The reason there are so many H1 visa people and foreigners in the IT business is because they are often the only people available. Here in the US, very few people are getting 4-year degrees in CIS, and there arn't that many self-taught competent engineers. Things are different in India, for example. In India, getting a CIS degree is in the same class as getting a law or medical degree in the US. There's an overabundance of lawyers in this country but very few people graduating with CIS degrees.
In almost every company I have worked most of the engineers are from India and a lesser extent from Taiwan and China. As a senior engineer who has interviewed many people, it is quite difficult finding talented programmers. It is even more difficult in that in the networking sector I work in most of the work is in embedded systems which requires far more skill than writing standard Unix or Windows programs. For one thing, a bad pointer will likely require a reboot since there's usually no safety net and the debugging tools are often quite limited (and very often home-grown).
Sure, anyone can implement a linked list, but there arn't many people who can implement interrupt handlers or use a logic analyzer to track down a system hang.
I have not worked on W2K drivers for about about 14 months. The last time I worked on one I was disgusted with Microsoft. Every build had new bugs, which, of course, they blamed on our driver. It sounds like the debugger hasn't improved any. I was lucky if I could keep windbg running more than 5 minutes without having to kill and restart the remote debugger. What made our driver difficult were all of the new features Microsoft required for certification, even though most of the features were worthless for our target market. Not only that, but Microsoft rendered our most important feature impossible. We had multi-port network adapters and provided failover and port aggregation (both proprietary and Fast EtherChannel). Microsoft's new requirements and APIs eliminated any way of configuring these features.
There are detailed descriptions about how these attacks are being executed at Dave Dittrich's web site. It looks like there are numerous vulnerabilities in both Red Hat Linux 6.0 and in Solaris that were exploited for this bug.
Linux, Solaris, or Windows is only secure if the system administrator constantly applies the latest security patches, and how many of you actually do that? The only way to prevent this sort of attack is being vigilant about security on all machines on your network.
This review is a bit late, as this book came out over a year ago and Mountain of Black Glass came out in September. Some other excellent Tad Williams books are Tailchaser's Song (an adventure story from the point of view of a cat), the Dragonbone Chair series, Caliban's Hour, and Child of an Ancient City. I am eagerly awaiting Sea of Silver Light to finish the Otherland series.
I worked with Jeff Hawkins back when he worked at GRiD Systems and developed his handwriting recognition system. He tried to sell GRiD on the Graffiti recognition since trying to recognize standard handwriting (even for a human) is *much* more difficult.
GRiD didn't want to do that so Jeff took off and founded Palm Computing, which was later bought out by 3COM.
Jeff developed the algorithms for handwriting recognition while in school during the late 80's and early 90's and his method was quite different that the methods used by everyone else. For one thing, his recognition algorithm would work well on an 8088 while everyone else needed at least a fast 386 or 486.
The main problem with trying to recognizing standard characters is that it is nearly impossible to recognize what is written without extensive context work. For example, an O (Oh) and a D can be nearly impossible to differentiate between depending on how the person writes. It's even more difficult to differentiate between, say, a 0 (zero) and an O (Oh).
GRiD worked by writing entry forms which limited the types of characters that could be entered. I.e. numeric fields looked for the closest numeric match and text fields looked for the appropriate text match (i.e. only one case). GRiD knew of the limitations of handwriting recognition, but their management was blind to the vast improvement graffiti made. I havn't read info on Xerox, but I know Jeff developed graffiti probably around '91-92, possibly earlier.
I worked on a palm computer with handwriting recognition back in '93, the Casio Zoomer. The thing took PCMCIA flash cards, ran MSDOS, had IR and serial links, and could take a PCMCIA option card. It even had digital audio so it would say 'You've got mail' when using it to connect to AOL. Oh, and the thing ran Geoworks with Jeff's handwriting recognition.
It never took off for several reasons. First of all, the handwriting recognition was not graffiti and thus had a high error rate. Second of all, it was a little too big to fit into a shirt pocket. It did have some cool features, though. Geoworks was pretty cool, but not as simple as PalmOS.
It had everything. It could tie into AOL Email, it had Pocket Quicken. It had Solitair (the most important Windoze application). It even had a program for caluclating what everyone owed at a restaraunt (including tip).
Various drugs should indeed be restricted to prescriptions. If every time we got a sniffle and grabbed anti-biotics, whether they are needed or not (i.e. a virus) then we just encourage antibiotic-resistent bacteria (which are very nasty, as my sister just had surgery since antibiotics no longer work).
I'm sorry, but for my own health I don't want my cube neighbor going down to the local pharmacy and buying antibiotics for the flu. There are specified doses of various drugs which vary based on age, weight, and family history. If drugs were unregulated we'd have a nightmare on our hands. There'd be lots of ODs and other problems.
As for antibiotic resistant bacteria, they are becoming more and more common, especially in countries that overprescribe antibiotics.
It is true that a negative is higher resolution than HDTV, HOWEVER, after all the mastering is done and the film is duplicated the resolution is lost. My information comes from SMPTE (Society of Motion Picture and Television Engineers) at a persentation done a few weeks ago at Sony in San Jose. Yes, film is higher resolution, but once it has been processed the advantage is lost. HDTV maintains its resolution throughout. Also, a CCD has higher resolution in low light levels than film since fast film must be used. In addition, the CCD has a higher dynamic range, i.e. there is much more low-level detail in dark areas with a CCD than is possible with film.
As far as grain, there currently is work progressing on digitally adding "film" grain to HDTV recorded video.
According to the SMPTE presentation, once the film is scanned to digital and goes back to film, the resolution is lost. The best film scanners cannot match what a good CCD can do at this time.
It's like photocopying a 1200 dpi grey-scale laser-printer image with a 1200 dpi scanner. Since the dots likely won't match up exactly the scanned image will not look as sharp or as good as the original. The same sort of thing happens with film. Let's face it, there are multiple generations of film to film transfers done before the film reaches the theaters. My guess is that there are a minimum of 3-4 film-film transfers. The first would be from the original film copied to a new film, since it has to be cut and edited. This would go from a negative to a positive. Now, this golden tape isn't about to be used to make duplicates for the theaters (unless somebody is really stupid) so this positive is duplicated again into one or more negatives. These negatives are then used to copy to the film that goes out to the theater. Now, that's four copies made. If the original resolution was 4000x2000 you're now probably down to 1500x800. If digital processing is done between the original film and the film that goes out to the theater then the resolution is probably higher, more likely around 2000x1000. This isn't far from HDTV which is 1920x1080. Not only that, but the HDTV can easily be recorded at 30FPS progressive or 60FPS interlaced. For the die-hard film bufs it can also be recorded at 24FPS progressive.
At the SMPTE presentation some A/B footage was shown with footage taken with an HDTV 24FPS camera and a 35mm film camera. The HDTV looked very close, except where there was a lot of contrast in which case it looked BETTER since there was more detail in the dark areas.
Expect to see reflective LCD display technology come out in the near future. Reflective LCD is not at all like your standard LCD, but is basically an LCD built over a semiconductor. Reflective LCDs can handle higher resolutions than DLP and should be cheaper to manufacture. In addition, they don't suffer from the annoying artifacts of DLP (i.e. color separation).
Film won't be used much in 10 years. HDTV actually has higher resolution than 35mm film and can have a higher frame rate. The 24fps HDTV standard came about not only to show films on HDTV, but for producers to record directly at 24fps. 24fps suffers during high motion scenes, but many film directors insist that they must have the properties of film (i.e. 24fps, grain, etc.)
One other advantage of using all digital is that currently there is degradation when digitizing film to add effects and going back to film.
HDTV also has a higher dynamic range than film. The new CCDs can handle darker material better.
In addition, within the next few years it will be less expensive to use tape instead of film. There's virtually no time spent on development, and if a shot is messed up, just rewind and start again.
I too am getting spammed like crazy from Barnes and Nobel. Perhapse this is because they purchased books.com:(
I had always used books.com because they never resorted to spamming (unlike Amazon in the past or B&N in the present). Books.com also had a feature where you could click and automatically compare their price with both Amazon and B&N and if either were lower then books.com would beat them.
Sadly books.com is now part of B&N who I now boycott. If I have to choose between Amazon and B&N I must hold my nose and go with Amazon (since they learned their lesson on spamming). I agree that their patent is silly and feel that it will be thrown out in an appeal.
Micron is coming out with a Pentium III compatible chipset that supports DDR SDRAM. I submitted an article to Slashdot at http://www.inqst.com/ddrvrmbs.htm which shows DDR smoking RAMBUS. I wouldn't be surprised if Micron also came out with a chipset for the Athlon, which can really take advantage of the higher memory speeds.
I suspect VIA also is coming out with a DDR compatible chipset.
You obviously did not read the report. Judge Jackson's main findings were about how Microsoft stifled free capitalism. IBM promoted Lotus Smartsuite with their PCs and hence Microsoft wouldn't grant them a Windows 95 license until just minutes before they released it, thus causing IBM to not be able to be ready for the Chrismas buying season. Microsoft basically told Netscape to get out of their territory or else and then cut off Netscape's air supply. As for comparing this to the breakup of AT&T, this is the best thing that could have happened. Each of the baby bells and AT&T spinoffs has been much more successful than they would had they still been a huge monolithic corporation. In addition, long distance rates have fallen dramatically due to the competition of the other carriers. I guess you don't remember when you couldn't even own your own phone. You had to lease it from AT&T.
Likewise, look at Standard Oil or the railroads who controlled everything. There is no such thing as a free market economy. An economy without checks and balances would be like a government without checks and balances. What would happen to our government if, say, the President wrote all the laws, enforced all the laws, and interpreted and executed all the laws? Can you say dictatorship? Capitalism leads to greed. Not that there's anything wrong with capitalism, but checks and balances are required such that one individual or corporation doesn't dominate huge aspects of the industry or our lives.
Can you imagine a world where MS used their capital to gobble up other companies left and right without any stopping of it? If MS had their way, no computer manufacturer would be able to sell or distribute any other operating system. Also, all of Microsoft's distributors (i.e. computer stores) would be prohibited from selling alternatives. MS would then require all ISPs to use some proprietary protocol for Internet connectivity (like AOL) and prohibit non-Microsoft platforms from accessing the Internet.
If you think Linux is a serious competitor, perhapse in a few years, but at this point it still has a long ways to go. The applications are still missing. I hate to say it, but supporting MS Office is essential in the corporate space. Even if nobody likes it, because everyone else uses it you are forced to use it just to read other people's documents. Star Office isn't enough (I know this from experience).
In the corporate space, Linux has barely made a dent. Go down to any business and look at their computers. I imagine that you'll find few Linux boxes in use. Let's face it, for a vast majority of users Linux has too much of a percieved learning curve. Someone can sit down to a Windows box and click start and start running apps. Installing apps is trivial due to their Install Wizard. Many people put up with the frequent crashes and lock-ups because they don't know any better. Give them a Linux box and there will be just as many crashes because they won't know how to properly configure it.
I just ran into a patent which basically puts the brakes on my work. For an upcoming networking product which will support Cisco's Fast EtherChannel it would be really nice to seamlessly interoperate through Cisco's simple PAgP (Port Aggregation Protocol). I found out that Cisco just got a patent on this simple protocol. I can't believe it. It's one of the stupidest patents I've seen. The result, however, is that Cisco can use it to continue to bolster their monopoly since they prevent anyone else from interoperating with them. They patented EIGRP as well, which is a popular routing protocol which is only supported by Cisco.
Software patents should be for things that are truely revolutionary or very different. Software patents also should be subject to a different time limit, i.e. 5 years, since an Internet year is so much shorter.
I just ran into a patent which basically puts the brakes on my work. For an upcoming networking product which will support Cisco's Fast EtherChannel it would be really nice to seamlessly interoperate through Cisco's simple PAgP (Port Aggregation Protocol). I found out that Cisco just got a patent on this simple protocol. I can't believe it. It's one of the stupidest patents I've seen. The result, however, is that Cisco can use it to continue to bolster their monopoly since they prevent anyone else from interoperating with them. They patented EIGRP as well, which is a popular routing protocol which is only supported by Cisco. Software patents should be for things that are truely revolutionary or very different. Software patents also should be subject to a different time limit, i.e. 5 years, since an Internet year is so much shorter.
Cowpland should have known better. When you work for a corporation and hold stock in that company and have advanced financial information you must be very careful about buying or selling stock. For example, at my last job managers were not supposed to trade company stock within 30 days of the quarterly earnings report. The higher up you are the more important this becomes. In fact, any large trade should be pre-announced well in advance.
1. The command shell is not really that complicated. All that is needed is an alias file and you can have a shell that behaves just like DOS (except using / instead of \).
2. Plug and play is now supported in the kernel for ISA and PCI hardware. This needs to be expanded such that Linux can automatically discover the mouse, modem, graphics card, and other components at installation time and in a driver configurator. I don't think this would be that difficult to do. The main thing needed is for drivers to include some descriptive file (such as Micros~1's.INF files). Also, the loadable module support needs to be improved so modules do not have to be the same version as the kernel.
When installing Linux, it should discover the sound card, mouse, graphics adapter, CRT, modem, SCSI adapter and other peripherals.
3. On the GUI front KDE and Gnome are quickly gaining acceptance. These GUIs offer as many features (and sometime more features) than Micros~1. All of the mainstream Linux distributions come with Gnome and KDE now.
4. Hardware vendor support will come when Linux gains enough market share. Right now Microsoft has total control over the hardware vendors. Microsoft tells them what features they must implement and often how to implement those features. As it is, I have found that Linux actually has pretty good hardware support.
Rambus is dead. Rambus has numerous drawbacks, such as a higher manufacturing cost and licensing fees. DDRAM has a nice window to kill it. DDRAM is basically SDRAM with double edged clocking so it can operate at speeds up to around 266MHz. It also costs little or nothing extra to produce than SDRAM and there are no licensing fees or royalties that need to be paid. Expect some non-Intel chipsets to appear shortly that take advantage of this.
Furthermore, most of the memory manufacturers are not supporting Rambus. Samsung just dropped Rambus and is going back to SDRAM.
Intel really shot themselves in the foot WRT Rambus.
While there may not be a shortage of web designers, MSCE sysadmins, or VB programmers, there is indeed a shortage of skilled programmers. Out of all of the engineers I have interviewed recently for an embedded networking position, there's very few engineers who stand out. With so few people in the United States graduating from college with a CIS or CE degree, it is becomming increasingly difficult to find qualified people. Many Silicon Valley companies I've seen or worked at are mostly employing people from other countries such as India, Taiwan, China, and so forth because that's where the good engineers are from. It really irks me to see many slashdotters complaining that they don't need a college education, but I bet many would fall flat on their face in an embedded environment. We don't need c00l d00dz who can wr1te c0de. We need engineers.
What was it, something like only 26,000 CIS/CE graduates in the United States last year, yet 340,000 positions open? I'm sorry, but many of these idiots who call themselves engineers arn't. Ask them about Big-O notation and they're lost. Ask them to write some basic algorithms during an interview and they're lost.
I recently interviewed an engineer from a well-known company and asked him to write an integer to ASCII conversion function. He managed to write something full of bugs. His response was that "the compiler will catch them and then I'll unit test it". Bzzzzzt, wrong answer. A good engineer will look through their code and double or triple check it if in doubt, since unit testing often misses things. When writing in an embedded system, such mistakes can be catestrophic in that if your code blows up due to a bad pointer or something the whole box crashes and it can be a nightmare to track the source.
Out of all of the engineers I've interviewed at my current employer, only one or two stood out. Many engineers may know how to do a few things very well but can't escape the box.
Due to our desparation to find good networking engineers we've been forced to go the H1B Visa route with a green-card sponsorship. It's expensive, legally. Other countries place a much higher status on engineering degrees than the United States. India, for example, holds engineering in the same esteem as doctors and lawyers.
In Silicon Valley, a good engineer will do very well, even considering the high cost of housing.
Currently I see little threat from Linux for Sun. Solaris is a whole different ball-game than Linux.
Solaris has many of the features needed for large enterprise installations which Linux currently lacks (i.e. journalling file system). Solaris is also much more scalable than Linux. How well does Linux run on a 64 processor machine?
Also, I must say that my experience with both Solaris and Linux that Solaris is much more stable. I'm even running KDE on Solaris which makes the user interface quite usable. I don't know about the stability of the x86 version of Solaris, but I imagine that Sun is far more interested in the Sparc version.
On my Solaris box I've filled up all virtual memory, filled up the disk, crashed KDE, had the network go down, and it just keeps going. No crashes! My Redhat Linux box, on the other hand, frequently crashes for no apparent reason with an uptime of a couple of days between crashes. The hardware is just a generic Pentium system, 64MB RAM, S3 graphics adapter, IDE hard drive, Intel 10/100 NIC. Nothing special.
Both Linux and Solaris have all the latest patches and fixes installed.
I have a Philips DVX 8000 which is a hybrid computer/DVD ROM player and line doubler AV controller. This box is notorious for not liking certain DVDs.
The Matrix DVD was almost flawless. There are a few places on the disk where the MPEG video screws up and it backs up a couple of seconds, but no hangs.
The software on the DVD cannot play the DVD, but that's not surprising given that the player has some rather special hardware (a very custom graphics adapter, for example).
I have a problem where it jumps out of the audio commentary track after each chapter, but I have this problem with other DVDs as well (i.e. Godzilla).
I love the fact that they include a separate audio-only track. I wish more DVDs would do this. Also, the white rabbit sequences are quite cool.
Linux still does this, however many pieces (i.e. the filesystem) cannot be removed. In fact, if we removed all the stuff we don't need there'd be almost nothing left. We don't need SMP support. We don't need a filesystem. We don't need security. We don't need virtual memory. We don't even want preemptive multitasking (our product uses cooperative tasking). The problem with Linux is that it's difficult to remove these pieces, since thing like the filesystem, virtual memory, and preemptive multitasking are integral to the operating system.
Linux was not designed to be a RTOS. A RTOS does not necessarily make the OS good for general purpose operation. I mean, a network switch, router, toaster, etc. has far different requirements than a PC. Real-time operating systems are more than just special scheduling. Often they're designed to be quite compact and have extremely modular functionality. They run in things like your VCR, DVD player, router, intelligent hub, palm pilet, etc. They are designed for very tight memory constraints. Some operating systems are very primitive, such as what runs in a VCR or your keyboard.
In our environment, for example, without all the things listed above, we use a unified memory space. All processes share the same memory space. We don't care about virtual memory since if a process dies it's probably catastrophic since most of the processes are interdependent.
In our product, a real-time process is often more like a TSR in that it is event driven and communicates with other processes via events.
Granted, there are places where Linux can work quite nicely, for example tivo, which, I must add, is a really cool product (and they'll burn a Linux CD with their kernel modifications upon request). But that's a different ballgame where a filesystem is important.
I researched replacing our current product with VxWorks to use either Linux or eCos. Neither would be an effective solution. eCos is far too limiting. The last I checked it didn't even have TCP/IP support, nor support the Mips processor (which is extremely popular for embedded environments [eCos does have an eval version for the NEC VR4300]). Another problem with Linux is the GNU public license. In an embedded environment it is often necessary to hack the kernel or pieces of it to do what you need.
I found RTEMS http://www.rtems.com is an excellent solution. It is open source and has the latest BSD TCP/IP stack. The license also allows us to hack up parts of the kernel without having to release the source code back.
For our current product, any RTOS we use we need to make some custom changes to the TCP/IP stack (which involve changes that would NOT be of benefit to the open source community unless the company wanted to compete with us). The eCos license would let us do this, Linux does not.
Don't get me wrong. I support open source and feel that any changes that enhance a product, fix a bug, or whatever that are not proprietary to one specific application certainly deserve to be released back to the community.
Another problem with the Linux kernel is that it is far too bloated. It would take a lot of work to bring it down to size. For example, everything we have is in flash. There is no file system. Our current image, symbols and all, is around 1.25MB compressed (using Zlib, which incidently is much faster than Gzip). We have a lot of code and not a lot of room. Having a small kernel is often necessary. We don't need a file system, virtual memory, security, loadable modules, or many of the other features of the Linux kernel. In fact, due to the fact that everything must be physically linked to the kernel, this totally rules out Linux. The 1.25MB is EVERYTHING, all of our code, protocol snooping, debugger, routing, console, SNMP, and firmware for supporting a second CPU (which runs our custom RTOS which is only a few pages of code).
eCos and RTEMS fill the void nicely in that one can pick and choose which features get compiled in. Unfortunately eCos is missing some very important pieces (i.e. TCP/IP).
VxWorks is a pain in the *ss since it has numerous bugs (i.e. improper cache operation on the NEC VR4300 CPU) and has terrible support (after paying thousands of dollars a year it's not unusual to have to wait weeks or months for a response). Not only that, but we had to buy their networking source code which costs many thousands of dollars (it's mostly just the BSD 4.3 TCP/IP stack, nothing special, and various daemons).
I am seriously looking at moving to RTems in the future. They have always responded promptly and have the features we need. My boss was all for Linux until I told him about what was involved to use it, and the GPL immediately killed it.
If we do move to RTems, I would certainly make a number of the changes we would make available back to the source code, but there are a number of others that we cannot release (since there's competition).
"I'd like to see, for once, highly technical jobs that judged you on the basis of your skills instead of "I'm an MSCE", or "I have a BA in *blah*". "
This has been my experience in Silicon Valley. They often could care less about how much education or seniority you have as opposed to your skill. Even Linus has commented on this.
I know people who have risen to the top of their fields as engineers who never graduated from college. It's also not unusual for someone with less education to earn far more than someone with a Phd. It all depends on skill.
I figure once I graduated from college future courses and further education is not that important as long as I keep current with the technology. In fact, when reviewing resumes, often I only skim the education part out of curiosity. I could really care less.
In Silicon Valley my experience has been that experience and skill has far more importance than education. If a company is going to give you trouble over your education (or lack thereof) then it's probably not a good company to work for to begin with.
Getting started without a degree can be more difficult, but with the shortage of qualified engineers, more and more companies offer internships. That's how I got started. My internship opened at least as many doors as my diploma did.
I'm not saying that an education is a bad thing either. I feel that immersing ones self in many different areas of knowlege is a good thing and produces a more well-rounded individual. Even those non-engineering related courses can prove invaluable from time to time. It's much more difficult to immerse ones self in many different areas when one has to work 40+ hours a week. Also, multi-talented engineers are far more valuable than someone who only knows a narrow field.
However, it is rather cool to have people call you doctor. Next thing you know Linus will have a blue police box and will be known as THE doctor.
I think that breaking up Microsoft would not do much to them. Instead, here's some other things that could be done to Microsoft to cripple it:
1. Force MS to release full specs on all application file formats (i.e. MS Office) and force them to write full export capabilities to the other top competing applications for Windows and other platforms (not just Mac).
2. Prohibit MS from bundling multiple packages together (i.e. Office 2000 integrating a photo publishing package to kill off Adobe).
3. Prohibit MS from deviating from standards set outside of MS (i.e. HTML, HTTP, TCP/IP) or from generating new internal protocols when standard protocols/file formats already exist.
Announcing layoffs has a much more widespread effect than many people realize. When I was at my last job, they had a 200 person layoff. As a consequence, morale went through the floor and many of those who were not layed off left for greener pastures. They lost a lot of intellectual knowlege and experience which cannot be easily recovered. Furthermore, here in California, if a company lays someone off they cannot hire someone else for the same job for at least 6 months. SGI is in for a very rough ride ahead. It will be very difficult for them to keep those who were not layed off.
When layoffs occur, it is some ways harder on those who were not layed off. You then start wondering, "when is it my turn to be layed off?"
Being layed off is not necessarily a bad thing around here in Silicon Valley. There are a lot of job openings elsewhere, and usually the company doing the layoff offers a good severense package. I know that when my division at my last company was dissolved, within a month everyone had found a better job than they had, and with the severense packages many people did quite well. Everyone from the admin to the managers found new jobs quickly. Most of the people are happier at their new job as well.
It is sad to see a company like SGI have a layoff. I fear that SGI may not be able to recover from this.
I must agree with you on one thing. I run KDE on both Linux (after Gnome crashed it hard one time too many) and on Solaris. Downloading the source and building it on Solaris wasn't too bad. Gnome, on the other hand, is a mess. Libraries must be downloaded from all over the place (and a number of the reported links are broken). I've given up trying to compile Gnome for Solaris, and given my experience of Gnome on Linux I've given up the thought for now.
Gnome looks great. It has a few things I really like, (lots of cool apps for the tool bar) and some things I don't like (instability). One thing I'll say about KDE is that it is clean. The UI is clean and to the point. The source tree is also clean. Just download qt and the KDE source files and that's it.
KDE has been running fine for months on my Solaris box with no crashes or lockups (once I downreved kwm to 1.1 from 1.1.1). I've even run out of swap and the system gracefully recovered.
I look forward to compiling KDE 1.1.2. I just wish the KDE folks would update their bug database better.
Slashdot Mirror
The reason there are so many H1 visa people and foreigners in the IT business is because they are often the only people available. Here in the US, very few people are getting 4-year degrees in CIS, and there arn't that many self-taught competent engineers. Things are different in India, for example. In India, getting a CIS degree is in the same class as getting a law or medical degree in the US. There's an overabundance of lawyers in this country but very few people graduating with CIS degrees.
In almost every company I have worked most of the engineers are from India and a lesser extent from Taiwan and China. As a senior engineer who has interviewed many people, it is quite difficult finding talented programmers. It is even more difficult in that in the networking sector I work in most of the work is in embedded systems which requires far more skill than writing standard Unix or Windows programs. For one thing, a bad pointer will likely require a reboot since there's usually no safety net and the debugging tools are often quite limited (and very often home-grown).
Sure, anyone can implement a linked list, but there arn't many people who can implement interrupt handlers or use a logic analyzer to track down a system hang.
I have not worked on W2K drivers for about about 14 months. The last time I worked on one I was disgusted with Microsoft. Every build had new bugs, which, of course, they blamed on our driver. It sounds like the debugger hasn't improved any. I was lucky if I could keep windbg running more than 5 minutes without having to kill and restart the remote debugger. What made our driver difficult were all of the new features Microsoft required for certification, even though most of the features were worthless for our target market. Not only that, but Microsoft rendered our most important feature impossible. We had multi-port network adapters and provided failover and port aggregation (both proprietary and Fast EtherChannel). Microsoft's new requirements and APIs eliminated any way of configuring these features.
There are detailed descriptions about how these attacks are being executed at Dave Dittrich's web site. It looks like there are numerous vulnerabilities in both Red Hat Linux 6.0 and in Solaris that were exploited for this bug.
Linux, Solaris, or Windows is only secure if the system administrator constantly applies the latest security patches, and how many of you actually do that? The only way to prevent this sort of attack is being vigilant about security on all machines on your network.
This review is a bit late, as this book came out over a year ago and Mountain of Black Glass came out in September. Some other excellent Tad Williams books are Tailchaser's Song (an adventure story from the point of view of a cat), the Dragonbone Chair series, Caliban's Hour, and Child of an Ancient City. I am eagerly awaiting Sea of Silver Light to finish the Otherland series.
Tad Williams' web site is located at http://www.tadwilliams.com/tadwilliams/ .
I worked with Jeff Hawkins back when he worked at GRiD Systems and developed his handwriting recognition system. He tried to sell GRiD on the Graffiti recognition since trying to recognize standard handwriting (even for a human) is *much* more difficult.
GRiD didn't want to do that so Jeff took off and founded Palm Computing, which was later bought out by 3COM.
Jeff developed the algorithms for handwriting recognition while in school during the late 80's and early 90's and his method was quite different that the methods used by everyone else. For one thing, his recognition algorithm would work well on an 8088 while everyone else needed at least a fast 386 or 486.
The main problem with trying to recognizing standard characters is that it is nearly impossible to recognize what is written without extensive context work. For example, an O (Oh) and a D can be nearly impossible to differentiate between depending on how the person writes. It's even more difficult to differentiate between, say, a 0 (zero) and an O (Oh).
GRiD worked by writing entry forms which limited the types of characters that could be entered. I.e. numeric fields looked for the closest numeric match and text fields looked for the appropriate text match (i.e. only one case). GRiD knew of the limitations of handwriting recognition, but their management was blind to the vast improvement graffiti made. I havn't read info on Xerox, but I know Jeff developed graffiti probably around '91-92, possibly earlier.
I worked on a palm computer with handwriting recognition back in '93, the Casio Zoomer. The thing took PCMCIA flash cards, ran MSDOS, had IR and serial links, and could take a PCMCIA option card. It even had digital audio so it would say 'You've got mail' when using it to connect to AOL. Oh, and the thing ran Geoworks with Jeff's handwriting recognition.
It never took off for several reasons. First of all, the handwriting recognition was not graffiti and thus had a high error rate. Second of all, it was a little too big to fit into a shirt pocket. It did have some cool features, though. Geoworks was pretty cool, but not as simple as PalmOS.
It had everything. It could tie into AOL Email, it had Pocket Quicken. It had Solitair (the most important Windoze application). It even had a program for caluclating what everyone owed at a restaraunt (including tip).
Various drugs should indeed be restricted to prescriptions. If every time we got a sniffle and grabbed anti-biotics, whether they are needed or not (i.e. a virus) then we just encourage antibiotic-resistent bacteria (which are very nasty, as my sister just had surgery since antibiotics no longer work).
I'm sorry, but for my own health I don't want my cube neighbor going down to the local pharmacy and buying antibiotics for the flu. There are specified doses of various drugs which vary based on age, weight, and family history. If drugs were unregulated we'd have a nightmare on our hands. There'd be lots of ODs and other problems.
As for antibiotic resistant bacteria, they are becoming more and more common, especially in countries that overprescribe antibiotics.
It is true that a negative is higher resolution than HDTV, HOWEVER, after all the mastering is done and the film is duplicated the resolution is lost. My information comes from SMPTE (Society of Motion Picture and Television Engineers) at a persentation done a few weeks ago at Sony in San Jose. Yes, film is higher resolution, but once it has been processed the advantage is lost. HDTV maintains its resolution throughout. Also, a CCD has higher resolution in low light levels than film since fast film must be used. In addition, the CCD has a higher dynamic range, i.e. there is much more low-level detail in dark areas with a CCD than is possible with film.
As far as grain, there currently is work progressing on digitally adding "film" grain to HDTV recorded video.
According to the SMPTE presentation, once the film is scanned to digital and goes back to film, the resolution is lost. The best film scanners cannot match what a good CCD can do at this time.
It's like photocopying a 1200 dpi grey-scale laser-printer image with a 1200 dpi scanner. Since the dots likely won't match up exactly the scanned image will not look as sharp or as good as the original. The same sort of thing happens with film. Let's face it, there are multiple generations of film to film transfers done before the film reaches the theaters. My guess is that there are a minimum of 3-4 film-film transfers. The first would be from the original film copied to a new film, since it has to be cut and edited. This would go from a negative to a positive. Now, this golden tape isn't about to be used to make duplicates for the theaters (unless somebody is really stupid) so this positive is duplicated again into one or more negatives. These negatives are then used to copy to the film that goes out to the theater. Now, that's four copies made. If the original resolution was 4000x2000 you're now probably down to 1500x800. If digital processing is done between the original film and the film that goes out to the theater then the resolution is probably higher, more likely around 2000x1000. This isn't far from HDTV which is 1920x1080. Not only that, but the HDTV can easily be recorded at 30FPS progressive or 60FPS interlaced. For the die-hard film bufs it can also be recorded at 24FPS progressive.
At the SMPTE presentation some A/B footage was shown with footage taken with an HDTV 24FPS camera and a 35mm film camera. The HDTV looked very close, except where there was a lot of contrast in which case it looked BETTER since there was more detail in the dark areas.
Expect to see reflective LCD display technology come out in the near future. Reflective LCD is not at all like your standard LCD, but is basically an LCD built over a semiconductor. Reflective LCDs can handle higher resolutions than DLP and should be cheaper to manufacture. In addition, they don't suffer from the annoying artifacts of DLP (i.e. color separation).
Film won't be used much in 10 years. HDTV actually has higher resolution than 35mm film and can have a higher frame rate. The 24fps HDTV standard came about not only to show films on HDTV, but for producers to record directly at 24fps. 24fps suffers during high motion scenes, but many film directors insist that they must have the properties of film (i.e. 24fps, grain, etc.)
One other advantage of using all digital is that currently there is degradation when digitizing film to add effects and going back to film.
HDTV also has a higher dynamic range than film. The new CCDs can handle darker material better.
In addition, within the next few years it will be less expensive to use tape instead of film. There's virtually no time spent on development, and if a shot is messed up, just rewind and start again.
I too am getting spammed like crazy from Barnes and Nobel. Perhapse this is because they purchased books.com :(
I had always used books.com because they never resorted to spamming (unlike Amazon in the past or B&N in the present). Books.com also had a feature where you could click and automatically compare their price with both Amazon and B&N and if either were lower then books.com would beat them.
Sadly books.com is now part of B&N who I now boycott. If I have to choose between Amazon and B&N I must hold my nose and go with Amazon (since they learned their lesson on spamming). I agree that their patent is silly and feel that it will be thrown out in an appeal.
Micron is coming out with a Pentium III compatible chipset that supports DDR SDRAM. I submitted an article to Slashdot at http://www.inqst.com/ddrvrmbs.htm
which shows DDR smoking RAMBUS. I wouldn't be surprised if Micron also came out with a chipset for the Athlon, which can really take advantage of the higher memory speeds.
I suspect VIA also is coming out with a DDR compatible chipset.
You obviously did not read the report. Judge Jackson's main findings were about how Microsoft stifled free capitalism. IBM promoted Lotus Smartsuite with their PCs and hence Microsoft wouldn't grant them a Windows 95 license until just minutes before they released it, thus causing IBM to not be able to be ready for the Chrismas buying season. Microsoft basically told Netscape to get out of their territory or else and then cut off Netscape's air supply. As for comparing this to the breakup of AT&T, this is the best thing that could have happened. Each of the baby bells and AT&T spinoffs has been much more successful than they would had they still been a huge monolithic corporation. In addition, long distance rates have fallen dramatically due to the competition of the other carriers. I guess you don't remember when you couldn't even own your own phone. You had to lease it from AT&T.
Likewise, look at Standard Oil or the railroads who controlled everything. There is no such thing as a free market economy. An economy without checks and balances would be like a government without checks and balances. What would happen to our government if, say, the President wrote all the laws, enforced all the laws, and interpreted and executed all the laws? Can you say dictatorship? Capitalism leads to greed. Not that there's anything wrong with capitalism, but checks and balances are required such that one individual or corporation doesn't dominate huge aspects of the industry or our lives.
Can you imagine a world where MS used their capital to gobble up other companies left and right without any stopping of it? If MS had their way, no computer manufacturer would be able to sell or distribute any other operating system. Also, all of Microsoft's distributors (i.e. computer stores) would be prohibited from selling alternatives. MS would then require all ISPs to use some proprietary protocol for Internet connectivity (like AOL) and prohibit non-Microsoft platforms from accessing the Internet.
If you think Linux is a serious competitor, perhapse in a few years, but at this point it still has a long ways to go. The applications are still missing. I hate to say it, but supporting MS Office is essential in the corporate space. Even if nobody likes it, because everyone else uses it you are forced to use it just to read other people's documents. Star Office isn't enough (I know this from experience).
In the corporate space, Linux has barely made a dent. Go down to any business and look at their computers. I imagine that you'll find few Linux boxes in use. Let's face it, for a vast majority of users Linux has too much of a percieved learning curve. Someone can sit down to a Windows box and click start and start running apps. Installing apps is trivial due to their Install Wizard. Many people put up with the frequent crashes and lock-ups because they don't know any better. Give them a Linux box and there will be just as many crashes because they won't know how to properly configure it.
I just ran into a patent which basically puts the brakes on my work. For an upcoming networking product which will support Cisco's Fast EtherChannel it would be really nice to seamlessly interoperate through Cisco's simple PAgP (Port Aggregation Protocol). I found out that Cisco just got a patent on this simple protocol. I can't believe it. It's one of the stupidest patents I've seen. The result, however, is that Cisco can use it to continue to bolster their monopoly since they prevent anyone else from interoperating with them. They patented EIGRP as well, which is a popular routing protocol which is only supported by Cisco.
Software patents should be for things that are truely revolutionary or very different. Software patents also should be subject to a different time limit, i.e. 5 years, since an Internet year is so much shorter.
I just ran into a patent which basically puts the brakes on my work. For an upcoming networking product which will support Cisco's Fast EtherChannel it would be really nice to seamlessly interoperate through Cisco's simple PAgP (Port Aggregation Protocol). I found out that Cisco just got a patent on this simple protocol. I can't believe it. It's one of the stupidest patents I've seen. The result, however, is that Cisco can use it to continue to bolster their monopoly since they prevent anyone else from interoperating with them. They patented EIGRP as well, which is a popular routing protocol which is only supported by Cisco. Software patents should be for things that are truely revolutionary or very different. Software patents also should be subject to a different time limit, i.e. 5 years, since an Internet year is so much shorter.
Cowpland should have known better. When you work for a corporation and hold stock in that company and have advanced financial information you must be very careful about buying or selling stock. For example, at my last job managers were not supposed to trade company stock within 30 days of the quarterly earnings report. The higher up you are the more important this becomes. In fact, any large trade should be pre-announced well in advance.
Many of your arguments are already addressed.
.INF files). Also, the loadable module support needs to be improved so modules do not have to be the same version as the kernel.
1. The command shell is not really that complicated. All that is needed is an alias file and you can have a shell that behaves just like DOS (except using / instead of \).
2. Plug and play is now supported in the kernel for ISA and PCI hardware. This needs to be expanded such that Linux can automatically discover the mouse, modem, graphics card, and other components at installation time and in a driver configurator. I don't think this would be that difficult to do. The main thing needed is for drivers to include some descriptive file (such as Micros~1's
When installing Linux, it should discover the sound card, mouse, graphics adapter, CRT, modem, SCSI adapter and other peripherals.
3. On the GUI front KDE and Gnome are quickly gaining acceptance. These GUIs offer as many features (and sometime more features) than Micros~1. All of the mainstream Linux distributions come with Gnome and KDE now.
4. Hardware vendor support will come when Linux gains enough market share. Right now Microsoft has total control over the hardware vendors. Microsoft tells them what features they must implement and often how to implement those features. As it is, I have found that Linux actually has pretty good hardware support.
Rambus is dead. Rambus has numerous drawbacks, such as a higher manufacturing cost and licensing fees. DDRAM has a nice window to kill it. DDRAM is basically SDRAM with double edged clocking so it can operate at speeds up to around 266MHz. It also costs little or nothing extra to produce than SDRAM and there are no licensing fees or royalties that need to be paid. Expect some non-Intel chipsets to appear shortly that take advantage of this.
Furthermore, most of the memory manufacturers are not supporting Rambus. Samsung just dropped Rambus and is going back to SDRAM.
Intel really shot themselves in the foot WRT Rambus.
While there may not be a shortage of web designers, MSCE sysadmins, or VB programmers, there is indeed a shortage of skilled programmers. Out of all of the engineers I have interviewed recently for an embedded networking position, there's very few engineers who stand out. With so few people in the United States graduating from college with a CIS or CE degree, it is becomming increasingly difficult to find qualified people. Many Silicon Valley companies I've seen or worked at are mostly employing people from other countries such as India, Taiwan, China, and so forth because that's where the good engineers are from. It really irks me to see many slashdotters complaining that they don't need a college education, but I bet many would fall flat on their face in an embedded environment. We don't need c00l d00dz who can wr1te c0de. We need engineers.
What was it, something like only 26,000 CIS/CE graduates in the United States last year, yet 340,000 positions open? I'm sorry, but many of these idiots who call themselves engineers arn't. Ask them about Big-O notation and they're lost. Ask them to write some basic algorithms during an interview and they're lost.
I recently interviewed an engineer from a well-known company and asked him to write an integer to ASCII conversion function. He managed to write something full of bugs. His response was that "the compiler will catch them and then I'll unit test it". Bzzzzzt, wrong answer. A good engineer will look through their code and double or triple check it if in doubt, since unit testing often misses things. When writing in an embedded system, such mistakes can be catestrophic in that if your code blows up due to a bad pointer or something the whole box crashes and it can be a nightmare to track the source.
Out of all of the engineers I've interviewed at my current employer, only one or two stood out. Many engineers may know how to do a few things very well but can't escape the box.
Due to our desparation to find good networking engineers we've been forced to go the H1B Visa route with a green-card sponsorship. It's expensive, legally. Other countries place a much higher status on engineering degrees than the United States. India, for example, holds engineering in the same esteem as doctors and lawyers.
In Silicon Valley, a good engineer will do very well, even considering the high cost of housing.
Currently I see little threat from Linux for Sun. Solaris is a whole different ball-game than Linux.
Solaris has many of the features needed for large enterprise installations which Linux currently lacks (i.e. journalling file system). Solaris is also much more scalable than Linux. How well does Linux run on a 64 processor machine?
Also, I must say that my experience with both Solaris and Linux that Solaris is much more stable. I'm even running KDE on Solaris which makes the user interface quite usable. I don't know about the stability of the x86 version of Solaris, but I imagine that Sun is far more interested in the Sparc version.
On my Solaris box I've filled up all virtual memory, filled up the disk, crashed KDE, had the network go down, and it just keeps going. No crashes! My Redhat Linux box, on the other hand, frequently crashes for no apparent reason with an uptime of a couple of days between crashes. The hardware is just a generic Pentium system, 64MB RAM, S3 graphics adapter, IDE hard drive, Intel 10/100 NIC. Nothing special.
Both Linux and Solaris have all the latest patches and fixes installed.
I have a Philips DVX 8000 which is a hybrid computer/DVD ROM player and line doubler AV controller. This box is notorious for not liking certain DVDs.
The Matrix DVD was almost flawless. There are a few places on the disk where the MPEG video screws up and it backs up a couple of seconds, but no hangs.
The software on the DVD cannot play the DVD, but that's not surprising given that the player has some rather special hardware (a very custom graphics adapter, for example).
I have a problem where it jumps out of the audio commentary track after each chapter, but I have this problem with other DVDs as well (i.e. Godzilla).
I love the fact that they include a separate audio-only track. I wish more DVDs would do this. Also, the white rabbit sequences are quite cool.
Linux still does this, however many pieces (i.e. the filesystem) cannot be removed. In fact, if we removed all the stuff we don't need there'd be almost nothing left. We don't need SMP support. We don't need a filesystem. We don't need security. We don't need virtual memory. We don't even want preemptive multitasking (our product uses cooperative tasking). The problem with Linux is that it's difficult to remove these pieces, since thing like the filesystem, virtual memory, and preemptive multitasking are integral to the operating system.
Linux was not designed to be a RTOS. A RTOS does not necessarily make the OS good for general purpose operation. I mean, a network switch, router, toaster, etc. has far different requirements than a PC. Real-time operating systems are more than just special scheduling. Often they're designed to be quite compact and have extremely modular functionality. They run in things like your VCR, DVD player, router, intelligent hub, palm pilet, etc. They are designed for very tight memory constraints. Some operating systems are very primitive, such as what runs in a VCR or your keyboard.
In our environment, for example, without all the things listed above, we use a unified memory space. All processes share the same memory space. We don't care about virtual memory since if a process dies it's probably catastrophic since most of the processes are interdependent.
In our product, a real-time process is often more like a TSR in that it is event driven and communicates with other processes via events.
Granted, there are places where Linux can work quite nicely, for example tivo, which, I must add, is a really cool product (and they'll burn a Linux CD with their kernel modifications upon request). But that's a different ballgame where a filesystem is important.
I researched replacing our current product with VxWorks to use either Linux or eCos. Neither would be an effective solution. eCos is far too limiting. The last I checked it didn't even have TCP/IP support, nor support the Mips processor (which is extremely popular for embedded environments [eCos does have an eval version for the NEC VR4300]). Another problem with Linux is the GNU public license. In an embedded environment it is often necessary to hack the kernel or pieces of it to do what you need.
I found RTEMS http://www.rtems.com is an excellent solution. It is open source and has the latest BSD TCP/IP stack. The license also allows us to hack up parts of the kernel without having to release the source code back.
For our current product, any RTOS we use we need to make some custom changes to the TCP/IP stack (which involve changes that would NOT be of benefit to the open source community unless the company wanted to compete with us). The eCos license would let us do this, Linux does not.
Don't get me wrong. I support open source and feel that any changes that enhance a product, fix a bug, or whatever that are not proprietary to one specific application certainly deserve to be released back to the community.
Another problem with the Linux kernel is that it is far too bloated. It would take a lot of work to bring it down to size. For example, everything we have is in flash. There is no file system. Our current image, symbols and all, is around 1.25MB compressed (using Zlib, which incidently is much faster than Gzip). We have a lot of code and not a lot of room. Having a small kernel is often necessary. We don't need a file system, virtual memory, security, loadable modules, or many of the other features of the Linux kernel. In fact, due to the fact that everything must be physically linked to the kernel, this totally rules out Linux. The 1.25MB is EVERYTHING, all of our code, protocol snooping, debugger, routing, console, SNMP, and firmware for supporting a second CPU (which runs our custom RTOS which is only a few pages of code).
eCos and RTEMS fill the void nicely in that one can pick and choose which features get compiled in. Unfortunately eCos is missing some very important pieces (i.e. TCP/IP).
VxWorks is a pain in the *ss since it has numerous bugs (i.e. improper cache operation on the NEC VR4300 CPU) and has terrible support (after paying thousands of dollars a year it's not unusual to have to wait weeks or months for a response). Not only that, but we had to buy their networking source code which costs many thousands of dollars (it's mostly just the BSD 4.3 TCP/IP stack, nothing special, and various daemons).
I am seriously looking at moving to RTems in the future. They have always responded promptly and have the features we need. My boss was all for Linux until I told him about what was involved to use it, and the GPL immediately killed it.
If we do move to RTems, I would certainly make a number of the changes we would make available back to the source code, but there are a number of others that we cannot release (since there's competition).
"I'd like to see, for once, highly technical jobs that judged you on the basis of your skills instead of "I'm an MSCE", or "I have a BA in *blah*". "
This has been my experience in Silicon Valley. They often could care less about how much education or seniority you have as opposed to your skill. Even Linus has commented on this.
I know people who have risen to the top of their fields as engineers who never graduated from college. It's also not unusual for someone with less education to earn far more than someone with a Phd. It all depends on skill.
I figure once I graduated from college future courses and further education is not that important as long as I keep current with the technology. In fact, when reviewing resumes, often I only skim the education part out of curiosity. I could really care less.
In Silicon Valley my experience has been that experience and skill has far more importance than education. If a company is going to give you trouble over your education (or lack thereof) then it's probably not a good company to work for to begin with.
Getting started without a degree can be more difficult, but with the shortage of qualified engineers, more and more companies offer internships. That's how I got started. My internship opened at least as many doors as my diploma did.
I'm not saying that an education is a bad thing either. I feel that immersing ones self in many different areas of knowlege is a good thing and produces a more well-rounded individual. Even those non-engineering related courses can prove invaluable from time to time. It's much more difficult to immerse ones self in many different areas when one has to work 40+ hours a week. Also, multi-talented engineers are far more valuable than someone who only knows a narrow field.
However, it is rather cool to have people call you doctor. Next thing you know Linus will have a blue police box and will be known as THE doctor.
I think that breaking up Microsoft would not do much to them. Instead, here's some other things that could be done to Microsoft to cripple it:
1. Force MS to release full specs on all application file formats (i.e. MS Office) and force them to write full export capabilities to the other top competing applications for Windows and other platforms (not just Mac).
2. Prohibit MS from bundling multiple packages together (i.e. Office 2000 integrating a photo publishing package to kill off Adobe).
3. Prohibit MS from deviating from standards set outside of MS (i.e. HTML, HTTP, TCP/IP) or from generating new internal protocols when standard protocols/file formats already exist.
Announcing layoffs has a much more widespread effect than many people realize. When I was at my last job, they had a 200 person layoff. As a consequence, morale went through the floor and many of those who were not layed off left for greener pastures. They lost a lot of intellectual knowlege and experience which cannot be easily recovered. Furthermore, here in California, if a company lays someone off they cannot hire someone else for the same job for at least 6 months. SGI is in for a very rough ride ahead. It will be very difficult for them to keep those who were not layed off.
When layoffs occur, it is some ways harder on those who were not layed off. You then start wondering, "when is it my turn to be layed off?"
Being layed off is not necessarily a bad thing around here in Silicon Valley. There are a lot of job openings elsewhere, and usually the company doing the layoff offers a good severense package. I know that when my division at my last company was dissolved, within a month everyone had found a better job than they had, and with the severense packages many people did quite well. Everyone from the admin to the managers found new jobs quickly. Most of the people are happier at their new job as well.
It is sad to see a company like SGI have a layoff. I fear that SGI may not be able to recover from this.
I must agree with you on one thing. I run KDE on both Linux (after Gnome crashed it hard one time too many) and on Solaris. Downloading the source and building it on Solaris wasn't too bad. Gnome, on the other hand, is a mess. Libraries must be downloaded from all over the place (and a number of the reported links are broken). I've given up trying to compile Gnome for Solaris, and given my experience of Gnome on Linux I've given up the thought for now.
Gnome looks great. It has a few things I really like, (lots of cool apps for the tool bar) and some things I don't like (instability). One thing I'll say about KDE is that it is clean. The UI is clean and to the point. The source tree is also clean. Just download qt and the KDE source files and that's it.
KDE has been running fine for months on my Solaris box with no crashes or lockups (once I downreved kwm to 1.1 from 1.1.1). I've even run out of swap and the system gracefully recovered.
I look forward to compiling KDE 1.1.2. I just wish the KDE folks would update their bug database better.