Bright Eyed Mister Zen
Kimbro Staken's personal view on the world.

My Sites
 
Interesting People


 

 

Tuesday, June 25, 2002
 

iMac Sales

Bushels of new iMacs piling up. Apple Computer, which initially couldn't meet demand for its new flat-panel iMac, now appears to have the opposite problem. [CNET News.com: Personal Technology]

This is rather disturbing to hear, but the article makes an interesting point that I'd never really thought about. It suggests that laptop sales are having a serious impact on the sales of desktops. Thinking about this, it really makes a lot of sense. I know everyone I know who's wanted to buy a Mac has wanted a laptop and I almost always recommend laptops. In fact, even though I have a Mac desktop I still do everything on my Powerbook. At this point I really doubt if I'd buy a new desktop.
7:00:09 PM    


Oracle 9i Release 2 XML DB

I've been planning to write an in depth review of Oracle 9i Release 2 XML DB for some time now, but the time just isn't emerging. So for now I just want to write a short piece to give my initial perspective on the new version of Oracle.

I've spent a fair amount of time in the last couple weeks studying the new Oracle XML features and I have to say I'm very impressed. Going in, I was highly skeptical of the hybrid approach that they were taking, but after actually using the product I believe they have a real solution. We'll ignore the fact that Oracle is massive with a 1.5GB download and a minimum 512MB RAM requirement and focus on the functionality.

Oracle 9i is now by every definition that I know of, a native XML database. And in my opinion is also the most complete and powerful native XML database around. It allows you many options for how XML is stored in the database. It can be stored in CLOBs, mapped to tables or any combination thereof that may be necessary for an application. Regardless of how the XML is stored it appears to SQL as an XMLType. This enables XMLType as a type within the relational model and opens the full richness of relational manipulation to XML data. They've provided some extension functions to SQL to enable accessing data within XMLType instances and then using that data within SQL queries. I expected this to be very cumbersome, and I guess it is, but the familiarity of SQL makes it surprisingly comfortable to use. Something I can't say about XQuery. In fact, I really have to question why Oracle would even want to add XQuery given how their SQL extensions work. There just isn't any need for it. Hell, with their SQL method you can even update XML, something XQuery can't do at all.

There are definitely flaws of course, for instance it apparently doesn't handle DTDs and they need to work on the updateXML function some more. But to me these aren't fatal and the reuse of the mature relational engine with full constraints, triggers, views and all the rest is pretty powerful. As is the ability to easily intermix data stored relationally and data stored as XML. The use of CLOB storage isn't great either, but there's absolutely no reason that other storage methods can't be provided in the future.

At this point I have to say there's little reason for products like Software AG Tamino to exist anymore. Even if there is some speed advantage to Tamino and brethren, it's simply lacking in almost every other regard. This is especially damning given that Tamino and brethren will cost more for most simple implementations. There's plenty of room for native XML databases to play in smaller apps and for apps requiring DTDs, but when it comes to large corporate repositories of data, there's really no reason now to go away from the existing relational databases. I expect IBM and Microsoft to be coming out with products of similar power in the near future.

Hopefully, someday I'll get around to writing a more detailed look at the XML features in Oracle 9i, but for now I'll say if you need to store XML it's definitely worth giving it a look.
5:35:14 PM    


Zoomed Interfaces

Ever since I read The Humane Interface by Jef Raskin I've been thinking about the concept of zoomed interfaces and where they may have applicability. In particular it seems that PDA software should have this facility built into the core platform. I constantly find my self wishing that I could zoom out to any arbitrary level while using a PDA. Some apps code this facility, but they always seems to choose preset factors that aren't quite what you need. A true scalable display is really needed. I guess a PDA environment based on vector graphics like Mac OS X would be able to provide very cool facilities in this regard. In fact being able to scale any arbitrary window in the OS could really be useful if it's able to render it smoothly enough.

It seems with PDAs expected to hit 1Ghz in the next couple years now is a prime time to really rethink the PDA operating system and come out with something more optimal for the size constraints. A vector based GUI with scaling built into the platform could be an important part of that.

BTW, this isn't a new idea either the Psion Series 5 has silk screened buttons to control the size of the data displayed. It's limited because there's only a few predetermined sizes but it is still a very useful feature of the platform. Arbitrary scalability would be outstanding.

There are a couple interesting research projects in the area. Pad++ and its Java counterpart Jazz. One interesting app built with Jazz is a photo management app very similar in concept to iPhoto. It doesn't have the polish of iPhoto of course, but the way it manages the photos through zooming to me seems to actually be superior. iPhoto provides zooming too, but only over the entire photo list at once.

Anyway, as machines get more powerful and vector based environments like OS X become more prevalent I believe there could be some real value including this capability as part of the OS. For PDAs it should almost be a no brainer. I'm really hoping to see a new PDA OS emerge.
4:54:07 PM    


Complaining Too Much

Wow, lately I've really been complaining too much. This blog is sounding more and more like a bashing platform for Microsoft. Definitely not what I really want. I guess I'm just really frustrated with the computer industry and its general lack of innovation right now. Computers used to fun, now it's all digital rights management, patents and protecting the interests of big companies. The little guy just doesn't seem to have a place anymore and the big guys are all too busy sniping at each other and playing follow the leader over way over hyped ideas. Ugh, there's that complaining again. I'll try to stop now. I'm going to start posting a lot more about research projects and try to ignore the more infuriating plays of the bigs.
4:29:56 PM    


Office, Tablet PC

From Microsoft, a new Office, Tablet PCs. Looking to expand its desktop software monopoly into new areas, Microsoft unveils a new version of its cash cow, Office, and the launch of the first Tablet PC systems. [CNET News.com]

Seems once again Microsoft is claiming Office will have better support for XML. This is at least the third straight version for which they've made this claim and they've never done it right. Allowing XML would mean, allowing anyone to read their file formats. And given their stated goal for XML they can't obfuscate it like I"m sure they want to. Maybe they'll finally get the balls to allow office to save proper, readable XML files that can be reused in other contexts outside of Office. I wouldn't hold my breath though.

Here's another great Microsoft innovation.

The highlight of Raikes' Tablet PC talk was a presentation of Microsoft's electromagnetic pen, which was met with applause by the audience.

Raikes demonstrated how the pen can move the cursor when it hovers above a Tablet PC and can create an ink-like line when it touches the screen.

Funny, I have an IBM thinkpad convertible tablet/laptop that must be about ten years old and can do exactly the same thing. This was always considered a kind of bad idea, because if you loose the special pen the device is useless. Mine has a little piece of string tying the pen to the screen. It's so nice to see Microsoft is still trying to pass off old ideas as innovations.

Of course this whole Tablet PC thing is just regurgitation of old ideas. Back when they decided to destroy Go Computing they stole all their ideas and then went through this whole pen windows exercise. My Thinkpad runs Pen Windows, or Windows for Pen computing whatever it was called. Pen Windows was crap then, and Pen Windows XP will be crap now.
2:19:20 PM    


Be and Palm

Be Inc. completes takeover of Palm. NeXT-style coup [The Register]

Many of these people are not only former Be, but also former Apple. They might just have a chance to do something great. At least I sure hope they do. The current PDA offerings are pretty pathetic.
1:50:08 PM    


Forest Fires

The wild fire situation in the western US is getting frightening. In particular, the rodeo fire hits close to home. Patty's parents and grandparents both own houses that are directly in the path of the fire, if not already consumed. They were evacuated last Thursday and she told me it was about a 95% chance that the fire would reach the area. Whether the houses get destroyed or not is another question, but the neighborhood will certainly be a different place after this. I've seen a lot of burned out forests around the area up there and it's amazing to think of the whole Showlow area looking like that. The entire area will just never be the same. There are a lot of people who live around there and even if their homes survive the loss will still be staggering.

The truly frightening thing about this, it's only June!
4:00:44 AM    


Databases Wrestle XML

Databases wrestle XML. IBM, Oracle, Microsoft wars hinder wide-scale Web services deployment [InfoWorld: Top News]
3:48:04 AM    


Sharp PDAs

Sharp hopes to regain PDA market share in Japan [PDABuzz.com]

If they hope to do this they really need to improve the software on the Linux based Zaurus models. I have a SL 5500, and the hardware is great stuff but the Qtopia environment that runs on it is typical Linux clunk. It looks good at first inspection, just like KDE and Gnome, but you rapidly run into all kinds of rough corners. It's very disappointing too because I really like the Zaurus and the fact that it has a full command line environment. Because of this, despite my criticism of the GUI, it's still the most useful PDA that I own (I own a lot). It just doesn't make a particularly good PIM, because the apps and GUI environment need a lot of work.

Really, the Qtopia environment has some real fundamental problems. Chief among them is that it's a Pocket PC copycat. Basically trying to cram a desktop style environment into a PDA form factor. This doesn't work for Pocket PC and it doesn't work for Qtopia either. While I certainly hope and expect to see Qtopia improve considerably as it matures, this fundamental problem is going to seriously impact the usability of the system.

My feeling is that PDA environments should be ground up rethinking of the computing metaphor that really take into account the platform limitations. Qtopia plays lip service to this, just like Pocket PC, but really it is still too similar to the desktop. For instance on the Zaurus most apps have a menu bar, tool bar and an OS task bar. So right off the bat you're giving up almost 1/3 of the screen space to UI. The use of the same interface widgets as QT makes app porting easier, but results in a less then optimal set of applications. I would much rather see an entire rethinking of the UI for improved usability rather then easy app porting. Because of the form factor, any truly usable app is likely going to have to be custom written anyway. Because of the current widgets, you see way too many annoying things like text edit boxes that force you to scroll horizontally. Yes app authors can fix this, but this kind of thing should be handled by the platform. It applies to all text input widgets, they should be aware of the size limitations and adjust accordingly like they do on the few real PDA environments such as Palm and the Newton.

The platform also provides very weak facilities for app integration. Basically a PDA is a portable database with apps to provide different views/manipulations of the data, the Qtopia environment completely fails to understand this. Each app has isolated data and while they made a good choice by using XML for the data format, they shot them self in the foot by choosing a completely inflexible design for the XML data. There's no platform provided way to share data between apps and the XML formats are so poor that adding it on is difficult. Now they're redesigning their data handling to fix this, but really this is a pretty basic idea and it's sad how bad it was handled.

Anyway, Qtopia will get better and this will make the Zaurus better. Unfortunately, the most serious problems with it are of a nature that I doubt will ever be resolved. They're problems of innovation, consistency, usability and a flawed philosophy of the overall approach. I love Open Source, but I've still yet to see a GUI project come out of the model that doesn't exhibit these issues. Complaining aside, I definitely do still like my Zaurus better then anything else out there, it could just be so much better. At least it is open source, so if I actually had the time I could improve it my self. I've poked around the code, maybe someday I'll do something more then complain.
3:02:26 AM    


X++ XML Programming Language

x++: The World's First Full XML-Based Programming Language Released!. Top XML Jun 24 2002 5:37PM ET [Moreover - XML and metadata news]

This is interesting, not sure if it's a good idea or not, but interesting.
2:09:59 AM    


Palladium

Microsoft to reveal Palladium source code. The software giant plans to release the source code to a key Palladium component to help foster trust, despite a history of arguing that doing so would hurt security. [CNET News.com]

Bah, this is just too sickening to really comment on. Microsoft's vision of the computing future just gets more and more unattractive. It is funny about showing people the source since they've used that so often as an argument against Open Source in regard to security.
2:02:48 AM    


Omniweb 4.1 - Better Late then Never

OmniWeb 4.1 released [MacCentral]

Omniweb has finally shipped the final Omniweb 4.1. I thought this was supposed to have happened a couple weeks ago, but seems the delay was because they had to create a new icon. Anyway, Omniweb is still my main browser so I'm happy to see this.
1:53:37 AM    



Click here to visit the Radio UserLand website.
Click to see the XML version of this web page.
Click here to send an email to the editor of this weblog.
© Copyright 2002 Kimbro Staken .
Last update: 7/1/02; 6:39:12 PM .
This theme is based on the SoundWaves (blue) Manila theme.
June 2002
Sun Mon Tue Wed Thu Fri Sat
            1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30            
Apr   Jul

Click to see the XML version of this web page.


Resume:

Yep, I'm available.
(Word/RTF) (Text)
kstaken@xmldatabases.org

Software:

Xindice (Formally dbXML)
Xindice XML-RPC Interface
Xindice HTTP Interface
XML:DB API Reference Implementation
Jive XML Database Port
FreeDB CD Database to XML Data Generator

Writing:

Introduction to Native XML Databases (xml.com)
Introduction to dbXML (xml.com)
An Introduction to the XML:DB API (xml.com)
Xindice Users Guide
Xindice Developers Guide
Xindice Administrators Guide
XML:DB API Specification
XML:DB API Use Cases
XUpdate Use Cases
ICEPick Peer to Peer Personalization

Presentations:

Native XML Databases Why Bother?
(HTML) (Powerpoint)


Click here to visit the Radio UserLand website.