Talk:Comparison of OpenDocument and Office Open XML formats

From Wikipedia, the free encyclopedia

You have new messages (last change).

Again, I'm trying to help reduce the length of the article OpenDocument. As with other forks of that I've made for this purpose, it needs to be wikified and have its references moved here from that article. --Thephotoman 04:49, 28 December 2005 (UTC)

Contents

[edit] MS Office runs on 2 platforms

"it was designed to only support one product which only runs on a single platform"

I think this is not accurate, for MS Office runs on OS X too. Not that it makes me happy, but nonetheless it should be noted. -- NachoKB, 12/31/2005 17:18 UTC

Fair for the "single platform", but the "only support one product" is still valid.

At the moment, this is somewhat irrelevant - we're talking about OpenXML, not Office, and there is no Mac product being offered (either release or pre-release) which contains support for this standard. AlexHudson 19:44, 19 August 2006 (UTC)

[edit] Non-Compliant?

Special:Contributions/71.117.0.250 had marked the article page as noncompliant. I’ve reverted it for these reasons, but am happy to discuss here:

  • This is the only contribution this unregistered user has made, so he lacks credibility
  • I don’t see any major non-compliance in the article
  • No reasons or examples for the tag have been given

Barefootguru 18:01, 20 February 2006 (UTC)

[edit] Does this article belong in an encyclopedia?

I'm not convinced that it does, especially considering that most of its content seems to be a (non-encyclopedic) paraphrase of a "Valoris report" which is not referenced. Also, as much as I love OpenDocument, this article seems to be written only from the perspective of an OpenDocument proponent. —donhalcon╤ 05:30, 6 March 2006 (UTC)

This article and a number of others were split from the OpenDocument article a while ago. Unfortunately most of the references are still back there. I’ve just moved across the Valoris report.
There’s been one addition from the MS POV, but yes, we need more. People are a lot keener to add tags at the top of the page than actually edit the article.
I think it does belong in Wikipedia, along with all the other comparison pages. It just needs some more work ;-} Barefootguru 18:09, 6 March 2006 (UTC)
Honestly I'm more inclined to add tags to the top than to edit the article because I don't believe this content can be encyclopedic any time within the next ten years. The encyclopedia already contains articles on OpenDocument and Microsoft Office; what information does this article provide that those don't? Moreover, this article contains some sections which apply only to OpenDocument and are not part of any comparison. Why are they here? —donhalcon╤ 18:36, 6 March 2006 (UTC)

Is this article a joke? - Adrian

This article may be encyclopedic, if it only discussed technical specifications. I made some editing today to contribute to that end. Louie 20:36, 29 May 2006 (UTC)

Yes it is, just in need of heavy cleanup and citation. --192.156.110.34 13:58, 14 June 2006 (UTC)

Added some advantages for OpenXML (with citation) and some example XML code to show the mixed/nonmixed content issue and the alternative tagging choices within the two documents. --62.58.36.57 09:21, 17 July 2006 (UTC)

Removed advert tag as it seems the artikel is now reasonably cleaned from te most biased comments. I'll leave the tone tag as the tekst could probably be formalised a bit better. So it probably could still be improved upon but as both formats are still relativly new and not yet the differences fully analized the artikel will become more indept in time.--HAl 15:33, 20 July 2006 (UTC)

I've removed a lot of the poorer text. I believe the references to the politics ought to be removed: this is a technical article, not a political one, and I'm sure the OpenDocument and/or OpenXML articles are better places to discuss the politics. I've improved the references to commentary - in particular, Microsoft itself has offered little to no comment on OpenDocument, so please don't attribute personal views to them as an organisation. The section on Custom XSD needs to be rewritten in its entireity - there are a lot of claims an uncited 'fact', and stinks of original research. I don't think it's NPOV either, but that's more to do with it being factually inaccurate.

The text on Valoris also needs to be totally rethought. Valoris is non-technical and was written in 2003 (?) and cannot inform a technical article about two standards in 2006. AlexHudson 19:58, 19 August 2006 (UTC)

[edit] no significant implementations in code not derived from OpenOffice exist yet

I don't understand this statement. There are already several "implementations not derived form OpenOffice". At least KOffice 1.5 has native OpenDocument support and is not derived from OpenOffice. Others, such as AbiWord, Gnumeric, Scribus and TextMaker, have filters for OpenDocument and are not derived from OpenOffice. Are these implementations not "significant"? If so, what does "significant" mean? —The preceding unsigned comment was added by 129.187.81.42 (talk • contribs).

I would suggest that a significant implementation would by able to do one one the following
A) implement near to full support for all Opendocument editting (complete office package)
B) implement near to full support for all Opendocument viewing (complete opendocument compatible viewer)
C) implement near to full support for editting a least one major office documenttype (as a dedicated wordprocessing or spreadsheet application)
D) implement limited support for editting most office Opendocument formats (light office package)--HAl 14:26, 20 July 2006 (UTC)
Okay, conditions met. Let's make sure it clearly states that OpenDocument meets those conditions, as an important differentiator. -- 129.246.254.14 22:14, 18 August 2006 (UTC)

[edit] Office Open XML and OpenDocument incompatiblities

Indeed, OpenDocument does not support VBA and OLE, although the importance of these features depends on who you are. The average Office user will not need them. I wish to post this to the article, and if there is no objections I will. - Yuhong 01:49, 13 June 2006 (UTC)

I think they're not supported precisely because they are proprietary. Support for either VBA or OLE depends only on whether you're running on MSWin; otherwise (*nix, Mac, or Palm) you don't need it. So that lack of support is not a bug: it's a feature. Louie 16:11, 13 June 2006 (UTC)
But Office for Macs support OLE and VBA, although these are some of the least updated part of it. Also, support for them is nesserary for full compatiblity with existing Office documents. Yuhong 14:41, 14 June 2006 (UTC)
Well, maybe that's why people want to move to Open Formats: to stop being locked in to one single company's OS, Office suite, and personal decisions... Louie 16:17, 14 June 2006 (UTC)
Fortunately, things such as VBA and OLE can be easily stripped by other MS Office Open XML readers. Yuhong 22:46, 14 June 2006 (UTC)
While Microsoft Office does not support UNO or Java. —The preceding unsigned comment was added by 212.202.36.24 (talk) 01:52, 31 January 2007 (UTC).

[edit] Artikel should be renamed !!!

The current name is: Comparison of OpenDocument and Microsoft XML formats It should be changed to:

Comparison of OpenDocument and Microsoft Open XML formats

or

Comparison of OpenDocument and Open XML formats

The formats are commently referred to as Opendocument and Open XML and that should be reflected in the title. The reference to Microsoft in the title could also be removed. We are not speaking of the OASIS OpenDocument format so why use the Microsoft Open XML format ???--62.58.36.57 10:25, 19 July 2006 (UTC)

I agree with the idea. I suggest using the titles of the respective articles: Comparison of OpenDocument and Microsoft Office Open XML. I don't think it's necessarily unbalanced to have Microsoft but not OASIS in the title: it seems Microsoft markets its format as Microsoft Office Open XML, while OASIS usually uses just OpenDocument. Wmahan. 05:15, 20 July 2006 (UTC)
Would people please stop playing Page Title of the Week games with OpenDocument-related pages? As I explained on the main OpenDocument article's talk page some time ago, tons of web sites link to Wikipedia articles on this very hot topic but the links return a 404 page not found error if you rename pages. For that reason, Wikipedia policy calls for starting a new page rather than renaming and then setting the original page as a URL forwarding page. Even were folks to actually start doing it the way they are supposed to, all of those links out there that use proper academic citation form and cite the article's title as well as its link have their citations trashed as a result because the article's title is no longer accessible. So even page redirects are a lousy workaround. But no one seems to have the energy to do even that in their frenzy to refactor these pages. My request that the people who made the mess go back and clean it up has been ignored. Wmahan, as I recall you are one of the offending parties on the other OpenDocument pages. If my memory is correct, please go fix the mess you made. Either way, please try to throttle back your refactoring hormones. The page titles and subtopic titles on these pages were good enough. For very actively edited pages, you might try setting yourself a rule that if the page title has anything to do with the article content, don't touch it unless it contains something as bad as an obscenity. Altering page titles equals link rot. People who do that are in reality writing bugs in the World Wide Web, writing 404 page not found error messages. Then people like me have to fix all those broken links on our web sites. Your creative moments make work for people like me. I have to come here, try to find the old content hiding under a new page title, then go back and edit my site to fix what you broke. I hope you will understand that I am not happy when you do that to me, especially when what I usually learn in the process is that the title changes were made for the pettiest of reasons if any reason at all. In fact, I am becoming quite irritated with people who do that to me. So please lay off renaming the OpenDocument-related pages or I just may proclaim a dispute the next time you do it. Especially if you lack the courtesy even to do a page redirect as required by Wikipedia policy. Marbux 05:44, 19 August 2006 (UTC)
The standard to rename an article is to use the move tab function in wikipedia and of course I used that method so there was a redirect. The old article reference is still there at Comparison of OpenDocument and Microsoft XML formats. So I am not sure what you are on about ??? hAl 06:49, 19 August 2006 (UTC)
Right. Also, "Open XML" is very confusing, it sounds like OpenOffice (which implements OpenDocument). The ECMA rules state that the results must be whatever Microsoft does, and more people are familiar with Microsoft than with ECMA, so I think "Microsoft Office Open XML" would be the clearest name to use. -- 129.246.254.14 22:17, 18 August 2006 (UTC)
If you believe that Ecma will put Microsoft's name in the standard's name, assuming Ecma adopts the proposal, your understanding of standardization processes needs some major surgery. And I would very much like to look at this Ecma rule saying "that the results must be whatever Microsoft does." Got a citation and a link? You are implying that Ecma and the other companies participating in the standard's development are just there to rubber stamp whatever Microsoft does. Wouldn't that be a headline!
If we're going to put the names of the development companies that support the format in the titles, then we're going to have to put a whole bunch of development compoany names in front of the OpenDocument name. Or if we're going to base it on who signed off on the covenant not to sue, then I suppose it should be "Sun Microsystems OpenDocument." Really! Wikipedia has a policy against advertising. "Office Open XML" is about as close as you can get to being accurate without turning Wikipedia into an adertising service for Microsoft. Most stylebooks call for using the full name the first time it appears in an article with the short form immediately after it, enclosed with parentheses and quotations marks, then the short form only in the remainder of the page. So first time around, Office Open XML ("OXML") is a specification for several file formats ... Thereafter, just OXML. Or whatever you guys decide on for the long and short forms. Marbux 05:44, 19 August 2006 (UTC)
I used Open XML in searching for the article but it came in down low for relevancy. Also it is not the full name as Office Open XML or Microsoft Office Open XML are current names for the format. Allthough I had a slight preference to Office Open XML adding microsoft in the name makes it consistant with the [[Microsoft Office Open XML] article which is much better for finding the article. hAl 06:49, 19 August 2006 (UTC)

[edit] MathML support

BTW, would whoever it is that thinks OXML supports MathML please read <http://blogs.msdn.com/brian_jones/archive/2006/08/16/700494.aspx>. Note that Jones says you can paste MathML to Office. That means it's going into the binary, not into the XML. I fixed the article, but that urban myth just won't stop going around. Marbux 05:44, 19 August 2006 (UTC)
That is a relativly new info by Brian Jones. Nice of you to refer to. You should read the comments of the article also before referring to it. In the comments of the article a MathML is created in MS Office 2003, OpenOffice.org 2.0 and MS Office 2007 beta. MS Office 2003 produces a binary. Funny enough OOo also produces a binary without extention in the fileformat when a MathML equation is entered. MS Office 2007 is the only one that produces pure XML. So your claim that OOXML will use a binary is nonsense. In fact there is no binary elements to the OOXML format as I can see except for things like embedded (media) files or compiled VBA which logically has to be binary. Those binaries will all be in separate files exactly as they will be in ODF.
Also in the comments it is noticed that ODF uses modified MathML in it's format. Is that enough to warrent it's standard support claim... hAl 06:49, 19 August 2006 (UTC)

[edit] Base comments on standards

We have copies of the standards available: wherever possible, please base comments on references to these standards. I removed "OpenXML supports MathML and PNG", because it doesn't - neither is in the standard at this time. OpenXML makes a single reference to the PNG standard, but does not reference it, OpenDocument does slightly better by recommending that graphics be saved as SVG or PNG. It's not clear that either standard _requires_ PNG.

If something cannot be cited with articles on the web - from either side - or found in the standards, please do not add it. AlexHudson 19:50, 19 August 2006 (UTC)

hAl 15:29, 20 August 2006 (UTC)

[edit] Citation unclear

I removed this citation about OASIS refuting comments on internalization as the url did not work:

These claims may be based on comments received by the OpenDocument TC during the ISO standardisation process, but are refuted by the TC, who claim that while support is not explicit, it does exist.[1]

The citation could be re-added in if someone finds the right url and it does show the ODF support for internationalization to be on par with OOXML. (That is not to be defined by implementation of odf)[urer:hAl]

That URL does actually work; but it is in OpenDocument format, which I grant you isn't clear. I've instead linked into the new appendix on bidi support in the OpenDocument specification; this is a purely informational addition to the OpenDocument 1.0 specification. Unfortunately, there doesn't appear to a copy in HTML available.
Alex, I still am not confinced about the comment you put in about the internationalisation issues found in ODF. I have not found a reference to this info in the offcial standards document (ODF specs). I am not sure if the support is in or not or if it will be added in a future version. At least I have no idea how one would implement internationalisation based on what is in the current standard. If it is not in the current standard than I would suggest you add it to the OpenDocument article as something for the future. If it is in the current standard can you give som details on where that support is referred in the actual standard itself ?? Also I thought the bidirectional support is not the only thing missing from the internationalisation support information in ODF. Isn't the list numbering an issue as well ? hAl 16:07, 20 August 2006 (UTC)
hAl - you probably need to review the document I'm including. This isn't a future addition to the standard, it's a future non-normative section: that means that it simply explaining what is already in the standard. Support for all those things you've mentioned is already included in the standard, since the standard uses Unicode - this covers bi-directional text, Arabic numbers and the other issues that have been listed. This why it is correct to call the assertions to the contrary "claims": the article says "OpenXML supports .. non-western languages better". It is painting a criticism as fact, which is unfair to OpenDocument. I'm going to try to rewrite that again in a more neutral form. AlexHudson 16:51, 20 August 2006 (UTC).
It's not correct to require that ODF shows equivalent i18n; we're not doing original research here. It is sufficient to show the claim and the counter-claim. Also, please don't attribute Brian Jones's comments to "Microsoft": nothing in his blog shows that these are official comments, and should be cited correctly. Lastly, it would be helpful if you could make separate types of changes in different edits: clearing up the naming of OXML was really useful, but putting in the other edits made it harder to clear them up. Thanks! AlexHudson 12:04, 20 August 2006 (UTC)
The comments of Brian Jones are not of a personal nature but made as a program manager on Office XML formats as he clearly states on the blog. That mean I can call him an official Micrsoft source. And you know what, so does Jean Paoli Microsofts senior director of XML architecture on the Micrsoft website [1].hAl 16:07, 20 August 2006 (UTC)
He is a program manager, but nothing states that the things he publishes are the official opinions of Microsoft. I think it's fair to attribute those comments to him in his role as program manager; I don't think it's right to represent them as the stance of Microsoft. You can call him a Microsoft source; you can't say he's speaking for Microsoft. AlexHudson 17:04, 20 August 2006 (UTC)
As his comments seem fully endorsed by a Micrsoft director who refers to his blog I think we asume his comment to be Microsofts. Micrsoft uses a lot of blogs for putting out information on certain aspects of its operations. and the blog has no personal content but only info on the topic of this article. Also the fact that the comments are mainly from his blog is established before the list of advantages and by the refences to the blog itself. I for instance would prefer to start the advantages of the ODF format with "The OpenDocument Fellowship published an article..." and leave the names of the authors to the references. For the viewer it is much better to place the info at the source using the organisation for which they speak than using their names. Only for very personal opinion like that of specialist or opinionmakers the name might be more relevant than the role that they are in.

[edit] Formula interoperability

User:HAl, I noticed that you reverted my edit on formulas without a comment or discussion. Please, don't do that. I'm not going to revert your change immediately, but I fear you're being factually incorrect. You've stated "Currently ODF cannot be considered interoperable for spreadsheet documents as it allows for vendor specific implementations of the formulas in spreadsheets".

This is a poor statement for a number of reasons. Firstly, vendor specific implementations of the formulas are actually a _feature_ of OpenDocument: even when OpenDocument receives a full specification for formula syntax, vendor-specific formula are still able to be used.

so that state about lack o interoperatbility is clear. Creating any ODF spreadsheet can lead to interoperability as you can use whatever formula definition you want. It is impossible for a infinite number of implementations to be interoperable without adding a further agreement on what to put in formula's. It just isn't there in the current standard. I think it is weird you added out my comments on that that without comment and now complain when I do simular to claim. Also what you do is state that I am actually right but that because it is a ODF 'feature that it can't be in the article. I do not agree and will refert your commentshAl 15:23, 20 August 2006 (UTC)
No, I don't think you understand the formats. For example, by your definition, Microsoft SpreadsheetML 2003 cannot be considered interoperable: but that's plainly silly, since people use that format. In fact, formula syntax and function availability has changed little since Lotus 1-2-3: it has never been specified properly in Excel, or OpenOffice.org, or any of the other spreadsheet systems. What is different about MS OXML is that this is now, for the first time, being specified. That's definitely a good thing, and the article should mention it: but saying that all other formats are uninteroperable flies in the face of logic; don't you see? AlexHudson 16:58, 20 August 2006 (UTC)
What you are describing is the difference between a standard and just a format. ODF has interoperability at it forefront. If Microsoft would add to it's ODF plugin certain specific Excel formula's to the format that no other spreadsheet software supports it would be fully standard ODF and they could easily claim that it is interoperable with the most used software but it would only work perfectly in MS Office. It is simular to the OOXML format not being able to support legacy Office formats as MS has put the compatibility with current documents at it's forefront. I guess if you feel strongly about the lack of formula's in ODF being so good you could mention it as a new advantage of the ODF format but since ODF already announced they will add the formula's 2nd half of 2007 they probalby noticed as well that it might belong into a standard. hAl 22:31, 20 August 2006 (UTC)
No, it's not really a standard vs. format thing. What I'm saying is that currently, the article is saying that ODF is somehow deficient in not having a formula syntax. That's not the case: no standard for spreadsheets has _ever_ defined that. OpenXML is completely new in defining it, and it's also somewhat arguable that they did it as a response to ODF. I think you might be misunderstanding my point: I'm not saying that's not a benefit of OXML that formulas will be defined, or that it's important such a syntax be defined - it's clearly a benefit. However, it _does not_ hamper interoperability - everyone implements spreadsheet formula virtually identically anyway, because this is one part of the file format you cannot control: it's the users who write the formulae. I'm saying that these needs to be written as a positive thing for OXML, not a negative thing for ODF - people have interoperated with 123/Excel for years without requiring _any_ of the format to be open, so it doesn't actually prevent interoperability. Formula syntax is well-known, well-defined and a de jure standard.

Second, the statement reeks of original research. The claim has been made by supporters of OpenXML: we must cite that claim correctly, and then give the response from the other side. It is not our place here to "state fact". —The preceding unsigned comment was added by AlexHudson (talk • contribs) .

First you state that the vendor specific forma's can be used and when that same thing is said in an article that actually cites the OASIS TC president you call it a claim ?? Frankly I do not understand why you not completly editted out all points in advantages of ODF as they were arguments by the Opendocument fellowship and not of them are factual but all all one sides conceptions on ODF. Your arguments seems completly biased in editting what you call claims by one side while you leave the other stuff that is pure claims from the other side alone.hAl 15:23, 20 August 2006 (UTC)
Which "points in advantages of ODF" have I edited out, and which are you calling biased? I would be grateful if you could point them out, because I've been mercilessly fair. I don't think it's necessary to state that each point in favour of OpenDocument is a claim, because they all arise from the same document - and in fact, I added text to point out the weaknesses in that document, and that it is out of date. Which parts do you think are biased? AlexHudson 16:58, 20 August 2006 (UTC)
Hmmm, were having a communcation problem I guess. Mayby it is my poor english. I did not mean to say you editted out any points. But the points of the ODF fellowship are not given a response to from the other side as you put it so they do not conform to your earlier comment and as such should not be in the article. But I do not think it is wise to add much reponse from the other side to any of htese comparisons. Just let each side name their strong point. For instance, both sides claim their solution to using links inside their format is the best so each has it named in their advantages allthough they are very conflicting. For instance if ODF also has a good internationalisation support but done n another way it could better go in their advantages as a positive comment than as a negative comment on OOXML advantages. And of course it could also go in the OpenDocument article.hAl 22:31, 20 August 2006 (UTC)
Oh, ok. I mainly agree with you - we don't need specific claim/response for everything. But where one side makes a claim that the other disagrees with, we need to ensure that dispute is documented. None of the things listed in the OpenDocument Fellowship's article have been disputed by Microsoft, and many of them are matters of opinion, not fact - eg., Microsoft agree that they have a different content model, and they believe theirs is better. It's not a matter of fact that ODF's model is better. A couple of Microsoft's claims, though, are not matters of opinion - e.g., saying ODF has poor support for i18n. That's a fact in dispute.
It may well be that the benefits of ODF over OXML section needs to be rewritten - I think some of it isn't very good - but because that section isn't great doesn't mean that we should correctly point out matters of fact that are in dispute. The thing about links is a great example of two different systems, where each side thinks their way is best, as is the content model. We can handle those things differently - the stuff about i18n is a different category though, and MS aren't the best people to say what ODF can/cannot do. AlexHudson 09:58, 21 August 2006 (UTC).

Ok, I've looked in more detail at this formula issue. I think this section is still inaccurate. In particular:

  1. it states OpenDocument is still working on the formula specification. This is true, but so is Ecma TC45. This makes it sound like OpenXML has finished the formula specification, and I don't see anything which supports that in the citation. We shouldn't level a criticism at just one standard when it is true of both.
  2. Brian Jones claims OpenDocument only has a "few lines" on formula specification. I just looked at the standard, and this isn't true - the definition of formula syntax, cell addressing, types (etc.) spans numerous pages (p114, p123- [formulas addressing databases], p142-, p185- [this is the main syntax, etc], p193-) not counting the places which simply reference formulas or the behaviour of formulas (e.g., copying ranges). I'm not totally happy with including this claim, because it's just not factually accurate. We should replace this with something about formula functions - this is the big difference.
  3. "ODF is [not interoperable because of] vendor-specific extensions". This doesn't correctly represent the view Brian Jones is putting forward. OpenDocument supports multiple formula syntax by explicit design and always will; similarly, OpenXML supports only one syntax by explicit design - but Jones isn't addressing those design differences. He's saying that he doesn't think that OpenDocument defines formula functions - which it doesn't - and that lack of those functions is bad.
  4. Related to the above, I'm not totally happy linking to that post by Jones. In it, he's not really even stating things about OpenDocument - he's saying "I looked at ODF, and didn't find much information on X, Y, Z - can someone help me here?". I'm not sure he's even making a criticism of ODF: he's asking questions about it. This same "can someone help me?" post is referenced 3 times - this isn't a very good citation. Can anyone find something a bit more solid, where specific claims are made, not just a few questions asked? Going back to point 2 above, I definitely think supporting factually incorrect information with a citation to someone asking for information is pretty bad.
  1. Open Office XML isn't finished so the formula specification isn't finished either
  2. Brian Jones was talking about spreadsheet formulas and p 184 is the only place where spreadsheet formulas are mentioned and essentially claiming that "The components used in the formula depend on the application being used." - so there is NO specifiation/standard and therefore hindering interoperability. And it's not only about functions. No operators, no error handling, parameter passing,...
  3. OpenOffice.org implements a superset of OpenDocument and has vendor specific extensions (besides formulas)
  4. http://mail.gnome.org/archives/gnumeric-list/2005-June/msg00134.html, http://blogs.gnome.org/view/mortenw/2005/06/16/0 especially the last one
--62.178.136.129 23:02, 29 August 2006 (UTC)


That's pretty confused. Point 2 is clearly wrong - I listed above a variety of pages which specify spreadsheet formulas. Perhaps you should read the specification again? Point 1 is irrelevant, 3 is unsubstantiated. Point 4 contains links we should use, though.

AlexHudson 13:57, 7 September 2006 (UTC)

What about http://www.oasis-open.org/committees/documents.php?wg_abbrev=office-formula filled with 371 pages of informations about spreadsheet formulas? 02:41, 2007-01-31
When talking about interoperability etc, I think it is really important to distinguish general characteristcs from version-specific characteristics. In the case of ODF, it means distinguishing between pre-ISO ODF drafts, ODF 1.0, 1.1, 1.2, ODF 1.3, ODF as implemented by particular vendors, ODF in general. For OpenXML it is simpler, with just draft ISO OpenXML and (with whatever changes are made) the resulting ISO OpenXML, OpenXML as inplemented by particular vendors, and OpenXML in general. Particular claims need to be referenced to particular versions, perhaps. Otherwise we end up with spurious comparisons, for example between Office 2003 XML and ODF 1.2 with OpenFormula, or between OpenXML and pre-ISO ODF. This is a challenge for readers, because they are not necessarily in the position to know from quotes which versions the comments are referring to. I suggest that comparisons should strictly limit themselves to being about ODF and OpenXML as given in the ISO standards or drafts at the current time, and with material that is about previous or in-the-pipeline or future versions being explicitly distinguished. 124.189.122.186 04:44, 2 April 2007 (UTC)

[edit] Critics

  1. Microsoft Office Open XML should be changed to Office Open XML or Ecma Office Open XML throughout the article
  2. The article should focus on comparing technical aspects/features maybe like Comparison of web browsers (i.e. using tables - I like tables)
  3. The article should not rate features - it shold simply list and compare features.

[edit] Advantages of OpenDocument over Microsoft XML format

Mixed vs. Non Mixed Content: Some people consider mixed content harmful and a bad thing. Some XML parsers can't even parse mixed content (only very few though). In quite some cases non mixed content is easier to work with. Most (nearly all) users will never see the markup and simply don't care. Technically there is no difference between these two models - it's simply a matter of taste. Unless your XML document stores structured data - in this case non mixed content clearly wins.

Similarity to XHTML: Creating XHTML using a mixed content XML document is as easy as creating XHTML out of a non mixed content XML document. Most people create office document using word processing applications or spreadsheets or ... and not some text editor. I don't see who someone could reuse his XHTML skills.

Better separation between content and presentation: Is it really better to always separate content and presentation? Any studies out there supporting this claim?

Hyperlinks Indirect addressing is commonly used pattern/method in computer technology and Wikipedia (i.e. http://en.wikipedia.org/wiki/Wikipedia). Sometimes it's better to store actual addresses separately.

Standards So does Office Open XML (where appropriate and possible). Most people familiar with XML are not familiar with either XLink, Dublin Core, SVG or MathML.

These are all your claims. This article isn't about original research on OpenDocument or Open XML. This is about criticisms and comparisons of both formats that have been made, and a summary of the debate. You might not see why mixed-content is relevant or why being close to XHTML is useful; that's not the point - this is a neutral summary of the debate. AlexHudson 14:19, 7 September 2006 (UTC)
These are all my claims and are my original research, indeed. Therefore I put it on this talk page and not into the article. This article should be a factual comparison of both formats and not a list of (false) claims by some questionable sources (opendocumentfellowship - promoting OpenDocument, Brian Jones - promoting Office Open XML, Groklaw - mostly citing opendocumentfellowship/hardly experts on file formats/rather anti MS, ...). Collecting such claims could by neutral but can never be encyclopedic. I would like to see a table that simply compares the features (content model, hyperlinks/references, standards reuse, ...).
Groklaw has for sure more experiences with fileformats then Brian Jones has. Also to name the disadvantages a fileformat has doesn't mean automaticly to be a MS-basher. —The preceding unsigned comment was added by 212.202.36.24 (talk) 02:01, 31 January 2007 (UTC).
The terms advantage and disadvantage are POV (in my opinion) and both "Advantages" sections are full of original research and hardly neutral.
All in all I still think this article is unencyclopedic -- 62.178.136.129 22:57, 22 September 2006 (UTC)

[edit] Advantages of Office Open XML formats over OpenDocument

I won't comment on Office Open XML as it is still in draft status.

Point (1) and (2) from Brian Jones are dead wrong. The perfect conversion of legacy-format Office documents into ODF format provided by the daVinci plugin and the Sun Microsystems plugin are demonstrations of this. [2] [3] Point (3) has also been addressed in the soon-to-be-available OpenDocument version 1.2.

[edit] Custom XML schema definitions (XSDs)

Unencyclopedic, mostly talking about opionion and speculating. Not talking about facts.

OpenDocument currently does not support custom XML schemas in way comparable to Microsoft Office

Actually, it does, but I agree - this section should be completely removed. It can't be saved; instead, we should re-write it when enough citations are found. AFAIK, no-one has really compared these features technically. AlexHudson 14:20, 7 September 2006 (UTC)

[edit] Cross-platform interoperability

see above

The paragraph talking about macros and OLE fully applies to OpenDocument as well (no standard for formulas and support for OLE objects).

Furthermore it is worth noting that currently no application or application suite fully complies to the OpenDocument standard. OpenOffice.org implements a superset and all the others have rather limited support and show heavy interoperability problems -- 22:30, 29 August 2006 (UTC)

Macros/OLE is a separate issue. OpenXML doesn't define a macro language, either syntax or API, or a system for OLE. Neither does OpenDocument. No-one is claiming that either issue is a problem, and we can't just say that it is because we think it is.
As for OpenDocument applications - there are at least three suites which have full OpenDocument support (OpenOffice.org, StarOffice, IBM Workplace). To criticise OpenDocument on that basis when OpenXML has *no* implementations would be slightly unfair, and I don't know anyone making that criticism. If you have some cites, please list them. AlexHudson 14:25, 7 September 2006 (UTC)
Wikipedia is not about fairness but about correctness. There are interoperability problems with OpenDocument right now (and IBM Workplace has only limited support OpenDocument by the way). I think there will be interoperability with Office Open XML as soon as there are implementations available (currently there are none). But this is speculation and hence unencyclopedic. -- 62.178.136.129 22:57, 22 September 2006 (UTC)
Could you name interoperability problems? and no, formulas are defined ( see http://www.oasis-open.org/committees/documents.php?wg_abbrev=office-formula ) and to support OLE, ActiveX and .NET3 can't be a valid goal.

[edit] Tags on article

62.178.136.129, you put Unencyclopedic and Advert tags on the article. Both these tags refer editors to resolve problems in Talk; we cannot resolve your problems unless you list them.

I've removed your tags for now. If you have problems with the article, please list them here first. Putting tags on the article without saying what's wrong where means people cannot fix the issue.

Ok, I missed your compaint about Unencyclopedia. That section should be removed anyway. I don't see anything which reads as an advert for either OpenDocument or OpenXML though. AlexHudson 14:21, 7 September 2006 (UTC)


[edit] XSD comparison

The original text follows. It's pretty rubbish. We need to find some sources which talk about the data support in both Opendocument and OpenXML. Valoris compares with Office 2003, which isn't really useful.

[edit] Custom XML schema definitions (XSDs)

The Valoris Report noted that Microsoft's Office Open XML format supported custom XML schema definitions (XSDs), while OpenDocument did not. XSDs made it "possible to attach one or more custom schemas to a given Word document. It allows the users to annotate the document with the elements found in the attached schemas.

In response to this, the OASIS OpenDocument TC added XForms to the OpenDocument specification. This is meant to provide equivalent functionality while reusing relevant standards and hence reducing the potential loss of interoperability.

It is extremely controversial, however, whether these embedded XSDs are actually an advantage or not. Proponents of OpenDocument claim that it is not valuable and possibly harmful, while some users of Microsoft's formats find it highly beneficial:

  • If the XSD is used as a "Save data only", it is simply saved as a normal XML file with standard packaging. OpenDocument files can embed XML files in a similar way (since both are essentially zip archives of a set of files), but there are no provisions in current implementations of OpenDocument to actually generate such XML.
  • If XSD is used to embed custom elements "throughout the tree", the result is a document that requires (as a matter of practice) specialized tools to process the additional data. However, according to some who have worked with the format, this criticism is overstated as it is a fairly trivial matter to load such an XML file into an XML DOM and access the custom schema directly by its namespace, mooting the issue of the two namespaces being intermixed in the file.[citation needed]

In many eyes, combining custom XML with the presentation XML impedes rather than aids interoperability, which would be a serious problem since the whole point of these standards is to promote interoperability.[citation needed]

A contrasting opinion is that such markup allows documents to take on semantic meaning that is not possible with "presentation-only" approaches. In this view, for example, an invoice can have both presentation as well as meaning for each of its elements such as quantity, item description, part number, total, etc. Such a document can be consumed by a process that wants to glean information from a database full of such documents. Since such data is often unique to a single entity such as an individual corporation with a specific database design, the XML needs to be free to be defined by the user and cannot be a priori determined by a standards body. On the other hand, this functionality is essentially "metadata" (data about data) and others believe it makes more sense to use a standard designed for metadata, like RDF.

The OASIS OpenDocument committee had considered adding this direct capability before public review of OpenDocument began, but decided not to, saying that this was "not essential for the current version of the specification." Conversely, others have claimed that this is a primary value of an XML format - to carry business data that is machine readable in context, rather than divorce it from the document or keep documents purely about generic presentation.

Instead, the OpenDocument Technical Committee developers embedded the ability to support XForms. XForms is a W3C standard, and supports the ability to handle custom schemas but add additional constraints that are believed to counter the perceived weaknesses of custom schemas. In addition, the OASIS OpenDocument committee has created a metadata sub-committee to expand metadata support (e.g. RDF) to OpenDocument. Adding RDF to OpenDocument would satisfy the use scenario above while avoiding the loss of interoperability that custom schemas present.

The above discussion of Microsoft's Office Open XML formats covers primarily the Office 2003 formats (and more specifically the Word2003 format). The formats produced by the upcoming "Office 2007" (currently in beta) are significantly different in approach. For example, a major difference is that the custom XML is stored as its own "part" in a ZIP container and the presentation XML contains references to that custom schema so values can be mapped to the document when it is loaded.

[edit] Intro

I have some objections against this adition to the article:

There is fierce debate about technical merit between supporters of each format. A significant issue in terms of the success of the formats is the politics of adoption. The technical arguments, as in other battles for standards, often turn out to be less important than customer perception. Fundamental differences between the two formats are that OpenDocument is an approved ISO standard (approved for release as an ISO and IEC International Standard in May 2006, designated, ISO/IEC 26300) and is controlled by OASIS, a foundation broadly made up from representatives of the ICT industry and its customers. Microsoft Office Open XML is defined by Microsoft and currently undergoing a standardization process by Ecma International, an ICT industry standardizations organisation. This Ecma standard will then be put through the process to gain ISO status. In the event of successful ISO adoption, control of the standard will then rest with Ecma International.

It contains an unverifiable clause stating that technical arguments are less important that customer perception and also somehow tries to link that same customer perception to an earlier sentence about the politics of adoption which is something totally different. In the rest of this paragragh there is a comparison about the standardization where it is strangly missed that the ISO standardization has been a much faster fasttracking proces than the current Ecma standardization where the technical committee seems to create a lot more significant changes to the actual standard format than happened during the ISO standardization process of ODF. What you can actually say about the standardization proces is that basically ODF is 12 months earlier in the standardization proces than OOXML but that at the cost of leaving out certain elements that wuill be added to the standard later on (something that is impossible for OOXML becuase it has to support a legacy of documents from the start as that is a claim encorporated in its standard hAl 14:32, 8 September 2006 (UTC)

I agree that the introduction definitely needs to be rewritten. This article should just be a comparison of the formats, not anything more.
You're misinformed about the standardisation process, though - both ODF and OpenXML would be subjected to the same process (assuming OpenXML continues as it is going). The equivalent to OpenXML's Ecma process is ODF's OASIS process, and if you look into that you'll see that large changes happened during that time (e.g., XForms). It's not the role of ISO to particularly modify a standard: either they accept it, or they don't - they may ask for small changes, but they don't redesign something. Both Ecma and OASIS have a fast-track at ISO, so in total the time it took ODF to go through the process will be the same as the time OpenXML takes (assuming OpenXML doesn't encounter problems; ODF did, but set it back only by a couple of months).
In total, OpenXML is more than a year behind, and it's not because ODF was somehow standardised quicker than OpenXML: the process is/was exactly the same. It's Microsoft's intention to get OpenXML standardised at ISO (*after* it goes through Ecma), and that would again be a process where ISO don't really make any changes to the spec - that's the whole point of fast-tracking. AlexHudson 12:26, 11 September 2006 (UTC)

[edit] Cross-platform interoperability

Valoris was comparing OpenOffice.org 1.x against Microsoft Office 2003 Reference Schemas not OpenDocument against Office Open XML (both not available at that time). I've removed all references to Valoris from this section since they where misleading. -- 62.178.136.129 23:18, 22 September 2006 (UTC)

[edit] Tags on article

this seems discussed times and again. Somehow, the tags were there but the article was okay. I couldn't understand what's wrong with the tone, and there are so many citations that it's more like a research paper than an encyclopedia entry... I read the talk page and I don't see how the criticism fits in the article. Towsonu2003 04:58, 8 December 2006 (UTC)

[edit] XMLs for Spreadsheet Comparison

Please be noted that in ODF, there're two attributes that ODF captured, and OOXML not.

Look at the ODF example, you'll see

  • office:value-type (e.g. office:value-type="float")
  • office:value (e.g. office:value="5")

According to http://books.evc-cit.info/odbook/ch05.html#table-value-table these are for good purpose.

  • value-type explicitly tells you the type of the value, it can be { float, currency, percentage, date, time } ... how OOXML tags in the example captures this type information ?
  • value will store the value that the application can use directly

Say, "45.6%" will have value-type="percentage" and value="0.456", and its textual representation ("45.6%") will be kept, for convienient uses, in the element text:p. Like this:

<table:table-cell office:value-type="percentage" office:value="0.456">
  <text:p>45.6%</text:p>
</table:table-cell>

I think these information should be incorperated into the main article in some ways, to give the reader a senes that while ODF for spreadsheet may longer, it captures more information, and for a good reason.

-- Bact 04:17, 23 January 2007 (UTC)

[edit] Miguel de Icaza's position

Hi,

is it possible to merge the position of Miguel de Icaza over the two formats? He points out some interesting points over the lacks of ODF, like for instance :

OOXML devotes 324 pages of the standard to document the formulas and functions. Depending on how you count, ODF has 4 to 10 pages devoted to it. There is no way you could build a spreadsheet software based on this specification. To build a spreadsheet program based on ODF you would have to resort to an existing implementation source code (OpenOffice.org, Gnumeric) or you would have to resort to Microsoft's public documentation or ironically to the OOXML specification.

http://tirania.org/blog/archive/2007/Jan-30.html

Great A'Tan

[edit] merge with Comparison of OpenDocument and Office Open XML licensing

I have tagged "Comparison of OpenDocument and Office Open XML licensing" to merge here. I think this would make more sense than merging this article or the licensing article with "Comparison of document markup languages". - Bcharles 20:53, 25 February 2007 (UTC)

The licensing article should just be removed altogether. hAl 23:33, 25 February 2007 (UTC)

[edit] Misc from Rick Jelliffe

The text "Office Open XML states that year 1900 must be considered as a leap year. This bug originated from Lotus 1-2-3, and was preserved by Microsoft Excel for backwards compatibility." is wrong. OpenXML uses vanilla XSD/ISO8601 date format. SpreadsheetML uses a numeric index to store dates in data cells, to reduce parsing/calculation time on load and save. One of the two choices uses dates based on the beginning of the last century: there are no dates allowed before 30 Dec 1899. AFAIKS, for the first two months in 1900, there can be the chance of an out-by-one error depending on whether you take 0 to be Jan 1 1900 or Dec31 1899. I suggest instead "In SpreadsheetML, one of the two date index format may have out-by-one issues for January and February 1900." Rick Jelliffe 01:53, 2 April 2007 (UTC)

The section on interoperability testing omits the use of schemas for validation. Acceptance and production of valid documents are the primary mechanism or bottom line of standards conformance: if the data in or out is not valid, nothing else matters. Open Document Foundation has recently put up an online validator, and it would be good if MS did the same. I suggest the following paragraph at the beginning of the section: "The OASIS and Ecma standards both supply schemas (XSD or RELAX NG) by which the XML documents can be "validated": validation involves testng that the element and attribute names are correct, that they have been used in the allowed positions, and that data values have the values or ranges required by the standards. Conformance to document standards from ISO is primarily couched in terms of schema conformance; tests for application conformance to the semantics of the standards is not dealt with by the standards." 124.189.122.186 04:29, 2 April 2007 (UTC)

Retrieved from "http://en.wikipedia.org../../../c/o/m/Talk%7EComparison_of_OpenDocument_and_Office_Open_XML_formats_fee2.html"