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 thumb apply:
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.
{ 122 comments… read them below or add one }
I reported this the first time at least 4 years ago. It applies to every search with “preference” of “new browser window” enabled. For example:
http://www.google.com/search?q=whatever
The resulting links have a target attribute value of “nw”, which is a single named window, NOT a new window (“_blank”).
The difference is HUGE, especially when you have multiple searches going on (or started working with a search hours ago in a minimized browser window and later need to do another). The results you expect to appear in a “new window” could appear in any of the 50 browser windows you have open. A horrible pain.
I haven’t used stored settings on Google in years because of this one setting.
Matt
Within Sitemaps, after finishing submitting a URL removal request and when you click on sign out, you get an error alert, like this one:
https://www.google.com:80/webmasters/tools
Maybe its just me. I wish to hear from other friends whether the have the same experience.
I have reported this one several times, including in the previous bug thread. It’s driving me mad
Google Calculator does not do temperature maths correctly (perhaps due to working internally in Kelvin and converting incorrectly?)
1 degree Fahrenheit added to 1 degree Fahrenheit should be 2 degrees Fahrenheit (or -16 Celcius). It should not be +239 celcius.
http://www.google.co.uk/search?q=1+f+%2B+1+f
Hi Matt.
Not an SEO bug, but a Google bug none-the-less. When you are on the secure version of the personal homepage:
https://www.google.com/ig
If you try a query into the search box and press enter or click ‘Google Search’ instead of take you to the SERP it takes you to the Google homepage:
http://www.google.com/
This is very annoying as I always forget, type in some long search and then have to type it all again.
Kind regards,
Andrew.
Analytics, new interface: Clicking on a traffic source from your dashboard gives a 404
Within Analytics the referring link detail page says ” Visit this referring link href=”javascript: _openHelp(‘answer.py?answer=67752′);”>)”
It still works though.
Also maybe you could explain this query so that people know what they shouldn’t be doing.
http://www.google.com/search?hl=en&q=john+chow
For Google Custom Search, I noticed that the results are not the same (as comprehensive) as when I search Google directly. I reported it to Google a while ago with no response.
I blogged about it recently for more details here
I’m not looking for hits through your site, feel free to trim the URL part if you like but do take a look at it.
@ Harith, I get that too but attributed it to my FireFox NoScript add-on.
My bug report is to do with Google Sitemaps/Webmaster Tools too.
On my dashboard I have 5 domains, all from the same top level domain.
http://blog.mydomain.co.uk/
http://faq.mydomain.co.uk/
http://forums.mydomain.co.uk/
http://videos.mydomain.co.uk/
http://www.mydomain.co.uk/
htaccess redirects http://mydomain.co.uk/ to http://www.mydomain.co.uk/
Here’s the bug:
In Webmaster Tools I click “Manage http://www.mydomain.co.uk/” then the [Statistics] tab. At the bottom of “Crawl Stats” under “Your page with the highest PageRank” it now lists:
Jun http://blog.mydomain.co.uk/
May http://www.mydomain.co.uk/
Apr http://www.mydomain.co.uk/
IMO the first one, blog.mydomain should not be listed under the Crawl Stats of http://www.mydomain considering it has nothing to do with it in principle
In a rather inconsistent fashion, when I check the Highest PageRank/Crawl Stats of the blog.mydomain.co.uk domain it says:
“Your page with the highest PageRank
We do not have data for this site.”
On Custom Search again. If I search for some extact text that shows on at least one of my pages Custom Search returns nothing. Yet, the same search on Google public shows the page as number 1.
Also, often when the term is something like: Big blue widgets, I get only a few results in Custom Search yet I know my site has 100′s of matches.
Since the fist week of June, we (a few fellow SEOs and I) have observed differences in results between Google.fr and all the other extensions.
These differences in ranking occur even when using the exact same query and changing only the extension from “.fr” to “.com”.
The same thing has been observed for .co.uk results.
Not sure this can be qualified as a bug, but it seems to be a change in the handling of country specific queries.
> much of the Sites reside und this
> barrier.
> Without PR higher than 4 nearly no natural
> Backlinks possible,
> without new natural Backlinks no PR4 for
> long times!
Karl: by your theory, one could never make a new site achieve some high PR. But that assumption’s proven to be false — new sites do get PR over time. Maybe they don’t get too many backlinks from people entering stuff into Google to find them, at first, but you can start out by letting people in your niche know about your site, and some of them may link to you. Not to say that Google shouldn’t give new sites a stronger start (it might make for more variety), but to call this a bug… hmm, seems off-topic.
We are seeing a geo-location issue (or if not sure what the problem is).
http://www.sofa.com is based in London – all geolocating stuff points to them being in the UK (IP address, Google local stuff, whois address on domain name).
They rank #1 for ‘sofa’ on google.com:
http://www.google.com/search?q=sofa
They are nowhere to be seen for the same search on google.co.uk:
http://www.google.co.uk/search?q=sofa
In our professional opinion, they should rank in the .co.uk search results for this and a lot of other things, but Google appear to think they aren’t UK-based (or something else weird is going on).
I have discussed this with Rand and the guys and seomoz and they had some useful suggestions (some of which are implemented, some are coming soon) but over the past months they have gathered new links (including a number from influential UK sites) and have gradually dropped down the UK rankings (while doing better in google.com). Some of the pages that link to them rank above them and really shouldn’t (in my opinion).
This sounds a lot like a “you’re not ranking my site” whine, but it really isn’t intended to be – we have spent a lot of time investigating this issue and are truly stumped and it seems to me that there might be a problem with the way Google determines where they are based (the domain name used to be owned by a US company).
[Disclosure: we are working with sofa.com, but we're very white hat and you'd approve of everything we've done to help them, everything we try has helped in google.com but not in google.co.uk - we are working with them on a new website that will fix many remaining problems (duplicate www / non-www, getting distinct titles on different pages, and a whole bunch of other things) - having said that, what we've seen is weird]
Yep – I brought this one up before with Susan M, but just in case she’s forgotten – in webmaster tools, if you use the robots.txt validator and forget to prepend your test url with http:// it gives a weird wacked out error message (sorry, I’ve forgotten the exact details) that leads you to believe that your robots.txt is faulty – when it’s not.
doc
Advanced Search page bug with multiple phrase queries
1. Search for: “ac motor” “servo motor” “electrical motor” (or any other query with multiple phrases).
2. Click “Advanced Search”. Notice that the query has been translated into something different: (all the words: servo-motor electrical-motor, and exact phrase: ac motor).
3. Click search. The results are different than the original query, although the differences may be minor.
Original posted at:
http://www.searchenginecaffe.com/2007/04/something-you-dont-find-every-day-bug.html
If you search for [chester nj], you get a map of Chester PA. I just noticed this yesterday, but not sure when it started.
Not sure if this is a bug, but, I have noticed that on some search queries where the first result is SiteLink, the second result is the same website. This seems like a bug to me….If the first result is a SiteLink, then it should be followed by a second website, not the same one.
Here is an example of a search query where I noticed this result:
http://www.google.com/search?hl=en&q=luxury+real+estate
In the teeny-tiny bug section: list items shown in snippets are seperated by a bullet in regular results (item1 · item2 · item3) and a semicolon in supplemental results (item1; item2; item3). I’m sure you’ve got bigger fish to fry, but it seems like an easy one to track down and fix.
In webmastertools, for a site you haven’t yet indexed, every time you log in to WMT, it says that the site was last crawled at (the exact time you logged into webmaster tools), when in fact it has never visitited except to out check robots.txt when asked to do so.
I am not near a computer with a WMT account to verify the exact behaviour, but thinking about it, I think I might have already mentioned it here, back in perhaps March or April, anyway.
It is a bug that you let companies like the one featured in this Toronto Star article (referred to in this SEO blog) get away with what they are doing for so long..
http://www.seomoz.org/blog/michael-yack-old-school-link-farmer-makes-it-big-in-the-toronto-star
http://www.google.com/search?&q=php
Do a search for PHP, PERL or AJAX
(look at the blue bar on top before the organics)
now do one for HTML, DHTML, JAVA, JAVASCRIPT, PYTHON, DHTML, C++ .Net, etc
Notice the difference and inconsistancies?
SearchEnginesWeb wants credit for revealing this, and will share other bugs!
A search for “Art Tv” did not return the canadian tv station “Artv”
The station’s brand name is typo-ish and I’m guessing it blows up Google’s regex
I noticed this on my friends computer that when he runs searches that have over 10 results on one page (he does 30), they are different then just running ten. The SERPs shows pages that weren’t in the top 30 when there are 10 results per page. Well that’s the only bug I am going to report on.
(I’m sorry this bug isn’t very specific) a site: specific search of a blogger (ftp published) blog returns far fewer (and less relevant) results than technorati’s blog search. The google search returns only very recent blog posts instead of returning all relevant blog posts whenever they may have been published.
thanks, I mentioned this once before but it was on an old post that you might not have looked at again.
Hi Matt,
Insofar as googlebombs are considered bugs, I noticed that they are alive and well in certain parts of Googledom, in spite of your team’s recent efforts to reduce the impact.
Specific query is “summerfield suites”, which returns wyndham.com as the number 5 listing, only from anchor text of affiliate links.
Hope that helps.
Best,
Bryson
See http://www.google.com/search?q=argentina (NSFW)
With a tip to http://www.wolf-howl.com/google/searching-for-argentina/
The universal search filter for safe searches seems to be akimbo.
The third result on this SRP goes to an https:// page. How could that have shown up in the search results instead of the normal http:// as it should be?
See:
“Recent Spotlights made by Footnote Members”
http://www.google.com/search?hl=en&q=dymaxion%27s+Profile+-+Footnote+Member&btnG=Search
Karl, I’m looking for concrete bugs, not a philosophicial discussion of how much PageRank new pages should start with. Pruning.
igor, a discussion of whether a particular page was/wasn’t indexed, or what the definition of a Googler is, also doesn’t count for what I’m looking for today. Pruning.
on webmaster tools you get errors on ie becase you mix secure and non secure items on the page.
The uk version of base some time throws certificate errors i mention this to a PM (chey) around chrismass time and I think theers a fix in the works for this.
I go to google.com/products and search for “hamster cage” the first result is
>
http://www.google.com/product_url?q=http://www.target.com/gp/redirect.html/ref%3Dtgt_adv_XSG10001%3FURL%3D/gp/detail.html%253Fasin%253DB0002AQWMS%2526AFID%253DFroogle%2526LNM%253DB0002AQWMS%257C2Story_Hamster_Cage__Yellow%26nAID%3D14110944&fr=AG8yBI2R72mSaB0Nh-o211p3lfjr_VdPQkZUCE_m1-_mAAAAAAAAAAA&gl=us&hl=en
>
This link works fine. But when I hit add to shopping list and then veiw my shopping list the link is changed and is broken it is
>
http://www.google.com/product_url?q=http%3A%2F%2Fwww.target.com%2Fgp%2Fredirect.html%2Fref%3Dtgt_adv_XSG10001%3FURL%3D%2Fgp%2Fdetail.html%253Fasin%253DB0002AQWMS%2526AFID%253DFroogle%2526LNM%253DB0002AQWMS%257C2Story_Hamster_Cage__Yellow%26nAID%3D14110944
>
And when I add it to my wish list same error. I will post that link @ the end.
I have posted this bug to the Google Check out Groups and the Google Public Support page to no avail (or repsonse). I have seen other people with same issue and no instances of this working correctly. It totally makes the wishlist useless and keeps me from recommending all my friends to set up wishlist (so I can do all my xmas shopping online). Here is the wishlist link
>
http://www.google.com/product_url?q=http%3A%2F%2Fwww.target.com%2Fgp%2Fredirect.html%2Fref%3Dtgt_adv_XSG10001%3FURL%3D%2Fgp%2Fdetail.html%253Fasin%253DB0002AQWMS%2526AFID%253DFroogle%2526LNM%253DB0002AQWMS%257C2Story_Hamster_Cage__Yellow%26nAID%3D14110944
>
Please help!!
Ryan and others that I didn’t approve, that behavior (show a url in SiteLinks and also show it in the #2 position) is intentional. I know it may seem like a bug, and we may change it later, but enough people don’t notice SiteLinks that we chose to show the url again at #2 if the scoring algorithms would have put it there. So that is intentional behavior, not a bug.
Also, enough people have affirmed Harith’s comment; I won’t approve any more of those, but thanks for mentioning it.
Matt, we could really use some guidence on whether this -950 penalty is a bug or not. It’s affecting lots of top notch sites (one really good one in particular ) an no one has any idea what’s causing it or how to correct it. Thanks
Why does:
site:puremobile.com
site:www.puremobile.com
both work the same way, but these do not:
link:puremobile.com
link:www.puremobile.com
I hope this qualifies as a bug…
After third-party review sites (CitySearch, InsiderPages, etc…) responsibly delete blackhat reviews, they still persist – for weeks – in Google Maps/Local business profiles.
I reported this four weeks ago in the Maps Group but received no reply from the Maps team – http://groups.google.com/group/Google-Maps-For-Business-Owners/browse_thread/thread/1ca4c29a4af8a960 . Unfortunately, yesterday’s update did not fix the issue.
Matt,
I have a bug/question about original content. How does google handle duplicate content STOLEN from your site. I’ve had incidents where, a few sites seem to have blatantly copy and pasted my content on their site and those copy pages rank higher in the serps. Can google tell where the content is originating from?
This is not a major issue, but I thought I would share it anyway since it appears to be buggy. If a search is performed with inurl:, and then preferences are changed to increase the number of results displayed, a 403 error is presented indicating that the query appears to be an automated request from either a virus or spyware.
Here’s an example:
http://www.google.com/search?hl=en&q=inurl%3Agoogle&btnG=Google+Search
and then click on preference, change number of results to 50, and save preferences.
The 403 error says it appears to be similar to an automated request.
All the static pages in the google directory have a modified date of 2005!
Modified by Google – ©2005 Google
Example: http://www.google.com/Top/Shopping/
The static pages have not been updates since then. One of my sites have been in DOMZ for more than two years, but I am not listed in any of the static pages. The only way to find my site is through a search. What is up with that?
I’ve noticed on several sites that cache:http://www.site.com/ shows a page as not being cached but then if you google the address it comes up with a cached version. If you then click on Cached it works with the only real difference being some sort of checksum in the URL ie 6q4x9e_sdzAJ:http%3A%2F%2Fwww.site.com%2F. Cheers!
Matt,
Fair enough. It just seemed a bit odd to me.
Thank you for the explanation.
Not sure if it’s a bug or just general weirdness, but we see completely off-the-wall titles every once in a while. Do a search for “Private Investigators Newark NJ” and look the title for http://www.magicyellow.com/category/Private_Investigators/Newark_NJ.html shows “Video Games in Alaska cities”
Yet the page (cached or live) has nothing about Alaska or video games.
And what I mean by search is this:
If i do this search: http://www.google.com/search?q=baby+carriers&sa=Google+Search&cat=gwd%2FTop%2FShopping&hl=en
my site shows up in the list. Then when I click the:
Shopping > Children > Baby > Equipment > Carriers
Link below my site name my site does not appear on that page.
Number of links listed do not reflect the true number of links
“Lonely garden pages” don’t get indexed ( ranked). These are a group of pages off a a sub page or directory
removed pages show up as broken links
Okay, I’m walking in to my noon bugs meeting, so I’ll probably lock this thread soon. Thanks for mentioning this stuff, everybody.
Perhaps this one is caused by my limited knowledge, BUT
In google webmaster, when you specify which is preferred, the ‘www’ or non-’www’ domain — I’m still seeing both returned in the google results page, one domain as a subbranch of the other, for example:
passion2publish.com yadyad
http://www.passion2publish.com yadyad
Why is this not resolved to one result? If I’m making a mistake, make this easier to setup.
Francois Faubert, it’s #2 on google.ca for [art tv].
Vancouver Island Daryl, example url?
Dave Jemison, calling it not-a-bug:
- on page we saw on July 9th, everything was fine
- viewing the cached page shows the video-game title
- [site:magicyellow.com private] looks good for other areas.
So I’m guessing a server bug/glitch on magicyellow.com. Ping me if you know of other examples on other domains though.
Hey Matt,
I’m not sure if this is a bug but, thought I’d pass it on to see if you had any ideas! It might help to know, A/ROSE is believed to be the name of an extension for Mac OS 7.5.
“I wanted to perform a search for a simple string containing a slash “/” character. The exact search was for A/ROSE
The problem is that Google kicks out the slash character and performs a search for “A ROSE” which is not at all the the same thing…”
“The closest I got to solving my problem was when I hit http://www.google.com/codesearch where one can enter REGEX expressions. Unfortunately it only searches c code.”
from:
http://forums.searchenginewatch.com/showpost.php?p=111887&postcount=1
Thanks for your time and have a great day!
Brian
a search for 08028 returns an interesting result at #5
probably not a bug now that i found this…
User-agent: *
Disallow: /
Hi Matt,
I’m not going to consider this a Google bug, but why is John Chow not ranking number 1 for his name? I believe that I sent you an e-mail about it, but wasn’t sure if it was the right address. It would be great to get specifics if John Chow is being punished for not ranking number 1 for his own name, and why he is not ranking number 1.
If you don’t reply to this post could you at least send me an email with some type of explanation, because I’m sure many people would find it interesting and it would be a great example of what not to do.
Thanks
Chris Hemphill
We’re running into an issue with geolocation pertaining to the map/address that is associated with our company in Google organic results. Query ‘citymax’ or ‘citymax.com’ and the results will display citymax.com with the following map:
Map of 4205 Broadway St # B, Pearland, TX 77581, USA
This location is being pulled from a page on a customer site hosted with us; their URI is a subdomain of ours, formatted as ‘customersite.citymax.com.’
If you search for ‘citymax’ in the maps tool the correct address shows:
1605 – 555 W Hastings St, Vancouver, BC V6B 4N4
We have attempted to fix this by manually submitting our correct address via Google’s Local Business Center, but the incorrect result persists.
Google Apps does not let you register an account for ANY domain with the letters “sex” in it. This includes stuff like “middlesex”, “sussex” and various others.
Yeah, I think there is a bug in google, because I am not ranked #1 for cheap hotels with or without quotes.
RSS feeds in the SERPS. I’m sick of them! Sometimes I click one unsuspectingly, so of course Safari hands me off to NetNewsWire, where I have to cancel and then head back to Safari, only to notice that Google also returned the non-RSS version of the search result, but ranked lower than the RSS. Why should a web search engine rank non-webpage content higher than a webpage? And if the only results available are RSS, could Google provide some kind of integration so clicking the search result doesn’t take me to the my feed reader? It’s extremely unlikely that I would want to subscribe to an RSS feed directly from a search result.
Does a Blogger bug count as Google bug?
Why does Blogger nofollow a reputable Google URL? I think the Blogger folks need training because they obviously don’t understand the nofollow-beast ;
BTW you should improve your ranking for [buy cheap viagra]
Oups, the “training” URL is http://sebastianx.blogspot.com/2007/06/blogger-abuses-rel-nofollow-due-to.html
Matt, I think the german umlauts (and maybe other special characters of other languages) need some attention. I have a large german blog hosting site and we get all kinds of encodings for umlauts (switzerland, austria and even dutch people) öäü .. and sometimes they submit in latin-1 in UTF-8 or even post their windows word text over to the WYSIWYG editors which comes in strangest encodings… MOST of the times we can catch the stuff, but if an umlaut goes unfiltered into the title of the page, you sometimes display it as a mess (like ö for the ö). It could be our bad coding and a bug on our config side, which I will check, but Google normally thinks better than I do and filters and corrects such errors… anything else: fine for me, thanks for asking!
Matt:
We have a Yahoo store site with approx 28,000 retail items. In automotive many many vendors have lines that have different titles (applications) but identical text descriptions. (ie T-rex billet grilles or Spintech mufflers) Google displays approx 9,400 items in the regular listings and has banished 9,800 items into the supplimental listings. We were one of the original Froogle bulk loaders and still bulk load our listings on a regular basis. How can we correct this as it has cut our sales in half?
Please advise
Thank you
Matt:
I don’t know if you consider the interaction of sitebox (onebox?) results with local to be a bug or not, but this query gives a single company most of the real estate on the page. I count nine (9) links on the SERP.
Also, splitting up 2 listing from a single site (Berkeley Parents Network) by dropping the local results in between is a really jarring UI experience.
http://www.google.com/search?q=berkeley+toyota
One bug that’s annoying me and everyone that I talked about it is the Flash Script bug on the date selection control of the new Analytics GUI.
Be it Mac, Windows, Linux, Firefox, Safari, Internet Explorer, it ALWAYS happens to freeze the browser when you click on the date/period control. Then the Flash Player pops an alert window stating that there’s a flash script which is making the machine to run slow, and asks if you want to abort the script or continue. Continuing the script makes the browser hang again.
Me and lots of guys never saw a glimpse of the looks for that period control. And so we’re forced to see all data in the default last 30 days views, which is a shame.
I did contacted the Analytics support team weeks ago, and they aknowledged the bug, but it still persists.
We’ve got a problem where our doman name is converted into a JIS encoded domain name.
The query:
site:wink.com -site:blog.wink.com
vs.
http://www.google.com/search?q=site%3A%EF%BC%B7%EF%BD%89%EF%BD%8E%EF%BD%8B.com
Yeilds a handful of results 8, however the query for the JIS encoded version of our domain name yeilds 287 results. The bug of course is why we’ve got multiple version of our domain name in the system (none of which should be in JIS encoding) and also it would be nice if site:wink.com would pull both domain names (since it’s JIS encoded ASCII).
There of course is a deeper issue about why we don’t have the ~30,000 pages in the index, but that’s not necessarily a “bug”.
Thanks.
I believe that is by design to maintain domain clustering.
I want to make sure if the post on “How Google is going to over come Opera mini’s difficuitly” really makes sense. http://www.mobilejgames.com/2007/07/18/googles-difficulty-in-over-coming-opera-mini-diffcuilty/
Hi Matt,
The info search is often telling my I might be a automated program, so it asks me verify that I am a human. When I do it tells me the search is not available.
The info search is extremely valuable when determining if PageRank is being faked (you could just fix the faking PR bug).
Thanks,
Sosh
Get a weird error when doing info: for a particular domain.
http://www.google.com/search?hl=en&rls=GGLJ%2CGGLJ%3A2006-36%2CGGLJ%3Aen&q=info%3Asecuritysaint.com%2F&btnG=Search
Returns: Sorry, no information is available for the URL securitysaint.com/
However the site: search appears normal.
http://www.google.com/search?hl=en&rls=GGLJ,GGLJ:2006-36,GGLJ:en&sa=G&q=site:securitysaint.com
Returns: Results 1 – 10 of about 223 from securitysaint.com
The link search also appears normal.
http://www.google.com/search?hl=en&rls=GGLJ%2CGGLJ%3A2006-36%2CGGLJ%3Aen&q=link%3Asecuritysaint.com%2F&btnG=Search
Results 1 – 9 of 9 linking to securitysaint.com/
As does the quoted search
Results 1 – 10 of about 1,680 for “securitysaint.com/”
Is this some kind of canonical search issue?
1. Google D&S’s “Save as PDF” doesn’t working on simplified Chinese characters (many characters missed.)
2. Gmail doesn’t recognize Word2007 file extension “.docx”
3. Can’t use Google video in some country.
4. Google Group’s search function don’t work very well on non-English posts, and many old non-English Usenet posts archives are gibberish.
5. Google Reader URL redirect problem:
http://reader.google.com/ redirects to http://www.google.com/reader/
but, httpS://reader.google.com/ redirects to http://www.google.com/
6. Google Groups’ “Manage Members” page truncated while using low screen resolution.
7. news.google.cn lost its ATOM/RSS output.
At the risk of sounding extremely egomaniacal, there may be a minor bug associated with my last name. One of my cousins pointed it out to me.
When you type in “Adam Senour”, results containing the word “Senior” and not “Senour” are included. Yet they don’t seem to show up that I could see for a “Senour” query.
(Personally, I find it bizarre that people are looking for me under my own name, but if people are that interested in me, who am I to question it?)
It would appear igor has been talking out both sides of his mouth. That is, bad mouthing many others for being black-hats, when he himself is one and has now be BUSTED
Reinclusion requests are now known as reconsideration requests.
M
The supplemental result search recently broke. Using it was useful for determining if a site had internal architecture issues which were causing it to wast PageRank and fill Google’s indexes with low information quality pages.
I consider it a bug that sites who receive the benefit of sitelinks are often shown at the #1 *and* at the #2 position.
Try this query: http://www.google.com/search?q=eurodns
Results are:
and at the #2 slot:
Perhaps, but I’m not the black-hat who spammed Google while all the time bad mouthing others for doing so
Hi all!
I noticed this issue: when typing “hotel milano” on google.it, some pages are displayed among the results both in the 1st and in the 2nd pages. They are the same pages and what is more surprising is that they are in the 8th and 10th position in the 1st page, while in the 1st and 3rd position in in the 2nd page. That doesn’t simply happen with the above mentioned keyword, but also for keywords such as “hotel londra” or “hotel parigi”: they all are related to the field of hotellerie.
The 8th and the 10th positions should of course “shift” in the second page where they are supposed to rank in the 1st and 3rd position, if we considered the first page with only 7 results dislayed (and other 3 at the top associated to a map).
Hello, Matt. I hope this is in the right place.
Some UK owned sites with .com domains (like this one – http://www.rousiland.com/) are not included in the .co.uk results even though they are definitely hosted on a UK IP. Could the list of reference IPs need updating? If there is anything that can be done by the website owner to alter this situation I would like to know about it.
Many thanks.
Is this a bug? Try doing the following link search: http://www.google.com/search?hl=en&q=link%3Awww.photostoday.co.uk&btnG=Search
27 results. Go to next page, 12 results (which is fine because similar pages are omitted), but click on “repeat the search…”, 37 results. Click to the next page, 18 results! How many results are there??
PS: I’m pretty sure this thread is intended to be a bug list. Please take your off-topic conversations elsewhere. You are just creating an admin nightmare. Bugfixing is enough of one of those as it is!
A few times a week when I do searches at home I get the following error:
Google Error
We’re sorry…
… but your query looks similar to automated requests from a computer virus or spyware application. To protect our users, we can’t process your request right now.
We’ll restore your access as quickly as possible, so try again soon. In the meantime, if you suspect that your computer or network has been infected, you might want to run a virus checker or spyware remover to make sure that your systems are free of viruses and other spurious software.
We apologize for the inconvenience, and hope we’ll see you again on Google.
I’m not doing anything automated and have no viruses or spyware on any computer within my home network. One search in particular was vacation ideas that triggered the error message. the only way I have figured out how to avoid it is to wait a few hours before doing another search or just reboot.
CY
http://www.google.com/search?hl=en&q=new+york+headline+news&btnG=Search
7 of 10 listings are for CNN.com this happens for a lot of other city searches, surely they are other relevant sites for this search query…
Matt:
Every now and again in my apache logs I see a page request from Google whereby the page size is ZERO (0) length when the bot makes a “GET” request.
Can you advise me as to how I might debug this problem?
Patrick
Nope, it’s also Google calling you one by banning you
I know your meeting was yesterday. Anyway I comment this just in case.
Blogspot has in place a robots.txt with
Disallow: /search
However, if I search for pages on my site, [site:noigo.blogspot.com] I see that the second result is
noigo.blogspot.com/search/label/vodafone
If you can go back in time then this might be worth looking at:
http://www.google.com/search?hl=en&q=are+meta+tags+dead+or+not
From my location I see Danny Sullivan’s article on SEW.com and Clickz.com ranking #1 + #2 respectively. I would have thought that would be a prime candidate for dupe content?
Bit noisy around here lately? Boy you are tolerant Matt.
Rgds
Richard
[quote]Boy you are tolerant Matt[/quote]
I do hope, Mama G has hired one supporter for him to pre-moderate the comments or pick special comments to answer! What he does here is IMHO very important for a (the?) pure Internet based company, but in that scope it is getting crazy!
OK thanks, I can investigate the idea of users in the community putting up duplicate content. People do that, secretaries start a profile and get interrupted then later start a new profile. At least I have a starting point now.
However, if nothing is being transferred, how does a search engine know something could possibly be duplicate content?
I’m not totally sure of this yet, but since you asked for bugs a few days ago now I thought I’d mention this before this page got too stale – Google Analytics seems to be having problems with URLs which contain semicolons. Like I said I haven’t really pipnned down the problem yet, but it looks as if it’s actually losing part of the URL. Seems like a bug to me unless I’m not understand some RFC or something…
I have just remember I hope you can do something…
http://groups.google.com/group/FFToolbar-group-advanced/browse_thread/thread/dcc792cfd0930739
The 302 Redirect issue.
I have my site listed at #3 for a particular keywords for 3 months.
All of a sudden a website url that is just a redirect to my site pops in as me in the list (no one else has changed, and technically the people clicking on the 3rd link will get to my site) but it’s their url.
Which means they can redirect my traffit elsewhere in the future.
(I have worked on the steps to correct this, and it has worked. My url is back at #3, but the bug remains.)
Hello Matt,
I would like to know if it is a ‘bug’.
I have a website and google has cached at least 5000 pages of the same.
Those pages are still inside the cache and one can see them by searching
through the cache. Unfortunately only 300 pages (out of 5000) are displayed
in the “search results”. All those pages are simply unique and different from
each other. All those 5000 pages were displayed during the first six months of my website and they slowly started disappearing!
If you have already explained this somewhere else then kindly give me the link so that I can solve this everlasting problem.
Best regards
Kolkata, India
google’s “nofollow” command has become a bug
wikipedia proved that by using this command on all outbound links — you can move up in google AND destroy the web at the same time.
Get rid of it, unless you want to call the “web” — the “string”
it is a dangerous command to the unity of the web and of search engines. Thanks. G.
Quite often blog feeds show up in google search, likely this is from google looking at all the internal links my feed gets and thinking that this page is more important. Obviously the page on my site which ends in .php or .html and has the same info is quite a bit better… As of typing this the search http://www.google.com/search?q=decisions+in+paradise has one item which has a blog feed. The url is ivicawd.com
How do people email you? I think you should write an entry replying to this article which I totally agree with – http://www.rustyrazorblade.com/index.php/2007/07/30/experts-exchange-should-be-removed-from-google-search-results/
Google webmaster tools behaves strangely for one of my site (www.klaus-unger.dk). When I try to set preferred domain, it says, that the site is not verified (although it is, otherwise I could not use the verifying-feature).
Everytime I try to (re-)verify the page, a new item automatically appears in my dashboard “http:///”.
For all other sites of mine, the verification takes place properly. This is annoying.
Adwords has a bug. When I go in to make the sites resume (after being paused), the “paused” message remains but the text says, “Your campaign is resumed.”
Also, the credit card area is having issues.
Bugs in Google Webmaster Control Panel
BROKEN LINKS: I don’t know whose bug this is, but the Wiki links are showing up as broken in my Google Webmaster Control Panel.
Other links are showing as broken that are non-existent such as http://www.domainname.com — I have no idea what this means.
RANKINGS: The search query results are often very, very off. It may say that my site is ranking #9, but rarely does it match the actual ranking of the keyword phrase.
Another bug… Google does not show all the backlinks coming into the website.
No idea if you’re still looking for bugs but I was using Google Maps today to see state maps and a search for pa (trying to get the Pennsylvania map) results in Panama and a search for va (Virginia) goes to Italy (maybe centered on the Vatican?).
What is the use of posting an entry here? Does Matt even bother to read any of these? Google is too busy building up new things thus ignoring the bugs which are persisting for a long time.
Today I found a bug: a content-thief – aktuellnews.de – steals parts (to the WordPress-more-Tag) or whole artikels (if there´s no More-Tag or something similar).
The Google-Bug for this is: Google Search Results lists only the stolen articles at aktuellnews.de, and not my original articals at Finanso.de – that concerns the articles for some weeks.
I think my original articles are segregated as double content.
Do I see that correct? And can you please repair the bug?
The whole story with pictures (German): http://www.finanso.de/blog/aktuellnewsde-content-dieb-und-google-ist-vllig-blind-geworden/
Supplement:
An example: The Google search
http://www.google.de/search?hl=de&q=%22Deutschland+hat+zu+hohe+Mobilfunkpreise+-+oder+die+Kunden+nutzen+zu+teure+Tarife%22&btnG=Google-Suche&meta=
according to an article of me: “Deutschland hat zu hohe Mobilfunkpreise – oder die Kunden nutzen zu teure Tarife”
http://www.finanso.de/blog/deutschland-hat-zu-hohe-mobilfunkpreise-oder-die-kunden-nutzen-zu-teure-tarife/
results in only two sides of aktuellnews.de an none of me the autor of this article.
In processing of the search request “Australia’s most renowned“, The Google begins erroneously highlight all “s” in URL of the documents found (marked green in SERP). See http://blog.dataparksearch.org/67
If you are interested in any sort of Google bug, here’s something I’ve noticed in Checkout.
Google Checkout bug
When I am signed in to my Google account, when I click on a Google Checkout button, the google.com/accounts/ManageAccount page displays, instead of the checkout.google.com/view/buy page.
When I am not signed in, a click on the Google Checkout button takes me to the correct page.
Google thought I was a bot, so it showed me the Captcha. However there was *no submit button* on the Captcha page. I knew hitting enter submits the form, but people like my mom would just get confused and search somewhere else.
I submitted this via the feedback form, and was given an explanation of what a captcha is =P
I’m not sure if Gmail is your area, but the spam bin is an absolute nightmare. Emails which are legitimate get put in the spam bin while spam emails get through to the real email account.
Even worse – *bounces* are ending up in the spam bin. If you send an email and it bounces you need to know so you can figure out why and try to resend it. Obviously some clueless people have thought they are spam, put them in the spam bin, and thus a problem is created.
Users should be able to over-ride the spam bin if they choose to. There seems to be no option for that.
Snoskred
Why don’t you use http://www.googlebugs.com to log the bugs?
Craigslist postings:
http://www.google.com/search?q=moving+help&hl=en&client=firefox-a&rls=org.mozilla:en-US:official&hs=DjG&pwst=1&start=10&sa=N
non-relavant postings on craigslist show up. Right now it looks like only a few are showing up on the second page… but every couple weeks 2 or 3 will popup on page 1.
Course the spammers/link sellers/etc always dominate this niche… too. But I’m sure you guys will get them one of these days!
Thanks!
Mike
Hello, my annoying is that when I used view “Cached snapshot of the page” from google toolbar, i get a CAPTCHA page, and after a input the text from image i’m still getting a page that tell me to get a virus/spamware remover. It’s very annoying because i often check cached snapsot of a page, but not enough times a day to be considered a “virus”, so please take that in consideration and if somebody want more specific details i could be conacted over email. Thanks
Hi Matt;
I’m not sure if this meet your criterias or not but i will be happy if you check http://groups.google.com.tr/group/Google_Webmaster_Help-Tools/browse_thread/thread/ad6cc82f90981962/240d4ebed7c912a5?hl=tr#240d4ebed7c912a5
I keep seeing near duplicates at the top of the listings, IE site 1 page a, site 1 page b, site 2 page a, site2 page b, site 3 page a, site 3 bage b etc. This doesn’t do a lot for user excperience, IMO so is this a bug or a deliberate attempt to make us all search deeper I wonder?
This search result displays a map placing the Natural History Museum on the east side of Manhattan on Park ave. instead of Central Park West where it has been for 100 years.
http://www.google.com/search?hl=en&rlz=1T4GGIH_enUS234US234&sa=X&oi=spell&resnum=0&ct=result&cd=1&q=natural+history+museum+ny&spell=1
This may be late but I wrote about a javascript error “nan” bug that appeared in Google Analytics.
http://www.webconsultingdc.com/2007/what-is-nan-google-analytics/
Here’s an interesting one…
Looking for the correct syntax in SQL for “Select Distinct Top 10″
‘Select Top 10′ promptly directs to ‘the page cannot be displayed’
This is consistent on at least 3 computers so far.
Yahoo! has the same result.
I am logged onto gmail. I log on to orkut in a new browser window. I log out of orkut -> I am logged out of gmail. Are the authentication mechanism’s linked?
A little bit above (here) one comment was dealing with Google geo-location issues, related to sofa.com.
What you mention that Sofa.com had better ranking in Google.com then in Google.co.uk is not a bug, but a quite natural thing as sofa.com is a .com site, and the name-server and web-server IP is absolutely irrelevant. (Actually currently the ranking is ok for sofa.com in both Google-site:perhaps the simple thing that you wrote the address of the company to the main page was enough).
Normally local Google sites give better rankings for local businesses so You have to make it clear in some kind of text format that where is your exact location.
For example I’ve just started to analyse sofaclassics.co.uk site and its rankings for different keywords and as it is normal it ranks far better in google.co.uk for any keyword then in google.com- as it is absolutely clear that it is a local UK based business: address, .uk domain, etc. But if it is not clear whether a .com site covers a local business or not it always will get worse position in local Google rankings. Mainly if you are searching for products.
I’ve had some odd problems with analytics. For a time, I was only able to look analytics for one page. I contacted Google and they took it very seriously. A representative called me a couple of times and the problem was fixed.
Hello Matt,
I had a good look through mattcutts.com / Google before I decided to report this one (there is one thread describing the problem at Webmaster World).
http://www.webmasterworld.com/forum80/620.htm
The issue is that some websites using https:// are not showing PageRank in the Google Toolbar when they clearly should. For example the query [CIA] returns the official CIA website as the first result which uses https, but the page doesn’t show any PageRank in the Google Toolbar.
This relates to google search, when logged into gmail.
When I log into gmail, and then go to google.com to do a search, for every search listing, there is this warning – “This site may harm your computer”.
If I am not logged in to gmail, this warning disappears.
This applies to Google Patent Search. When viewing a single patent, a button appears to “Download to PDF”. This used to append the .PDF extension to the downloaded files, but as of a few weeks ago the extension is removed. Now, every time a patent .pdf downloads the user must manually append the .pdf extension to the downloaded file.
Easy to overcome, but a bit of a hassle.
Google Calculator lg() Bug:
Google uses different notation for interpretation logarithm functions:
Google:
Input= Output
lg(5)= 2,321928090
ln(5)= 1,609437910
log(5)= 0,698970004
ld(5)= -
WolframAlpha:
Input= Output
lg(5)= 0,698970004
ln(5)= 1,609437912
log(5)= 0,698970004
ld(5)= 2,321928095
In my interpretation: WolframAlpha is the correct output, here the standard notation:
ln log_e natural logarithm
lg log_10 decadic logarithm
ld log_2 binary logarithm
This is not a major issue, but I thought I would share it anyway since it appears to be buggy. If a search is performed with inurl:, and then preferences are changed to increase the number of results displayed, a 403 error is presented indicating that the query appears to be an automated request from either a virus or spyware.
There is a relatvely new bug when surfing on the iPhone os version of Safari. More specifically when using iPod touch (it may be present on iPhone as well but I use iPod touch). When searching, you cannot click on any of the other search types (e.g. Image seach). Futhermore, if you try to work around this by add “image” to your search term and click on the search result that links to google image search, no images appear. Essentially it is now impossble to search images on iPhone os safari. It was not alway this way but more recently it has. Every time I update iPhone os, the problem persists.
Sam B.
Search result navigation from http://www.google.com/ is broken in FireFox (3.6.6) if JavaScript is enabled. Go to http://www.google.com/ and enter anything. The initial search works, but the “Next” and “Search” buttons and the numbered result page links don’t respond. They change the address, but the content of the page doesn’t change.
The buttons and links work if JavaScript is disabled in FireFox. Internet Explorer 8 and Opera (10.10) are also okay (even with JavaScript enabled).
Google web history sunmit button looks buggy!!
https://www.google.com/history/?hl=en
In the Blogger interface, sometimes the button to change the template is missing, and then it just reappears after a day or so. This happens with the new interface. I wouldn’t call this a bug really but perhaps the development team is just working on the new interface. I am sometimes notices this when logging in using admin email to Google apps. The link “Manage this domain” was not there. I wonder if anyone else had the same issue.