One of the problems faced by small software development houses in the scientific field is acquiring enough sample data that allows for meaningful testing of algorithms. Since most geophisical data is considered intellectual property of exploration companies it is closely guarded and hard to obtain unless you're already established in the market and have good reputation. Artificially generated data just doesn't have the same qualities real life samples have. I have seen the same algorithms work perfect on artificial SEG-Y data that crumbled when exposed to some real life heavily faulted seismic.
Re:Don't go too far in the other direction
on
The Venture Cafe
·
· Score: 2
I have to say that Ellison and McNealy are visionaries to some extent. They are however very practical 'visionaries' who simply improve upon existing systems instead of venturing into new areas they have no knowledge of. Bill Gates is the master of the 'copy and improve' business model. It's difficult to actually brand him a 'visionary' but his pragmatic approach to computing certainly paid off.
Ellison and McNealy tend to free-wheel a bit more than your average CEO but then again their companies have experienced relatively more up and down cycles. Obviously in case of Ellison it paid off handsomely.
I guess we're saying the same thing. It's ok to have a goal that you strive towards. At the same time it's even more important to be flexible and realise that sometimes you must adjust your ideas for the real world and often dramatically revise your initial plans. Obviously you might go extreme in this direction and catch a corporate attention deficit disorder (vide. Michael Cowpland).
I'm thinking about starting my own company (and a high tech one at that) and been thinking a lot about all these issues. But I just don't think this particular book has the answers I'm looking for or even enough quality content to justify spending my time reading it. And this conclusion I drew from reading the entire sample chapter.
Read the sample chapter - didn't like it
on
The Venture Cafe
·
· Score: 2
I read the sample chapter on the website and I will not be reading the rest of the book. Bold statements such as this:
"The ideal entrepreneur is not the kind of person that you'd want as a personal friend," Gill tells me. "The phrase we use in England is throwing Teddy out of the pram.' If they don't get their way, they get very upset. Without realizing it, they tend to be manipulative.
"Entrepreneurs have to be completely driven by vision, such that they only see what they want to see. Sometimes businesses go off the rails because they have a CEO who can't see some of the warning signs, but that's why there needs to be a team of at least two. You need the Genghis Khan' CEO and the safe pair of hands' CFO."
prove that those guys are the same dotcom crowd that ruined the tech industry by implementing their silly 'visions'. The very last thing I want is to work for another frigging 'visionary' at my company. All those 'visionary' CEOs were basically pipe dreamers whose ideas were either unimplementable or impractical to the point of being absurd. We all remember the 'visionaries' behind Petsmart, Dr Koop, Boo, March 1st, Viant and so on. They were so blinded by their own ideas that they wouldn't listen to their own (and others') common sense.
Stupid CEOs who think up those 'grand' ideas are the ones to blame for the dotcom crash. Take the example from this book. The paperless train ticket system... Nifty idea with a lot of ridiculous barriers that make it impractical. Most trains tickets are purchased at the station with no security gates (yet) no baggage check in and the whole hoopla that goes with air travel. For that purpose the paper ticket is the natural obvious and efficient solution. But the enterpreneur in the book obviously doesn't see it that way. He's already obsessing about his vision without giving it some clear headed thought.
I see a lot of people replying that there is no market for a videopohone. Video phones make sense but only in a limited number of scenarios.
While it's true that most people don't want to be seen when talking on the phone. It only makes sense for folks like me who want to be seen and see their families who are 5000 miles away. I am in the market for a long distance videophone. Webcams do the trick but the quality is definitely not there yet.
There is a market for videphones but it's not for short distance/local calls. You probably want to videoconference with someone you haven't seen for three years but video-talking to someone you saw last night really doesn't make much sense.
We'll never see another "Star Wars," no matter how much we want to. And we want to very much.
But like the cherished passions of first love, the fervor called forth by the landmark film is never coming back, and no amount of prequels or sequels is going to change that. Paradoxically, the fact that the latest prequel, "Star Wars: Episode II Attack of the Clones," is a bit better than its predecessor makes it clear how lacking in the things that matter these newcomers are.
Given its huffy 9-year-old protagonist and off-putting characters like Jar Jar Binks and Watto the junk dealer, "Episode I The Phantom Menace" was anything but a tough act to follow. Picking up the adventures of Anakin Skywalker 10 years later, "Clones" (which opens Thursday) has more menace and less Jar Jar, better battles and an impressive parade of eye-catching splendors. But like the Tin Man, "The Wizard of Oz's" C-3PO predecessor, it doesn't have much of a heart. Writer-director George Lucas' gift for animating the inanimate turns out to be paralleled by a tendency to deaden what should be completely alive.
As with "Phantom Menace," it is the pictorial element of "Clones" that makes the biggest impact. Production designer Gavin Bocquet, aided by four visual effects supervisors, three concept design supervisors, an animation director and a previsualization and effects co-supervisor (no, I don't know what that is either), has created some truly involving alternative universes, and costume designer Trisha Biggar has figured out what should be worn in each of them.
Some of the film's action is also well-done, especially a thrilling flying chase through the dizzying nighttime urban caverns of Coruscant, the "Blade Runner"-influenced capital city. But except for a climactic appearance by the venerable Yoda, whose computer-generated lightsaber skills got him on the cover of Time under a "Yoda Strikes Back!" headline, creating emotion is beyond this film's powers.
One reason is a script that feels, well, cloned, something Lucas and co-writer Jonathan Hales (TV's "Young Indiana Jones," story credit on "The Mummy Returns") threw together in their spare time. The plot is standard, and the dialogue, even for something intended for young people, is curiously flat. It ranges from the pious ("The day we stop believing democracy can work is the day we lose it") to the predictive ("Why do I get the feeling you're going to be the death of me," Obi-Wan Kenobi jokes to Anakin) to the pathetic, as when Anakin grumbles about Padmé Amidala, "I've thought about her every day since we parted--and she's forgotten me completely."
These stiff lines are matched by line readings so uniformly impassive that even such lively performers as Ewan McGregor (Obi-Wan) and Natalie Portman (Padmé) can't animate them. Only the veteran Christopher Lee, with experience of doing things on his own during his long career, gives a worthwhile performance as the villainous Count Dooku. For what Lucas gets out of his cast, the actors might as well be digital too, as is the rest of the film.
This dramatic stolidity underlines yet again how fortunate Lucas--and the world--was in the Harrison Ford-Carrie Fisher-Mark Hamill troika that animated the original "Star Wars." Ford especially brought the kind of wickedly nonchalant sense of humor to the proceedings that has gone missing this time around.
To be fair to the current "Clones" team, there's perhaps something more at work here. When that first film was being made, it meant less than zero to say you were part of "Star Wars"; the eyes of the world were not on the production, to say the least.
Now, everything has been close to sanctified, and those currently involved seem weighted down by the knowledge that they're part of a phenomenon. There's an unshakable self-consciousness about "Clones" that does not work to its advantage.
Still, the picture does start promisingly, with Senator (and former Queen) Amidala coming to Coruscant to try and preserve the Republic against a secessionist movement. She's quickly the target of multiple assassination plots, and the Jedi knight Obi-Wan and his Padawan learner-apprentice Anakin are called in to protect her.
Judging by his performance here (perhaps not a wise thing to do), young Canadian actor Hayden Christensen was picked for Anakin strictly on his ability to radiate sullen teen rebellion, something he does a lot. Anakin chafes like a grounded adolescent at the restrictions Obi-Wan places on him, grousing that the master is "overly critical. He never listens. He just doesn't understand. It's not fair."
This High School Confidential in Outer Space tone is continued in the forbidden romance (Jedis aren't allowed to fall in love) that develops between Anakin and the senator. As the young people hide from danger in an elegant Naboo retreat, they're burdened by a formidable lack of chemistry. (Where are Tobey Maguire and Kirsten Dunst when we really need them?) And they're saddled with dialogue that might have been ransacked from old Harlequin novels: "I'm haunted by the kiss you never should have given me."
Everything inevitably ends in a climactic battle, where the senator gets to fight bad guys while showing off a Britney Spears-like bare midriff. Impressive though the computer work is, it soon descends into video game overkill. Only a teenage boy could find this kind of stuff continually diverting, and only a teenage boy would not notice flimsy emotions and underdeveloped acting. It seems George Lucas, like Peter Pan, has never really grown up.
I thought most of them would. A 14" LCD is no bigger than a single A4 page. Then there usually is a fair amount of space around the edges in a typical overhead projector when you put your slide on it. I think it would fit...
Interesting. I still think my/SGI approach having more positive points than negative. Yes you will have to be more careful with it compared to a traditional projector but then you're saving at least ~$3000USD so some tradeoff has to be incurred. The diminshed lumination bit is something that bugs me but new overhead projectors have 2800 lumens illumination and more. I would have thought that should be enough to show crisp picture even through an LCD. Having said that I haven't tried building it yet so I simply don't know.
What you gain with my idea is high resolution. For ~800USD you should be able to get a good overhead projector and a quality 15" LCD that will give you 1024x768. Even many high end LCD projectors can only do 800x600 which makes them completely unusable as a replacement for a monitor.
...And would keep many Americans in (better) shape, which means fewer pounds commuting daily, which means less gas burned by American SUVs, which means less pollution! You're right, the benefits just keep on piling up.
Nonsense! Solitaire is mostly played by managers who don't really add any meaningful value anyway. Geeks are those who get the actual job done. The amount of productivity loss that slashdot has cost the economy is infinitely higher than that caused by Solitaire.
When documenting a software project keep one important principle in mind:
Write things Once And Only Once
This means that whatever documentation you produce it should not duplicate information that can be inferred from reading the code. Documentation should only compliment your source. Describing algorithms in a Word document is a terrible idea as any duplicate information will get out of date. It's only a matter of time. Usually documentation should be more detailed around the user requirements and sparse around how the code works. Documentation is not an asset. It's a liability. Any new document introduced during the development process is an additional maintenance headache and a delay in project's completion. Always think twice before adding every single type of document into your development process.
With apologies to the people of Bathurst, NB. It must be said we aren't ALL miners, mill workers and "tax leeches". Though I doubt many Bathurstians are/.ers...
Well how about a Saint Johnner, eh?
Talk about dull towns in the maritimes... Shit even Fredericton has more entertainment qualities than Saint John. And here you either work for the Irvings or you don't work at all. Sad but true. Can't beat the cost of living though.
Someone had to do it, sooner or later
on
ROX Desktop Update
·
· Score: 2
And now we have it. A brand new pr0n browsing desktop for Linux. Just like it always did, pr0n inspires yet another computer innovation.
I think this is the crucial moment for RMS where he either becomes more flexible or risk alienating the remaining few developers who still rally around him and his ideas. I mean for crying out loud the Ximian team is not even suggesting using non free code. They just want to base the next version on a spec that also has a nonfree implementation. There is no reason why there needs to be even a single line of non-free code in the Ximian implementation of dotNET.
You're only partially right. To say that Amazon is not a high tech firm is not quite fair. Amazon is the largest e-retailer and their traffic volume figures are quite staggering. I can't remember exactly but I believe that amazon websites handle more than 60 million unique sessions a month in the USA alone. That is a huge volume of traffic that few other enterprise applications have managed to achieve. It takes a lot of high tech know how in the company to build an application that scales to that kind of traffic.
All too true. But what's even worse is that if this thing gains any kind of publicity it is just going to give Linux a bad name. Joe Six Pack will try it and will see that none of his Windows apps are working and will conclude that that Linux thing is not a serious OS. These people should be stopped before they cause too much harm for Linux.
It's okay to stand your ground Alan, but in this economy even a kernel God may find alternative employment hard to come by. Something to think about before handing that resignation letter.
I played with RtP for quite a while. I love it. I learned so much about OS design just by reading the RtP manuals and I think it has a hell of a potential especially on internet appliances and web tablets etc.
The beauty of QNX and RtP is the microkernel design (let the flamewars begin). The OS is exteremly resilient because the core kernel just acts as a messaging bus for all other services that run in the user space. For example, should your filesystem crash you can just restart it like any other user space process!. Alternatively if you don't need multitasking capabilities but memory and hardware are at premium you simply don't run proc and don't have to put up with the overhead of a process scheduler. QNX is such a clean design it puts other microkernels to shame.
Rock on QSSL.
It doesn't help to be pretty good with UML because UML is useless. UML just slows down most development efforts because it's too static for a typical software development shop. It extends the feedback cycle between design and implementation and hence slows down the development process. And on the top of that it doesn't add value to the product, just produces more artifacts that need to be kept up to date. UML is business oriented rather than developer oriented and it consistently failed to aid those who actually end up implementing the system with any valuable information.
Extreme Programming is a methodology that is competitive to the Rational Unified Process BUT unlike Rational folks doesn't advise you to purchase their $10,000 modeling package or a $50,000 training course. It's a down to earth system of how to plan, design and implement a system with maximum predictability and on budget. And all this without the almighty Big Design accompanied by a zillion useless diagrams.
Go figure who I'm inclined to suspect of trying to sell me snake oil instead of helping coordinate my development process.
Yes, plugins are great but what you're touting here has been done long time ago by your competitors: Borland. JBuilder has its api for plugin development called OpenTools. With OpenTools you can make JBuilder do anything. And when I say anything I mean it. Open Tools is basically your own java code called by JBuilder when the user performs an action defined in the OpenTool itself. For example I wrote multiple plugins that add themselves as menu items to the Project menu in JBuilder and run some code parsing and display results in the info pane.
I don't know what you mean when you talk about an engineering application but there are some significant C++ applications written for the geoscience industry. I worked for a small company that does a 3D structural restoration and C++ was just fine for the purpose. We had our own scene graph/utility library and a GUI that exposed the toolkit's functionality.
The reason for using OO was code maintainability and compartmentalization. It is generally easier to divide work up once you're able to break your application down to classes.
OO also means extendibility. With the right class hierarchy it was easy for us to add a new node to our scengraph to support a new type of data. For example, initially the app supported horizon data only, later on we added support for faults just by extending one of the base classes. OO will help you write any application if your class structure is sane.
Slashdot Mirror
One of the problems faced by small software development houses in the scientific field is acquiring enough sample data that allows for meaningful testing of algorithms. Since most geophisical data is considered intellectual property of exploration companies it is closely guarded and hard to obtain unless you're already established in the market and have good reputation. Artificially generated data just doesn't have the same qualities real life samples have. I have seen the same algorithms work perfect on artificial SEG-Y data that crumbled when exposed to some real life heavily faulted seismic.
Ellison and McNealy tend to free-wheel a bit more than your average CEO but then again their companies have experienced relatively more up and down cycles. Obviously in case of Ellison it paid off handsomely.
I guess we're saying the same thing. It's ok to have a goal that you strive towards. At the same time it's even more important to be flexible and realise that sometimes you must adjust your ideas for the real world and often dramatically revise your initial plans. Obviously you might go extreme in this direction and catch a corporate attention deficit disorder (vide. Michael Cowpland).
I'm thinking about starting my own company (and a high tech one at that) and been thinking a lot about all these issues. But I just don't think this particular book has the answers I'm looking for or even enough quality content to justify spending my time reading it. And this conclusion I drew from reading the entire sample chapter.
Stupid CEOs who think up those 'grand' ideas are the ones to blame for the dotcom crash. Take the example from this book. The paperless train ticket system... Nifty idea with a lot of ridiculous barriers that make it impractical. Most trains tickets are purchased at the station with no security gates (yet) no baggage check in and the whole hoopla that goes with air travel. For that purpose the paper ticket is the natural obvious and efficient solution. But the enterpreneur in the book obviously doesn't see it that way. He's already obsessing about his vision without giving it some clear headed thought.
While it's true that most people don't want to be seen when talking on the phone. It only makes sense for folks like me who want to be seen and see their families who are 5000 miles away. I am in the market for a long distance videophone. Webcams do the trick but the quality is definitely not there yet.
There is a market for videphones but it's not for short distance/local calls. You probably want to videoconference with someone you haven't seen for three years but video-talking to someone you saw last night really doesn't make much sense.
But like the cherished passions of first love, the fervor called forth by the landmark film is never coming back, and no amount of prequels or sequels is going to change that. Paradoxically, the fact that the latest prequel, "Star Wars: Episode II Attack of the Clones," is a bit better than its predecessor makes it clear how lacking in the things that matter these newcomers are.
Given its huffy 9-year-old protagonist and off-putting characters like Jar Jar Binks and Watto the junk dealer, "Episode I The Phantom Menace" was anything but a tough act to follow. Picking up the adventures of Anakin Skywalker 10 years later, "Clones" (which opens Thursday) has more menace and less Jar Jar, better battles and an impressive parade of eye-catching splendors. But like the Tin Man, "The Wizard of Oz's" C-3PO predecessor, it doesn't have much of a heart. Writer-director George Lucas' gift for animating the inanimate turns out to be paralleled by a tendency to deaden what should be completely alive.
As with "Phantom Menace," it is the pictorial element of "Clones" that makes the biggest impact. Production designer Gavin Bocquet, aided by four visual effects supervisors, three concept design supervisors, an animation director and a previsualization and effects co-supervisor (no, I don't know what that is either), has created some truly involving alternative universes, and costume designer Trisha Biggar has figured out what should be worn in each of them.
Some of the film's action is also well-done, especially a thrilling flying chase through the dizzying nighttime urban caverns of Coruscant, the "Blade Runner"-influenced capital city. But except for a climactic appearance by the venerable Yoda, whose computer-generated lightsaber skills got him on the cover of Time under a "Yoda Strikes Back!" headline, creating emotion is beyond this film's powers.
One reason is a script that feels, well, cloned, something Lucas and co-writer Jonathan Hales (TV's "Young Indiana Jones," story credit on "The Mummy Returns") threw together in their spare time. The plot is standard, and the dialogue, even for something intended for young people, is curiously flat. It ranges from the pious ("The day we stop believing democracy can work is the day we lose it") to the predictive ("Why do I get the feeling you're going to be the death of me," Obi-Wan Kenobi jokes to Anakin) to the pathetic, as when Anakin grumbles about Padmé Amidala, "I've thought about her every day since we parted--and she's forgotten me completely."
These stiff lines are matched by line readings so uniformly impassive that even such lively performers as Ewan McGregor (Obi-Wan) and Natalie Portman (Padmé) can't animate them. Only the veteran Christopher Lee, with experience of doing things on his own during his long career, gives a worthwhile performance as the villainous Count Dooku. For what Lucas gets out of his cast, the actors might as well be digital too, as is the rest of the film.
This dramatic stolidity underlines yet again how fortunate Lucas--and the world--was in the Harrison Ford-Carrie Fisher-Mark Hamill troika that animated the original "Star Wars." Ford especially brought the kind of wickedly nonchalant sense of humor to the proceedings that has gone missing this time around.
To be fair to the current "Clones" team, there's perhaps something more at work here. When that first film was being made, it meant less than zero to say you were part of "Star Wars"; the eyes of the world were not on the production, to say the least.
Now, everything has been close to sanctified, and those currently involved seem weighted down by the knowledge that they're part of a phenomenon. There's an unshakable self-consciousness about "Clones" that does not work to its advantage.
Still, the picture does start promisingly, with Senator (and former Queen) Amidala coming to Coruscant to try and preserve the Republic against a secessionist movement. She's quickly the target of multiple assassination plots, and the Jedi knight Obi-Wan and his Padawan learner-apprentice Anakin are called in to protect her.
Judging by his performance here (perhaps not a wise thing to do), young Canadian actor Hayden Christensen was picked for Anakin strictly on his ability to radiate sullen teen rebellion, something he does a lot. Anakin chafes like a grounded adolescent at the restrictions Obi-Wan places on him, grousing that the master is "overly critical. He never listens. He just doesn't understand. It's not fair."
This High School Confidential in Outer Space tone is continued in the forbidden romance (Jedis aren't allowed to fall in love) that develops between Anakin and the senator. As the young people hide from danger in an elegant Naboo retreat, they're burdened by a formidable lack of chemistry. (Where are Tobey Maguire and Kirsten Dunst when we really need them?) And they're saddled with dialogue that might have been ransacked from old Harlequin novels: "I'm haunted by the kiss you never should have given me."
Everything inevitably ends in a climactic battle, where the senator gets to fight bad guys while showing off a Britney Spears-like bare midriff. Impressive though the computer work is, it soon descends into video game overkill. Only a teenage boy could find this kind of stuff continually diverting, and only a teenage boy would not notice flimsy emotions and underdeveloped acting. It seems George Lucas, like Peter Pan, has never really grown up.
Kenneth Turan
I thought most of them would. A 14" LCD is no bigger than a single A4 page. Then there usually is a fair amount of space around the edges in a typical overhead projector when you put your slide on it. I think it would fit...
What you gain with my idea is high resolution. For ~800USD you should be able to get a good overhead projector and a quality 15" LCD that will give you 1024x768. Even many high end LCD projectors can only do 800x600 which makes them completely unusable as a replacement for a monitor.
- Buy a typical overhead slide projector (the one your college has hundreds of)
- Get a flat panel LCD @ 14-15"
- remove the backlight from your flat panel
- slap the panel stripped of the backlight on top of your overhead projector
- Heh. There is no step 5!
Am I missing something or is this a better approach to the problem?...And would keep many Americans in (better) shape, which means fewer pounds commuting daily, which means less gas burned by American SUVs, which means less pollution! You're right, the benefits just keep on piling up.
Nonsense! Solitaire is mostly played by managers who don't really add any meaningful value anyway. Geeks are those who get the actual job done. The amount of productivity loss that slashdot has cost the economy is infinitely higher than that caused by Solitaire.
We all know that CIPA means 'cunt' in Polish, okay?
Write things Once And Only Once
This means that whatever documentation you produce it should not duplicate information that can be inferred from reading the code. Documentation should only compliment your source. Describing algorithms in a Word document is a terrible idea as any duplicate information will get out of date. It's only a matter of time. Usually documentation should be more detailed around the user requirements and sparse around how the code works. Documentation is not an asset. It's a liability. Any new document introduced during the development process is an additional maintenance headache and a delay in project's completion. Always think twice before adding every single type of document into your development process.
Well how about a Saint Johnner, eh?
Talk about dull towns in the maritimes... Shit even Fredericton has more entertainment qualities than Saint John. And here you either work for the Irvings or you don't work at all. Sad but true. Can't beat the cost of living though.
And now we have it. A brand new pr0n browsing desktop for Linux. Just like it always did, pr0n inspires yet another computer innovation.
Oh, and congratulations to the pair of ya from MSBob
The man is getting old and it shows.
Hope this helps.
You're only partially right. To say that Amazon is not a high tech firm is not quite fair. Amazon is the largest e-retailer and their traffic volume figures are quite staggering. I can't remember exactly but I believe that amazon websites handle more than 60 million unique sessions a month in the USA alone. That is a huge volume of traffic that few other enterprise applications have managed to achieve. It takes a lot of high tech know how in the company to build an application that scales to that kind of traffic.
All too true. But what's even worse is that if this thing gains any kind of publicity it is just going to give Linux a bad name. Joe Six Pack will try it and will see that none of his Windows apps are working and will conclude that that Linux thing is not a serious OS. These people should be stopped before they cause too much harm for Linux.
Hope this helps,
MSBob
The beauty of QNX and RtP is the microkernel design (let the flamewars begin). The OS is exteremly resilient because the core kernel just acts as a messaging bus for all other services that run in the user space. For example, should your filesystem crash you can just restart it like any other user space process!. Alternatively if you don't need multitasking capabilities but memory and hardware are at premium you simply don't run proc and don't have to put up with the overhead of a process scheduler. QNX is such a clean design it puts other microkernels to shame. Rock on QSSL.
Extreme Programming is a methodology that is competitive to the Rational Unified Process BUT unlike Rational folks doesn't advise you to purchase their $10,000 modeling package or a $50,000 training course. It's a down to earth system of how to plan, design and implement a system with maximum predictability and on budget. And all this without the almighty Big Design accompanied by a zillion useless diagrams.
Go figure who I'm inclined to suspect of trying to sell me snake oil instead of helping coordinate my development process.
Just some decent advice from a fellow programmer.
Yes, plugins are great but what you're touting here has been done long time ago by your competitors: Borland. JBuilder has its api for plugin development called OpenTools. With OpenTools you can make JBuilder do anything. And when I say anything I mean it. Open Tools is basically your own java code called by JBuilder when the user performs an action defined in the OpenTool itself. For example I wrote multiple plugins that add themselves as menu items to the Project menu in JBuilder and run some code parsing and display results in the info pane.
The reason for using OO was code maintainability and compartmentalization. It is generally easier to divide work up once you're able to break your application down to classes.
OO also means extendibility. With the right class hierarchy it was easy for us to add a new node to our scengraph to support a new type of data. For example, initially the app supported horizon data only, later on we added support for faults just by extending one of the base classes. OO will help you write any application if your class structure is sane.