Are these two "bugs" related?

302 "hijacking thing" and 1969 cache bug

[Receptional]

The 302 "hijacking thing" is discussed here:
[webmasterworld.com...]

The 1969 Unix "Bug" that affects Google's caching dates is discussed here:
[webmasterworld.com...]

Now... How does this hypothesis stand up?

The 302 hijacking really never used to work well as Google tended to index the first site it found in prefernece to the 302 unless the 302 hijacking URL had a large "importance" - which was rare. So the issues were rare.

But now.... pages cached with an apparant date of 31st Dec 1969 that have error 302s are seen as older and one bug may have amplified another - so that many more 302 "hijacks" are starting to appear - whether intentionally or not.

The result - Unix sites with no cache dates and 302 redirects are replacing the originals in the Google index.

What do you reckon - fact or flawed logic?

[DerekH]

Flawed logic

Well, I shouldn't be so brusque, but I don't think that that's the whole story...

I modified two of my sites extensively by changing their page templates. Virtually every page in the two sites went to either
URL only (we know you're there but you're not indexed) or
a Cache date of 1969

And there were no header return codes changed in any of this move...
DerekH

[dazzlindonna]

Derek, I don't see how that example represents flawed logic of the first post. Ok, 302's don't cause the 1969 problem, which your example clearly shows. However, the 1969 problem *could* be making the 302 problem worse than it used to be, according to receptional's theory. So, while the two problems are two separate issues, one of the problems could be making the other worse than before. It makes sense to me. It may or may not be true, but it does seem logical. Perhaps, then, if the 1969 problem is resolved, the 302 problem would lessen, if not go away. Does anyone know if the start of the 1969 problem coincided with a worsening of the 302 problem?

[Spine]

They seemed to become hot topics here at about the same time.

Google has to find a way to deal with this 302 crap.

[DerekH]

dazzlindonna
I do understand where you're coming from.
You're right to ask if the 302s make the 1969 problem worse, but my own experience is that the 1969 problem got worse all by itself (as the example I quoted you showed).

My feeling is that the 1969 problem become so immense (I'm aware of three sites of mine that all became trapped in a time-warp, none because of header problems) that it's going to be very hard to work out which other things have helped precipitate this.

But you're right, you shouldn't stop investigating.
DerekH

[dazzlindonna]

I think you are misunderstanding us Derek. Flip what you said around and you'll see. Instead of "the 302s make the 1969 problem worse", try "the 1969s make the 302 problem worse".

[DerekH]

dazzlindonna said

Flip what you said around and you'll see

!say you what see I

(I'll leave you to work out my warped sense of humour...)

Point taken, though - yes the presence of poorly performing, poorly cached pages on otherwise authoratative sites does indeed make it easy for others to hijack such pages. I'm not even sure whether it's simply that the cached page is apparently so old - in my experience, they ranked particularly awfully as well, just by being "supplemental"

Sorry to have wasted most of the first page of the discussion of this with me going off at a tangent!
DerekH

[DaveAtIFG]

claus's message #8 at [webmasterworld.com...] highlights problems with redirects since the summer of '03.

The "1969 cache bug" was first reported while the [webmasterworld.com...] thread was running, fall '04.

The "cache bug" suggests to me that Google recently began saving a cache date for every page, if no date is on file, the Linux default date is displayed. They may have collected cache dates for a limited set of pages prior to the "cache bug," perhaps it was limited to higher PR pages or...

The timing of the "cache bug" appearance and things I observed while testing redirect handling suggest that the "cache bug" is part of a fix for redirect problems. I expect it to disappear as Google full populates its database with cache dates.

The diminishing volume of fresh "I've been hijacked!" posts, here and elsewhere, suggests that Google has made some progress, and perhaps resolved most redirect problems.

Pre-Florida, during the monthly "Google dance" era, changes like this "cache bug" would be fully implemented in one monthly cycle. Today's Google seems to need 2-3 months, sometimes longer, for a major changes to percolate through the entire database. With today's Google, I think we need to watch trends in SERPs and complaints to keep up with them.

[walkman]

"The diminishing volume of fresh "I've been hijacked!" posts, here and elsewhere, suggests that Google has made some progress, and perhaps resolved most redirect problems"

or maybe they're seeing the already open threads. A few sites I'm following are still index with the 302 URL. I agree with you on the "google dance" thingy. Now it takes months and old pages hang in forever. I wonder if pages hit by the bug will suffer because of a preset (time wise) penalty even after Google fixes that.

[walkman]

self deleted. dupe

[edited by: walkman at 8:57 pm (utc) on Jan. 8, 2005]

[walkman]

"The result - Unix sites with no cache dates and 302 redirects are replacing the originals in the Google index. "

The problem I think is that Google now penalizes both pages, before it just ignored one. I have also noted that the rankings get worst as more time progresses with the dupe problem. You keep slipping and Googlebot will visit you less and less. Eventually I assume you'll lose pages from the index. This could explain why many are reporting URL only listings.

[Spine]

I'm seeing this first hand now.

I had worked in December to remove some forum pages google saw as dupes, and a 404 redirect I had that google foolishly thought was duplicate content also.

I saw my rankings come back, and all was well until..
This week I noticed my index lost it's fresh tag, less spidering, and no www in front of my domain name.

I ran a site: check and saw a directory that has me listed with a redirect. The site has had this link to me for almost a year, without seeing it show up like this with a site: command. I've contacted the site owner to have myself removed of course.
I had been checking regularly to make sure none of the false duplicates I removed had come back, what a nasty surprise this was.

Google changed something in late summer or early fall, otherwise these kinds of problems would have plagued my site for the past 4 years.

If my rankings slip to where they were from September to December because of this, I will be seeing red, and possibly go psycho.

[zeus]

Spine - no need to go psycho - I also lost all my visits from Google, because of there bugs, its said to see a SE go so quickly down.

Spine wait for MSN and everything will be good again, Google would NEVER could compite with MSN when it is in the new loghorn.

As I have said before this problem will get bigger and bigger.

[walkman]

"I saw my rankings come back, and all was well until..
This week I noticed my index lost it's fresh tag, less spidering, and no www in front of my domain name. "

I have the same thing ..minus the SERPS. Less spidering and no fresh tags even though all the 302s have been removed, at least until new ones start to show up. It keeps getting worst as time goes on and I looked at another DCs
216.239.53.99
66.102.7.99
216.239.53.104
66.102.7.104
66.102.7.105
66.102.7.147
that are supposedly holding the new SERPS...I'm not even in the 500. Very sad.

[DMan]

Please see my posting #37 here:
[forums.searchenginewatch.com...]

I would not be surprised at all if the two bugs were related.