My vote is Lone-tar This was one excellent product, well worth the money. I haven't used the Linux version, but assume it's as rock-solid as the SCO Unix one. The version I used was extremely customizable - you basically edit a shell script, although this may be different now.
I also found the support excellent: when I called once on a weekend, I was forwarded to the person who wrote the software, who came in from mowing the lawn to take my call. Now that's *support*!
Funny, what with this "paradigm shift" and all, everybody's forgotten about the PPC. The main difference is TM can run Windows software, it seems, through it's "code morphing" ability. Never heard of that till now, must be marketing speak for "emulation", like "service pack" instead of "patch".
The original poster is correct, taking a look at the TM comparisons, the TM chip looks good until you add in the "system power" column across the board, then the difference is pretty minor. This reminds me of when I was working at a company where the common view was our software was a "dog" because of our reliance on an ancient (7 year old, unsupported) compiler. But the real problem was network and disk latency, which really adds up.
Here, the situation is similar. To put this all in perspective, my ancient IBM L40 (386) lasted roughly the same time as my Thinkpad 600. Battery efficiency has improved, but not enough to offset the increased demands on the laptop from the display and os. Or another example: while processor speed has been vastly improved, why is the computer as slow as ever? It's the software.
The point is, it is better to focus on other areas to make the most improvement. IMO, a couple hours improvement is not going to cut it, nor is the low price. Ditzel is touting this as a mobile , internet appliance. "The internet changes everything..." blah blah. I think while Linus and Ditzel were hunkered down in their office for the past 4 years, the rest of the world roared by, in the form of the Palm, Psion, Windows Powered. The Palm runs on AA batteries for *2 weeks*.
No, but I saw the previous whine about how he was duped by all the y2k hype. Amazingly ignorant article, but perhaps the reason he's so popular is he reflects the view of the masses so acurately.
He's been a columnist for a long time, 10+ years, and I've always skipped past his rantings in favor of better columnists like Peter Coffee, or Nicholas Petreley (sp?). I figure he's sort of like a Howard Stern or Rush Limbaugh, by stating some outlandishness, he gets lots of reaction, thereby lots of readers.
I don't waste my time, and that's the main problem I think with flames, is that they *usually* clutter up the thread, rather than offering a well thought out counterpoint.
What kind of hardware were you installing on? Because when I used System V a few years back, it installed without a hitch on a no-name clone. It was definitely easier to install than BSD (we switched to SCO). Linux installs as nicely now (RHat, SuSE) as SCO did 5 years ago.
I didn't use their support much because first, the os was rock solid (luckily), and also support was expensive and so-so. But SCO also supplied a daunting stack of manuals.
I think what SCO brings to Linux is years of experience in sales and support for large customers, and partnering with other software companies to develop apps for SCO, similar to what DEC brought to Compaq.
Huh. Some horror story. I was expecting stories of big software failures etc, but just a meandering mish-mash of complaints, more against the software industry as a whole than against Java. A few examples:
So people put ads in papers looking for experienced J2EE people when J2EE is hardly defined. Er, this is a criticism against the ad poster, not Java. And since when is this solely a Java problem? Got 6 years C++ exp. but only 6 mos. of COM? Sorry, buddy - we need real COM exp (even though the spec is how old..?)
Then there's an odd comment that since IT managers are skeptical of Linux because it relies on beta software, they should also be afraid of Java. I'm not sure what to make of this comment - it seems to confuse beta software with the idea of the moving target of a quickly changing standard. Yet in a previous paragraph he states that the J2EE spec was standardized already. Anyway, agree that the software industry moves *very* quickly, but since when is only revolving around Sun?
Read the second to last paragraph: all about cross platform differences, changing and poorly documented spec, market hype, problems because of a general platform for everyone that's good for no one.
I don't know about you, but Windows Everywhere came to mind, Windows 3.1, NT, 2000, CE. Probably Mr. McAllister could've dusted off an old article complaining about Microsoft, ran a 'find/replace' with Java, tossed in some J2EE, and voila, a new article..
Yes, I try to skip reading InformationWeek. That magazine, like so many others (Application Development) is lots of fluff, little substance. One publication that does tell you about the projects which get cancelled is the Wall Street Journal - take a look there. They had an interesting article a few years back about the difficulty of building new systems at PG&E and other large companies.
I also used J++ quite a bit - I thought it was nice, and very flexible, but I am now using VisualAge, which is very good. I have also used Rational, although not for generating Java code, but this is where Rational will probably focus their efforts, to provide better code generation. They also appear to work closely with M$ - Rose is built using COM (so how close is the Unix version?), and VB comes with a simplified version of Rose (Visual Modeler). Therefore, I wouldn't be so ready to throw away your investment in J++.
and see also : http://www.cs.wisc.edu/niagara/
I also found the support excellent: when I called once on a weekend, I was forwarded to the person who wrote the software, who came in from mowing the lawn to take my call. Now that's *support*!
The original poster is correct, taking a look at the TM comparisons, the TM chip looks good until you add in the "system power" column across the board, then the difference is pretty minor. This reminds me of when I was working at a company where the common view was our software was a "dog" because of our reliance on an ancient (7 year old, unsupported) compiler. But the real problem was network and disk latency, which really adds up.
Here, the situation is similar. To put this all in perspective, my ancient IBM L40 (386) lasted roughly the same time as my Thinkpad 600. Battery efficiency has improved, but not enough to offset the increased demands on the laptop from the display and os. Or another example: while processor speed has been vastly improved, why is the computer as slow as ever? It's the software.
The point is, it is better to focus on other areas to make the most improvement. IMO, a couple hours improvement is not going to cut it, nor is the low price. Ditzel is touting this as a mobile , internet appliance. "The internet changes everything..." blah blah. I think while Linus and Ditzel were hunkered down in their office for the past 4 years, the rest of the world roared by, in the form of the Palm, Psion, Windows Powered. The Palm runs on AA batteries for *2 weeks*.
He's been a columnist for a long time, 10+ years, and I've always skipped past his rantings in favor of better columnists like Peter Coffee, or Nicholas Petreley (sp?). I figure he's sort of like a Howard Stern or Rush Limbaugh, by stating some outlandishness, he gets lots of reaction, thereby lots of readers.
I don't waste my time, and that's the main problem I think with flames, is that they *usually* clutter up the thread, rather than offering a well thought out counterpoint.
Is SCO still owned in part by Microsoft? At one point they owned 25%.
I didn't use their support much because first, the os was rock solid (luckily), and also support was expensive and so-so. But SCO also supplied a daunting stack of manuals.
I think what SCO brings to Linux is years of experience in sales and support for large customers, and partnering with other software companies to develop apps for SCO, similar to what DEC brought to Compaq.
So people put ads in papers looking for experienced J2EE people when J2EE is hardly defined. Er, this is a criticism against the ad poster, not Java. And since when is this solely a Java problem? Got 6 years C++ exp. but only 6 mos. of COM? Sorry, buddy - we need real COM exp (even though the spec is how old..?)
Then there's an odd comment that since IT managers are skeptical of Linux because it relies on beta software, they should also be afraid of Java. I'm not sure what to make of this comment - it seems to confuse beta software with the idea of the moving target of a quickly changing standard. Yet in a previous paragraph he states that the J2EE spec was standardized already. Anyway, agree that the software industry moves *very* quickly, but since when is only revolving around Sun?
Read the second to last paragraph: all about cross platform differences, changing and poorly documented spec, market hype, problems because of a general platform for everyone that's good for no one.
I don't know about you, but Windows Everywhere came to mind, Windows 3.1, NT, 2000, CE. Probably Mr. McAllister could've dusted off an old article complaining about Microsoft, ran a 'find/replace' with Java, tossed in some J2EE, and voila, a new article..
Yes, I try to skip reading InformationWeek. That magazine, like so many others (Application Development) is lots of fluff, little substance. One publication that does tell you about the projects which get cancelled is the Wall Street Journal - take a look there. They had an interesting article a few years back about the difficulty of building new systems at PG&E and other large companies.
You forgot Ivar Jacobson - it was his company, Objectory, that merged with Rational. The three are referred to as 'the three amigos.'
I also used J++ quite a bit - I thought it was nice, and very flexible, but I am now using VisualAge, which is very good. I have also used Rational, although not for generating Java code, but this is where Rational will probably focus their efforts, to provide better code generation. They also appear to work closely with M$ - Rose is built using COM (so how close is the Unix version?), and VB comes with a simplified version of Rose (Visual Modeler). Therefore, I wouldn't be so ready to throw away your investment in J++.