Talk:Single-instance storage
![]() | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||
|
Inaccurate Information
[edit]"However, Area Data System details a procedure they claim enables it on that release" -- The provided link does not back this up. They explicitly state in multiple places in the PDF that SIS is only available on Windows Storage Server 2008 such as the second paragraph of page 1: "And it is not available on Windows Server 2008 Standard and Enterprise Edition". —Preceding unsigned comment added by 74.98.209.224 (talk) 01:20, 3 December 2009 (UTC)
"in a mail server scenario disk space savings are transient and drop off over time"
- this is untrue. With regular use, savings of 50% or more are acheived with single-instance storage, mainly because a single copy of file attachments are kept when sent to many users. I thought I better add this to the discussion page since I don't have any sources to cite readily.
Bias
[edit]The whole article focuses on Windows systems. I think language could be used to be more neutral, and accomodate other OS's characteristics.--David Be (talk) 15:50, 13 August 2009 (UTC)
Only 2 of 8 paragraphs reference Microsoft. These paragraphs rightly do so as Microsoft is the offical entity of this technology. First paraghraph to mention is talking about how NTFS (which is a Microsoftoft technology) makes use of SIS. It also talks about the patents backing SIS technology. Which, if you bother to look them up, are held by Microsoft! It is surprising that the article doesn't start out "SIS, a technology developed by Microsoft, allows blah...". Imagine if the NTFS wiki didn't mention the word Microsoft or Windows. The second paragraph that mentions Microsoft (or Windows Server 2003) is simply stating a fact on how SIS works. Which the explanation comes from the expert, Microsoft, as they patented this technology. It would have been good of the original writer to cite where their quote came from, but at least they mentioned who it came from. Overall the article uses Micrsoft and Windows very few times for what is a supposed definition of a Microsoft technology. Clwood3 (talk) 21:38, 14 September 2010 (UTC)clwood3 —Preceding unsigned comment added by 131.107.0.86 (talk) 01:31, 20 August 2009 (UTC)
I agree with David and would like to add that the text I added to the page was removed TWICE by someone who had a problem with the fact that I mentioned my own product once. Specifically, I spoke about a patent of mine that preceded Microsoft's and which described a stronger form of SIS. I suspect that whoever it was who deleted my text was associated with Microsoft, but that's irrelevant: the point is, there needs to be a way to notify an author whenever his text has been changed or deleted or if comments have been made about his article.FractalBob (talk) 15:57, 6 March 2010 (UTC)
- I can't speak for the other editor, but I'm not associated with Microsoft beyond using some of their software. If you want to include stuff about your product in the article, consider finding a relevant secondary source like a review in a reputable publication that links your product with the subject. BCoates (talk) 08:16, 8 March 2010 (UTC)
- I agree with BCoates and would go further to suggest you should create a new article about the technology you reference (which is not SIS and doesn't reference SIS a single time in the patent). SIS is a very specific implementation of deduplication. I suggest that your patent references a different form of deduplication. You could create a new article and reference it in the see also section on the deduplication page which is an idea and not a specific implementation (as is SIS defined by a US patent). Clwood3 (talk) 21:38, 14 September 2010 (UTC)clwood3
I don't want to get involved in a patent war here. However, in doing my four-weeks-long (so far) major enhancement of this section (among others) of the Retrospect (software) article, last night I split a paragraph in two to create one that starts out "Retrospect does file-level deduplication, patented as IncrementalPLUS." Although I had added that sentence several weeks ago, last night I decided to add an internal link for "file-level deduplication" and found this article. This morning I double-checked the link and actually read this article—at which point I did a double-take upon reading about Microsoft's patent. In the Retrospect Macintosh 6 User's Guide, copyrighted 2004 by Dantz Development Corp. (presumably before it was acquired by EMC in 2004—see first paragraph of the Retrospect article), I found this paragraph on page 21: "Retrospect uses patented [my emphasis] technology to perform IncrementalPLUS backups. IncrementalPLUS intelligently copies only files that are new or have changed since the previous backup to the same backup set. You don’t have to specify whether you want a “full” or “incremental” backup. Retrospect, by default, copies only the files it hasn’t already backed up to the destination backup set."
I then rebooted my now-ex-wife's Digital Audio G4 Mac (which I've been storing for her since 2004), booting classic Mac OS 9.1 from my old Blue-and-White G3's drive which I moved into her machine last spring. I found a copy of the Retrospect User's Guide version 4.2, copyrighted 1999 by Dantz Development Corp.. On page 8 I found a paragraph that says the same thing, except without the mention of "patented technology". It says: "Unlike other backup software which only maintains backups of your current files, Retrospect adds new or changed files to your backup media without replacing the older files. Retrospect does incremental backups so that it is not copying the same files over and over.New backups copy everything from a source and subsequent backups only copy new files or those which have changed. This saves time and backup media. The technology to perform this operation is called IncrementalPLUS™ ,and it's at the core of what makes Retrospect 'better backup.'" DovidBenAvraham (talk) 11:06, 6 November 2016 (UTC)