Martijn,  thanks for your response.  I could see Google saying that on a given page the same source image shouldn't have more than one alt description, and either penalizing such or just picking one of them or ignoring it altogether.  It's the penalty I'm concerned with, of course. . But, I can also see that if they are seeing a blank.gif for the source and some code related to lazy loading they may go ahead and give credit to each alt as though a real src was loaded--and maybe even tying it to the real src image name for image search.  Just looking for a bit more feedback from real-world experience first..

Has anyone else worked with this and determined if it is a pro or con?

I didn't find an answer on a search on this, so maybe someone here has faced this before.

I am loading 20 images that are in the viewport and a bit below.  The next 80 images I want to 'lazy-load'.  They therefore are seen by the bot as a blank.gif file.  However, I would like to get some credit for them by giving a description in the alt tag.  Is that a no-no?  If not, do they all have to be the same alt description since the src name is the same?  I don't want to mess things up with Google by being too aggressive, but at the same time those are valid images once they are lazy loaded, so would like to get some credit for them.

Thanks!  Ted

Hi AMHC,

It makes sense that without hardly any backlinks built up Google wont find my upper case URLS since all the page links have been changed, however, I am writing out all of the urls that are redirected into email, and from that I can tell that Google is finding them--I guess they may have a list of urls from prior indexing that they crawl independent of what their crawler comes up with.

I'll keep looking to see what they have indexed and if it turns out they just aren't crawling certain pages, will put them in a sitemap to be crawled..It's a good idea for taking care of the problem quickly--so if it progresses too slowly I'll do that.

Thanks very much for your answers!

Thanks AMHC.  In my case, I just don't have many back links so I don't have the urgency that you faced with getting Google to see all the redirects.  But, I'm still not understanding--it sounds like you believe that once google sees the redirect it removes the old uppercase from its index.  It doesn't look to me like that is what happened in my case because Google is currently indexing BOTH, and so that means it has crawled my new lowercase and I know it isn't crawling any uppercase anymore (it cant--all are redirected).  So, that's why I wonder if I have to remove those uppercase urls...does that make sense or am I just not understanding it still?

EDIT: I just discovered I wasn't doing a 301 direct so it wasn't considered a permanent move.  That, if I understand it right, will remove the upper case from googles index permanently.

I gotcha.  Yeah, different thing going on here..these urls can be really difficult!  I have uppercase lowercase, https http, urls that have different content(not just formatting) for mobile as desktop and vice versa, mobile urls that dont even exist for desktop, and desktop urls that dont exist for mobile..all under the same domain.  1000s of internal pages....In the desire to create a good website for users I've created an SEO monster  because I didn't realize the many consequences with regard to search indexes.

If you know a true expert in these areas I need him/her.  4 years on this site, its live finally (2 months), and now I'm discovering all of these things have to be fixed, but i can't afford thousands of dollars..I'll do the work, I just need the knowledge!

Hutch, thanks.

The site is dynamic with thousands of pages that are now being redirected to lower case, so I'm not seeing how using canonical would work because the upper case urls aren't on the site anymore. I guess I think of canonical as being useful when you have ongoing content on the site that duplicates one or more other pages on the same site.  In my case none of the upper case urls exist anymore so they don't have 'ongoing' content.  I'm still new to this so if it sounds like I have it wrong, please correct me.

AMHC, thank you for your response.  I'm in the middle of quite a mess, as this is one of several issues, so really appreciate your help.  I must confess to not following everything you wrote exactly:

In your situation, I think i understand the redirect -- it is the same reason I am doing a redirect--it is so that anyone coming from to this site with uppercase in it will end up on the lower case page, and in the case of google will then index the page as a lower case page.  BTW, for me that has been easy as I am doing it via php -- if the url doesn't equal its strtolower of the url , then I redirect to strtolower.

I think I get what you are saying about the sitemap -- it speeds up google crawling the site and seeing that all those upper cases should be lowercase from your redirect.  In my case, i don't have the concern about Google discovering them as you did because my site is only a couple months old.  And, I never have given Google a sitemap so many of my pages aren't crawled yet (I am trying to clean up my entire url structure before i submit a sitemap to them--however they have already crawled perhaps 20% of the site, so I'm now trying to examine what google has crawled and how it has been indexed to figure out what needs to be done).

What I'm not understanding is this:  It seems to me that what you described should succeed for going forward to getting both Google and your users to the right ending page, but I don't see how it removes the prior uppercase urls from Google's index.  What is it that tells Google your prior upper case urls should no longer be in their index?  Is it the fact that they aren't in the sitemap you provide now?  Or, do they literally have to be removed using some kind of removal or disavow tool?  I discovered this (as you see in the op) because Google appears to never have removed the Uppercase ones even though they are indexing the lower case now.

Ted

This search   -  site:www.qjamba.com/online-savings/automotix

gives me this result from Google:

Automotix online coupons and shopping - Qjamba
https://www.qjamba.com/online-savings/automotix
Online Coupons and Shopping Savings for Automotix. Coupon codes for online discounts on Vehicles & Parts products.

and Google tells me there is another one, which is 'very simliar'.  When I click to see it I get:

Automotix online coupons and shopping - Qjamba
https://www.qjamba.com/online-savings/Automotix
Online Coupons and Shopping Savings for Automotix. Coupon codes for online discounts on Vehicles & Parts products.

This is because I recently changed my program to redirect all urls with uppercase in them to lower case, as it appears that all lowercase is strongly recommended.

I assume that having 2 indexed urls for the same content dilutes link juice.  Can I safely remove all of my UpperCase indexed pages from Google without it affecting the indexing of the lower case urls?  And if, so what is the best way -- there are thousands.

Thanks Cyrus.  Very good points!

Thanks Sheena.  In the second scenario good point--they are generated via user POST so in theory Google should never see them or index them, but since they can be shared Google ends up finding them, so I do need to make sure Google doesn't index them if possible.

Monica,

I'm going to open a new thread to ask a similar question, as I think I didn't ask it very well.

Thanks for your input,

Ted

Thanks.  If I understand you, the mobile bot won't crawl a url that the desktop bot has said needs to be fixed for it to work right for desktop. . Would you agree that doesn't really sound right on Google's part, since the url is fine for mobile use?  I don't know why it wouldn't crawl for mobile, but if that's the way it is I can try fixing it on desktop to see if that enables the mobile to get crawled.

Once I do that I guess I'll find out whether a 404 not found for desktop will disable it from crawling for mobile (yes that link is accessible from other pages)--I was hoping to avoid trial and error on that because the time lag seems like it would be hard to pin down.

In a nutshell here's what I'm concerned will happen:

Google mobile bot crawls my mobile page and indexes it:  Then the desktop bot crawls the same url and gets a 404 not found.  Because of the desktop not found, Google removes it from the mobile page index.

I don't see a good way to test that since it depends on when each crawler is crawling.  And, if this is what it is doing, I can't think of a good solution to having a responsive site with some content meant only for mobile indexing or only for desktop indexing.

Hi Monica-thanks for your reply:

Ok, for a page that is supposed to be mobile only within a responsive-like setup(ie one domain) here's what I see:

The desktop bot crawls the link and gives a soft 404 error -- presumably because the page is currently being redirected to the home page.

The mobile bot is not crawling that link despite it being prominent on the main site home page, as my dbase is tracking the bot crawling and is not showing that it crawled that link for mobile (but is for desktop), and a search on my smartphone doesn't show that link either (even though it does show other links for pages used by both)..  **Yet, if I fetch the mobile only page in webmaster tools using their mobile bot it finds it and renders it perfectly. ** So, why isn't it crawling it?  Is it because when the mobile bot crawls it first looks and sees that that link is already 'flagged'  as a soft 404 for the desktop?  Or, is it because the mobile crawler is getting hung up on a link on the home page for mobile that has nothing to do with this mobile-only link?

It appears that the mobile bot is influenced by the desktop bot results--which is my fear:  It seems to me their 2 bots should be independent of each other.  If they aren't independent then if I change it to a 404 not found for desktop, would that even help, or would that prevent the mobile bot from ever trying to crawl it?

I would think that anybody who has a responsive page design and has blocked out certain content so that it renders only for mobile or only for non-mobile has had to face this issue.

Not sure what to do--I could fix the soft errors--change them to 404 not found and just see then if Google starts indexing for mobile or not, but was hoping to get some feedback before experimenting.

Thanks again, and please share more if you have more thoughts!

Here are a couple of scenarios I'm encountering where Google will crawl different content than my users on initial visit to the site--and which I think should be ok.  Of course, it is normally NOT ok,  I'm here to find out if Google is flexible enough to allow these situations:

1. My mobile friendly site has users select a city, and then it displays the location options div which includes an explanation for why they may want to have the program use their gps location.  The user must choose the gps, the entire city, or he can enter a zip code, or choose a suburb of the city, which then goes to the link chosen.  OTOH it is programmed so that if it is a Google bot it doesn't get just a meaningless 'choose further' page, but rather the crawler sees the page of results for the entire city (as you would expect from the url),   So basically the program defaults for the entire city results for google bot, but for for the user it first gives him the initial ability to choose gps.

2. A user comes to mysite.com/gps-loc/city/results   The site, seeing the literal words 'gps-loc' in the url goes out and fetches the gps for his location and returns results dependent on his location.   If Googlebot comes to that url then there is no way the program will return the same results because the program wouldn't be able to get the same long latitude as that user.

So, what do you think?  Are these scenarios a concern for getting penalized by Google?

Thanks, Ted

I have a domain (no subdomains) that serves up different dynamic content for mobile/desktop pages--each having the exact same page url, kind of a semi responsive design, and will be using "Vary: User-Agent" to give Google a heads up on this setup.

However, some of the pages are only valid for mobile or only valid for desktop.  In the case of when a page is valid only for mobile (call it mysite.com/mobile-page-only ), Google Webmaster Tools is giving me a soft 404 error under Desktop, saying that the page does not exist, Apparently it is doing that because my program is actually redirecting the user/crawler to the home page.  It appears from the info about soft 404 errors that Google is saying since it "doesn't exist" I should give the user a 404 page--which I can make it customized and give the user an option to go to the home page, or choose links from a menu, etc..

My concern is that if I tell the desktop bot that mysite.com/mobile-page-only basically is a 404 error (ie doesn't exist), that it could mess up the mobile bot indexing for that page--since it definitely DOES exist for mobile users..

Does anyone here know for sure that Google will index a page for mobile that is a 404 not found for desktop and vice versa?  Obviously it is important to not remove something from an index in which it belongs, so whether Google is careful to differential the two is a very important issue.  Has anybody here dealt with this or seen anything from Google that addresses it?  Might one be better off leaving it as a soft 404 error?

EDIT: also, what about Bing and Yahoo?  Can we assume they will handle it the same way?

EDIT: closely related question--in a case like mine does Google need a separate sitemap for the valid mobile pages and valid desktop pages even though most links will be in both?  I can't tell from reading several q&a on this.

Thanks, Ted