CEOP show Maddie is missing on 30th April 2007

It's more likely (IMO) that the Wayback crawler just hoovers up everything on an unsecured server, linked or not.  Or at least, it did in 2007, because the WWW was a hell of a lot smaller then, and things weren't as sophisticated then as they are now.  Hackers, phishers and all sorts of cyber criminals are a lot more educated and clever than they were back then, and the mechanisms to hide and protect things were a lot more primitive too.  I should imagine it works in a completely different way now, like any other software that has evolved since 2007.

Of course the whistleblower theory is perfectly possible as well, and cannot be ruled out.  Did somebody really have the foresight to put a hyperlink in before 3 May?  Or did some junior, enthusiastic developer put it in thinking they were helping, and were then told to take it back off again?  And in the meantime, the crawler had grabbed it?  It's a very slim chance, but again, not impossible.

The only safe method, IMO, of keeping a webpage from being hidden from crawlers of any description, is to keep it on your local machine and don't put it on a server connected to the internet, until you are ready for people to see it.

After trying to read and superficially understand every argument in here, I don't get the argument WLBTS used to make it so highly probable that the date of crawl would be wrong.
Why take a look at the homepage on 30/04, and then at random captures to check for anomalies in the dates of the contents showing newslinks? Why not check other sites captured at 30/04 and then see if there's an anomaly anywhere.
Because, if all the other captures would only show contents from an earlier date than crawled, it would be strangely coincidental to see just the McCann-page being different.
And likewise, if you see the 30/04 of 2007 full of later newsfacts, then you would know the crawler was wrong. I think. Dunno.

In other words, strange logic has been used to try and prove the crawl-date would so highly probable be wrong.

If Wayback Machine fills in the various news headlines with dynamic content at the point of retrieval, then please explain how every other version of the page except for the 30th April 2007 does not. Every other page fills the news headlines with news that existed at the point of the crawl. There is clearly no 'insertion of dynamic content' for any other page.

Find me *one single page* that does that.

If there is insertion of dynamic content at the point of the retrieval, why then do you not get the same (current) news headlines when you look at a different version of the page? Or does the 30th April 2007 have code to retrieve dynamic content, while the same page on every other date does not?

[img][/img]

This a code btw? Just thought of saving it when I saw it go by on twitter a couple of days ago.

Burst wrote:After trying to read and superficially understand every argument in here, I don't get the argument WLBTS used to make it so highly probable that the date of crawl would be wrong.
Why take a look at the homepage on 30/04, and then at random captures to check for anomalies in the dates of the contents showing newslinks? Why not check other sites captured at 30/04 and then see if there's an anomaly anywhere.
Because, if all the other captures would only show contents from an earlier date than crawled, it would be strangely coincidental to see just the McCann-page being different.
And likewise, if you see the 30/04 of 2007 full of later newsfacts, then you would know the crawler was wrong. I think. Dunno.

In other words, strange logic has been used to try and prove the crawl-date would so highly probable be wrong.

Exactly.

If every other version of the home-page shows news that was prior to the crawl date, then it is 'strangely coincidental' to see news from the future on the 30th April 2007 version of the same page, and that page only.  It is exceptional.

And as I've pointed about previously, it demonstrates that Wayback Machine does not insert dynamic content for that page *ever*, except that one occasion on the 30th April 2007. It makes a few changes at the point of retrieval, but only to the code sent back in the HTTP response, not the stored file. It adds javascript for the Wayback toolbar, and a comment at the bottom detailing archive and retrieval date. It also changes some links so that they reference cached versions rather than live, but doesn't do this for all.

There is no evidence of future news being inserted excepted for that one date only, at all other times the news headlines would have existed at the time of the original crawl.

WLBTS wrote:

Burst wrote:After trying to read and superficially understand every argument in here, I don't get the argument WLBTS used to make it so highly probable that the date of crawl would be wrong.
Why take a look at the homepage on 30/04, and then at random captures to check for anomalies in the dates of the contents showing newslinks? Why not check other sites captured at 30/04 and then see if there's an anomaly anywhere.
Because, if all the other captures would only show contents from an earlier date than crawled, it would be strangely coincidental to see just the McCann-page being different.
And likewise, if you see the 30/04 of 2007 full of later newsfacts, then you would know the crawler was wrong. I think. Dunno.

In other words, strange logic has been used to try and prove the crawl-date would so highly probable be wrong.

Exactly.

If every other version of the home-page shows news that was prior to the crawl date, then it is 'strangely coincidental' to see news from the future on the 30th April 2007 version of the same page, and that page only.  It is exceptional.

And as I've pointed about previously, it demonstrates that Wayback Machine does not insert dynamic content for that page *ever*, except that one occasion on the 30th April 2007.

Outside the homepage, I mean. If the thing crawled with wrong timestamps writing away as the 30th, you'd see anomalies in other places than CEOP's. If that's not the case, it's too coincidental for the McCann-page to be written as if crawled on a wrong date. Then the crawler would probably not have been wrong.

Burst wrote:
Outside the homepage, I mean. If the thing crawled with wrong timestamps writing away as the 30th, you'd see anomalies in other places than CEOP's. If that's not the case, it's too coincidental for the McCann-page to be written as if crawled on a wrong date. Than the crawler would probably not have been wrong.

Normally I would agree, but the home page of 30th April 2007 has news from the future on it which isn't possible.

We don't know that other pages haven't suffered from the same issue. To prove that point you would need to go through all of Wayback's archives for every URL.

WLBTS wrote:

Burst wrote:
Outside the homepage, I mean. If the thing crawled with wrong timestamps writing away as the 30th, you'd see anomalies in other places than CEOP's. If that's not the case, it's too coincidental for the McCann-page to be written as if crawled on a wrong date. Than the crawler would probably not have been wrong.

Normally I would agree, but the home page of 30th April 2007 has news from the future on it which isn't possible.

We don't know that other pages haven't suffered from the same issue.  To prove that point you would need to go through all of Wayback's archives for every URL.

We don't know, indeed. I'm not capable and too lazy to prove such a point.

Yes Burst, the green bit is automatically generated code.

And the timer certainly seemed to be working just fine for a lot of other sites on 30 April 2007. I know, because I have been checking.

@Resistor Ah. Ty.

Sorry if this is not relevant. Was checking various urls on WBM and looking at ACPO there are crawls done on 5 Apr 07 and 9 May 07. Once on the page there is a link to latest news ( as opposed to this being written on the page) click 5/4 no mention of M obviously. Click 9/5 and there is. But both latest news sections have items dated after the crawl date. Does this mean that dynamic content has been entered even though the capture date is prior?
No requirement for answer if this not what you are looking for.

chirpyinsect wrote:Sorry if this is not relevant. Was checking various urls on WBM and looking at ACPO there are crawls done on 5 Apr 07 and 9 May 07. Once on the page there is a link to latest news ( as opposed to this being written on the page) click 5/4 no mention of M obviously. Click 9/5 and there is. But both latest news sections have items dated after the crawl date. Does this mean that dynamic content has been entered even though the capture date is prior?
No requirement for answer if this not what you are looking for.

Can you copy the URLs here please so that I can check?

I haven't read many of the posts on this thread as I'm a long way behind so it may have already been mentioned, and I apologise if it has, but wasn't Monday 30th April the day that was "glossed over" by the McCanns and Tapas friends that nothing much happened that day according to KM's diary and all of their statements.  It seems a coincidence that it is the same day that the Wayback machine has recorded Madeleine as "missing" on the CEOP website.

ETA: It was also the day that Robert Murat flew out to Portugal in the very early hours, so had something happened to Madeleine on the Sunday and everyone was brought in to "help", regardless of whether they knew (i.e. Murat for translation services) what they were letting themselves in for.

It was also the day that MBM was collected from creche by KM at 3.25pm after only being there 10 mins. (According to the creche sheets in the files).

WLBTS wrote:

chirpyinsect wrote:Sorry if this is not relevant. Was checking various urls on WBM and looking at ACPO there are crawls done on 5 Apr 07 and 9 May 07. Once on the page there is a link to latest news ( as opposed to this being written on the page) click 5/4 no mention of M obviously. Click 9/5 and there is. But both latest news sections have items dated after the crawl date. Does this mean that dynamic content has been entered even though the capture date is prior?
No requirement for answer if this not what you are looking for.

Can you copy the URLs here please so that I can check?

Looking at it again. When the black date box at the top states 5 apr then click latest news it shows news from 20 apr.and the blackbit remains on 5 apr
 I may be wrong about 9 May as now I see that when I access that date and click latest news the black bit jumps to june4.

www.acpo.police.uk

A very, very interesting thread.
I don't pretend to be clever enough to grasp everything you are all saying but the bits I can understand are fascinating!

chirpyinsect wrote:
Looking at it again. When the black date box at the top states 5 apr then click latest news it shows news from 20 apr.and the blackbit remains on 5 apr
 I may be wrong about 9 May as now I see that when I access that date and click latest news the black bit jumps to june4.

www.acpo.police.uk

The actual news headlines that you go to when clicking those links are about news that happened prior to the crawl date of the home page, but the article page shows a crawl date of later on.  All that really indicates is that the articles were crawled later on.

For example, this page - link - has a crawl date of October 4th 2007, but the news article is dated May 4th 2007.

What would be impossible is for the article to have a crawl date of May 4th 2007, but contain news concerning October 4th 2007, as that would indicate the ability to accurately predict the future.

Resistor wrote:

Thanks for this example of a page where all the news headlines are prior to the crawl date, not news from the future indicating some kind of dynamic content at retrieval. The Telegraph also no good at predicting the future it would seem.

More in the 'can organisations predict the future' conversation:

BBC news home page

Have a look through at a range of dates, satisfy yourself that even the BBC's main news page displays news headlines *in the past*, not *from the future*.  Also the market data is from the time the page was crawled, not the future.

So far, the only evidence of future predicting I've seen so far is the CEOP home page with the bogus 30 April 2007 time-stamp.

WLBTS wrote:

chirpyinsect wrote:
Looking at it again. When the black date box at the top states 5 apr then click latest news it shows news from 20 apr.and the blackbit remains on 5 apr
 I may be wrong about 9 May as now I see that when I access that date and click latest news the black bit jumps to june4.

www.acpo.police.uk

The actual news headlines that you go to when clicking those links are about news that happened prior to the crawl date of the home page, but the article page shows a crawl date of later on.  All that really indicates is that the articles were crawled later on.

For example, this page - link - has a crawl date of October 4th 2007, but the news article is dated May 4th 2007.

What would be impossible is for the article to have a crawl date of May 4th 2007, but contain news concerning October 4th 2007, as that would indicate the ability to accurately predict the future.

So if the crawl date says 5apr the latest news section cannot contain news from after that date? Do I understand correctly?

chirpyinsect wrote:
So if the crawl date says 5apr the latest news section cannot contain news from after that date? Do I understand correctly?

It's no more possible than if you went on to the BBC news page today and found news headlines from 2017.

Does the WBM archive the WBM ?
Im a member of DENSA

Bubblewrapped wrote:Does the WBM archive the WBM ?
Im a member of DENSA

I just tried that. Might have done it wrong though.

Bubblewrapped wrote:Does the WBM archive the WBM ?
Im a member of DENSA

Ha, best (and most intelligent) question ever - I doubt it, that would create infinite recursion which would break WBM's servers

WLBTS wrote:

chirpyinsect wrote:
So if the crawl date says 5apr the latest news section cannot contain news from after that date? Do I understand correctly?

It's no more possible than if you went on to the BBC news page today and found news headlines from 2017.

So how is this possible? The black highlighted section at the top right says 5 Apr 07


Wayback Machine



FEB Apr JUN
Previous capture 5 Next capture
2006 2007 2008

94 captures
26 Oct 05 - 20 Jan 11

sparklines



Close

Help



association title

home about us news policies national policing links contact us


of Chief Police Officers welcome

the voice of the police service

news 2007 - Quarter 2

frame frame

Latest news Press releases
angle


frame

arrow 2007 - Quarter 2
arrow 2007 - Quarter 1
arrow 2006 - Quarter 4
arrow 2006 - Quarter 3
arrow 2006 - Quarter 2
arrow 2006 - Quarter 1
arrow 2005 - Quarter 4
arrow 2005 - Quarter 3
arrow 2005 - Quarter 2
arrow 2005 - Quarter 1
arrow 2004 - Quarter 4






Latest news ACPO Press Office
angle


frame
10, Victoria Street, London SW1H 0NN
arrow Director of Communications
Hannah Gardiner
020 7084 8945
arrow Press Officers
Eve Shuttleworth
Lisa Vasco
Lou Bloomfield
020 7084 8946
020 7084 8947
020 7084 8948
arrow Map/Directions



frame policemen frame
April 20, 2007
ACPO RESPONSE TO THE ACMD REPORT
In response to the ACMD report on date rape, John Yates, ACPO lead on Rape and Assistant Commissioner of the Metropolitan Police, said:

"This report confirms the findings of the recently published report contained within Operation Matisse a j...
arrow Full story

March 27, 2007
POLICING REVIEW
Responding to the Government’s Crime and Justice Policy Review, ACPO President Ken Jones said:

“ACPO believes that the time has come for a wider debate on the future role and functions of the police service. The content of this welcome review...
arrow Full story

March 23, 2007
ACTION PLAN TO TACKLE HUMAN TRAFFICKING