When we were picking posts to tackle from our most recent Popular
Picks thread, I'd initially wanted to respond to beussery's question
about dynamic URLs and a recent change to our help documentation on
this subject. However, we had a blog post in the works about this
topic, so I refrained; but it's out now and I'd like to share with
you:
Yes, we deliberately removed the section of our help documentation
that recommended creating static URLs for dynamic pages. Hopefully
this blog post will clarify the issue.
> When we were picking posts to tackle from our most recent Popular
> Picks thread, I'd initially wanted to respond to beussery's question
> about dynamic URLs and a recent change to our help documentation on
> this subject. However, we had a blog post in the works about this
> topic, so I refrained; but it's out now and I'd like to share with
> you:
> Yes, we deliberately removed the section of our help documentation
> that recommended creating static URLs for dynamic pages. Hopefully
> this blog post will clarify the issue.
In a case like that, I would definitely recommend staying with the
dynamic version. To our crawlers, a URL like that is very hard to
understand. While I wouldn't recommend changing back to dynamic URLs
on a whim, I would recommend that you reconsider this decision with
your next redesign.
In any case, I wouldn't call URLs like that "SEF" (search engine
friendly); it's more like "SEUF" :)
> In a case like that, I would definitely recommend staying with the
> dynamic version. To our crawlers, a URL like that is very hard to
> understand. While I wouldn't recommend changing back to dynamic URLs
> on a whim, I would recommend that you reconsider this decision with
> your next redesign.
> In any case, I wouldn't call URLs like that "SEF" (search engine
> friendly); it's more like "SEUF" :)
If they are equal or the second is better, is a 301 redirect from the
first to the second enough to not loose visitors or we have to keep
static urls for old sections of our sites and dynamic urls for new
contents?
> In a case like that, I would definitely recommend staying with the
> dynamic version. To our crawlers, a URL like that is very hard to
> understand. While I wouldn't recommend changing back to dynamic URLs
> on a whim, I would recommend that you reconsider this decision with
> your next redesign.
> In any case, I wouldn't call URLs like that "SEF" (search engine
> friendly); it's more like "SEUF" :)
I've had a hunch that switching to the so called SEF urls were
actually hurting me in Google as I've fallen into obscurity for that
site ever since.
I've wanted to switch back to dynamic but now I'm scared that I'll get
hammered for switching back as well, and even though I don't really
have anything to lose in Google this site ranks very well in Yahoo and
MSN and I hate to lose that.
Anyway I said to heck with it and just switched the site back to plain
old dynamic URL's. Any advice on how to prevent any penalty that might
come with the switch back?
> In a case like that, I would definitely recommend staying with the
> dynamic version. To our crawlers, a URL like that is very hard to
> understand. While I wouldn't recommend changing back to dynamic URLs
> on a whim, I would recommend that you reconsider this decision with
> your next redesign.
> In any case, I wouldn't call URLs like that "SEF" (search engine
> friendly); it's more like "SEUF" :)
Hi sfraise
There is no "penalty" for switching URL formats, though it might take
a bit of time for everything to catch up. One thing I would recommend
is that you set up 301 redirects from all old URLs to the new ones
(this is something that you should do in any case when changing
URLs).
I would say the URLs that you provided are equivalent and generally
speaking, I would prefer to see the dynamic versions.
One of the reasons for that is that we can use the information
provided through the parameters to better understand what your site is
doing with those parameters. For instance, the URL
http://www.mysite.com/search.php?q=keyword can give us information
about what is happening, it could even allow us to recognize that this
is a search form and perhaps let us attempt other keywords that might
lead us to content that we haven't seen for your site. On the other
hand, a URL like http://www.mysite.com/search/keyword does not give us
any information at all about what the "file name" is used for.
I realize that this is a fairly big change in the way that webmasters
think about "search engine friendly" URLs. I think it's important that
we are open about the kinds of problems that we have seen frequently -
and that we can give you tips about how to resolve some of those
issues :).
> I realize that this is a fairly big change in the way that webmasters
> think about "search engine friendly" URLs.
Without trying to sound facetious or sarcastic - that would have to be
one of the understatements of the year.
I have spent the last 5 years rewriting dynamic URLs to eradicate all
the so-called issues that dynamic URLs are supposed to cause the
googlebot during it's crawl (not to mention the sheer number of
duplicate content issues a non-managed dynamic URL is supposed to
create).
This latest announcement has really put me back 5 years in the way I
think about site architecture with regards to dynamic data-base driven
sites and URL creation/management.
I appreciate the clarity you are giving to specific examples, but some
clarity on this apparent turn-around would be helpful.
I think a minor disadvantage is that people still do actually type in
a URL on occasion:) Pluses, equal signs, underscores, long strings
etc., are always a pain for bad typers like myself.
> I would say the URLs that you provided are equivalent and generally
> speaking, I would prefer to see the dynamic versions.
> One of the reasons for that is that we can use the information
> provided through the parameters to better understand what your site is
> doing with those parameters. For instance, the URLhttp://www.mysite.com/search.php?q=keyword can give us information
> about what is happening, it could even allow us to recognize that this
> is a search form and perhaps let us attempt other keywords that might
> lead us to content that we haven't seen for your site. On the other
> hand, a URL likehttp://www.mysite.com/search/keyworddoes not give us
> any information at all about what the "file name" is used for.
> I realize that this is a fairly big change in the way that webmasters
> think about "search engine friendly" URLs. I think it's important that
> we are open about the kinds of problems that we have seen frequently -
> and that we can give you tips about how to resolve some of those
> issues :).
Is the assumption that every time a ? is present in a URL that it's
dynamic?
We use parameters after a ? for reporting, to tell us about the
referral site or where someone clicked on our site. The content never
changes because of the parameters, but we end up getting a lot of
duplicate indexing. Googlebot thinks these are unique URLs:
> When we were picking posts to tackle from our most recent Popular
> Picks thread, I'd initially wanted to respond to beussery's question
> about dynamic URLs and a recent change to our help documentation on
> this subject. However, we had a blog post in the works about this
> topic, so I refrained; but it's out now and I'd like to share with
> you:
> Yes, we deliberately removed the section of our help documentation
> that recommended creating static URLs for dynamic pages. Hopefully
> this blog post will clarify the issue.
Hi seriocomic
In a sense this is not a change. We have been looking at a lot of URLs
over the years and internally it's always been much easier for us to
recognize a site's structure if we see the "real" URL which a page
has. If for instance a site injects irrelevant elements into a URL
through URL rewriting then it's much harder for us to determine which
parts of the URL is actually relevant and which parts can safely be
ignored.
If you think that it's important to keep the name in the URL, I would
recommend using something like http://domain.com/parts.php?item=green-auto-part which also includes the descriptive element in the URL but which makes
it absolutely clear as to what is happening here. All parts of the URL
are relevant, any additional elements (like a session-id) can safely
be ignored because we can try them out and notice that we're seeing
the same page.
An alternative would be http://domain.com/parts/green-auto-part --
again, all URL elements are relevant, no irrelevant parts are added.
However, the big problem for us is that the average webmaster often
has trouble doing that and if that is the case, we would prefer to
have slightly messy dynamic URLs than any kind of mediocre URL
rewriting. If you can perfect URL rewriting -- by all means go for it,
but for all other cases, we would prefer to be able to understand your
site than to have it hidden from us through pretty & problematic URLs.
> We use parameters after a ? for reporting, to tell us about the
> referral site or where someone clicked on our site. The content never
> changes because of the parameters, but we end up getting a lot of
> duplicate indexing. Googlebot thinks these are unique URLs:
> .html
> .html?click=rss
As long as that element is part of the parameter in the back, we can
learn more about it and understand that it might not be so important
for your site. It would certainly be preferred to see /page.htm?
click=xzy than to see something like /page.htm/click/xzy.htm :) So
while you may be causing duplication through URL parameters, we'll
still generally be able to figure out what you mean in a situation
like that.
Thank JohnMu, I guess what I was talking about as far as a penalty is
the search engines suddenly seeing duplicate content and deindexing
us.
Unfortunately I have well over 30,000 pages on this site so doing a
301 redirect on all of them would take way too long.
Since I've switched the urls back to dynamic we've lost all rankings
in yahoo, I kind of thought it would take a little longer than 12
hours lol. Hopefully the se's can all figure it out quickly because
until then there goes my income. I wish there was some way to tell
google and yahoo about these kinds of changes from the webmaster tools
so we wouldn't have to wait 6 months for them to naturally figure it
out.
> Hi sfraise
> There is no "penalty" for switching URL formats, though it might take
> a bit of time for everything to catch up. One thing I would recommend
> is that you set up 301 redirects from all old URLs to the new ones
> (this is something that you should do in any case when changing
> URLs).
I would really, really recommend setting up 301 redirects, even if it
is for 30,000 pages (there's usually a fast way to do that, somehow,
somewhere :) ). Besides search engines, users are otherwise going to
have a tough time finding your content. It's hard to say how long it
will take for a change like that to propagate completely, but without
redirects it will certainly take much longer.
Don't know who I'm addressing this to, as the widespread opinion
elsewhere seems to be that the Google team are out of the country and
you are both impersonators, or hackers. Alternatively, that you are
pursuing a hidden agenda, when a hidden need might be the worst light
to throw on this.
The change is a different point of view but shared and backed up by
technical detail, perhaps some would complain if you sent them a copy
of the algorithm every week, wrong format. The complaint of Google
secrecy is often heard, this thread and the blog try to achieve the
opposite and should be applauded for that, thanks for sharing useful
information.
> I would really, really recommend setting up 301 redirects, even if it
> is for 30,000 pages (there's usually a fast way to do that, somehow,
> somewhere :) ). Besides search engines, users are otherwise going to
> have a tough time finding your content. It's hard to say how long it
> will take for a change like that to propagate completely, but without
> redirects it will certainly take much longer.
> I would really, really recommend setting up 301 redirects, even if it
> is for 30,000 pages (there's usually a fast way to do that, somehow,
> somewhere :) ). Besides search engines, users are otherwise going to
> have a tough time finding your content. It's hard to say how long it
> will take for a change like that to propagate completely, but without
> redirects it will certainly take much longer.
Well I've got all versions of the home page 301 redirected as far as
the non www version pointing to the www version, the /index.php?
option=com_frontpage&Itemid=1 pointed to the regular main url, and
the /index.php redirected to the main url so that should take care of
the front page anyway. Other than that I have a video component to the
site, a profiles component of the site, a blog componenent, a forum, a
photo gallery in each profile, a groups component, and a whole slug of
other small components to the site with 30,000 registered members. So
here I would have to change the url to each photo, each video, each
forum post, and each blog post ect.
I just can't do that at this point, I'm just going to stick with the
idea of building a good quality website that people want to use and
hope Google can catch up and just rely on the traffic we bring in from
the other two search engines until that day comes. There's no doubt in
my mind that the only other chat site out there that should beat me at
this point is Yahoo, and given the fact that they are filled with spam
bots the only thing going for them is the fact that so many people
have used it for so many years. We have every single feature they have
minus the spam bots.
I've lost my rankings in Yahoo for the short term by switching back to
non sef urls which will cost me a couple thousand dollars a month
until they get us sorted back out, I just hope this move works to
finally get ranked in Google and that the short term loss pays off in
the long term.
I'm starting to get irritated with being burried by a bunch of penny
ante straight html sites with 2 or 3 pages and a simple cheesy generic
chat script with no features or profile systems that are all part of
the same site which just builds hundreds of crappy "directories" and
mirror sites to link back to themselves.... "cough" (chat avenue) in
Google. I mean these people have no content at all and my trouble
seems to be too much content. How does this coincide with Google's
phillosiphy?
I'm sure it will all work out, I just hope I don't have to eat spam
and romain noodles forever waiting on it.
> I would really, really recommend setting up 301 redirects, even if it
> is for 30,000 pages (there's usually a fast way to do that, somehow,
> somewhere :) ). Besides search engines, users are otherwise going to
> have a tough time finding your content. It's hard to say how long it
> will take for a change like that to propagate completely, but without
> redirects it will certainly take much longer.
It might be a good idea to start a separate thread here about your
site -- perhaps there are some tips and tricks that the users and
Googlers here can give you to improve things! Feel free to post the
URL of the new thread here so that we can find it :-)
I can understand that out of the millions of sites out there that use
dynamic URL's there are a large proportion that get URL rewriting
wrong (or rewrite 'unhelpfully' from Google's point of view) - so that
is where the focus is for this recommendation.
I think the original post should have been written to reflect that
difference, that, if done correctly, rewriting is OK, but for the
large part Google recommends leaving URLs alone. The post came across
largely as a denouncement of rewriting.
Nah I'm done, just wanted to go off on a rant for a bit lol.
I don't know if the url change was the cause of falling out in yahoo
today or just a coincidence durring an update but we're back up there.
I'll just keep waiting on Google and hope the url change helps.
> I can understand that out of the millions of sites out there that use
> dynamic URL's there are a large proportion that get URL rewriting
> wrong (or rewrite 'unhelpfully' from Google's point of view) - so that
> is where the focus is for this recommendation.
> I think the original post should have been written to reflect that
> difference, that, if done correctly, rewriting is OK, but for the
> large part Google recommends leaving URLs alone. The post came across
> largely as a denouncement of rewriting.
From spain,
It has being a lot of work to set the rewrites for every url.
I donīt think users will enjoy this movement, they want to see what
ther are looking for on the url. Actually, most of the sites without
rewrite are just dinamicaly generated content made just for SPAM and
users belive that.
I have being following the Google Rules for many years but.
As my rewrites are not generating duplicated content, and in the pages
they made I set up a robots with disallow
My users want to see the keywords they are looking for on the url
Accesibility and usability demands semantic url
Google is not the only search engine.
Sorry man, Iīm not going to make any change to the url
> Nah I'm done, just wanted to go off on a rant for a bit lol.
> I don't know if the url change was the cause of falling out in yahoo
> today or just a coincidence durring an update but we're back up there.
> I'll just keep waiting on Google and hope the url change helps.
> Thanks for all the help John!
> On Sep 23, 6:52 pm, seriocomic wrote:> Thanks John,
> > This clears things up somewhat.
> > I can understand that out of the millions of sites out there that use
> > dynamic URL's there are a large proportion that get URL rewriting
> > wrong (or rewrite 'unhelpfully' from Google's point of view) - so that
> > is where the focus is for this recommendation.
> > I think the original post should have been written to reflect that
> > difference, that, if done correctly, rewriting is OK, but for the
> > large part Google recommends leaving URLs alone. The post came across
> > largely as a denouncement of rewriting.
I can understand the whole hiding a session id in a folder structure.
But even when they are in a variable, I am constantly finding multiple
listings in the results. I don't believe in session id's in URLs at
all, but I am usually contracted to fix them.
Wouldn't a proper folder structure in a URL give the engine more of a
understanding of content structure. As in all of the information under
the folder "/About-Us/" is information about the company. All
information under "/Our-Services/" is about our services.
This new "Dynamic URL's" are better doesn't match up with the results
I have been receiving from your engine. There seems to be an
alternative motive for this announcement.
I make my URLs Human Friendly and Google seems to prefer that. Is this
a backlash on Wordpress? Is Blogger going to change their URL
structures?
Maybe I am coming into this a bit late and
don't have a total grasp of the thread but I would
like to make one point.
The words used in a URL while they may contribute
slightly to a search result on a page that has nothing
else going for it, I would not depend on it to last.
It is too easy to be outranked by the competition.
IMO, for the most part in and of itself the words in
a URL in general are of little benefit nowadays.
The SE's have become much smarter and complex.
Your efforts would be more productive if used
on more important items like page content
and quality relevant inbound links.
FWIW,
Abracadabra
On Sep 24, 9:23 am, Jeff Louella wrote:
> I can understand the whole hiding a session id in a folder structure.
> But even when they are in a variable, I am constantly finding multiple
> listings in the results. I don't believe in session id's in URLs at
> all, but I am usually contracted to fix them.
> Wouldn't a proper folder structure in a URL give the engine more of a
> understanding of content structure. As in all of the information under
> the folder "/About-Us/" is information about the company. All
> information under "/Our-Services/" is about our services.
> This new "Dynamic URL's" are better doesn't match up with the results
> I have been receiving from your engine. There seems to be an
> alternative motive for this announcement.
> I make my URLs Human Friendly and Google seems to prefer that. Is this
> a backlash on Wordpress? Is Blogger going to change their URL
> structures?
MY two cents worth is that maintaining static URLs is a much easier
task than working with dynamic URLs.
Static URLs are more abstract, and thus more difficult to hack. The
dynamic URLs are more transparent, allowing possible hackers to see
the technology used to build them and thus facilitating attacks.
Hackers will also be looking for dynamic urls to attack with Cross
Site Scripting. Any web page which passes parameters to a database can
be vulnerable to this hacking technique.
|
