They're trying to somehow justify why they silently rolled out http.sys, a rebranded IIS, as a kernel driver with Win XP SP2. With that on the machine, each and every XPSP2 and Vista machine has an HTTP server running as a kernel driver that ordinary user processes can use to publish whatever they want.
fsync semantic is needed whenever you want to implement ACID transactions. This lies at the core of database systems and journaling file systems, for example. No fsync, no data integrity.
Are the spooks running out of mathematicians?!
on
Pure Math, Pure Joy
·
· Score: 4, Funny
Why else would a major newspaper have a piece that describes maths in a positive light?
>> Why would I want ot give up my time tested, >> vendor supported database for a piece of >> propriatary technology from a company that >> might not be around next week?
(1) Faster development: The persistent classes from your application design become the database design. You don't have to write code that does the object-relational mapping. Less DB interface code means less opportunities for error injection. And less time to finish your program. (2) Performance: If your persistent data structures are hierarchical and you mostly do pointer-chasing, vector or hashtable look-ups, you'll see 10x-100x speedup versus the relational ORM approach. (I've wittnessed this myself with our own application using ObjectStore.) (3) Which situation leaves you in a better position: Being an influential customer of a small company, or being a faceless customer of a huge company?
Choice of a database shouldn't become a religious affair. Whatever get's the job done in the fastest and most convenient way should be used for the job.
Have you have looked at object-oriented databases? They give you ACID transactions, and also take care of mapping the data into your main memory so that you as a programmer only have to deal with in-memory objects. The leading OODBs are Objectstore (www.exln.com), Versant (www.versant.com) and Poet (www.poet.com).
- It shows that you have the stamina to work on a goal for a long time, and that you don't give up when you run into a problem - It's the biggest opportunity to make friends that you'll stay in contact for a lifetime, and that can help you out if you're in trouble - You learn how to learn new stuff. If you want to work in the IT field as career, you better get used to the concept that you never need [technological] knowledge that is older than four years. Trust me. - You learn how to attack problems - It helps you get a job in bad times - It makes you a more creative person And when you're finally in university: - Try to learn something that is not connected to your own field. I studied mathematics, but the most rewarding lectures were the ones in psychology, and the foreign language classes - Try to spend some time abroad - Every now and then get a reality check by working for money or as an intern
It boils down to where the complexity in your problem is:
- If the data structures are rich and manyfold, you model the data types first (Has-a and Is-a relationships). Then you spread your algorithms around it. That's what OOP is about in really big systems.
- If the data structures are trivial, but the algorithms working on it are complex, you put all your energy into the algorithms, decompose them, look for coupling and reuse and end up with structured programming. Then you spread the few data structures in that you have and move them around as parameters. That's what procedural programming is all about.
In my experience, commercial applications are rich in data structures but trivial in algorithms (only add and subtract) whereas engineering applications only have matrices as their data structures, and calculation and visualisation is complex.
The question is not on whose time you get trained (company's or your own). The question is: Can you afford to not be trained at all? If your employer is not willing to give you the time necessary for staying current, then hell do it on your own time. Otherwise you'll be without a job in two to three years!
Really difficult. For me as a programmer it's KDevelop, but what about KDE, Gnome, OpenLDAP, UnixODBC, and all the other important projects? Maybe Andover could divide the prize between the first three or five nominees?
Slashdot Mirror
They're trying to somehow justify why they silently rolled out http.sys, a rebranded IIS, as a kernel driver with Win XP SP2. With that on the machine, each and every XPSP2 and Vista machine has an HTTP server running as a kernel driver that ordinary user processes can use to publish whatever they want.
fsync semantic is needed whenever you want to implement ACID transactions. This lies at the core of database systems and journaling file systems, for example. No fsync, no data integrity.
Why else would a major newspaper have a piece that describes maths in a positive light?
>> Why would I want ot give up my time tested,
>> vendor supported database for a piece of
>> propriatary technology from a company that
>> might not be around next week?
(1) Faster development: The persistent classes from your application design become the database design. You don't have to write code that does the object-relational mapping. Less DB interface code means less opportunities for error injection. And less time to finish your program.
(2) Performance: If your persistent data structures are hierarchical and you mostly do pointer-chasing, vector or hashtable look-ups, you'll see 10x-100x speedup versus the relational ORM approach. (I've wittnessed this myself with our own application using ObjectStore.)
(3) Which situation leaves you in a better position: Being an influential customer of a small company, or being a faceless customer of a huge company?
Choice of a database shouldn't become a religious affair. Whatever get's the job done in the fastest and most convenient way should be used for the job.
Have you have looked at object-oriented databases? They give you ACID transactions, and also take care of mapping the data into your main memory so that you as a programmer only have to deal with in-memory objects. The leading OODBs are Objectstore (www.exln.com), Versant (www.versant.com) and Poet (www.poet.com).
- It shows that you have the stamina to work on a goal for a long time, and that you don't give up when you run into a problem
- It's the biggest opportunity to make friends that you'll stay in contact for a lifetime, and that can help you out if you're in trouble
- You learn how to learn new stuff. If you want to work in the IT field as career, you better get used to the concept that you never need [technological] knowledge that is older than four years. Trust me.
- You learn how to attack problems
- It helps you get a job in bad times
- It makes you a more creative person
And when you're finally in university:
- Try to learn something that is not connected to your own field. I studied mathematics, but the most rewarding lectures were the ones in psychology, and the foreign language classes
- Try to spend some time abroad
- Every now and then get a reality check by working for money or as an intern
It boils down to where the complexity in your problem is:
- If the data structures are rich and manyfold, you model the data types first (Has-a and Is-a relationships). Then you spread your algorithms around it. That's what OOP is about in really big systems.
- If the data structures are trivial, but the algorithms working on it are complex, you put all your energy into the algorithms, decompose them, look for coupling and reuse and end up with structured programming. Then you spread the few data structures in that you have and move them around as parameters. That's what procedural programming is all about.
In my experience, commercial applications are rich in data structures but trivial in algorithms (only add and subtract) whereas engineering applications only have matrices as their data structures, and calculation and visualisation is complex.
The question is not on whose time you get trained (company's or your own). The question is: Can you afford to not be trained at all? If your employer is not willing to give you the time necessary for staying current, then hell do it on your own time. Otherwise you'll be without a job in two to three years!
Really difficult. For me as a programmer it's KDevelop, but what about KDE, Gnome, OpenLDAP, UnixODBC, and all the other important projects? Maybe Andover could divide the prize between the first three or five nominees?