Slashdot Mirror
RTF Vs. OOXML
Rob Weir has an interesting essay comparing the viciousness of RTF and OOXML: "The [document format standard] concerns of 2004 (or 1995 even) are very similar to the concerns of 2007... 'RTF is defined as whatever Word saves when you ask it to save as RTF.' This should sound familiar. OOXML is nothing more than the preferences of Microsoft Office. Whenever Word changes, OOXML will change. And if you are a user or competitor of Word, you will be the last one to hear about these changes. The coding of Office 14 a.k.a. Office 2009 is well underway. Beta releases are expected in early 2008. But are file format changes needed to accommodate the new features being discussed in Ecma? No. Are they being discussed in ISO? No. Are they being discussed anywhere publicly? No. By owning the 'standard' and developing it in secret, in an Ecma rubber-stamp process, Microsoft rigs the system so they can author an ISO standard with which they are effortlessly compatible, while at the same time ensuring that their products maintain an insurmountable head start in implementing these same standards. Is this how an open standard is developed?"
Up front disclaimer: This article has a tangible odor of troll, so don't blame me and the other posters for responding in kind (flamebait, troll, offtopic, etc.)
FTS:
I wouldn't say this is entirely true (effortless) on Microsoft's part. Any user of any Microsoft product is well aware of how difficult it is to work in and out of various new vs. old formats. Yes, even Microsoft has a difficult time being compatible and interoperable with Microsoft (actually, I seem to have better luck overall with interoperability using OpenOffice...).
And, also FTS:
Actually no, usually Microsoft takes an existing open standard (e.g., sockets), implements it poorly (winsock), and puts it everywhere (95,98, NT, XP, etc.) forcing the technical community to re-adopt the standard in Microsoft's cast.
It's official, I've been on the internet way too much. I saw "RTF Vs. OOXML" with just a quick glance and read it as some new, bizarre acronym like "ROFLCOPTOR".
Those who believe the Internet is private,
find their privates are on the Internet.
No.
Here's a copy of the draft OpenISO.org "Problem Report" entry for this issue:
Microsoft's attempt to essentially unilaterally dictate office document standards is an abuse of their dominant position
Problem description:
Normally standardization is conducted by means all interested parties participating in a discussion of the desired features, so that all interested parties have an essentially equal opportunity to develop products implementing the standard.
By contrast, OOXML is simply documentation of the document format that Microsoft's products already use, and there is no indication that Microsoft would intend to make the details about future versions of OOXML available to competitors before Microsoft is ready to release their own implementation of the new features for public beta testing.
Expected impact:
To the extent that OOXML is accepted as a standard, all of Microsoft's competitors will be encumbered with a permanent economic disadvantage.
Possible solution:
Reject all claims about OOXML in some way being a standard, and take legal action, on the basis of national and international competition law, against Microsoft as well as against Ecma and all other organizations which are guilty of aiding and abetting Microsoft's anticompetitive actions.
I think I could beat OOXML, if I took a few weeks to train up with some old kung fu movies beforehand.
-- RTF
Slashdot Burying Stories About Slashdot Media Owned
It's definitely important that those who agree that OOXML is not a good standard should help organize a list of problems that can be easily seen by the members of the upcoming ISO OOXML ballot meeting in February 2008 and all the Internet in general.
OpenISO.org, an independent open organization much inspired by slashdot, is planning to include the issue of this post in the problem report document produces in its OpenISO.org Review of OOXML. OpenISO.org is asking for help to organize the comments of your country in a wiki at http://f29500.openiso.org./
Please have a look at all the problem reports at http://f29500.openiso.org/ and help to include more and organize the ones already included, even if only one or two. The more documented and organized the OOXML problems are for discussion in an easy accessible manner, the less likely it will be accepted as a standard.
ps: I'm not associated in any way with openiso.org, it just seems to be the right thing to do.
call it ROTFL
intellectual property law is philosophically incoherent. it is your moral duty to ignore it or sabotage it
My suggestion: Get a better title for the slashdot piece. How about "OOXML will not work just like RTF failed."
RTF? RTF what? What am I supposed to be reading to eliminate my ignorance of some set of operations?
While making a new standards body like OpenISO sounds like a good idea, I don't want to rain on that parade.
However, I think there is also a problem with the national standards bodies. They can vary from a formal technical committees answerable to democratically elected governments according to what their country needs, through to a ragtag bunch of nobodies who can dictate whatever they want according to their specific corporate interests. I think ISO needs to start with itself and standardise how national bodies work.
Also I think that if you are unhappy with the decision your national body made, then you need to either seek to get on it (or make a group that raises funds to get one of you on it), or setup a competing national standards organisation, get to work, and then try to replace the old one as ISO's National standards body for your country.
My little Linux and tech blog
On top of OOXML being developed in a closed environment, MS Office is not even using the proposed ECMA or ISO spec, they including all types of tie-ins. This article explains more: not even compliant
Computers do Input, Output, Processing and Storage. Its been that way for more than a half centuy.
At some point people thought it would be cool if that wasn't the case and dreamed up lots of crud to put in text books sold to college students and they made lots of money but hasn't changed a thing.
It still doesn't fix the problem that a word processor has an internal model of what the user typed. Its job is to output that in a way that is consistent with what it's showing the user and what the user told it to do. Now for some odd reason a large group of people come along and say "we want magic" and expect the input/output and storage models to be disassociated. How is that supposed to work? Remapping input? More levels of indirection? It sill doesn't fix the core problem.
I have worked on industry standards before. Writing spec is just half the battle. You then have the problem with implementation. Every company will implement it in slightly different ways. You would be surprised on how many ways there are to read a spec! Then you get in a yelling match over who is actually doing it correctly.
When you have a reference application to test with then you have less yelling.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
we must have ISO or ANSI write the standard and set up compliance tests for RTF and without input from Ms because Ms needs to take a good licking for their nasty behavior
There's more than one input and output.
The input and output on the screen and keyboard are part of the processing from the point of view of the storage.
The input and output of the file storage are part of the processing from the point of view of the user.
User - I/O - Processing - I/O - Storage
In addition there's a third I/O interface for printing, and a forth for online publishing.
There's no reason that the storage format needs to be tightly coupled to the display format. And, also, an editor doesn't have an internal model of "what the user typed", it has an internal model of a document and updates that according to the user's commands. In Word, that internal format doesn't even match what the user sees very well... there's no such thing as "nesting" or a "list" internally, for example, there's just a set of styles and rules about what paragraphs follow each other and lists are created and manipulated by dynamically updating the paragraph styles and next-paragraph rules.
If Word can manage to take this and generate nested lists on the screen and in HTML, and accept user's requests to change nesting depths and go through and update all the paragraph styles to match, there's no reason it can't do the same thing in its storage format.
> Now for some odd reason a large group of people come along and say "we want magic" and expect the input/output and storage models to be disassociated. How is that supposed to work?
So you saying that standards can not possibly work? That people want "magic?"
Hate to break it to you, but standards already work. Consider ASCII. Also ODF is already incorporated in several word processors.
Of course input/output and storage models can be disassociated, it's done all of the time.
There are two kinds of vendor standards -- the one in which the vendor publishes what they've done, and the other in which the vendor publishes what they will do. This article contends it's the first. Is OOXML the first or the second? We'll see.
Don't piss off The Angry Economist
Microsoft's published RTF specs for a quite some time now -- the latest version of the spec is 1.9 and you can download it from Microsoft in your choice of binary .doc or MS-OOXML .docx, sorry no .rtf!
The spec is actually not bad, though the continued efforts to shoehorn in new features gets a little laughable. Here's an example of an RTF-reencoded XML tag from the spec:
As far as I know they've never tried to have RTF ratified by any standards body, but it's still very widely used. People have a lot of files named .doc around that are actually RTFs, and some word processors (AbiWord for one) actually use .doc-named RTFs as their "Word" format, since, having a spec, it's a lot easier to write than the binary .doc format. By design, old Word versions and non-Word software ignore any tags they don't understand, and I'd guess that most modern third-party RTF parsers and encoders are designed around the 2000 RTF spec (version 1.6) without all the new stuff.
comma delimited data, for example
as i noted in the other post we need ISO and ANSI to write the spec and set up the acceptance test
how many foos i have falling over commas and or quotes that are enclosed within quoted text -- or who cannot write a number properly
AGGRAVATING! these foos need to get a big fat F on their report card and join Hillary in re-hab
Will the next version decide to save as ".doc9"?
Between the file format change and the GUI change it feels like entirely different App co-branded by MS.
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
And the one that Microsoft typically uses: the standard which is never published.
MS-Office has been called a "standard" for so long, people believe it. And they conflate a "standard" piece of software with a document "standard."
In any kind of engineering, there is only one true standard-- that which is agreed-upon by the manufacturers and/or engineers themselves. Bridge architects will never use a non-standard size bolt, nor a non-standard metal. Why? Because the weld strength is too important. Because the dielectric interaction of different metals and alloys results in a weaker bridge. Because lives depend on it.
Corporations who push their own format as a "standard" harm the industry. If they never publish their formats, or publish a bowdlerized version of their standard (MS-Office 2007 doesn't even conform to the OOXML specification as published), they are selling their customers a trojan horse.
Microsoft is to software what Budweiser is to beer.
Having been involved in two standards bodies, those bodies found enormous benefit to having reference implementations being built at the same time the standards were being developed. It certainly helps drive out issues during the standardization process that would otherwise/sometimes make the standard unusable.
Granted, what Microsoft is doing may not be a 'reference implementation' but still... there are benefits to doing *some* implementation in parallel with standardization.
Ask and ye shall receive. It was released in 1999, so it even meets your "long time ago" requirement.
One of the more amusing lines of the specification;
Because of the way Microsoft word processors implement tables, and the table-driven approach of many Microsoft RTF readers, it is very easy to write tables in RTF that will crash Microsoft word processors when you try to read the RTF. And of course, how much value there is in the specification for a format more than a decade obsolete, I'll leave as an exercise for the reader."I've got more toys than Teruhisa Kitahara."
But given the MS of embrace and extend, I must resign myself to a world in which MS products are just too unreliable to use for real work of any significant magnitude. I know that RTF is not sufficient to make the fancy memos people like, but it does seem to work.
"She's a scientist and a lesbian. She's not going to let it slide." Orphan Black
Well put, I think its hypocritcal and unconstructive to try to undermine Microsoft through the courts, legislation, and standards bodies. All of these underhanded tactics will come back to haunt FOSS, IBM, and everyone else using them because new competitors will adopt the same tactics rather than creating something truly better than the leading technology. These are dangerous precedents that hinder inovation and creativity.
we need a document format that is both simple and open so that documents can easily be exchanged
.rtf
.rtf and we need a certification test so that vendor products can be evaluated and given a pass/fail grade
clearly this CANNOT be trusted to a corporation; it has to be owned publicly.
HTML might be used, but while HTML is fine for browser documents it lacks some features needed for print format.
PDF is the exact opposite, it prints perfect but browses poorly and requires specialized compilers to build documents.
and so at least for now there is still a place for something like
but we need an open standard for
But the key point is that the whole process was open.
The ODF standard was designed by an comitee (OASIS), where several of the various office suite maker collaborate.
The procedure has been openly documented and everyone was able to know what was being done.
The standard was available "in advance" of the products, not the other way around. The standard will be followed by the various maker as you said.
The critics made to microsoft in TFA are that Microsoft is designing the standard alone without consulting the concurrence or even letting them know what they are doing.
The next iteration of OOXML is probably going to be made available "afterward" : they're going to first build MSOffice 14 and then publish "What we've done new in MSOffice 2010" or some other king of list of modification they did (notice past tense) to the standart. As you say, it's the product which will define de standard.
Yes, in both case the standard are published.
Yes, in both case they started life as internal representation of specific softwares.
BUT, OOXML is still an internal representation of word, and is best defined as "whatever the next version of word spills when you hit "Save" ", if Office change, OOXML will change with nobody knowing it in advance and being able to take part into the process. Want to make cross-operating software ? Please wait until Microsoft takes their next product to the market and makes it mind about what they'll throw next into it. Too bad that this will introduce delays into your own product.
The "standard" is still a moving target, the only difference with reverse engineering is that nobody needs to decypher cryptic binary data but only read 1700 pages, appart from that it's the same "play catch up".
WHEREAS ODF has been beated into a standard by a body where different vendors/makers could give their opinion and everyone can be informed of potential modification of the standard as it's a public procedure.
Want to take part in the development of the next standard ? You can !
Want to write software compatible with it ? Just stick to what is published in the ISO standard no need to track a single specific vendor and it's proprietary product.
ODF may be a bad standard for some people but it's still an OPEN standard, as in "the procedure of the creation of this standard was open".
OOXML is just a "we let you read the text we print to document what we've thrown in Office 2010" closed standard.
Yup. I agree with you.
Must probably all the noise comming from Microsoft's marketing department "But see, our is a standard too : we publish the specs too !!!"
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
You wouldn't even need a few weeks if you used a proper training montage. Those only take like five minutes.
- None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
Just not Word.
Ask anyone in publishing or anyone who writes for a living what tool they use and there's little chance that they're going to tout the advantages of Word. Word is borked in so many ways, from a writer's point of view, that it's hard to know where the start when cataloging the problems. You can begin with the broken tokening used in lists, move to the Master Documents "feature" and finish up with the fact that what's displayed on the screen not only differs depending on which computer you're viewing it on, but it isn't even the same as what's printed. (Trust me, you never EVER want to have to try and fiddle page breaking in Word, where the breaks shown on screen are seemingly unrelated to the way the pages break when output to the printer.
A "real" writing tool either completely divorces the content and presentation (a la TeX) or presents on-screen an EXACT representation of what will be printed (for example, FrameMaker). Word tries to do the second, but fails. Miserably.
Interested in a Flash-based MAME front end? Visit mame.danzbb.com
It is important to remember that the whole point of TFA is that MS only publishes the documentation *after* they have implemented it, so there's no time for competition to come up with similar products in the same timeframe, MS is always ahead by definition. 'RTF is defined as whatever Word saves when you ask it to save as RTF.'
The documentation is available only in DOC or DOCX, why not in HTML or TXT or some international document standard like PDF or ODF?
From a comment in TFA, there seems to be a hidden list of previous versions:
"A while back Brian Jones, a Microsoft Office project manager, claimed good interoperability pedigree for Office applications since they can read and write "standards such as RTF and CSV". I took him to task over that one because neither of these is a standard; they are standins for standards. CSV is extremely unstable across versions and languages of Excel, and as far as I can tell is undocumented except to Microsoft developers. RTF is simply a representation of every version of Word. I took the trouble a while ago to compile this list, since such a list does not exist online from Microsoft:
* March 1987: An article by Nancy Andrews of Microsoft.
* 1.0 1987: Word 3.0 for Macintosh
* 1.0 June 1992: Word for Windows v2
* 1.1 Unknown, unavailable
* 1.2 Unknown, unavailable
* 1.3 January 1994: Word v6
* 1.4 September 1995: Word v7 (Word 95)
* 1.5 April 1997: Word v8 (Word 97)
* 1.6 May 1999: Word v9 (Word 2000)
* 1.7 August 2001: Word v10 (Word 2002)
* 1.8 April 2004: Word v11 (Word 2003)
* 1.9 January 2007: Word v12 (Word 2007)
The worst part about this "standard" is the license: it is packed in a Windows-only executable package and is licensed for noncommercial use on Windows machines only.
# posted by Blogger Sean : Thu Dec 20, 04:33:00 PM EST
OOXML is defective by design
If you find a typo, you may keep it.
I have a vague idea about what OOXML is, but not RTF
Imagine hell with more backslashes.
It goes from God, to Jerry, to me.
I came here to ridicule the people who tagged this article "flamebait", who, in their delicate minds, confuse righteous anger and unabashed criticism with meaningless name-calling. The article is factual and its point is razor sharp. Since when are we worried about hurting Microsoft's self-esteem?
Now, my post and the above is more accurately accessed as "flamebait". His point is lost inside a wall of meandering text, which is really just a a vehicle for his ineffectual expression of contempt for "FOSSies". Here's some advice - don't hide like you're afraid of losing your precious karma, and don't be so stupidly wrong. If you want evidence of consumer choices and Microsoft's growing irrelevance, look at the list here: http://www.amazon.com/gp/bestsellers/pc/
It's a lot less about 'FOSSies' mentality that you think.
It's not about forcing a standard onto Microsoft. It's more about having a standard with publicly available documentation that anyone can adopt or use. Having Microsoft (or any private entity for that matter) be the primary controller of a standard will definitely create a climate where Microsoft gets the upper hand. This is the entire reason that standards consortiums are created. I, personally, would have no issue if Microsoft were to create a standards consortium for it's OOXML, as long as other organizations could be directly involved as well. Anything less is definitely anti-competitive.
To me, it makes no sense to force a proprietary format, or an 'open standard' that is controlled by a single private organization on the general public. One could make the argument that the standard may be created without Microsoft's involvement. The unfortunate problem is that any standard created this way wouldn't be adopted. Make no mistake, Microsoft has the market share, and this is almost entirely due to the fact that many new computers come pre-installed with Windows/Office (This is also the reason that Vista will become the OS leader despite it's many shortcomings). All Microsoft has to do is not adopt the open standard, and push it's own formats.
Imagine, for instance, if you had to purchase wrenches and screwdrivers of different form-factors for each and every different car manufacturer. Or more appropriately, what if there was no HTML standard? Imagine having to purchase software (or an API/Protocol Stack) for each internet document server out there.
Spork.
P.S. Spork.
Yeah, I think the article is a little off base about RTF. It's not IP-encumbered. As you pointed out, you can download the spec from MS. There's a ton of OSS implementations. There are perl modules that read and write it. There's an O'Reilly book on it (free short version of it here).
Okay, so RTF changes when a new version of Word comes out. That means that, e.g., it shouldn't be used for archiving government documents, and it's not suitable as a universal format for people to collaborate on extremely complex documents using different software. It doesn't mean RTF is useless or evil. I write fiction, and RTF turns out to be a very useful lingua franca for magazines that accept electronic submissions. For a fiction manuscript, you don't need anything very fancy --- basically just the ability to underline, and put a header on each page. RTF works just fine for that, and I'm really, really glad that RTF is the de facto standard for this purpose, and not doc.
It's unfortunate that OOo's RTF support is so horrible. E.g., if you save a document from OOo in RTF format, open it, edit it again, and save it again, you lose the whole document. Yeesh!
Find free books.
If you filter the A/C's rant, there's a point of view in there which essentially says "just follow what Microsoft wants you to do". That's fine if you want stagnation of capabilities and to pay through the nose for the privilege.
Example, the Browser War was more than just a browser - it was hijacking the Internet Ecosystem which was supposed to be open and available to all. Microsoft saw open standard browsers and servers, particularly with Java, as a mortal threat to their platform.
To counteract, Microsoft developed IE and IIS to be a client-server relationship instead of a stateless browser as intended. Tools were widely distributed to create web sites for that system which were wholly incompatible with anything else. The goal was for anyone NOT using a complete Microsoft chain of technology to see a blank page on the Internet.
They almost pulled it off. The result was IE gained market dominance and, with the exception of exploits and treachery of completely hijacked computers, no other technical advances in browsers came about for many years. The fly in the ointment was they didn't have server dominance. Had they been able to overcome Apache, you can bet we'd be paying Microsoft for every page view on the Internet.
That's why we shouldn't just do what Microsoft says.
Most of the stuff on
I'd rather have a rubber stamped version of Microsoft's office formats than no version at all: pathetic as ECMA may be, ECMA does force some additional disclosures and documentation, and it fixes a version for a while.
Of course, people need to keep in mind that an ECMA rubber stamp isn't the same as a traditional ISO/ANSI standard and it doesn't make the format "open". But with that caveat in mind, it's still better than nothing.
Food for thought: I was hoping someone would come out with one document type to rule them all. It would be kind of nice if web pages could run from 'integrated' document like a PDF which, would adjust itself based on screen res. You still have your master document, and have your app do all the background code (but there could be add-ons for the language like Latex so you can add technical fonts for math).
If this were to become standard, we could have simpler web-pages, a standard document type with WYSIWYG compatibility, embedded fonts and changeable style sheets (especially for the visually impaired). It would also be nice since *one* language could help create portable, always "readable" documents and would code for any type of document and webpage. And it would probably be a lot easier for web-developers (one file to deal with) and web-browser programmers (one standard format and not a mingling of them all).
I don't know, but it seems with all the complaints that there's too many competing standards, now would be a great time to come up with *one*.
Personally and professionally speaking, with the amount of documents I've amassed in WPD, Word and god-know-what other formats and not many good "translation" scripts to one standard document format, it would be ideal to get this one standard format. Otherwise, it does make it hard to choose which to use.
How can these two words be used in the same paragraph?
What great fortune for rulers that men do not think.
RTFD :)
Is this how an open standard is developed?
I couldn't say for sure what is the norm for open standards development, but just about all _good_ standards are merely a codification of existing technology and implementations. Standards that are ratified before the real-world working experience exists are usually hell to work, or never get off the ground at all. Commonly they are hodegepodges of compromise between conflicting interests serving none of them well, are impossible to implement correctly, implementation efforts reveal gaping holes of under-specification, and the official 'reference implementations' tend to rather be useless, buggy, incomplete 'toy' technology demos than what their name would suggest.
Now, I agree that getting their standard ratified should not be just a rubber stamp process. There should be some vetting that might require MS to change certain things. But accusing the advantage of having the pre-standard implementation of being unfair is in itself unfair. There always will and should be someone who has that position, because the alternative is for the standard to suck.
You could argue that it should not be Microsoft who gets that position all the time, and that having someone there still doesn't guarantee that the standard won't suck, and I might agree. But argue those points, not for a 'clean-slate' standard.
The last thing we need is more paper tigers that are unusable but slow development of alternatives because "there is a standard".
sudo ergo sum
I don't even know why he's harping on Lunix. I can't imagine that a significant amount of people would actually want to use a Commodore 64.