Put the knife down and take a green herb, dude. (c) Ruffin Bailey 2001-2021

Anyone who has bothered to read this blog more than twice would likely know I like to use Java to build client-side, cross-platform apps, but would also know that I have a bone to pick with Swing, Sun's recommended UI widget toolset for Java apps with GUIs.

In short, Swing stinks. They tried to shoot the moon and make a "look and feel" (which is a well-made phrase) for Java that would act exactly the same no matter what platform was running the app. It's a great idea in theory, and theory's always been Sun's strong suit. Trouble is, when we boot up Windows, we run Windows apps. When we log into Mac OS X, we run Mac apps. Java has not by any stretch dominated the client-side market the way Sun planned, making the OS running the apps relatively inconsequental, and likely never will. The niche for the xplat LaF (look and feel) is, short-run at the very least, "d-e-d dead". People, understandably, want all their apps to look and feel roughly the same, and since they run more native apps than, say, Java, they want their Java apps to seem to be native apps. Until they're running more Java than native, until Java hits that tipping point [that it never will], Java will need to continue to look, act, smell, & eat native.

Instead of moving back to AWT, the Java UI widget set that essentially exposes native UI widgets to Java coders, with native looks, actions, and reactions, Sun decided to make Swing, originally made to look "Java", look native. So now we have a native look built into Swing, itself built on top of AWT which is, you guessed it, based on native widgets. Tell me that's not a little off. Instead of being one step removed from native, you're now emulating native on top of a toolset itself built on a toolset that really does hit native widgets. Boy, I seem to like beating that dying horse.

Anyhow, much longer intro than I wanted, but you get the point if it's your first time here. Of course emulated native never quite acts like real native -- sorta like Asimov's robots that, even when they looked like people, could still trip off a sixth sense in "true humans", and Swing's native looks continue to have real issues that, ironicly, AWT (with its unfortunately small set of widgets) doesn't, like mousewheel issues on Windows or the inability to use brushed metal windows on OS X. In AWT, neither are (the former *never was* a) problems.

Which is why this Javalobby survey drives me crazy.

Everyone has their favorite L&F that they use every time they start a new application. In this week's poll, we'd like to know which ones you guys prefer.

Personally, I prefer the JGoodies ones, especially with the various themes that come with it. What are your favorites?

Look, the answers you get are going to be from Java coders who write quick little apps for in-house or personal use only. Nobody picks a third-party LaF for a large-scale, commercially released, *cross-platform* app unless they want the worst sort of QA/QC testing issues imaginable. Seriously, how many things do you want to have to double-check when your UI isn't working for ten customers running Mac OS X 10.0? The answers are silly -- like the person who answered the Napkin look and feel that I've discussed here before. Its whole raison d'etre is to make a prototype app *look* like a prototype so that your customer doesn't get the psycological impression that you're almost done once you've finished the UI. Nice, but hardly a favorite LaF -- unless you never get to the point that you've actually gotten an app far enough along to entered final testing. Nobody -- and I mean nobody -- should ever have unit tests for their UI run while in Napkin LaF. That's a complete waste.

The right answer for 99.44% of Java coders is one of either Metal or native, by the way.

But the lesson here is that questions like that -- even the original author's answer of JGoodies, which I'd also hesitate to recommend using, though their PlasticXP on WindowsXP (see a theme? Single platform release, trying to look 'more native than native'; why are you using Java anyhow?) doesn't look that bad, for a Java app anyhow -- show a circular problem from which the Java Desktop is going to have a hard time emerging. People writing the apps don't understand Java's strengths nor their customers' expectations for their apps, and until they do, we'll continue to have a hard time even pretending to have the potential to approach that tipping point. The only place even Metal (and now Synth and Ocean), that is, anything *but* native, really could belong is on Sun's own Sun Java Desktop System, where they could potentially make the native look and feel like Java.

Now if I'd just make myself program in my free time instead of blog... And while I'm dreaming, somebody buy me an iPod shuffle.

Might need to return to the world of the professional typists before too horribly long, so I've been keeping my eye out on Monster. Found a decent local advert last Friday and nibbled. Talking to the headhunter today, I found myself praising Microsoft, particularly SQL Server, and my disappointment with 3rd party libraries in general. Being something of an xplat fan on this site, I figured that required some fessing up, and some explaining.

Trying to be brief, in my years of experience, I've found MS IT software (excepting Windows Forms) to be incredibly rock-solid. Over those same years, I've found third party libs, from Blackbaud's Raiser's Edge to ESRI's ArcView, horribly hit-and-miss. For an IT career, I just don't want my success to be tied to how far down the object model some mid-sized corporation's programming team decided to go during development. MS seems to turn out good stuff, which means I can turn out good stuff. I can't even say that about Oracle. Headless Java apps, sure, but that's about it. Please, give me MS-SQL Server (well, anything *after* 6.5) and please heavens don't give me 3rd party libs whose object model can't be easily read off of an 8.5"x11" sheet of paper, because past that level of complexity, it appears [somewhat understandably] difficult for a mid-level co. to deliver a product with consistent high quality.

In other news, I like the new iPod idea, if the rumors are true. For $149, it sounds like you'll get an iPod that'll hold significantly more than a CD of songs and play 'em randomly. Hopefully you'll be able to skip to the next in the queue, but past that, that's really all I'd need. I enjoy popping my CD of mp3s in the car; that's about all I need to have what amounts to my own radio station. Add the ability to interface with iTunes and play AACs and you're getting somewhere. It's a glorified USB flashstick. ;^) And, of course, it'll get the cache from being called an iPod. Good move, Apple.

title: Put the knife down and take a green herb, dude.	descrip: One feller's views on the state of everyday computer science & its application (and now, OTHER STUFF) who isn't rich enough to shell out for www.myfreakinfirst-andlast-name.com Using 89% of the same design the blog had in 2001.
FOR ENTERTAINMENT PURPOSES ONLY!!! Back-up your data and, when you bike, always wear white. As an Amazon Associate, I earn from qualifying purchases. Affiliate links in green.
x MarkUpDown is the best Markdown editor for professionals on Windows 10. It includes two-pane live preview, in-app uploads to imgur for image hosting, and MultiMarkdown table support. Features you won't find anywhere else include... MarkUpDown Multiline Table & Bootstrap Grid support. Beautiful Easy Actions that keep the Markdown flowing. HTML paste to paste HTML source into your documents. You've wasted more than $15 of your time looking for a great Markdown editor. Stop looking. MarkUpDown is the app you're looking for. Learn more or head over to the 'Store now!

Sunday, January 30, 2005
Stefan Bodewig's Weblog: "javac: target release 1.1 conflicts with default source release 1.5" revisited Stefan Bodewig's Weblog: "javac: target release 1.1 conflicts with default source release 1.5" revisited: "[javac] The -source switch defaults to 1.5 in JDK 1.5. [javac] If you specify -target 1.1 you now must also specify -source 1.2." posted by ruffin at 1/30/2005 05:36:00 PM

Saturday, January 29, 2005
Wired News: My IPod, My Self From here: IPod and user form a cybernetic unit," said Giesler. "We're always talking about cyborgs in the context of cultural theory and sci-fi literature, but this is an excellent example that they're out there in the marketplace.... I have seen the future, and it is called the cyborg consumer." The cyborg consumer, Giesler said, is one that uses several different technologies -- from cell phones to Viagra -- and is highly connected, technically and socially. The iPod, for example, isn't just an MP3 player. It's an extension of the memory: storing the soundtrack of a lifetime, as well as names, addresses, calendars and notes. Oh, he's seen the future, has he? He thinks he's got some great cyborg metaphor, does he? Apparently he missed (though smart money is that he hasn't) Donna Harraway's 1984 "A Cyborg Manifesto: Science, Technology, and Socialist-Feminism in the Late Twentieth Century". You can read that shindig here, though it's admittedly a real challenge to understand without checking out some of her other writings. (Her book Primate Visions helped me out a good deal). I realize he's trying to pay tribute when he says, "context of cultural theory"; that's a bit of a key phrase there. But she does talk about exactly this -- what's he's got is a natural extension of her manifesto. To call her work cultural theory and his some sort of new fangled consumerism is to, in some ways, have created a scandal similar to Levi-Strauss' "incest scandal" that Derrida exposes in his [Derrida's] "Structure, Sign, and Play in the Discourse of the Human Sciences". The only scandal here is a system that would make a distinction between cultural sciences and whatever jive this fellow is proposing he's doing. Haraway's work says exactly what he's saying, and said it when he was seven years old. Anyhow, my $0.02. Labels: harraway posted by ruffin at 1/29/2005 11:25:00 AM

jTDS JDBC Driver 1.0 Hey, lookit there! jTDS has finally hit version 1.0! From the forum: The jTDS Project has released version 1.0 of the jTDS JDBC driver for SQL Server and Sybase. After 3 1/2 years of development, 21 releases and 100K downloads, jTDS is finally considered stable enough and JDBC feature complete to grant the first official production release. All important bugs fixed, jTDS is still the most performant JDBC driver for SQL Server and Sybase. It passes the J2EE 1.3 certification and Hibernate test suites and is the preferred SQL Server/Sybase driver for JBoss, Hibernate, Atlassian JIRA and Confluence, DbVisualizer and ComPiere. posted by ruffin at 1/29/2005 10:49:00 AM

Sunday, January 23, 2005
Dead horse beating So the JavaLobby survey results are in. JGoodies wins by a landslide. So why does the author of the survey, himself a JGoodies fan, feel compelled to spend his next column talking about "Per Platform Polishing of Your Look and Feel with Patches"? WinLaF is a project at http://www.java.net that fixes some look and feel problems in the default look and feel on Windows. By problems I am specifically referring to areas where the emulated look and feel is not faithful to the native appearance. Quaqua is a similar project to WinLaF, but for the OS X look and feel suite. Nice. Not only is Sun working their tail off to get Windows right and Apple's Java team throwing resources into having a platform-specific Aqua LaF, there are now two projects to slap atop those efforts to try and close the deal. I mean crimminy, look at what they're bothering to fix. The looks aren't close to done. Remember how far away native looks already were in Swing and how things like mouse wheel implementation lagged like mad? Well, now we have yet another QA issue nightmare. "Maybe that's the native LaF on that platform." "Perhaps. Are they using with Java 1.4.01." "Not sure. Think it might because they're trying to run WinLaF on 1.3?" The headaches. Here's the deal -- people are used to deploying Java apps in controlled environments, where if you don't have 1.5.1 installed, the IT group can force it on you tomorrow, no matter what that breaks, and can force the uninstallation of whatever else useful you might have installed, all so that their software finally works. As long as that mentality prevails, where the customer is always wrong, Java coders won't be ready for true client side apps. That said, JGoodies' Book Finder application does look fairly good. ;^) I wouldn't recommend using it, obviously, but at least they're trying. posted by ruffin at 1/23/2005 06:24:00 PM

Friday, January 21, 2005
Javalobby.org's telling survey: Your Favorite Swing L&F Anyone who has bothered to read this blog more than twice would likely know I like to use Java to build client-side, cross-platform apps, but would also know that I have a bone to pick with Swing, Sun's recommended UI widget toolset for Java apps with GUIs. In short, Swing stinks. They tried to shoot the moon and make a "look and feel" (which is a well-made phrase) for Java that would act exactly the same no matter what platform was running the app. It's a great idea in theory, and theory's always been Sun's strong suit. Trouble is, when we boot up Windows, we run Windows apps. When we log into Mac OS X, we run Mac apps. Java has not by any stretch dominated the client-side market the way Sun planned, making the OS running the apps relatively inconsequental, and likely never will. The niche for the xplat LaF (look and feel) is, short-run at the very least, "d-e-d dead". People, understandably, want all their apps to look and feel roughly the same, and since they run more native apps than, say, Java, they want their Java apps to seem to be native apps. Until they're running more Java than native, until Java hits that tipping point [that it never will], Java will need to continue to look, act, smell, & eat native. Instead of moving back to AWT, the Java UI widget set that essentially exposes native UI widgets to Java coders, with native looks, actions, and reactions, Sun decided to make Swing, originally made to look "Java", look native. So now we have a native look built into Swing, itself built on top of AWT which is, you guessed it, based on native widgets. Tell me that's not a little off. Instead of being one step removed from native, you're now emulating native on top of a toolset itself built on a toolset that really does hit native widgets. Boy, I seem to like beating that dying horse. Anyhow, much longer intro than I wanted, but you get the point if it's your first time here. Of course emulated native never quite acts like real native -- sorta like Asimov's robots that, even when they looked like people, could still trip off a sixth sense in "true humans", and Swing's native looks continue to have real issues that, ironicly, AWT (with its unfortunately small set of widgets) doesn't, like mousewheel issues on Windows or the inability to use brushed metal windows on OS X. In AWT, neither are (the former never was a) problems. Which is why this Javalobby survey drives me crazy. Everyone has their favorite L&F that they use every time they start a new application. In this week's poll, we'd like to know which ones you guys prefer. Personally, I prefer the JGoodies ones, especially with the various themes that come with it. What are your favorites? Look, the answers you get are going to be from Java coders who write quick little apps for in-house or personal use only. Nobody picks a third-party LaF for a large-scale, commercially released, cross-platform app unless they want the worst sort of QA/QC testing issues imaginable. Seriously, how many things do you want to have to double-check when your UI isn't working for ten customers running Mac OS X 10.0? The answers are silly -- like the person who answered the Napkin look and feel that I've discussed here before. Its whole raison d'etre is to make a prototype app look like a prototype so that your customer doesn't get the psycological impression that you're almost done once you've finished the UI. Nice, but hardly a favorite LaF -- unless you never get to the point that you've actually gotten an app far enough along to entered final testing. Nobody -- and I mean nobody -- should ever have unit tests for their UI run while in Napkin LaF. That's a complete waste. The right answer for 99.44% of Java coders is one of either Metal or native, by the way. But the lesson here is that questions like that -- even the original author's answer of JGoodies, which I'd also hesitate to recommend using, though their PlasticXP on WindowsXP (see a theme? Single platform release, trying to look 'more native than native'; why are you using Java anyhow?) doesn't look that bad, for a Java app anyhow -- show a circular problem from which the Java Desktop is going to have a hard time emerging. People writing the apps don't understand Java's strengths nor their customers' expectations for their apps, and until they do, we'll continue to have a hard time even pretending to have the potential to approach that tipping point. The only place even Metal (and now Synth and Ocean), that is, anything but native, really could belong is on Sun's own Sun Java Desktop System, where they could potentially make the native look and feel like Java. Now if I'd just make myself program in my free time instead of blog... And while I'm dreaming, somebody buy me an iPod shuffle. posted by ruffin at 1/21/2005 02:22:00 PM

Wednesday, January 19, 2005
PC Advisor - Expert Advice you can trust Now here's an interesting development on the email front... (On a side note, I'm using Google Alerts to keep tabs on 'competitors' (quotes b/c it's not like I'm really selling just yet), as suggested by some consultant to small business schmoes (can't find the link), and it really ain't a bad idea. If you're a small business schmoe, do it.) Anyhow, interesting that Novell is going to try and out-Outlook M$ -- on Windows! Also tried out Entourage ("Outlook on Mac", so to speak) and it's actually an excellent client as far as initial impressions go. If you have Office and haven't at least taken a look, hook up a hotmail address for testing and do so. Novell has thrown its weight behind a fledgling effort to develop a Windows version of the Evolution groupware client. The software company hopes the move will give Windows desktop users an open-source alternative to Microsoft Outlook. Evolution, like Outlook, is a suite including email, calendaring and address book software. A version for Linux desktops is already available, and work is now under way on porting it to Windows. posted by ruffin at 1/19/2005 04:40:00 PM

Wednesday, January 12, 2005
xhtmlrenderer: Flying Saucer: a 100% Java XHTML CSS renderer I've probably blogged this as a note-to-self already, but here it is again... xhtmlrenderer: Flying Saucer: a 100% Java XHTML CSS renderer Afaict, this is a sort of JEditorPane replacement that works just as poorly, but trades a relatively obsolete HTML 3.2 spec (that JEditorPane supports) with XHTML 1.0. Doesn't support everything you'd like, but apparently works fairly well. I'm beginning to get interested in the JDesktop Integration Components project, though it does everything Java shouldn't and goes against my theorhetical preferrence for an updated AWT over SWT. Everything about it is platform specific, but the promise of a native browser xplat is pretty compelling. Problem is it's one more possible weak link in your apps, and a likely one at that. I can certainly see an update to XP changing the way IE works, killing JDIC in a way that JEditorPane wouldn't... thus my interest in the Flying Saucer bit. Anyway, it's amazing that this far along in the life of Java we still don't have a good, xplat, reliable, relatively up-to-date html viewer as part of the jre. posted by ruffin at 1/12/2005 10:06:00 PM

Monday, January 10, 2005
Sell-outs and random iPods Might need to return to the world of the professional typists before too horribly long, so I've been keeping my eye out on Monster. Found a decent local advert last Friday and nibbled. Talking to the headhunter today, I found myself praising Microsoft, particularly SQL Server, and my disappointment with 3rd party libraries in general. Being something of an xplat fan on this site, I figured that required some fessing up, and some explaining. Trying to be brief, in my years of experience, I've found MS IT software (excepting Windows Forms) to be incredibly rock-solid. Over those same years, I've found third party libs, from Blackbaud's Raiser's Edge to ESRI's ArcView, horribly hit-and-miss. For an IT career, I just don't want my success to be tied to how far down the object model some mid-sized corporation's programming team decided to go during development. MS seems to turn out good stuff, which means I can turn out good stuff. I can't even say that about Oracle. Headless Java apps, sure, but that's about it. Please, give me MS-SQL Server (well, anything after 6.5) and please heavens don't give me 3rd party libs whose object model can't be easily read off of an 8.5"x11" sheet of paper, because past that level of complexity, it appears [somewhat understandably] difficult for a mid-level co. to deliver a product with consistent high quality. In other news, I like the new iPod idea, if the rumors are true. For $149, it sounds like you'll get an iPod that'll hold significantly more than a CD of songs and play 'em randomly. Hopefully you'll be able to skip to the next in the queue, but past that, that's really all I'd need. I enjoy popping my CD of mp3s in the car; that's about all I need to have what amounts to my own radio station. Add the ability to interface with iTunes and play AACs and you're getting somewhere. It's a glorified USB flashstick. ;^) And, of course, it'll get the cache from being called an iPod. Good move, Apple. posted by ruffin at 1/10/2005 05:27:00 PM

Saturday, January 08, 2005
25 Smackers? You're kidding. Pricing software is crazy. I look at something like MarsEdit and think it's a neat app, but that I'll never spend $25 on it (even if I had a use for spellchecked blog entries, which I obviously don't), so why bother? I've noticed a number of sites says $19.95 is the least amount to charge if you want to be taken seriously, but it really seems to be pushing it. I've seen a number of apps recently with very reasonable prices -- $8, even $2 or so a plugin at cfx or whatever that iMovie plugin place is called (I bought one). These things say, "Try me, and if you use them at all, the price is low enough you'll be practically guilted into buying." The best rationale I've seen for a higher price is that you can, as a sole programmer, provide better support to 1000 people who bought at $20 than 2000 that bought at $10. Now why the amount grossed is the same in the examples I'll leave as an excercise for the reader. Let's just say I'm not sure that's the case -- 3 registrations at $8 is better than none at $18. It takes customers to build customers, and it's likely better to have a grand version 2.0, where you plan to charge more, than to take the time to build 2.0 straight up. Anyhow, it's a mess. Luckily I haven't made enough time for Digest Handler 1.1, and that's the milestone before I really start worrying about that sort of thing. posted by ruffin at 1/08/2005 05:37:00 PM

Friday, January 07, 2005
Printing multiple images quickly In case MacOSXHints.com doesn't deem it worthy, here it is, for free, to all of you freakname regulars (so anyone who's been here more than once -- which likely means me and one other guy, who luckily also uses a Mac): Printing multiple images quickly in OS X I'd scanned a number of pages into jpegs from a book, and was going crazy printing image by image in Preview. Heck, even Windows lets you right click selected files in its "Finder-equivalent" and Print. I had to be missing something. Sure enough, I was. To print multiple files at once in OS X land, just open up terminal and use the `lpr` command, using a space between file names! `lpr img023.jpg img024.jpg img025.jpg img026.jpg` Couldn't be easier. Even scaled each jpeg to an 8.5"x11" sheet for me. Thanks heavens. posted by ruffin at 1/07/2005 04:01:00 PM

Tuesday, January 04, 2005
Lessig and $499 Macs Good ole Prof Lessig tries to answer the question, "Why extend the copyright on works that no longer have commercial value?" Here he argues to only extend the copyright, Bono-style, on things that still have value, so to speak, to the original copyright owners. Allowing someone -- anyone -- to make such a distinction is A Very Bad Idea. In other news, I just checked my RSS reader (JNN, fwiw) and saw one of the $499 headless Mac rumors. You know, now we're getting somewhere. Here's some hopeful text in the explanation: Indications are Apple has been working mostly on finding the right mix of price, performance and features that would motivate Windows users to consider a Mac, and less on the actual engineering of the product. (emphasis mine, of course) That's good stuff. At $500, I think I'd have a great chance at convincing friends and relatives to give the Mac a shot. They cringe at a $1300 iMac, act interested with an $800 eMac (but want something smaller -- not ironically more iMac like), and if they're in the laptop market, often bite at $999 for the iBook. Half the iBook price for the same hardware (which is just enough to run OS X well) in an ultra-small desktop means Apple's finally hitting the economy digital hub niche -- the luxury niche they've been doing quite well for some time. Still happy I shelled out for my [now old] G4 iMac with 17" screen and Superdrive, but if I hadn't, this would be beyond tempting; it'd be a great buy. posted by ruffin at 1/04/2005 05:10:00 PM

<< Older \| Newer >>