November 01, 2005

All Right, That's It... No More "Mr. Nice Fence Post Sitter"!

� Performance analysis of OpenOffice and MS Office | George Ou | ZDNet.com

In my last blog, where I did a high-level technical evaluation of Microsoft Office 2003 and OpenOffice.org 2.0, I showed that OpenOffice was a memory and resource hog. Contrary to popular belief (among Open Source advocates), Microsoft Office came out very lean and fast while OpenOffice.org Office Suite was just the opposite. Some couldn't accept the numbers and complained that the Task Manager numbers may be inaccurate and hiding memory usage. They demanded more proof, so here it is.

I'm not even go to worry too much about performance although anybody who has used Office... I mean truly used Office, not just opened up Word, typed a few words, and then closed it just so you could honestly state "I've used Office, and I just don't like it..." knows that the above linked article is spot on in every way.

But I'm going to set this aside and focus on something all together different than performance. Argue performance all you want. But its tough to argue whats not even comparable, so this is where I plan to push my focus for this post.

In the paragraph just above the phrase "used office" absolutely must be viewed to mean MUCH more than the editing of a text document once and a while, or an Excel spreadsheet from time to time...

Why?

Well, there's a lot of reasons. But if all you have "used it for" is to bold some text, write and update your static, paper-based resume, and other activities of this sort... then you really have never "used" MS Office and should probably use MS Works or, if you'd like, OO.o, although please don't take that to mean I am comparing the two and suggesting they're the same in function, and instead price (that is of course if, like most Windows users, you got a copy for "free" when you purchased you're PC)

So, for those still with me, the others reason's are plenty and they all come down to this slightly colored description of Microsoft's Office 2003 Application Suite...

It flat-out Kicks-A$$.

It's faster [BY A LOOOOOOONNNNNNNG SHOT!] than OO.o, more reliable, more user friendly, more collaboration friendly, more programmable (speaking from a Macro-level, COM/ActiveX/OLE(if we have to go that far back into the cross-application object/component-based acronym heap)-level, etc... etc... etc...)

Oh, and yes, when I say more programmable I mean this even taking XForms support in OO.o into account. Don't get me wrong. XForms is cool. In fact I love XForms! Micah Dubinko (W3C XForms Spec Editor/XML Hacker extraordinaire) is someone in whom I consider both a good friend and colleague in whom I have a HUGE amount of respect for. In fact we have and will hopefully continue to work together on several fronts... some you may know about and some you many not. But's thats not what's important (at least not for this post anyway). Whats important is the understanding that XForms wasn't designed to be a full-featured replacement for WinForms... it was designed from an internet/browser-based cross-platform replacement for HTML Forms. So tweaking it, building an editor, and then trying to use it to take on Microsoft from a extensibility standpoint?

Come-on...

You're not even in the same ballpark, or even playing the same game. Trying to suggest "well, you can build cross-platform, web-enabled applications with it and you can't with Office" is the most sorry attempt at trying to belittle MS and MS Office I have ever heard.

First off, building web-enabled applications is absolutely possible with MS Office. Will they work across every browser on every platform.

No.

But MS has never been focused on building Office as cross-platform mediocrity. Instead they've focused on being the absolute best available office application suite on the Windows and Macintosh platforms.

Disagree? Sorry, you're wrong.

There's NOTHING even close.

While I will have to verify these numbers to be absolutely certain, Windows accounts for about 95% of the desktop market. Mac? About 3%. So you've got 98% of the desktop market covered, with the remaining 2% of the desktop population who wouldn't touch MS Office if you paid them...

So why put the resources into people who won't use your product?

Instead, this same 2% group of folks have been scraping by with KDE Office knock-offs, Star office freebee's, and now OpenOffice.org which, while definitely a good start, still lags behind MS Office by, many, many, MANY years.

Don't buy it? Then don't listen to me. Jono Bacon has a great post from last month last month entitled "Opening the Potential of OpenOffice.org" and, while I may be mistaken, I'm pretty sure he never states "This is it!!! We've got the Golden Ticket" and instead "We've got problems to fix, let's fix them." This is of course a paraphrase and a short one at that. Maybe you feel the message he was sending out was different. If you haven't read it, then please do, and then come back and continue reading and tell me if flat-out, totally, and completely disagree.

--
All right, either you're back, or you haven't left... either way, I'll continue and then see where this leaves us. :)
---

Being one who has truly used both products (MS Office and OO.o) A TON I can attest that whomever continues to claim that OO.o is an adequate replacement for MS Office is either a lier, doesn't know what they're talking about, has never and probably will never use a MS product long enough to truly state they know what they're talking about, or doesn't use enough of MS Office to justify the expense. My guess is that a majority (by a long shot...) of the OO.o supporters are of the last variety... They simply don't use or have a need for the extended features that MS Office offers them. For this group of folks... Go for it... If OO.o gives you what you want and need in an office software suite, then use it. I'm not suggesting OO.o is bad or completely unusable. For a lot of tasks, OO.o is fine. It gets the job done.

But there's a BIG, HUGE, GIGANTIC, E-FUCKING-NORMOUS!!! difference between an office-type software suite that adequately supports a decent feature-set of office-type software related tasks, and having a fully-fledged Office-Suite of EXTREMELEY fast, EXTREMELY usable, EXTREMELY programmable/extensible, EXTREMELY collaborative-enabled applications that are well on their way to moving directly into the next realm of "software documents" where the documents themselves are the application... or better said, the documents describe the application using Meta-Data that states "this is the kind of data I am" and "this is the component I need to be able to perform this task." Don't have the component? You're software subscription mechanism of the not-to-distant future will take care of that. You only need that component once... you only pay the price to use it once. They're called compound documents and they've been under development for longer than most of us have even owned a desktop computer (no, not a laptop... a tried-and-true "heavier than a ton of bricks with a monitor twice that weight and half its size to go with it" desktop that we all took pride in being the first on the block to own.)

You see, this is what concerns me about the OASIS Open Document Format. It's a great document format, don't get me wrong! It's open, it's well thought through, it covers the gamut when it comes to describing how a document should look, what it can contain, it allows for user meta-data and even describes how an application must handle that meta-data if it doesn't know what it means (preserve it! don't muck with it... whatever you do, if you don't understand it, don't just delete it or insert anything into it!!! LEAVE IT ALONE! < My own words of course, but that the general idea.) It has XForms as part of it, and even allows for scripting.... GREAT!!! So we've basically got HTML, CSS, XForms, and Javascript, that's XML-based and as such its an eXtensible Markup Language-based format nicely tucked into a well written specification by some really talented best-of-breed developers and spec writers.

So whats the problem?

If the current "DOM" way of thinking is just fine by your standards, then there isn't one. The OASIS Open Document specification heavily embraces the existing document model where thats all it is... just a document. While its extensible (a good thing) and could technically be "hacked" to handle things of this nature, using this specication there's no way for me(speaking figuratively), XYZ Software Company, to build productivity enhancing, collaborative-based software components that based on a common component-based communication protocol can "speak-up" to the Operating System handling the document-wide transactions taking place and sat "hey, I can do that... thats what I'm designed to do and what I've been licensed to handle."

And the funny thing is is that this isn't "cutting-edge" stuff... It's COM. It's DCOM. It's been the focus of software engineering for decades. So why is this then not the focus for developing a truly open document specification?

Well, it is. Just not by the same group of folks.

If you're answer to this is "its two different ways of thinking... who's to say one is right over the other." Sure. Fine. If you're cool and happy with "heavy" systems that use scripting for automation, security, extensibility, etc... then there's nothing to argue.

But I'm not OK with such a systems and I know a lot of other developers who have no desire what-so-ever to continue in this "application heavy, document-lite" way of thinking. We don't want definition to how "creative" we can be. We don't want bounds of "document expression" that we must be confine ourselves to. Speaking the same language is one thing. Understanding one another is a necessary component of the COM-based way of thinking.

But COM is not DOM. Document Object Models are quite different than Component Object Models. Document Object Models don't mean "Object" as in component objects. Document Object Models suggest the document IS THE object. We traverse the Objectified-document looking for these pre-defined, pre-determined "style-elements" such that the systems can then apply these style characteristics to the users insterface, whatever that might be.

Component Object Models don't focus on describing the document as far as style is concerned, and instead on the communication side of things (as outlined earlier.) COM (at least the spirit in which COM was developed) is an open ended model where the style definitions can be as far and as wide as they want to be. But if you develop a "style-definition" that's different in any way, then you have to make sure that you either develop the component(s) to handle this new style definition, or open things up and convince others to develop components to process your style language. SVG would be a perfect example of just such an endeavor.

Now don't get me wrong. There are some pieces of the DOM way of thinking that are good, and should be carried on into a cross-breed of DOM and COM. This leads us right back to compound documents and is exactly where the focus needs to be.

Don't build a fence and tell me not to leave its confines. I can build my own fence, I don't need you to build it for me.

Now teaching me how to speak the native language such that I can communicate with my neighbors, become a part of a community, developing my own relationships based on my own personal likes, desires, needs, self expression, or other areas of my own choosing, possibly building a fence if I desire, or not if I don't...

Now that's a world that I would like to live in... both physically and virtually.

In fact, I do live in that world... and I have no plans to be leaving this world anytime soon.

Paul Graham hit on something in "Hackers and Painters" and he hit on something big. In fact, I would almost go as far as suggesting that not even Mr. Graham knows just how significant of an impact he has already had, continues to have, and will continue to have for many, many years to come. But then again, I don't know Paul Graham nor can I speak to what it is that he truly understands or would even be willing to admit in regards to his own beliefs as to the impact both his books and his way of thinking have, continue, and will continue to have had on the generations of computer programmers to come.

But I do know one thing... As Mr. Graham lays out in words better than I can even speak, software is art and art has no definition, has no bounds, and has no borders.

But it is confined... but its confined by only one thing:

Your imagination.

I think I'll just leave it that. Enjoy! :)