de. And with http://www.google.com/ncr it’s sometimes the first results with german, sometime with thai and sometimes really from USA.
Thanks.
Thomas Soudaz Said,
October 24, 2007 @ 7:00 am
http://www.google.fr/search?hl=fr&q=vin
“vin” french for wine
Show us a nonrelevant map of Wien city (austria) on first result.
Ryan Said,
October 24, 2007 @ 7:00 am
With more people using analytics and it’s “Source=” for tracking.. I’m seeing a lot of URLs show up in Google now with a “?Source=” in them. Granted, it’s a wrong usage by the webmasters… but enough are doing it that I can see it becoming a duplicate content / totally improper tracking issue.
L. Cosio Said,
October 24, 2007 @ 7:12 am
http://www.google.com/search?q=%28.*%3F%29
Try this one… I think is a bug
Gerry Said,
October 24, 2007 @ 7:12 am
gmail won’t download on my Nokia E65 …
Omar Yesid Mariño Said,
October 24, 2007 @ 7:28 am
I live in Colombia now… many times when I type http://www.google.com in browser, it is redirected to http://www.google.com.ar (Argentina) when the correct thing must be http://www.google.com.co (Colombia)
CHRISTIAN Said,
October 24, 2007 @ 7:40 am
my own 2 cent…, the gmap mobile j2me won’t allow to add attachments, maybe is a missing feature, but i consider as a bug
Also, it’s restricted to @google, and it’s not open to GFD accounts, another missing future.
So you ask for a bug, and i list all the missing feature, but hope still help.
netmeg Said,
October 24, 2007 @ 7:44 am
I have recently run into what I consider to be a bug in Google Checkout, which has prevented me from using it on three separate client ecommerce sites. For each site, we’re using a feed from UPS to get the shipping charge. Those feeds get passed to Google Checkout just fine. But no matter what we do (and we had three people spend three days trying to get this to work) it defaults to showing the fastest and most expensive shipping method first (and selected) We don’t want the angry calls from people who forgot to change it back to the standard shipping method, so we just decided not to use Google Checkout. Nothing we could come up with would change the sort order, and it’s not UPS, because it works fine in the shopping cart and with other payment providers.
Stephen Newton Said,
October 24, 2007 @ 7:50 am
If your site is hosted in the USA, you won’t appear in your country’s search results. (Not a problem for Yahoo!, although I don’t know how they figured out I’m in the UK.)
Google appears to use physical server location, which is silly given that the internet is an international medium.
I geo-tag my sites e.g.
It would be great if Google honoured one of these.
Stephen Newton Said,
October 24, 2007 @ 7:55 am
Sorry! Here’s those tags without the brackets:
meta name=”ICBM” content=”53.43491, -2.27434″
META NAME=”geo.position” CONTENT=”53.43491, -2.27434″
META NAME=”geo.placename” CONTENT=”Chorlton-cum-Hardy”
META NAME=”geo.region” CONTENT=”GB-MAN”
If Google honoured something like geo.position you could even hook up with Google Maps to see where the website owner was located.
Multi-Worded Adam Said,
October 24, 2007 @ 8:01 am
I’ve long noticed this as a bug with the Toolbar and now that you were nice enough to give me the chance to report it, I’m gonna take full advantage.
Type the following into the Toolbar search:
W32.proxy.ar (sans quotes)
Then type it with quotes. Ideally, the “With Quotes” behavior should be the same on the Toolbar (or at least give us the ability to toggle this). It works as expected if you use the browser’s default search engine, but not within the Toolbar.
This next one isn’t so much a bug as an “it would be nice”. I’ve noticed that, when you search for addresses within Google Maps, it pinpoints the location in the middle of the street instead of on the side of the street the building is actually on.
For example:
http://maps.google.com/maps?sourceid=navclient&ie=UTF-8&rls=GGLF,GGLF:2006-19,GGLF:en&q=17705+Leslie+Street,+Newmarket&um=1&sa=N&tab=wl
This building is on the east side of the street (right about where the pointer shadow ends). It would be nice if a G map would indicate it.
I think I’m done bitching for now. Thanks, Matt.
Maurice Said,
October 24, 2007 @ 8:26 am
Stephen Newton
Testify brother
though its UK not GB and if anyone from Cisco read s this take note I had a aironet wireless controller insist on GB had great fun when i entered the correct cc code and it trued to set up the ap’s for some country in south America as a result.
Having sad that today after just after I wrote my comment about geolocation I noticed that a .com hosted in Germany is now ranking in .co.uk.
Has there been a geolocation related change to the algorithm ?
Shawn Moniz Said,
October 24, 2007 @ 8:31 am
Only thing I’ve noticed is with Google Webmaster Tools is when you go to “external links” and click on the number of links you have, instead of it going to a listing of backlinks it says the system is busy. Been going on for a few days now.
Alan Perkins Said,
October 24, 2007 @ 8:50 am
You might want to check out what is going on here:
http://www.google.co.uk/search?q=site%3Awww.ibm.com gives 307,000 results.
http://www.google.co.uk/search?q=site%3Awww.ibm.com&meta=cr%3DcountryUK%7CcountryGB gives 3740 results
With most domains, you would expect the second search to give either around 307,000 results or around zero results … not somewhere in between.
I can guess what is happening, but you should make sure…
a) you know what is happening
b) you want it to happen
c) that it is working as you want it to
Harith Said,
October 24, 2007 @ 9:19 am
Andy Beard
All paid links are Spam-Bugs and need to be reported within Google Webmaster Tools
But in your case, I would suggest filing a reconsideration request explaining your case to the friends at Google WebSpam Team. Give it a test drive, you have nothing to lose and everything to win
Shashi Said,
October 24, 2007 @ 9:29 am
Spell check/suggestions in email search in Gmail.
jonny_no_pr Said,
October 24, 2007 @ 9:30 am
Matt-
I found a bug when I got up this morning, just about every website I go to has dropped in PR by 2 or 3 points. What’s up with that? The toolbar is a joke and has been for a long time, when will you guys get rid of it?
Jonny_no_pr
Sam Said,
October 24, 2007 @ 10:02 am
Blogsearch sometimes returns results that include personalization for someone other than the searcher. For example, the link in the results may be displayed as something like “News for John Doe - http://www.etcetcetc.com.” This gives the impression that users’ personal data (search history, maybe?) is being indexed and displayed to the public. The issue has been reported to the Blogsearch team (see support incident #110120570), but with no apparent effect.
David Payne Said,
October 24, 2007 @ 10:12 am
I get this error at least once a day. I am not doing anything sneaky. Just doing site: link: inurl: intitle: types of searches.
Why is that happening? I don’t have any viruses (I had the IT Dept check).
David Payne
Business Development
OneCall
David Payne Said,
October 24, 2007 @ 10:13 am
Hmmm, looks like images aren’t allowed.
Here is the link to the screenshot:
http://www.brentdavidpayne.com/URaVirus.gif
David Payne
Business Development
OneCall
Convent Station Guy Said,
October 24, 2007 @ 11:20 am
Bugs Bugs, Hmmm have not really seen any as of late. Only thing I have seen is a total lack of information on how to use google maps to create a map with multiple points of interest on it. I know it is possible to do as I have seen many but in the limited amount of time I have been able to devote to it I have not found the directions.
And what is the deal with requiring a Google API key for things??
Hugo Guzman Said,
October 24, 2007 @ 12:03 pm
I’ve noticed that, occasionally, Google SERPs will serve up the same exact listing twice. Like for example, the homepage listing will show up on page 1 and then again on page 2.
Interestingly enough, if I change the search parameter from the default (10 results) to 20,50,100, etc.. the duplicate listing disappears. Also, it appears as if this only happens with urls that are also listed within Google Maps.
I posed about this phenomenon at WebmasterWorld and got some confirmation of my suspicion: that there is some sort of glitch with universal search, which causes both the regular indexed version of a url and it’s corresponding Google Maps version to appear simultaneously.
If you have questions or want a specific examples I can provide that as well.
Matt Cutts Said,
October 24, 2007 @ 12:06 pm
Jeremy, I’m in the meeting now, so I’ll ask about this one.
Scented Candles Said,
October 24, 2007 @ 12:10 pm
Gmail does not download on my sprint phone. Any advise please ?
Bob Myers Said,
October 24, 2007 @ 12:51 pm
I consider it a bug that when searching at google.com for an all-kanji term, even if the context is clearly Japanese (e.g. 筑波, the name of a city in Japan), it returns Chinese results.
Sam Sondelila Said,
October 24, 2007 @ 12:57 pm
How about any of the tens of thousands of (actual) rss feeds of pages that have made it into the index and shouldn’t have. I get so sick of clicking on search results and seeing my browser fill with near jibberish (mixed text and tags). So do my Mom, Dad, sisters, brothers, and friends. This should be easy to eliminate since most have /rss/ in the url.
Errioxa Said,
October 24, 2007 @ 1:01 pm
This loss of the PR is for the algorithm or is human?
Good strategy if your intention is to scare the people
Good work
teddie Said,
October 24, 2007 @ 1:24 pm
Matt, you know we’ve been here before. People really want an override for non regionally specific domains. As more come online .mobi .biz .info .travel… .whatever, the requirement to decouple physical location for the local filter will increase.
Jay Westerdal Said,
October 24, 2007 @ 1:32 pm
Bug 1:
When I type in “google.com/adsense” in the address bar I am prompted with a SSL cert warning me “google.com” is not “www.google.com”. Google needs to redirect to “www.google.com/adsense” before turning on the HTTPS.
Suggestion 2:
I can’t spell analtics. So when I go to “google.com/analtics” I get a 404. So instead I preform a search query on Google for “analtics” and the first link is for “www.google.com/analytics” url. It would be great if google created a typo directory that redirect to the proper URL. I know you have a list misspellings for each word…
Tom Said,
October 24, 2007 @ 1:52 pm
signing out of adwords signs me out of igoogle.
Marcin Sochacki Said,
October 24, 2007 @ 2:05 pm
There is a serious bug with national characters decoding in Google Groups Usenet archives. This exists for months, I have reported it via normal support, sent the screenshots and details, but it seems to be ignored.
If you visit the URL
http://groups.google.com/groups/search?q=ironia+ugryz%C5%82a&start=10&sa=N&ie=utf-8&oe=utf-8&
you’ll notice many instances of incorrect decoding of diacritics (Polish in this case). See e.g. “Grali¶my w co¶”, which should in fact be “Graliśmy w coś”. This is probably related to treating the message as ISO-8859-1, while it’s ISO-8859-2 in fact. After clicking the link, the message text is decoded properly. However, looking at the message source here: http://groups.google.com/group/pl.pregierz/msg/38cf664af6e1a468?dmode=source again reveals the same problem with decoding.
Martin Porcheron Said,
October 24, 2007 @ 2:29 pm
Not sure if this comes under your realm, but it is a SERP issue.
The new Google web search help is at http://www.google.com/support/?ctx=web and has been for many years however at the bottom of every web SERP, you have a “Search Tips” link which goes to http://www.google.com/intl/en/help.html, the old Google help centre circa. 2004.
Really, all the old content from the old help centre should be merged into the new one and the link updated (and the old help centre, which is (c) 2004, redirected).
Jonathan Sutherland Said,
October 24, 2007 @ 2:45 pm
In Blogger it is very iretating to have to change the language setting every time I am logged in automatically or manually. It would be really nice if the language setting could also be saved in the cookie.
Aidan Said,
October 24, 2007 @ 2:50 pm
Hi Matt
When navigating to the login page of adwords and already signed in under my primary account I am always redirected at the login page to the page saying:
Welcome to the Google AdWords interface.
We noticed that you are an AdSense publisher. Were you trying to log in to your AdSense account instead?
It would be nice to be able to sign in as a different user, as some people have more than one adwords account. Even a simple link “No Thanks Sign In As A Different User” would be useful.
Cheers
Aidan
Matt Cutts Said,
October 24, 2007 @ 3:31 pm
Hugo Guzman, I’ve noticed that as well recently, and I believe there’s something in the pipeline to take care of that. Thanks for mentioning it though!
Matt Cutts Said,
October 24, 2007 @ 3:34 pm
Sam Sondelila, Google has been getting better at processing RSS feeds recently, and you shouldn’t see them showing up in our search results as often going forward. If you see specific examples, feel free to post them here..
Alexis Kauffmann Said,
October 24, 2007 @ 4:12 pm
It is noticeable how Google Brasil (www.google.com.br) is much more vulnerable to spamming than Google in english. Making tests for some important search terms (which includes my wife’s name!) one can find very well-ranked crappy pages all around. Is there a difference between the spam-detection technologies employed in Google U.S. and other systems? Or are brazilian spammers more witty than american ones? It’s hard to believe the second theory…
Best wishes,
Peter Ortner Said,
October 24, 2007 @ 4:31 pm
Personalised Google Home (Australia) can display a theme at the top, and the image varies with the time of day. I have the tea house theme selected. At present, it’s 9am, but the theme displayed would be more suited to night-time. The image displayed at night time (my time) has the sun in it!
This feature has previously worked properly. I have checked to see that the time zone in “My Account” is set correctly.
lunov Said,
October 24, 2007 @ 4:34 pm
Hi Matt,
Thanks for promoting Serbia… ” If we don’t handle the top-level domain of Serbia correctly..”
Yes. You don’t handle the top-level domain of Serbia correctly. Lunov should be the top search result when googling Serbia. It’s not a Bug, it’s that your servers are outdated for webpages likes ours. Namely Google’s robot (and Yahoo & MSN ) are unable to analyse our content bringing 31 pages in one page. It’s quite not a problem now as Lunov’s configuration is unique, but it could be a Big problem for your system in a near future (something like a streisand effect you know).
Thanks again for promoting Serbia!
Sam Sondelila Said,
October 24, 2007 @ 4:55 pm
Matt, re: the rss feeds, just a quicky I ran across moments ago - wordpress rss feeds are still in the google results. To bring a bunch of examples up quick search for:
inurl:/wordpress.com/feed
M.A. Said,
October 24, 2007 @ 5:51 pm
Don’t know if you were looking for this kind of things, but Google.com redirects me to some other language host when I’m not logged in (or don’t have a cookie set). Even though I have English set as my preferred language in the browser. I have no idea what this switch is based on than doing some half-assed IP scan and guessing on where I’m from [and apparently what language I want to browse in], but it shouldn’t. Accept-Language exists for a reason in the HTTP protocol.
Kris Olivera Said,
October 24, 2007 @ 6:59 pm
Not sure if this is a bug but I have seen it a fews time a on diff searches. When you do a search on “forex review” you get the same site for 1 and 2 sorry I don’t recall the others off the top of my head.
cheers
kris
JLH Said,
October 24, 2007 @ 7:41 pm
Good news on the RSS feeds. I’ve always nofollowed and robots.txt blocked them not for any PageRank funneling exercise but because I hate clicking one in the search results and I didn’t want my visitors to have to suffer that awful back clicking inducing trauma.
ListenArabic Said,
October 24, 2007 @ 8:13 pm
Check SERPS of http://www.listenarabic.com , most pages include Arabic , English an French titles but Google decided to begin the title SERP by … 3 points. Any specific reason for this issue that occurs since beginning of October?
Thanks Matt.
McMohan Said,
October 24, 2007 @ 9:13 pm
I can cite many examples when I try to see the cache of a webpage via Google TB, it shows no cache, while clicking the cache link off the SERPs, it shows.
Andrew Said,
October 25, 2007 @ 12:34 am
International bug - Google is DOUBLE INDEXING some websites. Here is an example.
Google.co.uk
Search term: sheffield jobs
look at result 8 and result 11
http://www.google.co.uk/search?hl=en&q=sheffield+jobs&meta=
http://www.google.co.uk/search?q=sheffield+jobs&hl=en&pwst=1&start=10&sa=N
If this is a bug it needs fixing as a double indexed seearch engine is not a good thing! cheers
MattUK Said,
October 25, 2007 @ 2:48 am
Neither are bugs, though I feel that they’re worth a mention.
Firstly - UK spellings, on Google.co.uk when you serach for a query using a UK version it’ll correct you with the US spelling.
OptimiSation - Did you mean: optimization? NO I’m just English! :p
I don’t notice it with colour/color though?
Secondly, in webmaster tools, it would be really good to specify which local version of Google a site should appear in. You guys seem to be getting better recently at deciding on the correct local serach a site should be in, but it would be great to have a manual override that we could use.
Gustave Stresen-Reuter Said,
October 25, 2007 @ 2:58 am
Matt,
I’m not sure if this is a bug, missing functionality, or just wishful thinking on my part, but here’s the question (and I’ll be blogging about this too): In 2002 I started tedmasterweb.com. It was my professional face on the web. In 2004 I became a partner in a small IT firm in the Canary Islands. Some time around the beginning of 2007 I migrated tedmasterweb.com to a subdirectory of clevernet.biz (clevernet.biz/tedmasterweb/) to see if I could improve the page rank of clevernet.biz since that is now my primary interest (and because I prefer to manage just one web site, not two).
I employed a 301 redirect, as recommended by you and others as best I could, but I did not ask the people linking to tedmasterweb.com to update their links. Now, tedmasterweb.com has a page rank of 0 and there did not appear to be any noticeable increase in the page rank for clevernet.biz.
I realize this isn’t exactly a search-related issue and I’m probably sounding like a whiner (but hey, it’s cool, it was an experiment, after all), but thought you might be able to shed some light on why the page rank for tedmasterweb.com did not follow the move.
Thanks in advance,
Gustave S-R
Trebor Said,
October 25, 2007 @ 3:21 am
The ‘did you mean’ mispelling function on Google search does not work correctly on Google.co.uk - It is american focussed. For example, in the UK we spell optimisation with an ’s’ but if you type this into google.co.uk i get the message ‘did you mean optimization?’. This happens with many other keywords.
Richard Hearne (RedCardinal) Said,
October 25, 2007 @ 6:12 am
Is there any official change on geo-location criteria for sites?
Some reports of people hosting in US with .com TLD and appearing in country level searches e.g. ‘pages from UK’ and ‘pages from Ireland’.
cyberjack Said,
October 25, 2007 @ 6:35 am
Traveling in Mexico with only a Dell Axim x51v for wi-fi access.
Google mobile assumes Spanish based on my ip address and provides no way to change language to English on mobile interface - same with Google mobilizer..
Errioxa Said,
October 25, 2007 @ 9:02 am
And this?
http://google.com/views?hl=en&q=+linux+view%3Atimeline&btnG=Google+Search
Linux in 1911, and the first person to have set foot on the Moon was in 1492.
Aaron Xavier Said,
October 25, 2007 @ 10:27 am
Please make it easier for sites to target specific pages to specific countries. My client is a large company that does business worldwide. It is ridiculous that they would need to acquire all of the foreign domains and host their site in multiple locations to be considered relevant for that region. It would be nice to have a meta tag that could denote where each page is targeted so that the UK pages aren’t showing up in the US SERPS and vice versa.
The biggest issue here is that there are a lot of different English, Spanish, French……..speaking countries however; my client’s pricing and sales offices are different for each country. This is especially problematic with brand terms.
This would be a fantastic fix and certainly make the search experience better for international searchers.
Thanks for taking the time to address this.
Nathan Haas Said,
October 25, 2007 @ 11:43 am
There’s a strange mapping bug: go to maps.google.com, and search for vancouver, washington.
in the address field, type in 8100 ne parkway drive, vancouver, wa
hit enter.
it comes up with a result,
Address: 8100 NE Vancouver Plaza Dr
Vancouver, WA 98662
Make this my default location
which is due south of the street it’s supposed to show. the information box that pops up is usually covering the street it’s supposed to show.
james Said,
October 25, 2007 @ 12:24 pm
link:www.example.com vs the inbound links shown from webmaster tools. Is the disparity on purpose or is it a bug?
Subikar Said,
October 25, 2007 @ 1:47 pm
This is not the bug I am discussing as I got better platform to speak some thing about google. Gmail is giving enough space to store mail but if there is an option to take backup of all the mails time to time just clicking one button is easier because some day some how my mail id crashed or may be due to any reason banned my all important data get lost. So getting good backup system of mail will be one of the best option I hope.
Matt Cutts Said,
October 25, 2007 @ 2:16 pm
Subikar, I agree that a one-click backup would be really nice. You can use the POP download feature to back up your email, but it’s not a brain-dead easy kind of thing. I like that the new Apple OS will include really easy backup.
Errioxa, What we’re looking for is years mentioned in the documents. That’s always going to be a little noisy and is to be expected. I take your point that it would be nice if the proximity of the “year mention” to the user’s word in the document would be taken into account.
Hey all, just to let you know, another Googler offered to prowl through the comments and try to submit obvious bugs to the right teams. I really appreciate everyone that has commented with bugs.
Simon Waters Said,
October 25, 2007 @ 3:14 pm
Probably a well known one but I did this search earlier this week….
Google search for “-2147467259″ (presumably other negative numbers as well, but I didn’t check them all ;-).
Clearly it should either return all results for “-2147467259″ or results not containing “2147467259″ (of which I suspect the Internet has many billions).
Yes it is 0×080004005 (the COM object didn’t report its error because the programmer couldn’t be bothered) in disguise as a decimal.
But the result is you get a handful of answers, searching for “2147467259″ gets you more results for “-2147467259″ so it isn’t a good search for “-2147467259″, but all the results you do get have “2147467259″ in, so it isn’t excluding them either.
Okay maybe I could have phrased the search a bit better (”214767259 site:microsoft.com” for example), but it wasn’t the magic “do what I mean” that Google usually achieves.
Do I get pagerank/job/meal at Google office or something for bug reports, or just a better search engine?
Subikar Said,
October 25, 2007 @ 3:58 pm
Thanks for the update Matt. But the thing is that I am developer so I know different way to do the backup. But the main concern to those who are the layman and have little knowledge on computer but running a big business all the important mails are maintain there in google mails. This will be helpful much for them that I think.
And again thanks for answering the quries.
Subikar
earlpearl Said,
October 25, 2007 @ 5:12 pm
Matt: This has nothing to do with bugs but have you seen the latest toon pic….vis a vis PR reductions.
http://www.seorefugee.com/seoblog/2007/10/24/lorena-cutts-bobbitt
Its something else to be “the face of google”
Lunov Said,
October 25, 2007 @ 6:53 pm
Quote: lunov Said,
October 24, 2007 @ 4:34 pm
Hi Matt,
Thanks for promoting Serbia… ” If we don’t handle the top-level domain of Serbia correctly..”
Yes. You don’t handle the top-level domain of Serbia correctly. Lunov should be the top search result when googling Serbia. It’s not a Bug, it’s that your servers are outdated for webpages likes ours. Namely Google’s robot (and Yahoo & MSN ) are unable to analyse our content bringing 31 pages in one page. It’s quite not a problem now as Lunov’s configuration is unique, but it could be a Big problem for your system in a near future (something like a streisand effect you know).
Thanks again for promoting Serbia! - End Of Quote
I apolozige for the (mis)leading link.Here it is http://www.lunov.com/Serbia-Macedonia-Montenegro.htm
Please, take a look.
Thanks.
Jack from the Netherlands Said,
October 25, 2007 @ 6:57 pm
When translating a website using Google Translate, where -on the original website- plain text with several paragraphs is placed between “PRE”-tags, all whitelines between paragraphs are removed in the resulting page.
Google Translate changes the formatting of the original text. The result is a pretty unreadable mess of very long lines of text and a huge horizontal scrollbar.
Pls. contact me for more details if desired.
Tristan Said,
October 25, 2007 @ 6:58 pm
There’s a big problem with googlebot bandwidth usage ( in france ).
GGbot is crawling very fast since 4 or 5 days, making as a “slashdot effect”.
French hosting service “OVH” is almost out of order, are there any problems with ggbot settings ?
Excuse my poor english, I’m french
Leonard Bartholomew Said,
October 25, 2007 @ 7:32 pm
Bugs? I would say so. All those organic creatures running around the Googleplex pretending to be experts in their field and getting paid for it. Especially when they make the kinds of decisions they did this week with Google’s most recent slapdown on paid links when everyone knows that Google continues to maintain paid links on it’s site..
We don’t tell you how to run your business, do we? Maybe we do and you just don’t listen.
Bad form Google!
dockarl Said,
October 25, 2007 @ 8:39 pm
This ‘bug’ relates to Google Australia (assuming is the same for other country specific googles).
Whenever you do a search, EVEN if you don’t specify ‘Pages from Australia’ a disproportionate number of pages returned are from Australia. This occurs across a broad range of searches.
The geotargetting is ’subtler’ than that which you get when you specifically ask for geotargetted results, but it is definitely there.
I don’t want to see geotargetted results if I haven’t asked for them - I don’t want to have to keep using an overseas proxy to see what the rest of the world sees.
doc
Jim Said,
October 25, 2007 @ 10:25 pm
There seems to be a Google PageRank bug, my blog has reduced to PR3.
To eat and drink I had to sell a few links, now I will have to sell my body!
Marty Said,
October 25, 2007 @ 11:08 pm
Click the Next link at the bottom of the page. It should take you to page 11, but instead loads page 1. Eeenteresting.
Dave (original) Said,
October 25, 2007 @ 11:19 pm
doc, why not use Google.com and not Google.com.au ?
Doubt this is a bug, rather a useful feature.
Marco Said,
October 26, 2007 @ 2:21 am
I’m seeing a bug on “bleu widget” a domain (index) is on every serp page on place 10.
So on start=20 its on 29, start=30 its on 39 etc. It goes on to start=170
Nacho Said,
October 26, 2007 @ 2:29 am
I don’t know if it can be call as a bug, but..,
If you go to the spanish version of Google Zeitgeist you only get until November 2005 list.
http://www.google.es/press/zeitgeist.html
You have to go to the International Site of Google Zeitgeist to get an updated list of Top Gaining Queries in Spain.
http://www.google.com/intl/en/press/intl-zeitgeist.html#es
Halfdeck Said,
October 26, 2007 @ 2:32 am
Matt, you or someone else over at Google needs to make a statement regarding the recent TBPR drops. (Yes its off-topic, but a few other people already brought this up). I personally think its bad PR (public relations). If you want to send a strong message, you need to penalize link sellers’ rankings and traffic, not just what they see in the toolbar. Otherwise people won’t take you seriously.
Johan Said,
October 26, 2007 @ 5:11 am
Seemingly related to my earlier comment:
http://www.google.co.uk/search?hl=en&q=http&btnG=Search&meta=cr%3DcountryUK%7CcountryGB
(Almost) All of the 141 million results are only listed because of this sentence in your cache header:
“These terms only appear in links pointing to this page: http”
Surely, Google can remove http and www from the links for the sake of indexing and ranking. That way, real pages about the HTTP protocol can be ranked first page in Google UK, not millions of pages which happen to have http ina link pointing to them.
Nick Shepherd Said,
October 26, 2007 @ 6:19 am
Not sure if you are still checking the comments for this post, but I have an interesting bug that I’ve been trying to work out for several weeks now and no one can seem to help me with figuring out just exactly why I’m seeing these results.
The query in question is:
“info:www.thedevlounge.com”
That shows the results for another domain (www.casadelewis.com) on a completely different server and completely different IP address than my domain. What’s more is using the webmaster tools, a lot of the keyword phrases it says I’m ranked for, I can’t find myself anywhere on that page for those keywords (is this normal?), and I’m wondering if this has anything to do with it. Also if you try:
“cache:www.thedevlounge.com”
It shows the page for casadelewis.com and has it as recently spidered, but there is no way that that page could have been spidered in association with my site. I have a few ideas of why it’s doing this, but have no clue how to resolve it. I’ve sent in emails to google on multiple occasions with no response, so I’m attempting to see if you can give me any insight on this problem.
Thanks tons in advance,
Nick Shepherd
dockarl Said,
October 26, 2007 @ 7:08 am
Dave (Original) google.com redirects to google.com.au from Oz, so you’ve got no choice.
re: useful feature - absolutely agree - in which case why have the ’show local results only’ feature when it’s going to be geotargetted to an extent regardless.
It isn’t really a complaint or a bug, rather just something that seems quite wacky.
doc
rené Said,
October 26, 2007 @ 7:09 am
google reader search and non-ascii-letters:
[Ivanić] finds articles that include “Ivanić”
[Ivanic] does not return articles that include “Ivanić”
whereas web search would return websites that contain “Ivanić” using either search (sometimes the result order differs, sometimes not)
but there is no easy way to enter those specialletters with a american keybord in google reader…
René
Bob Said,
October 26, 2007 @ 9:13 am
I’d like Google to address higher SERP’s by registering 50+ domains with different addresses, IP’s, etc then cross linking the “affiliated” sites with the ecommerce sites to get high rankings. One search result has 5 of the top 15 spots (including 2 of the top 5) using this tactic.
obble wobble Said,
October 26, 2007 @ 9:21 am
Hi Matt!
Two bugs:
1. The following characters:
% and $ and !
return no results when searched alone or with keywords.
2. No feedback is presented when no results are returned. (No explanation of what made that query invalid, etc.)
Examples:
“&!” returns results with names that include an ampersand. The expected result would include “Yotsuba&!” a book title including that string.*
“$BIGNUM” returns no results, even in google groups, when I would expect at least one result returned for alt.sysadmin.recovery.
“DNA$” returns results about DNA. The expected result might be equivalent to “dnas”, Sony’s playstation security protocol.
Suggested fixes:
List of forbidden characters in the advanced search help.
List escape sequences to use, and/or provision of escape sequences.
Provide feedback when no results are returned.
*a very cute, innocent manga. Recommended reading..
David Sickmiller Said,
October 26, 2007 @ 10:29 am
I’ve noticed a bug with the Google Toolbar for Firefox. When I enter a keyword and click “Search Site,” it works right the first time. However, after that, the toolbar erroneously prepends “site:” to every search I do until I restart Firefox.
Frederick Gimino Said,
October 26, 2007 @ 10:36 am
Matt:
I am not sure if this is a bug but the results for Google UK appear to only list UK sites when selecting search the web. I thought that searching sites from the UK should produce these kind of results but it seems that is not the case. Almost identical results are obtained when the search is performed for the entire web or just UK sites. Is this releveant? Why have 2 search options with identical results? If the design is to allow UK sites exclusive preference when searching the web why bother having the UK exclusive radio button?
Richard Hearne (RedCardinal) Said,
October 26, 2007 @ 10:37 am
http://www.google.ie/search?hl=en&q=free+games&btnG=Google+Search&meta=cr%3DcountryIE
Country search for ‘free games’ returns a .com site hosted in the US for #1 & #2. Where’s the relevancy there?
Not like there aren’t enough results on .ie domains alone:
http://www.google.ie/search?hl=en&q=free+games+inurl%3A.ie&btnG=Search&meta=cr%3DcountryIE
Never mind sites physically hosted in Irish IP space.
Alex Said,
October 26, 2007 @ 10:57 am
Hey Matt,
this is not about bug but I would like to thank you for mentioning Serbia. It may help Serbia to get more visibility
Keep up doing good work,
Alex
mick Said,
October 26, 2007 @ 11:20 am
Would say an american site with american ip returning first for a specific item in a country filtered results page{ say filtered for ireland} be regarded as an international bug , or just as relevant to the query ?
Kind regards Mick
G Said,
October 26, 2007 @ 2:38 pm
We need more data from the “Allowed Urls” feature.
How many hits have been on the link (1 or 1000)
Unless you know then how can you decide its a ’safe’ link, or an attack on your adsense?
Automatically include all google data centers and regional search engines.
Mary Said,
October 26, 2007 @ 4:46 pm
Yes, the site I maintain (www.legionofmarytidewater.com) is listed as a PR 0. That is down from a PR 4. I don’t have any paid links on or to the site. This must be some sort of bug?
Dave (original) Said,
October 26, 2007 @ 6:40 pm
I can confirm that is NOT true.
michell Said,
October 26, 2007 @ 7:02 pm
Sites appearing in both google maps and google search results.
Effectively, gives that site 2, or even 3 positions in the top 10.
For that matter, the fact that google map results in search don’t ever change order, and are not based on worth.
IN australia, it is especially bad.
michell Said,
October 26, 2007 @ 7:04 pm
In Google tools, there is nothing for the many site webmaster.
eg.
a. Search for a domain…. need to scan through them.
b. Verify ALL sites.
c. Load sitemap location for ALL sites.
Peter Ortner Said,
October 27, 2007 @ 1:47 am
I can confirm it’s definitely true, and by design. If you don’t want the country specific redirect, access http://www.google.com/ncr
If you’re in Australia and not being redirected then I suppose it’s a legitimate issue. I’d guess the problem is caused by outdated geographic data.
dockarl Said,
October 27, 2007 @ 2:57 am
Thanks Peter - yup - Occasionally you’ll happen across an IP that doesn’t automatically redirect (especially if you’re with Optus who recently started using a big slab of previously unallocated IP addresses) but in the majority of cases you’ll be redirected.
I notice Frederick Gimino (above) raises the same issue with Google UK (ie getting geotargetted results whether you ask for it or not) - so whilst it is probably totally by design it is still a little unusual.
Thanks Peter for the link to http://www.google.com/ncr - that’s useful.
Cheers,
Matt
Cindy Turrietta Said,
October 27, 2007 @ 5:15 am
Hi Matt: Does the “supplemental” query we talked about at SES SJ count? I recall you saying it was a “bug”… ; )
teddie Said,
October 27, 2007 @ 6:07 am
Matt. Chatting to my team something that definitely appeared to be a bug was that one of them was searching on Google and got a Squid Proxy http://www.squid-cache.org/ error come back in the results page? We don’t use Squid and doubt British Telecom do, so figure it must have been coming from Google infrastructure? Perhaps something worth checking at your end.
Guillaume Voyer Said,
October 27, 2007 @ 6:22 am
Hi Matt,
The currency conversion in french has a little bug. For example : http://www.google.ca/search?hl=fr&q=50+dollars+canadiens+en+dollars+am%C3%A9ricains&meta= won’t return the result of the conversion. If I use http://www.google.ca/search?hl=fr&q=50+dollars+canadien+en+dollars+am%C3%A9ricain&meta= instead, it will work. So, the first query do not work because of the “s” at the end of canadiens and américains. In french, adjective related to a noun that is plural (in this case dollars) will take an “s” at the end. It would be great if your search engine could understand the right syntax.
Best regards,
Guillaume Voyer.
Tina Said,
October 27, 2007 @ 7:52 am
All my sites have been hit heavily on the toolbar PR droping from 5 to 3 or 4 to 3 a couple from 3 to 1 and I don’t sell links. I have advertisers, you click a link, make a purchase and I earn some cash but then isn’t this what its all about?
I work hard on the Internet, several hours a day trying to make a living and if this tb decrease is leading to a drop in SEO then I am going to have serious financial losses.
I can accept gentle and slight changes, a fall of a few places or a rise but what I find hard to accept are massive changes that punish the innocent. Everyone at Google who are responsible for the big decisions need to realise that they are playing with peoples livliehoods. Can we have some sort of announcement, do those that have big losses on TB PR need to brace themselves for major search term losses?
John Lockwood Said,
October 27, 2007 @ 11:17 am
Matt,
On the webmaster tools request reconsideration page, it strikes me that the page should have a section for those who have NOT violated the quality guidelines in the past, but who believe Google THINKS they have. I have never sold a text link ad in my life, yet I have reason to believe (cf http://www.problogger.net/archives/2007/10/27/the-google-page-rank-pendulum-swings-again/) that I’ve just been penalized for selling them.
Maybe this is way out of line and related to the whole business of feeling like I’m not getting the rank I deserve, but it seems to me your reconsideration page shouldn’t force a confession before letting one back in church and should admit the possibility that the church may have erred.
Multi-Worded Adam Said,
October 27, 2007 @ 3:11 pm
You can do that all the way from Alaska, Dave?
Steven Said,
October 27, 2007 @ 5:17 pm
In Google webmaster tools, it says that my index page was last crawled Jan 29th. That’s not anywhere correct. Googlebot crawls it almost everyday. No errors show up in Google Webmaster Tools and no other search engines have problems crawling the page.
The site has been around for more than 2 years but was suddenly dropped from Google’s index a few weeks ago. I strongly suspect the two things are related.
Isaac Sunyer Said,
October 28, 2007 @ 1:40 am
Embed maps are really slow on Firefox,and not on internet explorer.At east, in Spain.
There is another problem with Sitemaps. I have a redirection to any url’s, for example: http://www.mapfactory.net/red/map_factory_redir.php?i=1&id=230 is redirecting to: http://www.casas.nl/ I’m the owner of mapfactory and my sitemaps webmasters tool is showing me all http://www.casas.nl/ For me, is fantastic, but I guess it’sa bug.
Isaac
Neva Howell Said,
October 28, 2007 @ 6:16 am
Matt:
In my adsense account, under diagnostics, I see several non-existent urls that appear to have been blocked on my domain. Problem is, they are all from China, (.cn url extensions) and they do not exist.
I’ve written google about this but gotten no response.
Here are the non-existent urls that is listed in my account. There presence signaled a huge drop in adsense impressions at the same time so it’s not a coincidence:
http:/ / gas-furnase. cn/ feed/
http:/ / 2flash-games. cn/
http:/ / valeentines. cn/ feed/
http:/ / schoolshooting. cn/ feed/
Paolo Groppo Said,
October 28, 2007 @ 3:14 pm
Analytics > Map Overlay > Detail Level: City
The city of Rome (Italy) does not exist! A city called Augusta (???) is displayed instead of Rome.
Dave Said,
October 28, 2007 @ 3:33 pm
Is this a bug or a way to cheat the system using multiple IP’s?
http://www.invoicedealers.com shows up when searching “Only from Australia” when they are clearly US based and not relevant to people here. This is the only example I’ve seen of this…
Thanks
Dave
elvirs Said,
October 28, 2007 @ 3:55 pm
my website has a link from http://www.domain.GOV.tr (turkey), got it few months ago and still has it, its not nofollow but Goole webmaster tools does not show it.
jim Said,
October 28, 2007 @ 6:10 pm
Not sure if this is a “bug” but it’s a problem.
Doing a search on Google Brasil. http://www.google.com.br/
Do a search for “cd usado comprar sao paulo” Basically Used Cd’s in Sao Paulo.
The first several pages of results are dominated by viva street. They are using a combination of sub-domains and directories in an effort to dominate the results. Isn’t there a better way? Is this correct? It doesn’t seem like this use of both sub domains and directory structure helps the user.
dockarl Said,
October 28, 2007 @ 7:03 pm
b2b PR is updated and changed on a constant basis.
Just because it has appeared in the form of a little green bar increase recently doesn’t mean it’s only just come into existence. If folks with proxies are now 301′ing little green bars across to their mates then they’re bloody idiots - that would raise red flags acres in size.
Reverse the PR update? dude - again - it’s not an update, it’s an export. Use webmaster tools to report them as spammers and be done with it - this thread is about bugs not spam.
dockarl Said,
October 28, 2007 @ 8:10 pm
hmm.. ok dokey.
PR is mostly to do with the links coming to your site. If the proxies are getting more incoming link action then it’s quite possible they’re also involved in some form of link trading designed to manipulate the system - again, something a spam report will help.
My hijacking? champ - I’m a blogger that owns 3 websites, not the owner of an evil black hat empire. Chill out.
In my jurisdiction that would be considered a threat and your hunting down days will probably be short if you make threats like that.
umm.. yep. I reiterate. This isn’t a bug - this is a case of spam, and it doesn’t automatically become worse overnight when Google has a PR export. Report the spam, if necessary also file a DMCA report and take a deep breath.
Multi-Worded Adam Said,
October 28, 2007 @ 10:12 pm
dockarl: did I miss something? Who exactly are you respondi