You can have the best of both worlds with symbolic links. A tool like Pkglink can do most of the heavy lifting and also give you the ability to have multiple versions of the software installed.
Of course, that presumes that you have a decent set of functional tests that cover the code well in order to get a good idea of what is going on. Of course, the permutations involved on a moderate to large code base could be far more than you could keep in your head...
Views are meant to isolate applications from the physical representation of the data (ie. the tables). Their flexibility is in their ability to give a (typically read-only) application a logical view of the database that doesn't fit third normal form.
Maybe its time to move on to the really next generation of Star Trek -- Time Trek! They've already laid the groundwork for this on Voyager. It would be the way to do an American Dr. Who without really stepping on Dr. Who.
Once the (bulk of) Solaris is GPLed, it will be absorbed into Linux (and *BSD?) and that will be that. At that point, Sun can continue to support it until they ramp up something else (Plan9, Hurd, ?), but leave the bulk of support to the community. It's a winning situation for Sun in that:
* they can move to linux * they can downsize Solaris * they can downsize their support staff
The writing is on the wall -- Solaris is being put to bed.
Several problems with this theory. OSS means that Microsoft can simply copy your software (ie. embrace) and incorporate it into their environment (ie. extend). The lawyers cannot do a thing about this.:(
The future of transportation is improved mass transit, not flying cars.
Actually, I could see the Moller SkyCar (with VTOL capability) being the backbone of a "taxi" service in many city locations. Using "vertiparks" (I think that's the term Moller uses), commuters could live in the suburbs, commute to the vertipark, hop a skytaxi into the city, and use a Segway to get to work. At each step, you have individualized transportation, so do not have to spend time waiting for intermediate commuters to get on and off the "mass transit". The skytaxi service could (be programmed to) follow well defined skylanes that might not be more than a few (hundred) feet off the ground -- the new parachute ejection systems would ensure that problems could be handled.
1. Let the computer do it! Artificial Intelligence is coming. 2. If flying cars become viable, the sky won't be "your backyard" anymore. 3. Depends on how far you roll... 4. Yeah, but so is the automobile when compared to the bicycle. 5. Use a bigger fly swatter.
Huh? The Tucker Torpedo has been around since 1947. Unfortunately, the automobile was ahead of it's time and the Big 3 (GM, Ford, Chrysler) ran Tucker out of business before it could really get going.
How do you know that "here open standards/open source have a far better track record" in security? Could it be that the track record is due to number of systems available to attack more than the type of systems they are? That is, perhaps M$ has had problems in security because they are popular more than because they are closed source...
I remember Bill Joy giving a presentation at a place I won't name some 20 years ago wherein he talked about about innovations to come in the computer industry (like "mega-cubed" [1M x 1Mhz x 1Mpixel]) and how Sun would conquer all. He was also talking about how his budget for R&D was bigger than the net worth of Microsoft (at that time), so he really didn't feel that Bill Gates and Microsoft would be much of a problem in the computer market of the future. I guess Bill Gates gets the last laugh...:-(
A complete coding standard is difficult to come up with in one shot. Also, developers generally have their "de-facto" standards that they like to follow, so getting buy-in can be difficult. Perhaps it's easier to just allow the coding standard to evolve as needed by using a collaboration facility to come up with it.
Oh, really? Then why is it that it's only the worker's jobs that get offshored? American companies could save many millions of dollars per year by offshoring management jobs, but that never happens.
What level of management jobs are you talking about? Low-level managers generally (but not always) work where their workers work, so, if their workers are offshore, they usually are as well. High-level managers generally work where the executives in the company work as they've got to keep the executives appeased about the state of the company. Executives are generally big stakeholders in the company and, therefore, get to work wherever they want.
You don't know enough people to make a dent. The costs of SPAM (like bulk mail) are so low that one hit in thousands (or even millions) would probably justify the expense.
The problem with filtering is that, in essense, you've already paid for the email. Think about it, the email has already been received by your system, so you've paid the cost of the spammers email.
Using a "token" or "tag" based system means that spammers would have to request a token from you before they could send you an email. Their costs now go up (as does yours) as they have to filter your email (and the thousands of others replies to token requests) to get the token to send you. In the long run, companies could provide the service of token processing so that your personal email would only be accessible to white-listed addresses (thus moving your costs to the token-processing server).
But you only get to look under the hood and see the rust on the engine after you buy the car. Or, in this case, you only get to see the source after you buy the binary.
Slashdot Mirror
You can have the best of both worlds with symbolic links. A tool like Pkglink can do most of the heavy lifting and also give you the ability to have multiple versions of the software installed.
You can emulate both approaches with symbolic links (or even hard links). See Pkglink.
Of course, that presumes that you have a decent set of functional tests that cover the code well in order to get a good idea of what is going on. Of course, the permutations involved on a moderate to large code base could be far more than you could keep in your head...
Why should we be tethered to the ground? Let's move into the sky with the Moller Sky Car!
Views are meant to isolate applications from the physical representation of the data (ie. the tables). Their flexibility is in their ability to give a (typically read-only) application a logical view of the database that doesn't fit third normal form.
Maybe its time to move on to the really next generation of Star Trek -- Time Trek! They've already laid the groundwork for this on Voyager. It would be the way to do an American Dr. Who without really stepping on Dr. Who.
To kill Solaris -- why else?
Once the (bulk of) Solaris is GPLed, it will be absorbed into Linux (and *BSD?) and that will be that. At that point, Sun can continue to support it until they ramp up something else (Plan9, Hurd, ?), but leave the bulk of support to the community. It's a winning situation for Sun in that:
* they can move to linux
* they can downsize Solaris
* they can downsize their support staff
The writing is on the wall -- Solaris is being put to bed.
Several problems with this theory. OSS means that Microsoft can simply copy your software (ie. embrace) and incorporate it into their environment (ie. extend). The lawyers cannot do a thing about this. :(
What are you talking about? They all look spelled correctly to me...
Although there may have been problems in the *BSD community, could the core problem be the open source model?
Whatever happened to Project Stargate to put Usenet on a satellite channel?
Sure. Just label it a "deflection device" and nobody will be the wiser... :-)
Actually, I could see the Moller SkyCar (with VTOL capability) being the backbone of a "taxi" service in many city locations. Using "vertiparks" (I think that's the term Moller uses), commuters could live in the suburbs, commute to the vertipark, hop a skytaxi into the city, and use a Segway to get to work. At each step, you have individualized transportation, so do not have to spend time waiting for intermediate commuters to get on and off the "mass transit". The skytaxi service could (be programmed to) follow well defined skylanes that might not be more than a few (hundred) feet off the ground -- the new parachute ejection systems would ensure that problems could be handled.
The only one that's gotten "off the ground" is the Moller SkyCar and he has done most of the funding himself.
1. Let the computer do it! Artificial Intelligence is coming.
2. If flying cars become viable, the sky won't be "your backyard" anymore.
3. Depends on how far you roll...
4. Yeah, but so is the automobile when compared to the bicycle.
5. Use a bigger fly swatter.
This Taero 4000 reminds me of the Sokol A400 Flying Car. I haven't looked deeply, but I'll bet some of the same people are involved.
It's ashame that the Solotrek died so fast (see more about it here).
Huh? The Tucker Torpedo has been around since 1947. Unfortunately, the automobile was ahead of it's time and the Big 3 (GM, Ford, Chrysler) ran Tucker out of business before it could really get going.
How do you know that "here open standards/open source have a far better track record" in security? Could it be that the track record is due to number of systems available to attack more than the type of systems they are? That is, perhaps M$ has had problems in security because they are popular more than because they are closed source...
I remember Bill Joy giving a presentation at a place I won't name some 20 years ago wherein he talked about about innovations to come in the computer industry (like "mega-cubed" [1M x 1Mhz x 1Mpixel]) and how Sun would conquer all. He was also talking about how his budget for R&D was bigger than the net worth of Microsoft (at that time), so he really didn't feel that Bill Gates and Microsoft would be much of a problem in the computer market of the future. I guess Bill Gates gets the last laugh... :-(
A complete coding standard is difficult to come up with in one shot. Also, developers generally have their "de-facto" standards that they like to follow, so getting buy-in can be difficult. Perhaps it's easier to just allow the coding standard to evolve as needed by using a collaboration facility to come up with it.
Take a look at Twiki.
What level of management jobs are you talking about? Low-level managers generally (but not always) work where their workers work, so, if their workers are offshore, they usually are as well. High-level managers generally work where the executives in the company work as they've got to keep the executives appeased about the state of the company. Executives are generally big stakeholders in the company and, therefore, get to work wherever they want.
You don't know enough people to make a dent. The costs of SPAM (like bulk mail) are so low that one hit in thousands (or even millions) would probably justify the expense.
The problem with filtering is that, in essense, you've already paid for the email. Think about it, the email has already been received by your system, so you've paid the cost of the spammers email.
Using a "token" or "tag" based system means that spammers would have to request a token from you before they could send you an email. Their costs now go up (as does yours) as they have to filter your email (and the thousands of others replies to token requests) to get the token to send you. In the long run, companies could provide the service of token processing so that your personal email would only be accessible to white-listed addresses (thus moving your costs to the token-processing server).
See TMDA as one approach to this problem.
See the TMDA website.
But you only get to look under the hood and see the rust on the engine after you buy the car. Or, in this case, you only get to see the source after you buy the binary.