The whole motivation for the Emperor turning against House Atreides, according to Thufir Hawat (in his conversation with the Baron Harkonnen), is that Duke Leto had trained a small military force to the same level of competence as the Emperor's Sardaukar, and was "in a position" to enlarge it. This was viewed as a direct threat to the Emperor's supremacy, and mandated the Emperor's conspiracy with the Harkonnens to wipe out the Atreides.
Until that conversation with Hawat, the Baron had not really understood why the Emperor had helped him on Arrakis. In fact, a large section of the book (the encounter between Count and Lady Fenring and the Harkonnens on Geidi Prime) is dedicated to showing the Baron's political naivete.
Then the Lynch movie wrecks all this subtlety with "weirding modules" and rain-making. Christ.
Interesting that you should mention the possibility of Geac buying Baan.
The Geac bottom-feeder strategy follows Nature's successful model of a parasitic worm that eventually kills its host. It works like this. First, you find a struggling company with a large existing base of customers. You buy it on the cheap (there are plenty of such companies for sale). You fire everyone you can, especially all the admin types, to cut the run rate as low as possible. HR, accounting, etc., gets consolidated into Geac headquarters in Markham, where you flog a bunch of underpaid admins to try to service the increased load. You put hiring and wage freezes into place at the acquired company, which pretty much guarantees that high-salary engineers leave by attrition.
You get rid of most of the field service people. You consolidate what you decide to keep into Geac field service. Of course, this means you most often send Geac techs to service the software/equipment from the acquired company. Naturally they don't do a very good job at first, but don't worry -- they eventually learn via on-the-job training.
Finally, you raise the cost of service to the existing customers so it becomes extremely profitable. This causes every customer who was even marginally unhappy with the acquired company's service or product to drop out in disgust, but so what. What remains is a (declining) service base that generates an enormous amount of cash.
You're happy if the acquired company makes new sales, but even if they don't, they generate cash from service -- lots of cash -- for a number of years.
Naturally, Geac companies managed this way always have flat or declining sales. Eventually they die.
But in the meantime, you've generated a huge bunch of cash onto your balance sheet, which you use to acquire more struggling companies that you strip in the same way to generate more cash.
Loop.
Unfortunately, Geac seems to have forgotten this highly successful, though disgusting, model. These days Geac appears to believe that it's a real software company, especially after the D&B acquisition. Classic case of a parasite forgetting its mission objective!
That's OK, though. The guy who thought up the original model retired several years ago, a multi-millionaire.
Hey, those of us who can't invest the time to get *good* at the goddamn game have no choice. Hiding in the shadows is about all I can accomplish without getting my nuts shot off.
I ran the development group for a product that generated over $20M in sales using QNX4. The product is installed in over 2000 locations, and it is rock solid. Our experience with QNX was positive. The pricing was very competitive.
However, the following must be kept in mind:
1) QNX is supported by QNX. If you have a problem, you have to talk to them. There are a few smart people in Kanata, Ontario, who have a heck of a lot to do, and not enough resources to do it, just like every software organization. So before you start comparing QNX to Linux, ask yourself what you're supposed to do when you hit a bug, and who's going to help. We got answers from QNX, but we were shipping loads of QNX licenses.
2) QNX is very proud of the fact that everything lives outside of proc. This is a very elegant design. Nevertheless, if the filesystem has a bug in it, it doesn't really matter whether it's running in kernel space or as a separate process. You're still screwed. You have to wait for Bill Flowers, or whoever is supporting fsys now, to fix it. At the end of the day, you're queueing on one guy, maybe two, and the kernel structure, except for bragging rights with your friends, doesn't help you much. Moral? Better design doesn't necessarily mean better support or a more reliable product. I'll take 4000 hackers swarming on the bloated Linux kernel over one Dan Dodge tweaking proc when he has time, any day.
3) QNX always claims they're more reliable than Linux. Maybe they are, these days. All I know is, back in QNX4 days, our death-screen ratio during development was about the same as the death-screen ratio I had on the last NT project I did (which is pretty good -- NT was relatively stable for us). But I've never crashed Linux. Even once.
I'm fond of QNX, and I'm glad it's being given away, even if they're keeping the code proprietary, and doing it only for non-commercial use. But QNX has a murky future. It's not clear whether anyone besides Microsoft is going to be able to get a license fee for an operating system for very much longer. Until the whole ball of wax is open-sourced, and QNX mutates into an applications house (like Spyglass), they'll be counting on OS revenue to keep the lights on during that dark Canadian winter. There are some darker winters in store, I fear.
This is extremely sound advice. As a hiring manager, I've found that you have to push hard on HR numb-nuts who ask you for lists of "keywords" to put in their ads. My standing rule is that no ad goes in unless I have personally written it, then reviewed the final copy. Any HR weenie that puts an ad in that was _not_ written and reviewed personally by me gets called on the carpet, along with his/her boss, and furthermore HR pays for the ad out of its operating budget, not out of my hiring budget.
But a lot of hiring managers don't have (or don't realize that they have) this degree of power in the organization, so the HR drones rule. This means the ads have exciting and attention-getting titles like "Programmer III" and "Junior Support Engineer II", along with silly lists of keywords that someone might have barfed up in a spare moment, but never believed in their wildest dreams would actually find their way into print.
So, this person is 100% right. Apply anyway. You'll probably find that the hiring manager is secretly appalled at how bad his/her ad looked. And, if s/he turns out to be a keyword bonebrain, you don't have to work there. But at least get the interview.
I traipsed around and visited a bunch of Boston area VC's over the span of about 14 months, a while ago. Ultimately, our venture was not funded, and we gave up. But I learned an awful lot.
1) VC firms are in the business of meeting with investors. That's how they find new opportunities. They are *not* in the business of stealing your idea; most of them won't have the faintest idea how you do what you do. It may be difficult to get a first appointment, but there are a number of ways to do so. In Massachusetts, there is an organ of the state government that functions as a pseudo-VC; they are extremely helpful in getting you in to see other VC's.
Getting a meeting with a VC means nothing. Don't get excited when it happens. They're PAID to meet with you.
And don't expect them to sign an NDA, for the reasons that others have enumerated below.
2) VC's generally have a "hidden agenda." They like to invest in certain companies with a certain profile. It helps to identify this profile before you see the VC. Ask every VC you get to meet if s/he can think of a VC who invests in companies with the same basic market/structure/whatever as yours. A VC who decides you don't fit the profile dismisses you, and you'll often never know why.
3) Some VC's invest in zero-stage companies; most don't. Most VC's would rather be second- or third-money in; nobody wants to be first. So the first investor is the tough one, because everyone else just piggy-backs off #1's due diligence. It's usually easy to find a zero-stage VC -- just ask. For example, "Zero-Stage Capital" was such a VC in Boston. Duh.
4) For some ventures, only "angel" investors are appropriate. Our venture, for example, fell somewhere between two VC "categories", and everyone from category #1 was scared off 'cause the venture smelled like category #2; and vice-versa. Angels are tough to find; several of the ones who met with us were funneled our way by a VC who couldn't fund us, but who believed in our idea.
5) Patents are appropriate for four reasons: first, they smoke out whoever else has already patented your idea. "Wait," you say, "my idea is totally original and wonderful and nobody could have possibly patented it." Wrong. You should see the claims in some of these patents. Patents from left field are uncovered by the patent search, and are somehow applicable to your idea! You have to navigate carefully through this minefield.
Second, a patent (pending or otherwise) is a valuable commodity when dealing with some investors. Hip VC's know that a patent is useless, and that speed to market is the key to success; but angel investors (and even some VC's) sometimes aren't that bright. Neither are bankers.
Third, a patent is an important *asset* for your new company. Why? Because patents can be CAPITAL ASSETS. Developed software in an LLC, for example, IS NOT A CAPITAL ASSET. What does this mean? It means that when you sell out your little LLC to some bigger fish, you pay ORDINARY INCOME TAXES on the sale of the asset. Can you say ouch? On the other hand, you can sell the PATENT to someone (along with the software) and get capital gains treatment for most of the transaction. This could amount to huge $$.
Fourth, a patent *protects* you from some other asshole who comes along later and patents something similar. Until we slap ourselves upside the head and make software patents illegal, there's always going to be some moron who tries to patent, say, a heapsort. And there's probably some patent examiner who will let it through. So you *have* to patent your idea, in self-defense.
You should expect to pay about $5-7K, spread over several years, to your patent attorney. S/he should be able to tell you whether you are likely to exceed that amount or not, depending on what the patent search turns up.
6) VC's are generally looking for reasons NOT to invest in your company. That's helpful to them, because it culls the herd, so to speak, quickly. I can remember one meeting with a Boston VC which was dominated by a woman who had read an article the day before (with only tangential applicability to our idea) that pissed all over certain technologies. She trotted out her article, and that was the end of us.
Maybe she was just having a bad day. Whatever. We were screwed. Get used to this phenomenon -- VC's try to pigeonhole you so they can understand you. Truly new ideas are difficult to get across to them. They would, by and large, much rather invest in something conventional.
7) VC's tend to like big deals better than little deals. If you need $20M, you are more likely to find it than if you need $2M. Paradoxical, but true. Try to find reasons why you need boatloads of cash, but don't think you can fool anyone. They have to be legitimate reasons. These guys are finance dweebs and they are idiot savants when it comes to budgets and dollars.
Yes, there is a serious problem with Gartner group (and other) reports. Unless you are in the trenches with the techies, you really don't know what you're talking about. The Gartner people are smart, but many of them are seriously uninformed. Their senior analysts, who usually do have a track record in the industry they cover, generally haven't developed software in years (the Bob Metcalfe syndrome).
jht makes an important point: even if you ARE technically proficient, it is very tempting to make predictions that fall flat. I love #1 -- who didn't read about 100 different smarty-pants analysts, including Gartner, who predicted the end of Apple? OK, in fairness to Gartner, they tried to temporize their Apple predictions with "probabilities of decreased market share" and "recommendations against" for businesses, but it amounted to same thing.
CEO's (and other executives) really do have a serious problem. They must make technical decisions for their companies, and they have nowhere to turn but to the Gartners of the world. When the CEO is called in front of the board of directors, s/he's got to 'splain him/herself. Years ago, the question from the board was always "Why not IBM? Why did you buy this other crap?" So they bought IBM, defensively. Now they buy Microsoft, defensively. The so-called "Linux hype" has broken up this cozy little defensive arrangement, and the Gartners of the world must scramble to provide their clients with new justifications to provide to their bosses.
Slashdot readers should also understand that there are lots of dirty secrets surrounding these "reports." For example, if your company PAYS for industry coverage by Gartner, generally speaking your company gets an opportunity to present ITS side of the argument in a way that it normally would not, just because of its ACCESS to the analyst(s) writing the report. Gartner will deny this to their last breath, but it's true.
Another dirty little Gartner secret is the quality of the "analysts." I happen to know a Gartner analyst personally, who has been quoted many times in the mainstream press making pronouncements about Micros~1 and others. This person's background? Liberal arts. Has absolutely no clue -- and I should know, I've had to answer a lot of this person's stupid questions over the years.
So the bottom line for Gartner is that they need to rotate out the fuddy-duddies from the 70's, whose last remembrance of Unix is v6 in college on a PDP-11/70. Then they need to ask some important questions about why the majority of competent technoids is so damn excited about Linux and the Open Source movement. Simultaneously, they need to recruit (and pay!) technical people who actually know that they're talking about. For starters, how about people who have hands-on Linux experience and who can explain why we can accomplish
- about 100x more, - about 100x faster, - with greater stability, - in certain problem domains,
with Linux than with Microsoft.
As long as Gartner fails to understand WHY there is a nerd revolution, and WHY Linux is so exciting to the people like us who, increasingly, actually control the technical infrastructure of the world, they will totally miss the point. Their liberal-arts analysts can only see the world through their Microsoft laptops (which constitute, in many cases, the only computer they've ever actually used).
I worked with QNX for several years, '92-'95. During that time period, QNX dropped its "QNX Windows" product in favor of its new "Photon" product. The old product was junk; unfortunately, we used it heavily. Photon's schedule kept slipping and I went on to better things before Photon became a reality.
The OS is pretty neat, but it could use some open-source thrashing instead of just a few (admittedly smart) guys in Ottawa hacking with it. We experienced a reasonable number of "Blue Screens of Death" (yes, QNX kernel panics bear a marked resemblance to the NT dies we are all familiar with). Probably many of those have been fixed over time, but equally probably new ones have come along. Exposing the source code to thousands of eyes would be productive, I suspect.
We also had some problems with the file system (corrupted files). At that time basically only one guy (Bill, a really smart fellow) was working on the file system, and unfortunately you had to hit him in the (proverbial) nuts with a baseball bat to get his attention. To QNX's (and Bill's) credit, they did fix the problem. However, this illustrates a big disadvantage that closed-source vendors have -- namely, in many cases only one dude can do the work, whereas open-sourcers can call on help (and patches) from around the world.
The OS does fit in a remarkably small memory footprint, and its messaging system is very fast, even between machines. It was a good and stable framework on which to build a meta-messaging applications layer. Our customers were astounded at the up-time of their QNX-based systems, as opposed to the Micros~1 competition.
QNX suffers from the same problem as all proprietary operating systems -- you have to call them when you have a problem, and their attention to you seems to depend on how much of their business you represent. In 1995, it appeared that they were focusing heavily on the embedded market (set-top thingies, etc.), and were more-or-less in standby mode on the general-purpose OS side. I haven't seen anything to indicate a change in direction in the last few years, although Photon appears to be real now. Also, I should admit that I haven't been watching that closely.
As far as QNX pricing is concerned, it was very reasonable for OEM quantities. I would encourage anyone to contact them to get a price quote. They seem to be willing to structure a deal, unlike some larger companies we know.
Last time I installed a game ("Legacy of Time"), a bunch of stuff stopped working on my machine, no doubt due to some DLL incompatibility introduced by the install procedure. Fortunately, I always have a complete copy of windows\system and of the Registry squirreled away. So I just copied them back, and voila, everything started working again.
The game is still on the shelf. Want it? I *never* install anything on this pig without a complete copy of everything journalled aside. That's the *real* Micros~1 install procedure, as far as I'm concerned.
I guess maybe it's OK to close your eyes and pray that the next install won't gronk your machine if all you do is play games on it. Me, I gotta keep this Win95 pig running, so I can eat next week, because I have to maintain code for my clients.
Which means I don't install nothin' on it, baby. Nothin'.
Good point. Let's turn it around, then: the icon still means Open Source Friendly, but we maintain a "good guys" list only. Any ideas how to promote this widely, other than the usual newsgroups/mailing lists?
This is a terrific idea. The basic problem, as Scott points out, is that there could be serious abuse of the blacklist. Most companies have no chance of controlling all of their pointy-haired middle managers, so the company could end up blacklisted just because of one person, even though their "official" policy might be reasonable.
So let's give companies a really easy way to get off the blacklist: they can put an "Open Source Friendly" icon on their web page. I'm imagining something like a penguin holding hands with or hugging a baby gnu (the BSD daemon won't go over with the Christian Right, sorry), with a caption reading "Open Source Friendly." The icon links to *OUR* page, which explains the pledge that the icon represents.
I think that this will work wonders. Recruiters are extremely sensitive to issues like this, and "Open Source Friendly" (and the icon) will start appearing on recruitment advertisements very quickly. Hiring managers will quickly learn that if they're not "Open Source Friendly," their chances of filling positions will decrease dramatically.
I volunteer to draft the wording of the web page. Anybody want to host it? GIMP the icon?
This story is fascinating to me, and I say/. should follow it to the bitter end. It has all the elements of "Edge of Night" or "As The World Turns," or any of the old daytime soap operas. Like a soap, the story keeps morphing around, reminiscent of my boss trying to explain at review time why I don't make as much money as the idiot down the hall.
There are elements of heartbreak: the happy story of the QNX engagement, followed by Jim Collas' devious Victorian trick of disappearing out the opposite door of the four-wheeler on the way to the church, leaving our Canadian friends holding their Canadian private parts at the altar.
There are elements of humor: the vain attempt to annoint my toaster oven's Internet interface as, somehow, "Amigan." Amiga in the fridge, Amiga in the microwave -- gimme that bouncin' ball!
There is betrayal: the sudden departure of once-friendly and communicative executives, to be replaced by mindless drones of a mechanized society, spouting biz-speak from both sides of their mouths.
And there is overall "Dallas"-style incompetence, as the idiot children of a once-great empire recklessly spend all the carefully-built-up good will of the legions of loyal fans who cut their teeth on these wonderful machines.
I've looked hard at porting applications to Linux from Microsoft platforms, mostly VB apps. It's a bitch. There's no VB equivalent for Linux (I know there's some activity on this front), and I'm not aware of an easy porting path (if anyone knows one, comment, please).
There is no question that Linux is a more stable platform than any Microsoft platform. We've been installing Linux machines for our clients whenever we can -- Linux is much easier to work with. However, for GM to convert all those machines over to Linux doesn't make much sense to me. Not only do all of the VB apps have to be rewritten, but also most of the third-party software packages that this guy *thinks* he'd like to add are probably written for Windows (WINE advocates: I can't see my local GM dealer putting up with an "80% working" application under WINE, can you?).
Finally, are this guy's motivations (running out of disk space) silly, or is it just me? He could solve his problem by spending $200 at each store to upgrade to a larger hard drive and maybe to install more memory (DriveCopy does a very nice job of cloning hard drives, small --> large). It would probably cost him another $200 per store to hire a nationwide third-party service organization to visit all his dealerships and perform the upgrade. But at the end of it, he'd have what he needs: working systems with plenty of elbow room to grow.
(Yes, he could take the $400 and probably buy a better PC than he has right now. But I'm assuming he'd rather not IPL them all, install software on them all, burn them in, and so on, which will cost at least another $400 per machine to stage, not including shipping and installation at the dealer.)
If he converts to Linux, he has a huge problem on his hand: not only does he need to rewrite all his applications and do all that expensive staging above, but then at the end of the whole process (assuming it all works) he'll have to explain on a daily basis to his bonehead management how come the XYZ Inventory Control System doesn't run on his machines (his boss spots the XYZ ad in Useless American Cars Weekly, for example).
I have a friend who works at Liberty Mutual. According to him, if you have a PC on your desk, and you install any software (repeat: ANY SOFTWARE) on it whatsoever, and you are caught, this is grounds for dismissal. Every PC's configuration comes from a "gold" CDROM that is maintained by IS. If a PC is found to have been "corrupted" by foreign software, it is immediately reloaded from scratch from the "gold" CD.
Some folks have been quite impassioned in this forum about the freedom and productivity that results from having one's own PC. No doubt there are many desktops at Liberty Mutual that are running "verboten" software, and no doubt there are copies of the "gold" CD floating around that are used to reload PC's that have "gone south," and no doubt all of this activity is occurring without anyone in IS knowing about it.
But from the *company's* standpoint, their policy is working. They don't have to run around supporting PC users (because everyone is afraid of getting caught, so all everyone ever does is reload from the "gold" CD instead of calling for help). The company believes that everything is wonderful. And the IS department believes they are in control of this wonderful imaginary world.
All of which is not unfunny. But here's the rub: IS is in charge of all the procurement decisions. And if IS is the customer, then the SunRay sure sounds like a terrific idea, doesn't it?
I guess I can relate this to my own experience as VP Engineering for a small technology company. One day a whole metric f**ckload of low-end Compaq PC's showed up in Marketing, Sales, and Administration. Some Compaq sales dude had sneaked in the back door and sold a bill of goods to guys who had absolutely no clue what they were buying (my personal theory as to why Compaq sales are off -- technical guys are pissed off at them for selling directly to the suits, and so as our power increases, we buy elsewhere for revenge). Well, those same dudes are buying SunRays.
Which begs the question, of course, as to whether SunRays are good or bad. I can see arguments both ways, most of which have already been made by others.
In my experience, a troll message follows these general rules:
1) It is a relatively short message. 2) It contains poorly-spelled words or odd (or no) capitalization. 2a) It contains multiple exclamation marks or question marks. 3) It contains obscene language or four-letter words. 4) It generally does *not* contain a URL. 5) It often (but not always) comes from ACs.
Based on these criteria, it seems to me that an automatic troll filter could be built fairly easily that automatically deprecates messages that fall into these categories. Perhaps the spell-checking would be awkward from a CPU time perspective, but the other stuff should be straightforward.
No, *you* don't get it. Friends are allowed to help friends become less nerdy and more interesting. The quality of stories and articles on/. is high; the editorial judgement is good; the comments are often amusing; and therefore Rob and Jeff are clearly capable of high quality effort. We should urge them to apply those same high standards to their broadcasts.
2) Lose the third guy. It's way too confusing to listen to. 3) If you have nothing to say, say nothing. Does the "broadcast" *have* to be 13 minutes? 4) Take Hemos' head out of the water jug it sounds like he's talking from (seriously, what the $#&^% is wrong with the guy's mike??). 5) Stick to the news. I don't care what you guys did at the Linux conference, or how many t-shirts you got, or how much pinball you played, or whatever. I DON'T CARE. OK? 5a) OK, if someone got laid, I wouldn't mind hearing about that. But that's unlikely, considering what you guys sound like. 6) I also don't care about what imaginary spot you're broadcasting from. Just tell Hemos to stop broadcasting from the water jug. 7) Stop giggling.
Slashdot Mirror
This comment says it all, and it's hilarious! Mod it up!
The whole motivation for the Emperor turning against House Atreides, according to Thufir Hawat (in his conversation with the Baron Harkonnen), is that Duke Leto had trained a small military force to the same level of competence as the Emperor's Sardaukar, and was "in a position" to enlarge it. This was viewed as a direct threat to the Emperor's supremacy, and mandated the Emperor's conspiracy with the Harkonnens to wipe out the Atreides.
Until that conversation with Hawat, the Baron had not really understood why the Emperor had helped him on Arrakis. In fact, a large section of the book (the encounter between Count and Lady Fenring and the Harkonnens on Geidi Prime) is dedicated to showing the Baron's political naivete.
Then the Lynch movie wrecks all this subtlety with "weirding modules" and rain-making. Christ.
Interesting that you should mention the possibility of Geac buying Baan.
The Geac bottom-feeder strategy follows Nature's successful model of a parasitic worm that eventually kills its host. It works like this. First, you find a struggling company with a large existing base of customers. You buy it on the cheap (there are plenty of such companies for sale). You fire everyone you can, especially all the admin types, to cut the run rate as low as possible. HR, accounting, etc., gets consolidated into Geac headquarters in Markham, where you flog a bunch of underpaid admins to try to service the increased load. You put hiring and wage freezes into place at the acquired company, which pretty much guarantees that high-salary engineers leave by attrition.
You get rid of most of the field service people. You consolidate what you decide to keep into Geac field service. Of course, this means you most often send Geac techs to service the software/equipment from the acquired company. Naturally they don't do a very good job at first, but don't worry -- they eventually learn via on-the-job training.
Finally, you raise the cost of service to the existing customers so it becomes extremely profitable. This causes every customer who was even marginally unhappy with the acquired company's service or product to drop out in disgust, but so what. What remains is a (declining) service base that generates an enormous amount of cash.
You're happy if the acquired company makes new sales, but even if they don't, they generate cash from service -- lots of cash -- for a number of years.
Naturally, Geac companies managed this way always have flat or declining sales. Eventually they die.
But in the meantime, you've generated a huge bunch of cash onto your balance sheet, which you use to acquire more struggling companies that you strip in the same way to generate more cash.
Loop.
Unfortunately, Geac seems to have forgotten this highly successful, though disgusting, model. These days Geac appears to believe that it's a real software company, especially after the D&B acquisition. Classic case of a parasite forgetting its mission objective!
That's OK, though. The guy who thought up the original model retired several years ago, a multi-millionaire.
Hey, those of us who can't invest the time to get *good* at the goddamn game have no choice. Hiding in the shadows is about all I can accomplish without getting my nuts shot off.
I ran the development group for a product that generated over $20M in sales using QNX4. The product is installed in over 2000 locations, and it is rock solid. Our experience with QNX was positive. The pricing was very competitive.
However, the following must be kept in mind:
1) QNX is supported by QNX. If you have a problem, you have to talk to them. There are a few smart people in Kanata, Ontario, who have a heck of a lot to do, and not enough resources to do it, just like every software organization. So before you start comparing QNX to Linux, ask yourself what you're supposed to do when you hit a bug, and who's going to help. We got answers from QNX, but we were shipping loads of QNX licenses.
2) QNX is very proud of the fact that everything lives outside of proc. This is a very elegant design. Nevertheless, if the filesystem has a bug in it, it doesn't really matter whether it's running in kernel space or as a separate process. You're still screwed. You have to wait for Bill Flowers, or whoever is supporting fsys now, to fix it. At the end of the day, you're queueing on one guy, maybe two, and the kernel structure, except for bragging rights with your friends, doesn't help you much. Moral? Better design doesn't necessarily mean better support or a more reliable product. I'll take 4000 hackers swarming on the bloated Linux kernel over one Dan Dodge tweaking proc when he has time, any day.
3) QNX always claims they're more reliable than Linux. Maybe they are, these days. All I know is, back in QNX4 days, our death-screen ratio during development was about the same as the death-screen ratio I had on the last NT project I did (which is pretty good -- NT was relatively stable for us). But I've never crashed Linux. Even once.
I'm fond of QNX, and I'm glad it's being given away, even if they're keeping the code proprietary, and doing it only for non-commercial use. But QNX has a murky future. It's not clear whether anyone besides Microsoft is going to be able to get a license fee for an operating system for very much longer. Until the whole ball of wax is open-sourced, and QNX mutates into an applications house (like Spyglass), they'll be counting on OS revenue to keep the lights on during that dark Canadian winter. There are some darker winters in store, I fear.
This is extremely sound advice. As a hiring manager, I've found that you have to push hard on HR numb-nuts who ask you for lists of "keywords" to put in their ads. My standing rule is that no ad goes in unless I have personally written it, then reviewed the final copy. Any HR weenie that puts an ad in that was _not_ written and reviewed personally by me gets called on the carpet, along with his/her boss, and furthermore HR pays for the ad out of its operating budget, not out of my hiring budget.
But a lot of hiring managers don't have (or don't realize that they have) this degree of power in the organization, so the HR drones rule. This means the ads have exciting and attention-getting titles like "Programmer III" and "Junior Support Engineer II", along with silly lists of keywords that someone might have barfed up in a spare moment, but never believed in their wildest dreams would actually find their way into print.
So, this person is 100% right. Apply anyway. You'll probably find that the hiring manager is secretly appalled at how bad his/her ad looked. And, if s/he turns out to be a keyword bonebrain, you don't have to work there. But at least get the interview.
I traipsed around and visited a bunch of Boston area VC's over the span of about 14 months, a while ago. Ultimately, our venture was not funded, and we gave up. But I learned an awful lot.
1) VC firms are in the business of meeting with investors. That's how they find new opportunities. They are *not* in the business of stealing your idea; most of them won't have the faintest idea how you do what you do. It may be difficult to get a first appointment, but there are a number of ways to do so. In Massachusetts, there is an organ of the state government that functions as a pseudo-VC; they are extremely helpful in getting you in to see other VC's.
Getting a meeting with a VC means nothing. Don't get excited when it happens. They're PAID to meet with you.
And don't expect them to sign an NDA, for the reasons that others have enumerated below.
2) VC's generally have a "hidden agenda." They like to invest in certain companies with a certain profile. It helps to identify this profile before you see the VC. Ask every VC you get to meet if s/he can think of a VC who invests in companies with the same basic market/structure/whatever as yours. A VC who decides you don't fit the profile dismisses you, and you'll often never know why.
3) Some VC's invest in zero-stage companies; most don't. Most VC's would rather be second- or third-money in; nobody wants to be first. So the first investor is the tough one, because everyone else just piggy-backs off #1's due diligence. It's usually easy to find a zero-stage VC -- just ask. For example, "Zero-Stage Capital" was such a VC in Boston. Duh.
4) For some ventures, only "angel" investors are appropriate. Our venture, for example, fell somewhere between two VC "categories", and everyone from category #1 was scared off 'cause the venture smelled like category #2; and vice-versa. Angels are tough to find; several of the ones who met with us were funneled our way by a VC who couldn't fund us, but who believed in our idea.
5) Patents are appropriate for four reasons: first, they smoke out whoever else has already patented your idea. "Wait," you say, "my idea is totally original and wonderful and nobody could have possibly patented it." Wrong. You should see the claims in some of these patents. Patents from left field are uncovered by the patent search, and are somehow applicable to your idea! You have to navigate carefully through this minefield.
Second, a patent (pending or otherwise) is a valuable commodity when dealing with some investors. Hip VC's know that a patent is useless, and that speed to market is the key to success; but angel investors (and even some VC's) sometimes aren't that bright. Neither are bankers.
Third, a patent is an important *asset* for your new company. Why? Because patents can be CAPITAL ASSETS. Developed software in an LLC, for example, IS NOT A CAPITAL ASSET. What does this mean? It means that when you sell out your little LLC to some bigger fish, you pay ORDINARY INCOME TAXES on the sale of the asset. Can you say ouch? On the other hand, you can sell the PATENT to someone (along with the software) and get capital gains treatment for most of the transaction. This could amount to huge $$.
Fourth, a patent *protects* you from some other asshole who comes along later and patents something similar. Until we slap ourselves upside the head and make software patents illegal, there's always going to be some moron who tries to patent, say, a heapsort. And there's probably some patent examiner who will let it through. So you *have* to patent your idea, in self-defense.
You should expect to pay about $5-7K, spread over several years, to your patent attorney. S/he should be able to tell you whether you are likely to exceed that amount or not, depending on what the patent search turns up.
6) VC's are generally looking for reasons NOT to invest in your company. That's helpful to them, because it culls the herd, so to speak, quickly. I can remember one meeting with a Boston VC which was dominated by a woman who had read an article the day before (with only tangential applicability to our idea) that pissed all over certain technologies. She trotted out her article, and that was the end of us.
Maybe she was just having a bad day. Whatever. We were screwed. Get used to this phenomenon -- VC's try to pigeonhole you so they can understand you. Truly new ideas are difficult to get across to them. They would, by and large, much rather invest in something conventional.
7) VC's tend to like big deals better than little deals. If you need $20M, you are more likely to find it than if you need $2M. Paradoxical, but true. Try to find reasons why you need boatloads of cash, but don't think you can fool anyone. They have to be legitimate reasons. These guys are finance dweebs and they are idiot savants when it comes to budgets and dollars.
OK, that's the brain dump. Hope this was helpful.
Yes, there is a serious problem with Gartner group (and other) reports. Unless you are in the trenches with the techies, you really don't know what you're talking about. The Gartner people are smart, but many of them are seriously uninformed. Their senior analysts, who usually do have a track record in the industry they cover, generally haven't developed software in years (the Bob Metcalfe syndrome).
jht makes an important point: even if you ARE technically proficient, it is very tempting to make predictions that fall flat. I love #1 -- who didn't read about 100 different smarty-pants analysts, including Gartner, who predicted the end of Apple? OK, in fairness to Gartner, they tried to temporize their Apple predictions with "probabilities of decreased market share" and "recommendations against" for businesses, but it amounted to same thing.
CEO's (and other executives) really do have a serious problem. They must make technical decisions for their companies, and they have nowhere to turn but to the Gartners of the world. When the CEO is called in front of the board of directors, s/he's got to 'splain him/herself. Years ago, the question from the board was always "Why not IBM? Why did you buy this other crap?" So they bought IBM, defensively. Now they buy Microsoft, defensively. The so-called "Linux hype" has broken up this cozy little defensive arrangement, and the Gartners of the world must scramble to provide their clients with new justifications to provide to their bosses.
Slashdot readers should also understand that there are lots of dirty secrets surrounding these "reports." For example, if your company PAYS for industry coverage by Gartner, generally speaking your company gets an opportunity to present ITS side of the argument in a way that it normally would not, just because of its ACCESS to the analyst(s) writing the report. Gartner will deny this to their last breath, but it's true.
Another dirty little Gartner secret is the quality of the "analysts." I happen to know a Gartner analyst personally, who has been quoted many times in the mainstream press making pronouncements about Micros~1 and others. This person's background? Liberal arts. Has absolutely no clue -- and I should know, I've had to answer a lot of this person's stupid questions over the years.
So the bottom line for Gartner is that they need to rotate out the fuddy-duddies from the 70's, whose last remembrance of Unix is v6 in college on a PDP-11/70. Then they need to ask some important questions about why the majority of competent technoids is so damn excited about Linux and the Open Source movement. Simultaneously, they need to recruit (and pay!) technical people who actually know that they're talking about. For starters, how about people who have hands-on Linux experience and who can explain why we can accomplish
- about 100x more,
- about 100x faster,
- with greater stability,
- in certain problem domains,
with Linux than with Microsoft.
As long as Gartner fails to understand WHY there is a nerd revolution, and WHY Linux is so exciting to the people like us who, increasingly, actually control the technical infrastructure of the world, they will totally miss the point. Their liberal-arts analysts can only see the world through their Microsoft laptops (which constitute, in many cases, the only computer they've ever actually used).
The OS is pretty neat, but it could use some open-source thrashing instead of just a few (admittedly smart) guys in Ottawa hacking with it. We experienced a reasonable number of "Blue Screens of Death" (yes, QNX kernel panics bear a marked resemblance to the NT dies we are all familiar with). Probably many of those have been fixed over time, but equally probably new ones have come along. Exposing the source code to thousands of eyes would be productive, I suspect.
We also had some problems with the file system (corrupted files). At that time basically only one guy (Bill, a really smart fellow) was working on the file system, and unfortunately you had to hit him in the (proverbial) nuts with a baseball bat to get his attention. To QNX's (and Bill's) credit, they did fix the problem. However, this illustrates a big disadvantage that closed-source vendors have -- namely, in many cases only one dude can do the work, whereas open-sourcers can call on help (and patches) from around the world.
The OS does fit in a remarkably small memory footprint, and its messaging system is very fast, even between machines. It was a good and stable framework on which to build a meta-messaging applications layer. Our customers were astounded at the up-time of their QNX-based systems, as opposed to the Micros~1 competition.
QNX suffers from the same problem as all proprietary operating systems -- you have to call them when you have a problem, and their attention to you seems to depend on how much of their business you represent. In 1995, it appeared that they were focusing heavily on the embedded market (set-top thingies, etc.), and were more-or-less in standby mode on the general-purpose OS side. I haven't seen anything to indicate a change in direction in the last few years, although Photon appears to be real now. Also, I should admit that I haven't been watching that closely.
As far as QNX pricing is concerned, it was very reasonable for OEM quantities. I would encourage anyone to contact them to get a price quote. They seem to be willing to structure a deal, unlike some larger companies we know.
Last time I installed a game ("Legacy of Time"), a bunch of stuff stopped working on my machine, no doubt due to some DLL incompatibility introduced by the install procedure. Fortunately, I always have a complete copy of windows\system and of the Registry squirreled away. So I just copied them back, and voila, everything started working again.
The game is still on the shelf. Want it? I *never* install anything on this pig without a complete copy of everything journalled aside. That's the *real* Micros~1 install procedure, as far as I'm concerned.
I guess maybe it's OK to close your eyes and pray that the next install won't gronk your machine if all you do is play games on it. Me, I gotta keep this Win95 pig running, so I can eat next week, because I have to maintain code for my clients.
Which means I don't install nothin' on it, baby. Nothin'.
Good point. Let's turn it around, then: the icon still means Open Source Friendly, but we maintain a "good guys" list only. Any ideas how to promote this widely, other than the usual newsgroups/mailing lists?
This is a terrific idea. The basic problem, as Scott points out, is that there could be serious abuse of the blacklist. Most companies have no chance of controlling all of their pointy-haired middle managers, so the company could end up blacklisted just because of one person, even though their "official" policy might be reasonable.
So let's give companies a really easy way to get off the blacklist: they can put an "Open Source Friendly" icon on their web page. I'm imagining something like a penguin holding hands with or hugging a baby gnu (the BSD daemon won't go over with the Christian Right, sorry), with a caption reading "Open Source Friendly." The icon links to *OUR* page, which explains the pledge that the icon represents.
I think that this will work wonders. Recruiters are extremely sensitive to issues like this, and "Open Source Friendly" (and the icon) will start appearing on recruitment advertisements very quickly. Hiring managers will quickly learn that if they're not "Open Source Friendly," their chances of filling positions will decrease dramatically.
I volunteer to draft the wording of the web page. Anybody want to host it? GIMP the icon?
- Eric Strovink
strovink@acm.org
This story is fascinating to me, and I say /. should follow it to the bitter end. It has all the elements of "Edge of Night" or "As The World Turns," or any of the old daytime soap operas. Like a soap, the story keeps morphing around, reminiscent of my boss trying to explain at review time why I don't make as much money as the idiot down the hall.
There are elements of heartbreak: the happy story of the QNX engagement, followed by Jim Collas' devious Victorian trick of disappearing out the opposite door of the four-wheeler on the way to the church, leaving our Canadian friends holding their Canadian private parts at the altar.
There are elements of humor: the vain attempt to annoint my toaster oven's Internet interface as, somehow, "Amigan." Amiga in the fridge, Amiga in the microwave -- gimme that bouncin' ball!
There is betrayal: the sudden departure of once-friendly and communicative executives, to be replaced by mindless drones of a mechanized society, spouting biz-speak from both sides of their mouths.
And there is overall "Dallas"-style incompetence, as the idiot children of a once-great empire recklessly spend all the carefully-built-up good will of the legions of loyal fans who cut their teeth on these wonderful machines.
I say, bring on the next chapter! More! MORE!
I've looked hard at porting applications to Linux from Microsoft platforms, mostly VB apps. It's a bitch. There's no VB equivalent for Linux (I know there's some activity on this front), and I'm not aware of an easy porting path (if anyone knows one, comment, please).
There is no question that Linux is a more stable platform than any Microsoft platform. We've been installing Linux machines for our clients whenever we can -- Linux is much easier to work with. However, for GM to convert all those machines over to Linux doesn't make much sense to me. Not only do all of the VB apps have to be rewritten, but also most of the third-party software packages that this guy *thinks* he'd like to add are probably written for Windows (WINE advocates: I can't see my local GM dealer putting up with an "80% working" application under WINE, can you?).
Finally, are this guy's motivations (running out of disk space) silly, or is it just me? He could solve his problem by spending $200 at each store to upgrade to a larger hard drive and maybe to install more memory (DriveCopy does a very nice job of cloning hard drives, small --> large). It would probably cost him another $200 per store to hire a nationwide third-party service organization to visit all his dealerships and perform the upgrade. But at the end of it, he'd have what he needs: working systems with plenty of elbow room to grow.
(Yes, he could take the $400 and probably buy a better PC than he has right now. But I'm assuming he'd rather not IPL them all, install software on them all, burn them in, and so on, which will cost at least another $400 per machine to stage, not including shipping and installation at the dealer.)
If he converts to Linux, he has a huge problem on his hand: not only does he need to rewrite all his applications and do all that expensive staging above, but then at the end of the whole process (assuming it all works) he'll have to explain on a daily basis to his bonehead management how come the XYZ Inventory Control System doesn't run on his machines (his boss spots the XYZ ad in Useless American Cars Weekly, for example).
Ouch.
I have a friend who works at Liberty Mutual. According to him, if you have a PC on your desk, and you install any software (repeat: ANY SOFTWARE) on it whatsoever, and you are caught, this is grounds for dismissal. Every PC's configuration comes from a "gold" CDROM that is maintained by IS. If a PC is found to have been "corrupted" by foreign software, it is immediately reloaded from scratch from the "gold" CD.
Some folks have been quite impassioned in this forum about the freedom and productivity that results from having one's own PC. No doubt there are many desktops at Liberty Mutual that are running "verboten" software, and no doubt there are copies of the "gold" CD floating around that are used to reload PC's that have "gone south," and no doubt all of this activity is occurring without anyone in IS knowing about it.
But from the *company's* standpoint, their policy is working. They don't have to run around supporting PC users (because everyone is afraid of getting caught, so all everyone ever does is reload from the "gold" CD instead of calling for help). The company believes that everything is wonderful. And the IS department believes they are in control of this wonderful imaginary world.
All of which is not unfunny. But here's the rub: IS is in charge of all the procurement decisions. And if IS is the customer, then the SunRay sure sounds like a terrific idea, doesn't it?
I guess I can relate this to my own experience as VP Engineering for a small technology company. One day a whole metric f**ckload of low-end Compaq PC's showed up in Marketing, Sales, and Administration. Some Compaq sales dude had sneaked in the back door and sold a bill of goods to guys who had absolutely no clue what they were buying (my personal theory as to why Compaq sales are off -- technical guys are pissed off at them for selling directly to the suits, and so as our power increases, we buy elsewhere for revenge). Well, those same dudes are buying SunRays.
Which begs the question, of course, as to whether SunRays are good or bad. I can see arguments both ways, most of which have already been made by others.
In my experience, a troll message follows these general rules:
1) It is a relatively short message.
2) It contains poorly-spelled words or odd (or no) capitalization.
2a) It contains multiple exclamation marks or question marks.
3) It contains obscene language or four-letter words.
4) It generally does *not* contain a URL.
5) It often (but not always) comes from ACs.
Based on these criteria, it seems to me that an automatic troll filter could be built fairly easily that automatically deprecates messages that fall into these categories. Perhaps the spell-checking would be awkward from a CPU time perspective, but the other stuff should be straightforward.
Ahhhhhhh, that's better.
No, *you* don't get it. Friends are allowed to help friends become less nerdy and more interesting. The quality of stories and articles on /. is high; the editorial judgement is good; the comments are often amusing; and therefore Rob and Jeff are clearly capable of high quality effort. We should urge them to apply those same high standards to their broadcasts.
2) Lose the third guy. It's way too confusing to listen to.
3) If you have nothing to say, say nothing. Does the "broadcast" *have* to be 13 minutes?
4) Take Hemos' head out of the water jug it sounds like he's talking from (seriously, what the $#&^% is wrong with the guy's mike??).
5) Stick to the news. I don't care what you guys did at the Linux conference, or how many t-shirts you got, or how much pinball you played, or whatever. I DON'T CARE. OK?
5a) OK, if someone got laid, I wouldn't mind hearing about that. But that's unlikely, considering what you guys sound like.
6) I also don't care about what imaginary spot you're broadcasting from. Just tell Hemos to stop broadcasting from the water jug.
7) Stop giggling.