Talk:Windows thumbnail cache

Article completeness
I expanded this article a bit and took off the stub since pretty much all the information needed is there.

68.237.128.69 / Jard Dooku / Jard Yan Dooku 01:50, 16 April 2006 (UTC)


 * What about the format that it uses? Or what it scales the picture to? What is the process of encoding this file? In general, it could be helpful to be more technical. —Preceding unsigned comment added by 69.129.55.57 (talk • contribs) 18:08, 8 July 2007


 * Having just checked on a screengrab, it's 96x96 (or 94x94 for what you see, as the images are given a 1-pixel border on display, however it would be strange to save them to a size not a multiple of 8), or for folder thumbnails, the first four images in the folder at 40x40 (with a 1-pixel divider, so 81x81 overall)... or of course, for non-square images, 96 on the longest edge and the shorter one sized proportionally. I'll put in a simpler version of this on the article. 77.102.101.220 (talk) 00:42, 21 July 2008 (UTC)

Why doe's thi's article have so many incorrect apo'strophe's? --89.243.185.148 (talk) 00:25, 8 July 2008 (UTC)


 * Th'ey were just add'ed a few ' hours'befo're you 're'ad the article. ''  Warren -talk- 02:22, 8 July 2008 (UTC)

File format
I thought these were CFBF files saved by OLE IStorage, but looking at the header for Vista ones with a hex editor it doesn't seem to match the magic number. -- bitplane 18:26, 8 July 2009 (UTC)
 * Vista Thumbacaches are in an entirely new format. It's a simple indexing system in which the index points to the 4 secondary files which contain the actual thumbnail images. They use a mixture of hashes of the thumbnail, image and a ThumbnailID mixing the MFT Record Number, Date and Drive GUID. I know the entire format and can work with it from personal knowledge so this is all unsourcable. - Jimmi Hugh (talk) 01:22, 9 July 2009 (UTC)

"preventing system wide use of the data"
What does that mean??? —Preceding unsigned comment added by 203.206.162.148 (talk) 04:38, 7 August 2009 (UTC)
 * Well, it means (but is incorrect) that thumbnail data cannot be used in other parts of the system, aside from those related to the actual image. What it is should, and is trying, to say is that the thumbnail image data cannot be accessed arbitarily by any program, because the files are fragmented around the file system. - Jimmi Hugh (talk) 08:44, 7 August 2009 (UTC)

EXIF thumbnails
Hi.

There seems to have been a dispute between User:Capmo and I in the article; a kind that is usually resolved very quickly but unfortunately more reverts than I hoped has occurred. It is one of those situations that one thinks "with this revert and explanation, everything will be okay" and suddenly it is not. Today, Campo made a very suspicious claim that EXIF thumbnails of JPEG files in Windows take precedence over thumbs.db ones. It was unreferenced and left many questions unanswered, so naturally I removed it per WP:V. But Campo did something that I didn't expect of a 9-years veteran: He counter-reverted with a pointy edit summary, "Ever heard of Citation needed?" (Sure. And I explained when I use it.)

The sources that are later given ( and ) are questionable: First, they are self-published sources, i.e. posts in an open forum by a person called "Kresho123". Second, they do not talk about thumb.db's precedence; one of them is about Windows 7, which does not use thumbs.db files.

Of course, I didn't expect a counter-revert, but it came anyway. It reads "Sources are from the official Microsoft Support website, maintained by Microsoft personnel". It is false in two points: Microsoft Support Website is at support.microsoft.com, these are answers.microsoft.com, an open forum in which anyone can post. Second, there is no evidence Kresho123 is even associated with Microsoft. Third, Microsoft forum moderators are well-known for being incompetent and possess insufficient knowledge.

I invite a third person (or more; the more the merrier) to check my claims above. Hopefully, this issue is resolved quickly.

Best regards, Codename Lisa (talk) 15:53, 6 April 2015 (UTC)


 * Lisa, you are insulting me with your insinuations here and on my talk page. Who do you think you are to point your finger at me as if I were vandalizing the article? From your last comment that "Microsoft forum moderators are well-known for being incompetent and possess insufficient knowledge." it is patent that you also do the same with others as well. This is not the expected behavior from a wikipedist with your experience. The reason why I edited this article was because I faced exactly the problem cited by me; after solving it I came to Wikipedia and noticed that this article didn't say anything about the subject, so i decided to improve it so that others wouldn't be at a loss as I was momentarily. With your insistence in removing my contribution, it's the Project that's losing, not me.
 * With regard to the sources provided, Microsoft Answers forums are, as you confirmed yourself, monitored by Microsoft personnel who strive to resolve the issues the best they can. Links to the various Microsoft forums are being used in several other articles and accepted as valid sources. The simple fact that two different threads there deal with this subject is enough indication that the paragraph I added covers a real issue. Anyway, I can provide other sources if necessary.
 * PS: You are wrong about Windows 7 not using Thumbs.db. Both Windows 7 and 8 create it, besides XP and Vista. —capmo (talk) 17:51, 6 April 2015 (UTC)
 * Hi,
 * I am not going to reiterate everything I said above; only point out that in your last message, you have thrice said "as you confirmed yourself", which is false; not only did I not confirmed, I vehemently contradicted. The answer by "Kresho123", which is your source is clearly unadulterated and not chosen as the answer either.


 * A summary of the problem with your contribution is: (1) Using self-published sources, (2) failure of verification (source does not say what you wrote), (3) and not adhering to even thought you have nine years of reputation in Wikipedia.


 * Best regards,
 * Codename Lisa (talk) 08:35, 7 April 2015 (UTC)

The both of you have points in your favour, but the both of you need to settle down before we need to take a trip over to AN/I. Some issues:


 * 1) Microsoft Answers would not normally be considered a reliable source, but the information provided by Kresho123 does appear to be correct. Best way to deal with that for now is to allow details in the article stating that Windows will default to extracting the Exif thumbnail, but mark it cn until more reliable sources can be found to support that. However, I highly doubt that Windows ignores the stored data in Thumbs.db; it's far more likely that it extracts the Exif thumbnail and stores it in Thumbs.db and updates it whenever it notices the original file has changed its internally stored thumbnail.
 * 2) I can vouch that Windows 7, 8, and 8.1 do generate Thumbs.db files and that Explorer will on occasion hold open handles to those files. However, that's anecdotal, and it's much more likely that they're created merely for interop or back-compat functionality&mdash;the thumbcache_*.db files tend to be used instead on those versions.
 * 3) Both of you are displaying some nasty bad faith towards each other, and you both deserve a healthy trouting for this. This isn't a case of one editor being pointy and rude, it's the both of you.

It's probably safe to add back that Windows will prefer thumbnail data already in a file over generating its own thumbnail to store, but not the claim that it'll ignore Thumbs.db in favour of in-file Exif thumbnails. However, the claim should not reference Microsoft Answers as a source. IF there's information on the actual Microsoft Support site, or on an MSDN blog run by someone who actually works/worked on the shell team for Microsoft, those could be considered reliable. Discussion forums, even if moderated by Microsoft employees (which, IIRC, MS Answers actually isn't), can't be considered reliable unless the identity of the person whose comments are being used can be proven to be an expert in the topic under discussion. And even then it's probably iffy. // coldacid (talk&#124;contrib) 20:23, 6 April 2015 (UTC)


 * I'm just going to note that I'm going to watchlist the article and your talkpages for a while, and if I do see more fighting, I'll likely elevate this to ANI for the admins to handle. BRD is fine, but this behaviour isn't. // coldacid (talk&#124;contrib) 20:26, 6 April 2015 (UTC)


 * Hello,
 * Normally, I start by saying "thanks for the third opinion" but all you did was insulting us both. You accused both of us of bad faith, which is such an immense crime that no Wikipedian usually commits. You don't even know what assuming bad faith is, or else you wouldn't have accused anyone (especially those who do it) of it. The only person who is fighting here is you. And I don't intend to fight you.


 * "I can vouch that..." Wikipedia is not interested in your vouching. Only your source..., so that I can tell you that they are still useless, because that's not what Capmo claimed. He claimed that EXIF metadata takes precedence over cache. This does not even appear in his own sources. There is a lot of WP:SYNTH here.


 * Best regards,
 * Codename Lisa (talk)


 * The edit log and the comments between you and speak for themselves. And if you think that no Wikipedian usually assumes or acts on bad faith, you should visit the noticeboards more often. In any case, I see that my help or opinion is not wanted here, so I'll wash my hands of this article and the both of you. Have fun continuing to be incivil towards each other. // coldacid (talk&#124;contrib) 12:39, 7 April 2015 (UTC)


 * No, it does not speak for itself; we speak through it for ourselves. And neither I or capmo have yet accused the other of bad faith. (, if you are reading this, I don't think you are acting in bad faith. I think we have a dispute and we can talk it over.) I don't think you are acting in bad faith either, even though your behavior look a lot like Joseph Gregory. Best regards, Codename Lisa (talk) 17:31, 7 April 2015 (UTC)


 * Hi, thanks for jumping in and for your considerations on the subject, they're most welcome. I didn't take offense at what you said, and will follow your suggestion on finding reliable sources. I may be not able to provide a source that Windows ignores thumbs.db in specific situations, but it does: I have a bunch of images recovered from a defective hard disk, and many of them are truncated (though their Exif thumbnails are not); on asking Windows to display thumbnails, it will display those extracted from the Exif, and may not even create a thumbs.db file for such images with Exif. Even when a thumbs.db file is present, Windows resorts to the Exif thumbnails every time the folder is opened. This can be confirmed by selecting one or more images and instructing Windows to "refresh thumbnail": the thumbnails will now appear truncated just as the images themselves. But then again, this is not permanent: next time, the perfect thumbnails from Exif will be shown again. (i.e., thumbs.db was not updated.)
 * , even though you say that you've not accused me of bad faith, that's exactly what your message on my talk page sounded like ("...In fact, it looks like you've written them yourself."), besides your affirmation there implying that my contributions were of a substandard quality. And now you suspiciously suggest that I may be a sock-puppet of a certain "Joseph Gregory"(??); maybe that's the reason why you were acting so aggressively towards me? I don't know... what I know is that I get really mad when one of my contributions, for which I may have spent a lot of time trying to find the correct words (I'm not a native English speaker) is mercilessly reverted as if it was garbage. There are better ways to deal with such situations, as I said. If you had inserted a citation needed tag I would have understood it and immediately looked for sources to back up my assertions. Anyway, I'll take your message in parentheses as an apology. I just hope you won't revert me again when I rewrite my text with better sources. —capmo (talk) 03:12, 8 April 2015 (UTC)


 * The Joseph Gregory comment was toward Coldacid, not you. And sock-puppet? Thank God I didn't say Albert Einstein or Issac Newton. You know those two I hope?
 * Reverting is a natural thing in Wikipedia that most people do and don't take offense because of it (although the same cannot be said about edit warring). Example: One time, reverted me  and I didn't get offended. In another,  reverted   and the latter didn't get offended either. In yet another, I reverted Dsimic  and he didn't act offended at all. Also  and ViperSnake151 once reverted each other without getting all huffy and offended . Even once Jeh, reverted  with a somewhat of a displeased remark  and you don't see them trying to kill each other. These people are all civil lovely hard-working and are not in a hurry to take offense; hence, they don't see the phantom of insult or sock-puppets either. In fact, I became friends with Jeh, Damaster98 and Jasper Deng when they reverted me and talked it over them.
 * When you found a good source, let it be reviewed first... not by me, if you are uncomfortable; other eminent members of Wikipedia can be called to see it. Microsoft has implemented thumbnail caching in all operating systems since Windows 2000; that mean Windows XP, Windows Server 2003, Windows Vista, Windows Sever 2008, Windows 7, Windows Server 2008 R2, Windows 8, Windows Server 2012, Windows 8.1 and Windows Server 2012 R2. (And I didn't mention their Windows Embedded variants.) It is unwise to assume no change in its mechanism has ever occurred.


 * Best regards,
 * Codename Lisa (talk) 09:52, 8 April 2015 (UTC)

I'm not going to take sides, saying one or both of you are in the wrong (or not) with the "truth" or insults (have only read the above quickly). I only "know" one of the parties (Codename Lisa), but I generally trust Lisa knows the rules better than I do. I value WP:V, but the truth even more and just noticed:
 * WP:TRUTHMATTERS (that is only an essay): "Verifiability in reliable sources is absolutely critical. But elevating that to a religion which rejects truth is a huge mistake." - Jimbo

Hopefully stuff that is true isn't being reverted. In my own case ("a somewhat of a displeased remark") I didn't take it personally, and while I still *believe* what I put in, is the truth, I let it go, as some things are just very hard to verify beyond a reasonable doubt. If someone reverts, hopefully they really think you are wrong (or have a point/clarification, I tend to use reverts that way..). comp.arch (talk) 13:37, 8 April 2015 (UTC)
 * Hello, . Haven't seen you in a while!
 * Aren't you forgetting something? This certain area of computing is my area of expertise. So, I am equipped with the knowledge and resources to tell when something is worth looking for the source, or when the writer has mistaken the cause for the effect and the effect for the cause. (More accurately, I have Process Monitor.) But no matter what, I am not going to reject the contribution on the ground of wrongness; that's what a close-minded person do. But by removing on the grounds of lacking source, I enable the contributor to prove me wrong with a source. My advice for the seeker of the truth is to study the comment of Kresho123 more carefully and be sure to read the comment of Axel Rietschin below it. Best regards, Codename Lisa (talk) 09:07, 9 April 2015 (UTC)


 * Just to be clear and to answer your original question. If you revert based on (non-reliable) self-published sources I think you are absolutely within the rules. And it seems you are right that the sources are self-published. What I do not know is if the sources however are telling the truth. It would be hard for me to know or verify.. I was mostly pointing out the general principle I live by.. comp.arch (talk) 10:29, 9 April 2015 (UTC)


 * I understand. Editing Wikipedia has at times been hard for me, when I had to hold back because I didn't remember the source that told me what I would have otherwise written. This is the first time that I directly asserted my major. But again, never before had my disputing party made such a strong claim on his information being correct in spite of the source. Best regards, Codename Lisa (talk) 14:12, 9 April 2015 (UTC)

[While "Verifiability in reliable sources" isn't speaking about the source code.. It is a question if it should have some value as a reliable source.. In case of "open source"/free software it is actually available (and "human" readable, strictly object code can also be..) to all.. Just running the code would not be 100% conclusive, but could it have some value?] comp.arch (talk) 13:48, 8 April 2015 (UTC)

Hey everyone! As Codename Lisa already described it, reverts shouldn't be taken personally or as something offensive. I've been reverted more than once, and in 95% of such instances it was that I've made a mistake that required to be rectified. For the next 4.5%, discussing the things in a civil way is the key. The remaining 0.5% goes to a few editors that actually haven't come here to build an encyclopedia. Actually, I like when I see that other editors are improving what I've put into articles, as that shows me that it isn't only me looking at the same lines. :) Becoming aware of one's mistakes is the only way to become better, at least in my humble opinion. &mdash; Dsimic (talk | contribs) 23:42, 9 April 2015 (UTC)


 * To be honest, I just saw ViperSnake151's revert. But I support the non-personal position and I am not going to kill him. ;) (Hollywood uses the phrase "it's not personal" when someone is about to kill another. If it is not personal, then don't kill his/her person! Or maybe they are implying that killers are misguided in all departments?)
 * I might have fallen for your "assuming bad faith" trick if you had used it against anyone other than Codename Lisa. She is famous for thinking everyone except vandals are definitely acting in good faith. (I'd love to see you trying to argue that with her! I tried!)
 * Microsoft forums are acceptable only when citing a person who has "Microsoft" tag under his name, only when you want to say "That Microsoft employee said such and such". If there is another source opposing it, that must be included too. Now, even under these conditions, the source is self-published and must be treated as the self-published statement of a Microsoft employee, only when such a citation is relevant and has due weight. Other than that, even an MVP's word should not be taken as gospel truth.
 * Fleet Command (talk) 11:01, 13 April 2015 (UTC)