Are you a Google Analytics enthusiast?
More SEO Content
Google Broken Article
Posted 12 September 2004 - 12:45 AM
Posted 12 September 2004 - 05:09 AM
I find it all quite funny really
I read about this way over a year ago, initially it made me go hmmmmm, but then I just thought whatever.
I don't care one way or another, all I do know is that the pagination increase for client searches seems to be climbing, so new pages are coming from somewhere. Sure they might be newly optimised pages that were already in the index but are now showing up.
OTOH, I have not experienced a problem getting new sites indexed in Google, so if my new pages result in some other buggers pages getting removed, then so be it. I really don't care about nameless pages in a far away land.
Posted 12 September 2004 - 09:02 AM
I'm not convinced either way though, and probably fit into the "So what?" group.
The fact is all those pages don't mean much to me. I only compete for the top ten positions anyway. and couldn't care less about how many thousands of pages are below me for a given term. That sounds a little self-centered perhaps, but business is business. If they DID have to remove pages to make room for new, common sense tells me they wouldn't cull top ranking pages, just the dead wood. If I have a page that gets dropped because it's 4,285,199,775, I'd consider it "justifiable homicide", and be thankful I'm no longer having to battle for position 4,285,199,774.
I think most here would agree that if you do your job correctly there is no problem in getting your pages above the 4,285,199,774 mark, be it a real, or perceived limit.
Posted 12 September 2004 - 12:24 PM
Posted 12 September 2004 - 03:50 PM
Another thing to consider is the discussion of Y2K. The ease of crossing boundaries is directly proportional to the skill and insight of the programmer. If I wrote a program it 1980 that stored 4-digit years and only showed 2 digits then the "upgrade" would only take a day and I'd be a hero. If I hadn't the foresight to plan for it then it would take 2 years and I'd be a rich jerk.
I'd be willing to give the PhD's the benefit of the doubt that they would expect it to get big one day and use a #define ID=4 so when the day came to update it they could just flip the switch, make a couple of code changes and roll with it. It's not that mysterious that one day G would have more than 4 billion pages - just look at McDonalds - they used to say how many were served too.
Bottom line, IMO? Google is made up of pros even if they make the occasional bad decision. It's hard to believe that they would be stuck at 4bil for a couple of years without changing it in anticipation of future growth. Furthermore, even if they are stuck I'm sure there's a bunch of trash that needs cleaning out. For example, I really wonder the relevance of the page at www.radix.net/~dglenn/words/glenn-bio.html which is entitled "Glenn's alt.transgendered Bio" and says at the top:
Posted 14 September 2004 - 02:30 PM
I disagree completely with this statement. It is easy in hindsight to call all the Y2K authors just plain dumb. But these were the top programmers of the day and probably still are. They were probably better programmers than most today because they were trained in programming unlike the vast majority of web programmers who think they know all there is to know about programming because they can put up a db driven website.
It is also easy to assume something that looks trivial is trivial to change. Anyone that has worked on large scale programming knows that changing something that is used everywhere in the application, like a date or a docid, is not a flip of the switch and is not a trivial change. Yes it looks easy to plan for some changes, and it is the changes you dont plan for that are the hardest. But even with planned changes, there is always more to it than you planned for. After 4 years and 100+ master chefs that all know the ONE right way, the code you started with is NOTHING like what you originally envisioned.
Remember what they say about assuming: when you assume you pull an emu out of your ass .
Posted 14 September 2004 - 02:42 PM
The concensus appears to be that NO new sites, regardless of their page rank or inbound linking, are showing in SERPS. What is not agreed upon is whether old sites with new content are also seeing the same effect.
Has anyone on this board put up a new site in 2004 and had it show up in SERPs?
Posted 14 September 2004 - 02:46 PM
Posted 14 September 2004 - 03:11 PM
Posted 14 September 2004 - 04:19 PM
Sensory Support Services Bristol
The site is now in Google, and so everyone can be certain there is no scullduggery, here is a link showing when the domain was registered.
So can we now agree that new pages are going into Google?
P.S. please don't comment on how the site is a bag o bones, I didn't buil it, and I am well aware of the accessability problems it has, this is one of the areas I am helping them with.
Posted 14 September 2004 - 04:39 PM
Posted 14 September 2004 - 05:39 PM
That's the first time I've heard that said. I think I've put up at least 3 brand new sites this year - no problems with Google indexing them at all.
Posted 14 September 2004 - 05:59 PM
The question wasnt whether new sites get indexed by google. The question is whether they show up in the SERPs. That seems to be the gist of the issue.
Does this show up for any keywords?
Posted 14 September 2004 - 06:51 PM
It is trivial with a capital TRIV. There are bazillions of ways to code. IF, and I am making an assumption here, IF Google used a database, it is as trivial as changing the table schema. "SELECT * FROM all_urls WHERE word='dog'" will always work whether or not the primary key is 8 bits, 16 or 33. The logic is on the database side, not the coding side.
Now this is, with all due respect, just totally wrong.
The social issues of progarmming, the version control, the standards and the scalability (especially object oriented methodologies) are all light years ahead of the 1960s. We also are not talking web programmers, but well trained, knowledgable programmers operating in C.
I am 99.9% sure that Google employ great coding standards, using a modular design that is either object oriented or built around functions. this would mean that, even if there is no database and all access to data is written from scratch in C, will require changes to be a dozen some functions / classes, not thousands of lines of code. That could be as easy to fix as a recompile, in virtually real toime on one server @ a time.
As a real world example of how fast Google can fix a known issue, way back when, Google had a problem displaying right-left language snippets for languages like Hebrew and Arabic, http://www.webmaster...orum3/14110.htm. In this case, they recompiled the display code in less than a day. Problem fixed. There is no reason to believe that a 4 byte primary key / index would be that much harder to fix, if the code is well written and modular.
As for the Y2K problem, that was a very, very different problem. We are talking about legacy systems built circa 1960-70 vs Google, a product of the late 1990s. The problems of Y2K, that of legacy systems designed in an area of Million dollar computers, $400 storage PER BYTE and miniscule processing speeds (by todays standards), are not relevant to Google. What's more, the advancements in programming and the ability to use object oriented methodologies more widely should lessen the time required to change such a design decision. this is really just speculation of the highest order by both sides, so really not evidence suppoirting anyone.
IMHO, and I don't think you are a conspiracy theorist dmart, the pro 4 billion max side has very little evidence that is current to support it, and the burden of proof lies with those making a claim. The hypothesis is @ best unproven and not discreditted, and in any case not really an issue worth worrying about. There is nothing any of us can do to change the situation, and as such it simply is not worth worrying about, AFAI am concerned.
Posted 15 September 2004 - 11:52 PM
With all due respect, you have completely ignored the memory and hard disks requirements of the docID in the inverted index. They are incredibly massive. No one would use a 5-byte docID in 1998-1999, when the core Google software was written, because it would involve a massive amount of extray memory/hard disk space in the inverted index. In 1999 Google had less than 1 billion documents indexed. Even with a 4-byte docID, there were a lot of zeros doing nothing in the inverted index. However, all the C and C++ libraries are efficient at 4 bytes, because it's a 32-bit processor in the boxes that Google uses, and all the standard libraries assume a 32-byt (4-byte) integer.
No, Google doesn't use SQL. Don't make me laugh. This is not a trivial issue.
And don't give me this stuff about +the.
Try allinurl with the
Try allinanchor with the
Try allintext with the
They all give the same result. Incredible coincidence! It's bogus.
And Jill, I fully understand your point about how this makes no difference to the average user. But it should make a difference to the average Google shareholder, and it should make a difference to SEOs, who can be expected to take the long view of Google's prospects.
If someone told me in April 2003 that update Cassandra, when Google had to throw out an entire crawl and revert to an earlier index, was something that would continue in one form or another for the next 17 months, I would have said that Google is better than that.
If someone told me in August 2003, when the Supplemental Index kicked in, that this was a permanent innovation that makes their results better, I would have said that Google is more sophisticated than that. Yet the Supplemental Results continue today, and they don't do a thing for anyone. The key to success in Google is how to get out of the Supplemental Index.
If someone told me in November 2003, when the Florida filter hit, that Google would crank back the knob but still not turn it off as of 10 months later, I would have said that Google is smarter than that.
If someone told me a year ago that the URL-only listings would be even worse 12 months down the road, I would have said that they're a conspiracy theorist and Google will get their act together long before then.
But here we are. The SEOs are raking in the bucks the faster Google goes downhill. You don't have to wear a tin foil hat to see the connection here. The SEOs have a vested interest in covering this up. Otherwise, you have to conclude that many of them are just not programmers, and have no idea what they're talking about.
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users