Check out AnalystScan, a DB and JSP intensive site running exclusively on Tomcat and PostgreSQL. It's survived several mentions in the mainstream press, as well as a mini slashdotting.
The only caveats I have for you are to be sure that you use the groovy built in connection pooling for your DB resources, and that you tune your JSPs before going into production.
Remember when Gateway, et al, were subject to a class action lawsuit regarding the labeling of monitor sizes? Gateway was targeted because they were selling "17-inch" monitors that only had 15-inch viewable area. In that case, while the 17-inch designation was technically correct, the courts found that the practice was deceptive.
That could be an interesting precedent for this case.
Actually, that totally depends on the VM. For 1.1 AWT under Windows, Microsoft's VM is actually the fastest.
As for the Swing vs. AWT issue, that really is a toss up. The main reason for this is twofold:
Swing now has access to native video accelleration throw Java 2D, eliminating most of the advantage native peers had over swing widgets
VMs have matured greatly over the last several years, and continue to improve. The most noticable difference came when the Hotspot VM was introduced in JDK 1.2.
So, unfortunately, the question is difficult as the answer varies from problem to problem.
An interesting compromise between AWT and Swing has emerged in the form of Eclipse's SWT. For more info on that product, check out the eclipse home page.
Absolutely. Java has come a LONG way since Corel tried to port Office to Java. Corel was trying to port to JDK 1.1, which was totally pre-swing and pre-Java 2D. This meant that there was no font support outside of "monospaced", "serif" and "sans-serif", and it also meant no access to acceleration tools provided by Java 2D.
Any effort to create an office suite today would have a tremendous chance of success, although it would still be a challenge.
Actually, I would find it almost impossible to live without my floppy drive, which seems absolutely crazy. Most (if not all) vendor provided hard drive diagnostic tools run ONLY from DOS, and the same appears to be true for most BIOS upgrade tools as well. Some vendors have begun to provide Windows based BIOS reflashing tools, but they don't work under WINE, and it seems easier to keep a floppy drive around rather than a Windows partition.
Hint to vendors: Provide tools that run under Linux, or provide bootable CD images PLEASE!
As many have noted, the Random NYT Login Generator is not working. The block they seem to have implemented is based on the referer (yes, I know the right spelling. Trying looking at the HTTP header).
To get around this problem, simply save the page to your hard drive, and open it from there. Your referer will now be some file:// URL, and it will work.
2 8oz steaks (I use ribeye, but have also used new york steak), or 4 4oz chicken breasts 1oz Sesame Seed Oil 3oz of your favorite vegtable oil (I use canola) 2 tsp. fresh ginger, grated (you can substitute 1 tsp dry ground ginger) 1 tbsp roasted sesame seeds
Heat the oils over a medium heat until they are just shy of smoking (usually about 4 minutes). Add the ginger and sautee briefly. Cube the meat into 3/4 inch cubes. Add the cubes to the oil, cooking to desired finish. Add the sesame seeds and toss before serving over fresh rice, or pasta.
WebSphere 3.5x is one of the most painful, convoluted, and just plain evil products ever developed. First off, you are out of luck if you need to use any JDK1.3 or higher features, as WS 3.5x ONLY run's on a specially enhanced version of the IBM JDK 1.2x. While this may not seem like a huge deal, it eliminates the possibilities of using proxy objects in your J2EE apps, which in my experience makes things much more difficult.
Additionally, you will have a hard time maintaining deployment procedures since WS 3.5 uses its own deployment tools and descriptors.
Finally, there are more "undocumented" features in the J2EE support than IBM cares to admit. The best piece of advice I can give regarding that is to first assume the problem is with the app server, then your code. Google will be your best friend during debugging.
If you are stuck with WS3.5, good luck.
Of course, I'm an unemployed Java architect, so what would I know?:-)
Oh my goodness... this is simply the most amazing codec since the release of WMA8. I have compressed some of the most difficult samples available at 48kbit and it sounds just as good as 128kbit MP3.
Okay... I'm going to say one more thing, then I give up.:-)
BIOS - all that does is control a secure boot up process that loads a few system components
Exactly. You are correct that saying the BIOS is responsible for bootstrapping, but under Trusted PC, only signed components, whose public key certificate is issued by a CA whose root cert is stored in the unmodifiable area of the BIOS, will be allowed to be "booted". Who decides which CA certs are in the bios, and how does the CA decide which bootstrap component vendors should be issued BIOS certs?
This is a total chicken and egg problem (as is this point-counterpoint). Eventually, someone other than the user will have to make a trust decision. That is the fundamental problem that I have with Pallidium and Trusted PC.
Dude... chill. This isn't a personal vendetta against you.
My point is that according to the Trusted Computing Platform Alliance, only motherboard manufacturers may include binary code in the protected space of a trusted BIOS.
The spec, available here in PDF, prohibits the end user from updating this code. So the question is, whose binary will run in this space? My binary compiled from the publically available shared source? Probably not.
Granted, this is not Microsoft's Palladium, but it is logical to assume that the so-called Trusted PC will be an important part of the actual Microsoft implementation.
Yeah, that'd be great, except that MS is releasing all the source to Palladium.
It would appear that the "jack ass", as you referred to the original poster as, has a point.
Even though Microsoft is releasing all of the source to Palladium, what proof do we have that the binaries that are distributed are built verbatim from the sources provided? None.
The reason that source is distributed is so that you have, if you so desire, the tools necessary to verify the functionality of the code, then build a binary from the code that you just completed verifying.
Microsoft is under no obligation to provide you a binary that is built from the sources they provide. You could decompile the binary and compare opcodes against those of a binary built from the provided source, but even that is a faulty solution, since Microsoft is under no obligation to compile the sources using a publically available compiler.
The original poster may have gone overboard, but I hardly think he is a jack ass.
Man, does that suck. Once more, with feeling is one of the best episodes of any show ever aired. Whether you love or hate Buffy creator Joss Whedon, you have to appreciate the sheer chutzpah required to create a musical episode in the style of the classic Rodgers & Hammerstein musicals.
If you have not seen the show, do yourself a favor and *ahem* acquire the music. Hearing the cast sing is alone worth the download (Well, maybe not Alyson Hannigan, but...).
Sigh... to be expected... I guess the Slashdot editors are only interested in preserving their rights when it doesn't involve any action on their part.
News.com just printed my response to a letter published by former counsel to John Ashcroft on this very topic.
We, as the informed users, need to get the word out. The cable companies are doing a great job convincing the policy makers that there really is competion.
Oh yeah... that's a great way to encourage productivity and creativity from a software development team. At the last company I worked out, we would blast MP3s through the entire dev team area... it really helped creativity, especially during the 80 hour weeks required right before RTM.
Smart companies should look at the potential situation of an RIAA raid as a simple cost of employing creative types. I personally wouldn't work in an environment where I couldn't have MP3s on my PC, and many of the better programmers I know feel the same way.
Yet more proof that analysts do not have a clue as to what is going on in the market. I particularly like JP Morgan's coverage, which was initiated a long-term buy at around 30, then upgraded to a buy at around 8.
Slashdot Mirror
Don't you mean "Disney(R) Presents Goofy"?
The only caveats I have for you are to be sure that you use the groovy built in connection pooling for your DB resources, and that you tune your JSPs before going into production.
Good luck!
That could be an interesting precedent for this case.
Just find a nearby minor to press the "I Accept" button for you. Done and done!
See my journal for a discussion of this very topic, including a domain much more exciting than freerooftile.com. :-)
As for the Swing vs. AWT issue, that really is a toss up. The main reason for this is twofold:
- Swing now has access to native video accelleration throw Java 2D, eliminating most of the advantage native peers had over swing widgets
- VMs have matured greatly over the last several years, and continue to improve. The most noticable difference came when the Hotspot VM was introduced in JDK 1.2.
So, unfortunately, the question is difficult as the answer varies from problem to problem.An interesting compromise between AWT and Swing has emerged in the form of Eclipse's SWT. For more info on that product, check out the eclipse home page.
Any effort to create an office suite today would have a tremendous chance of success, although it would still be a challenge.
Talk about the pot calling the kettle black...
Hint to vendors: Provide tools that run under Linux, or provide bootable CD images PLEASE!
To get around this problem, simply save the page to your hard drive, and open it from there. Your referer will now be some file:// URL, and it will work.
2 8oz steaks (I use ribeye, but have also used new york steak), or 4 4oz chicken breasts
1oz Sesame Seed Oil
3oz of your favorite vegtable oil (I use canola)
2 tsp. fresh ginger, grated (you can substitute 1 tsp dry ground ginger)
1 tbsp roasted sesame seeds
Heat the oils over a medium heat until they are just shy of smoking (usually about 4 minutes). Add the ginger and sautee briefly. Cube the meat into 3/4 inch cubes. Add the cubes to the oil, cooking to desired finish. Add the sesame seeds and toss before serving over fresh rice, or pasta.
Serves 4
WebSphere 3.5x is one of the most painful, convoluted, and just plain evil products ever developed. First off, you are out of luck if you need to use any JDK1.3 or higher features, as WS 3.5x ONLY run's on a specially enhanced version of the IBM JDK 1.2x. While this may not seem like a huge deal, it eliminates the possibilities of using proxy objects in your J2EE apps, which in my experience makes things much more difficult.
Additionally, you will have a hard time maintaining deployment procedures since WS 3.5 uses its own deployment tools and descriptors.
Finally, there are more "undocumented" features in the J2EE support than IBM cares to admit. The best piece of advice I can give regarding that is to first assume the problem is with the app server, then your code. Google will be your best friend during debugging.
If you are stuck with WS3.5, good luck.
Of course, I'm an unemployed Java architect, so what would I know? :-)
AMAZING job, guys.
Please be gentle.
It would be interesting to see how one would use a train to crash into a skyscraper.
As for the last airing, that is dependent on your local syndication partner.
Oh, and in the tradition of Comic Book Guy, the quote is "80 past 2 on April 47th" :-)
BIOS - all that does is control a secure boot up process that loads a few system components
Exactly. You are correct that saying the BIOS is responsible for bootstrapping, but under Trusted PC, only signed components, whose public key certificate is issued by a CA whose root cert is stored in the unmodifiable area of the BIOS, will be allowed to be "booted". Who decides which CA certs are in the bios, and how does the CA decide which bootstrap component vendors should be issued BIOS certs?
This is a total chicken and egg problem (as is this point-counterpoint). Eventually, someone other than the user will have to make a trust decision. That is the fundamental problem that I have with Pallidium and Trusted PC.
My point is that according to the Trusted Computing Platform Alliance, only motherboard manufacturers may include binary code in the protected space of a trusted BIOS.
The spec, available here in PDF, prohibits the end user from updating this code. So the question is, whose binary will run in this space? My binary compiled from the publically available shared source? Probably not.
Granted, this is not Microsoft's Palladium, but it is logical to assume that the so-called Trusted PC will be an important part of the actual Microsoft implementation.
It would appear that the "jack ass", as you referred to the original poster as, has a point.
Even though Microsoft is releasing all of the source to Palladium, what proof do we have that the binaries that are distributed are built verbatim from the sources provided? None.
The reason that source is distributed is so that you have, if you so desire, the tools necessary to verify the functionality of the code, then build a binary from the code that you just completed verifying.
Microsoft is under no obligation to provide you a binary that is built from the sources they provide. You could decompile the binary and compare opcodes against those of a binary built from the provided source, but even that is a faulty solution, since Microsoft is under no obligation to compile the sources using a publically available compiler.
The original poster may have gone overboard, but I hardly think he is a jack ass.
And "I'm Under Your Spell" made EVERYONE fall in love with Tara's character even more.
Okay... time to break out the VCD and watch again. :-)
If you have not seen the show, do yourself a favor and *ahem* acquire the music. Hearing the cast sing is alone worth the download (Well, maybe not Alyson Hannigan, but...).
Sigh... to be expected... I guess the Slashdot editors are only interested in preserving their rights when it doesn't involve any action on their part.
We, as the informed users, need to get the word out. The cable companies are doing a great job convincing the policy makers that there really is competion.
Smart companies should look at the potential situation of an RIAA raid as a simple cost of employing creative types. I personally wouldn't work in an environment where I couldn't have MP3s on my PC, and many of the better programmers I know feel the same way.
Yet more proof that analysts do not have a clue as to what is going on in the market. I particularly like JP Morgan's coverage, which was initiated a long-term buy at around 30, then upgraded to a buy at around 8.