Re: Claim by 'Stevo' - "CEOP show Maddie is missing on 30th April 2007"

by Guest 21.06.15 21:48

HKP wrote:@BB, that retrieval / folder allocation doesn't make sense (to me anyways)

It's from the redirected one.

They all have a dynamic text block added at the end if you "view source".

by **Richard D. Hall** 21.06.15 21:49

Where is the evidence that other files in that folder were created after 30 April?

by Guest 21.06.15 21:52

Richard D. Hall wrote:It doesn't prove it is derived from the folder, it could be derived from within the file. But that could be the case that. Where is the evidence that other files in that folder were created after 30 April?

It doesn't.

But you can't claim it's in the file either.

Do you not think that the fact that the October homepage ended up in april means the WBM is screwed?

You have to admit something is wrong with it regardless of what you think about the other pages.

I completely get why you want this to be so, so do I. However, reality has kicked in for me.

by **Richard D. Hall** 21.06.15 21:56

Are we seriously saying that WBM server clock was incorrect ? - That the clock read 30 April, when it should have been a few months or weeks later??? Normally if clocks do reset they are Decades wrong. Also the integrity of server clocks is high. Because there system is pulling date from the clock all the time and is "bread and butter" so to speak, I would be very surprised if it was a clock issue.

by **Richard D. Hall** 21.06.15 21:57

Can you show me the October home page in the 30 April folder?

by **Tony Bennett** 21.06.15 21:58

BlueBag wrote:
Richard D. Hall wrote:It doesn't prove it is derived from the folder, it could be derived from within the file. But that could be the case that. Where is the evidence that other files in that folder were created after 30 April?
It doesn't.

But you can't claim it's in the file either.

Do you not think that the fact that the October homepage ended up in april means the WBM is screwed?

You have to admit something is wrong with it regardless of what you think about the other pages.

I completely get why you want this to be so, so do I. However, reality has kicked in for me.

@ BlueBag @ Richard D. Hall

In another place, WLBTS has issued this challenge, in case you want to comment:

I'll just add that the claims that Richard D. Hall is making now - that the mccann page must have been archived on 30th April 2007 because that date is in the comment at the bottom of the file - demonstrate a total lack of understanding on this subject.

It should be quite clear to anybody with technical expertise that the comment at the bottom is most likely added at retrieval time, along with the Wayback Toolbar UI. The retrieval date and time is also part of that comment!

It beggars belief that Mr Hall could possibly think this reflects the state of the archived page in the hard store. Does the WBM system update the archived page with the new comment every time it is retrieved? Why on earth would it do that?

If the 30th April archive hadn't have been removed, this would have been easy to demonstrate with those later articles, which I am certain would have had the 30th April archive date in the comment - something I can't prove at this time because of their removal.

by Guest 21.06.15 22:00

I agree, the comment is dynamic content added at retrieval time.

by **Richard D. Hall** 21.06.15 22:04

Yes I am happy to accept the timestamp could be added at retrieval time and is derived from the folder name. The folder being created on 30 April for files found on 30 April. I am asking now just to see files in that 30 April folder that were created later. Can you point me to them? That file mccann.html as I said has no references to future dates as far as I can see.

by **Richard D. Hall** 21.06.15 22:05

Can you point me to the October home page and show it was in the 30 April folder. Cheers.

by HKP 21.06.15 22:06

@BB so if it is at retrieval time the comments are useless?

by rustyjames 21.06.15 22:09

Tony Bennett wrote:
STATEMENT

Look at the above entries. All except the top one (30 April) are of a news event (7 June 07) indexed after the news event.

A news event cannot possibly be indexed before the news event itself.

Therefore it is obvious that there must have been an error or glitch in the system

It looks that way, but I'm not claiming to be an expert on web archiving and I'd never make such a statement unless I was 100% sure; however, it mirrors what BlueBag was reporting. The slight difference is the CDX info I was extracting is one level below the presentation of the page on the Wayback machine - i.e. it is the information Wayback uses in order to efficiently get the relevant bits of information in order to recreate, ("replay"), the page.

The CDX is still just an index though created from the source (w)arc files. Those files are huge captures of the content the crawling has found, but they are random and the info for a given site or date could be spread over lots of them on different servers. The index provides an efficient way of finding the bits that are of interest just like with a book -- it takes time to create it, but once you have it then you don't have to trawl through everything to find what you want.

So if the index is wrong, it still says nothing about the source data, and to add to the difficulty, if WBM are trying to correct the "issue" then who knows what state the index is in.

by **Richard D. Hall** 21.06.15 22:09

Why is it useless just because it is copied from the folder name? - Of course it is not useless, it has been used in court cases.

IF there genuinely were files created by CEOP AFTER 30 April, which were sitting in that particular folder on the WBM server, I would agree there is a bug somewhere. I am just asking for you to point me to the file(s) that were in that folder which were created after 30 April. I am not saying you are wrong. I want to verify what is being said.

by Guest 21.06.15 22:14

Richard D. Hall wrote:Can you point me to the October home page and show it was in the 30 April folder. Cheers.

I asked for the same thing earlier today and member LG provided it (thanks LG).

That's what did it for me.

It's even discussed right at the beginning of this thread where more astute people already spotted it.

I wish I had known earlier.

[You must be registered and logged in to see this image.]

by **Richard D. Hall** 21.06.15 22:17

Sorry I cannot read that. Can you send a higher definition image. Thanks.

by rustyjames 21.06.15 22:17

Richard D. Hall wrote:Why is it useless just because it is copied from the folder name? - Of course it is not useless, it has been used in court cases.

IF there genuinely were files created by CEOP AFTER 30 April, which were sitting in that particular folder on the WBM server, I would agree there is a bug somewhere. I am just asking for you to point me to the file(s) that were in that folder which were created after 30 April. I am not saying you are wrong. I want to verify what is being said.

There aren't "folders" as such. The web crawling creates huge arc/warc files containing the crawled data. The data is indexed into a database creating what are reffered to as CDX records. When you request a page those CDX records are used to efficiently retrieve the information from the archives to "replay" the page. WBM adds stuff to the page whilst it does this, and rewrites the URLs contained within it, (e.g. images), to point into the web archive rather than to their original location.

Hope that's clearer.

by **Richard D. Hall** 21.06.15 22:19

Do we not need to see the source code of what is held on their server rather than a screen shot. To see what dynamic content is in the page?

by **Richard D. Hall** 21.06.15 22:21

rustyjames wrote:
Richard D. Hall wrote:Why is it useless just because it is copied from the folder name? - Of course it is not useless, it has been used in court cases.

IF there genuinely were files created by CEOP AFTER 30 April, which were sitting in that particular folder on the WBM server, I would agree there is a bug somewhere. I am just asking for you to point me to the file(s) that were in that folder which were created after 30 April. I am not saying you are wrong. I want to verify what is being said.

There aren't "folders" as such. The web crawling creates huge arc/warc files containing the crawled data. The data is indexed into a database creating what are reffered to as CDX records. When you request a page those CDX records are used to efficiently retrieve the information from the archives to "replay" the page. WBM adds stuff to the page whilst it does this, and rewrites the URLs contained within it, (e.g. images), to point into the web archive rather than to their original location.

Hope that's clearer.

Thanks

by Guest 21.06.15 22:21

Richard D. Hall wrote:Sorry I cannot read that. Can you send a higher definition image. Thanks.

That's what I said as well initially.

You can if you try if you know what to look for.

The date 30 Apr can be seen large in the tool bar at the top.

That is definitely the 2nd October Latest News further down.

Sorry Richard but it that was it as far as I was concerned. The WBM evidently did something wrong.

by **Richard D. Hall** 21.06.15 22:21

BlueBag wrote:
Richard D. Hall wrote:Sorry I cannot read that. Can you send a higher definition image. Thanks.
That's what I said as well initially.

You can if you try if you know what to look for.

The date 30 Apr can be seen large in the tool bar at the top.

That is definitely the 2nd October Latest News further down.

Sorry Richard but it that was it as far as I was concerned. The WBM evidently did something wrong.

The October date could be dynamic content, no?

by Guest 21.06.15 22:22

Richard D. Hall wrote:Do we not need to see the source code of what is held on their server rather than a screen shot. To see what dynamic content is in the page?

They deleted it that's the point.

But there are screenshots right at the start of this thread before serious questions were asked of Chris Butler.

by Guest 21.06.15 22:24

Richard D. Hall wrote:
BlueBag wrote:
Richard D. Hall wrote:Sorry I cannot read that. Can you send a higher definition image. Thanks.
That's what I said as well initially.

You can if you try if you know what to look for.

The date 30 Apr can be seen large in the tool bar at the top.

That is definitely the 2nd October Latest News further down.

Sorry Richard but it that was it as far as I was concerned. The WBM evidently did something wrong.

The October date could be dynamic content, no?

No it's actually part of the archived page.

You can check Octobers real archived pages and see it.

by **Richard D. Hall** 21.06.15 22:24

That screenshot is not proof this page was created after 30 April. The code in the page could be just pointing to another URL get the "October" text.

by rustyjames 21.06.15 22:26

Richard D. Hall wrote:Do we not need to see the source code of what is held on their server rather than a screen shot. To see what dynamic content is in the page?

Yes and that is what is stored in the w(arc) files as per my previous post a couple of minutes back, and they also contain lots of metadata such as timestamps and ip addresses etc. The problem is the index is the only view onto it.

by **Richard D. Hall** 21.06.15 22:26

Yes, can you send me a screeshot of the source code? cheers

by Guest 21.06.15 22:27

Richard D. Hall wrote:That screenshot is not proof this page was created after 30 April. The code in the page could be just pointing to another URL get the "October" text.

No... if you open a page with a redirect, the redirect date is in the toolbar at the top.

I've been playing with WBM for hours.

This is an October page in an April folder.