This isn't a foreign language, it's a computer language. Especially for business programming, it should be about week and a competent programmer should be able to write decent code with a manual or a web page open.
Most of the code I write is in python or java. But I'm writing a small windows application, and chose C#: I installed the IDE, and I was writing code in a few days, especially with google helping out.
When needed I write C and C++ as well, again not a big deal except it takes a lot more time to get the same amount of work done, but sometimes there's a library you need to link to.
The caveat in all of this is you need to be competent.
1. Get the OS to compile and boot. 2. Reproduce windows bugs. 3. Fix bugs faster the MS does.
[...]
6. Profit!
Sounds entertaining... but
on
Software Exorcism
·
· Score: 5, Insightful
Here's a short version of what you need to know when you're working for someone.
Do you know the difference between a cost center and a profit center?
A cost center's something the business needs to do but doesn't make any money. Think accounting, or maintaining print servers -- the goal is to make its function as cheap as possible. One attractive way is to offshore it, provided things work out as cheaply as possible.
A profit center makes the business money. Like software development, or whatever it is that the business does: doing a good job will make the company money.
Spread seems to offer some interesting MOM functionality. Pub/Sub systems are fun!
Spread is a toolkit that provides a high performance messaging service that is resilient to faults across external or internal networks. Spread functions as a unified message bus for distributed applications, and provides highly tuned application-level multicast and group communication support. Spread services range from reliable message passing to fully ordered messages with delivery guarantees, even in case of computer failures and network partitions.
Spread is designed to encapsulate the challenging aspects of asynchronous networks and enable the construction of scalable distributed applications, allowing application builders to focus on the differentiating components of their application.
So here's what's stopping me from really considering a mac (as opposed to 2nd tier Thinkpad, which I normally get every 2-3 years).
How does it work w/ corporate networks? I do a lot of consulting, and more than once I've ended up connecting to VPN's using Win2k's client, or even worse a 3rd party client like Nortel w/ a SecureID token generator. My big fear is that I'd be SOL there, and worse, I'd get the "oh a mac" groan from IT.
Sounds really good, MP3's even from flash would be awesome, and being able do the hardware hack to wire in an IDE Hard Drive would make this a winner. (And it's cheapcheap).
This review says it doesn't work with VBR MP3 -- is this still the case with the firmware hacks?
Let someone else take the risk of screwing up a motherboard and needing to buy a new one. Retail vs. wholesale's a pretty small margin, it's not worth it to me -- especially since I've got a small work area.
Dianne Feinstein was up for re-election in 2000. Despite being from San Francisco, she's probably one of the least friendly congresspeople towards technology and the internet. Her opponent was Tom Campbell, an imminently sensible pro-technology Republican. He got his clock cleaned.
Since Feinstein's up in 2006, I'd volunteer Boxer in 2004. There's plenty of time, she's in favor of the Hollings bill, and she's a perrenially weak candidate. If the California Republicans would re-run Campbell, or someone else similar from Silicon Valley, it's be a great way of getting our voices heard.
I'd use Yahoo to read POP3 mail (Check other Mail) from work when I couldn't log into my home box. Then I'd fetchmail everything so it ends up in one place.
Thankfully, I don't have too much invested in using Yahoo mail -- I'm basically using it as a POP3 mailreading web app (I set the reply-to correctly and told people never to address it to me@yahoo.com).
Anyone know of an alternative? Hushmail w/ POP over SSH?
I graduated 9 years ago with a pure math / CS double major. I must say that my biggest regret is that I didn't study more of the applied stuff. The pure stuff was a challenging exercise, and certainly did wonders for my thinking skills, but it's just not that useful in day-to-day.
When I find some free time I'll be auditing the local college's courses.
In the San Francisco area, the Biotech companies are on a hiring swing. It's a notoriously hard area for even the strongest programmers to get a job in, unless they've worked in biotech before.
Any indications if this book (or any of the others noted here) would be enough to get someone in the door?
Incremental Compilation: VAJ or Eclipse
on
Java IDEs?
·
· Score: 1
Incremental Compilation is something I've found incredibly useful -- get a running app, add classes, change methods, have the app take advantage of it while it's running.
This is incredibly useful as a system grows and 5 minute startup times become normal. It's also something I became used to from my Lisp and Smalltalk days.
As far as I can tell, only VAJ and Eclipse can do this.
And in light of this, JBuilder, Forte, etc seem like glorified wrappers around jdk functionality. Might as well use emacs.
.. Because it's possible for person of average intelligence (IQ = 100) to write passable code in it.
LISP and to a lesser extent Smalltalk are power tools. If you're good and smart, you'll write something which is really beautiful, elegant and fast. But most people aren't that good, and they produce crap.
And crap in Java is still more maintainable (I should know, I'm untangling a 700+ class ball of spaghetti as we speak) than crap in LISP or Smalltalk.
The fact that MIT and Berkeley (I think) and other schools with great CS programs start with Scheme illustrates the point -- it's not for well... the dumbass in the other cube.
It seems that both management and prima donna types are failing on this..
Each bit of code written is a liability for the organization. It's code which needs to be maintained, debugged, and potentially ported. That's why the best code strives to be mundane. Coding as a profession should be.. routine.
Documentation is a force -- if the code's too hard to document, it's a hint that refactoring needs to happen.
Unit tests are another force -- what I'll do when handing off a project is go through the unit tests, I may end up writing them myself. That's a contract which the implementation has to abide by.
Finally, it needs to be pointed out that if a prima donna writes some code which nobody understands... and turns out some spotty documentation.. it's no better than a 3rd part component. If it took a month to develop, it easily cost the organization 10k (salary, opportunity cost, support). Does a 3rd party component cost less than 10k? Good choice. Otherwise, the in-house development needs to add more value than just working.
MIT and Berkeley were pretty well known for using Scheme as an intro computing language. It gives a pretty easy transition into computuational issues and if you "get it", you learn powerful metaphors which can be implemented in many many languages.
For O-O, I think Smalltalk (or right now Squeak) are utterly appropriate. For the same reason -- you get used to metaprogramming concepts, and Smalltalk's probably the best environment I've seen where you really grasp the implications of the concept -- "everything is an object."
Java, C, COBOL, and C++ are industrial languages -- they were designed with industrial tradeoffs which have nothing to do with educational forces.
I've got a huge problem with treating CS programs like job training: if you want Java for a resume, go to a Sun training class. In SF there are "learn IT in 3 weeks!" posters in bathrooms. Blah.
Reading philg's piece [quickly] it seems one reason that aD's hand was forced to take VC money is because they were experiencing 1000% growth -- which presented some serious organizational issues. (There's also the need for a pile of money, but let's play with this one)
I'm curious if this constraint could have been relaxed: could aD have "spun off" aD1, aD2, etc.. and kept the informal management style which made them as effective.
Slashdot Mirror
This isn't a foreign language, it's a computer language. Especially for business programming, it should be about week and a competent programmer should be able to write decent code with a manual or a web page open.
Most of the code I write is in python or java. But I'm writing a small windows application, and chose C#: I installed the IDE, and I was writing code in a few days, especially with google helping out.
When needed I write C and C++ as well, again not a big deal except it takes a lot more time to get the same amount of work done, but sometimes there's a library you need to link to.
The caveat in all of this is you need to be competent.
1. Get the OS to compile and boot.
2. Reproduce windows bugs.
3. Fix bugs faster the MS does.
[...]
6. Profit!
Here's a short version of what you need to know when you're working for someone.
Do you know the difference between a cost center and a profit center?
A cost center's something the business needs to do but doesn't make any money. Think accounting, or maintaining print servers -- the goal is to make its function as cheap as possible. One attractive way is to offshore it, provided things work out as cheaply as possible.
A profit center makes the business money. Like software development, or whatever it is that the business does: doing a good job will make the company money.
It's always better to work for the profit center.
Well... they're MS, so odds are it would be more like:
while(1) {
blue.screen();
}
So here's what's stopping me from really considering a mac (as opposed to 2nd tier Thinkpad, which I normally get every 2-3 years).
How does it work w/ corporate networks? I do a lot of consulting, and more than once I've ended up connecting to VPN's using Win2k's client, or even worse a 3rd party client like Nortel w/ a SecureID token generator. My big fear is that I'd be SOL there, and worse, I'd get the "oh a mac" groan from IT.
Looks very nice, though..
Interesting -- it seems then that a decent strategy would be to mix disks from different manufacturers.
That and adequate alarming.
Sounds really good, MP3's even from flash would be awesome, and being able do the hardware hack to wire in an IDE Hard Drive would make this a winner. (And it's cheapcheap).
This review says it doesn't work with VBR MP3 -- is this still the case with the firmware hacks?
Way too short -- public and private keys reside in the same space and should have the same length.
They all look alike!
(I'm kidding)
Is your Websphere app unit tested? Use JUnit, and potentially add-ons like Cactus and HttpUnit so you've got some way of verifying your port.
If you decide you need a consultant, remember that the only thing worse than an expensive consultant is a cheap one.
Let someone else take the risk of screwing up a motherboard and needing to buy a new one. Retail vs. wholesale's a pretty small margin, it's not worth it to me -- especially since I've got a small work area.
I'll use pricewatch or Yahoo shopping to get upgrade parts, though.. no problem.
Open-source anti-virus would be very cool, but it's really labor intensive and the signature databases are the vendor's crown jewels.. as it were.
The Virus Bulletin's VB100 test rates AVG fairly low. Do other tests rate it higher?
Dianne Feinstein was up for re-election in 2000. Despite being from San Francisco, she's probably one of the least friendly congresspeople towards technology and the internet. Her opponent was Tom Campbell, an imminently sensible pro-technology Republican. He got his clock cleaned.
Since Feinstein's up in 2006, I'd volunteer Boxer in 2004. There's plenty of time, she's in favor of the Hollings bill, and she's a perrenially weak candidate. If the California Republicans would re-run Campbell, or someone else similar from Silicon Valley, it's be a great way of getting our voices heard.
Is there a similarly weak Republican?
I'd use Yahoo to read POP3 mail (Check other Mail) from work when I couldn't log into my home box. Then I'd fetchmail everything so it ends up in one place.
Thankfully, I don't have too much invested in using Yahoo mail -- I'm basically using it as a POP3 mailreading web app (I set the reply-to correctly and told people never to address it to me@yahoo.com).
Anyone know of an alternative? Hushmail w/ POP over SSH?
Installing linux and making a beowulf cluster of them.
-1: stupid
I graduated 9 years ago with a pure math / CS double major. I must say that my biggest regret is that I didn't study more of the applied stuff. The pure stuff was a challenging exercise, and certainly did wonders for my thinking skills, but it's just not that useful in day-to-day.
When I find some free time I'll be auditing the local college's courses.
In the San Francisco area, the Biotech companies are on a hiring swing. It's a notoriously hard area for even the strongest programmers to get a job in, unless they've worked in biotech before.
Any indications if this book (or any of the others noted here) would be enough to get someone in the door?
Incremental Compilation is something I've found incredibly useful -- get a running app, add classes, change methods, have the app take advantage of it while it's running.
This is incredibly useful as a system grows and 5 minute startup times become normal. It's also something I became used to from my Lisp and Smalltalk days.
As far as I can tell, only VAJ and Eclipse can do this.
And in light of this, JBuilder, Forte, etc seem like glorified wrappers around jdk functionality. Might as well use emacs.
.. Because it's possible for person of average intelligence (IQ = 100) to write passable code in it.
LISP and to a lesser extent Smalltalk are power tools. If you're good and smart, you'll write something which is really beautiful, elegant and fast. But most people aren't that good, and they produce crap.
And crap in Java is still more maintainable (I should know, I'm untangling a 700+ class ball of spaghetti as we speak) than crap in LISP or Smalltalk.
The fact that MIT and Berkeley (I think) and other schools with great CS programs start with Scheme illustrates the point -- it's not for well... the dumbass in the other cube.
It seems that both management and prima donna types are failing on this..
.. routine.
Each bit of code written is a liability for the organization. It's code which needs to be maintained, debugged, and potentially ported. That's why the best code strives to be mundane. Coding as a profession should be
Documentation is a force -- if the code's too hard to document, it's a hint that refactoring needs to happen.
Unit tests are another force -- what I'll do when handing off a project is go through the unit tests, I may end up writing them myself. That's a contract which the implementation has to abide by.
Finally, it needs to be pointed out that if a prima donna writes some code which nobody understands... and turns out some spotty documentation.. it's no better than a 3rd part component. If it took a month to develop, it easily cost the organization 10k (salary, opportunity cost, support). Does a 3rd party component cost less than 10k? Good choice. Otherwise, the in-house development needs to add more value than just working.
MIT and Berkeley were pretty well known for using Scheme as an intro computing language. It gives a pretty easy transition into computuational issues and if you "get it", you learn powerful metaphors which can be implemented in many many languages. For O-O, I think Smalltalk (or right now Squeak) are utterly appropriate. For the same reason -- you get used to metaprogramming concepts, and Smalltalk's probably the best environment I've seen where you really grasp the implications of the concept -- "everything is an object." Java, C, COBOL, and C++ are industrial languages -- they were designed with industrial tradeoffs which have nothing to do with educational forces. I've got a huge problem with treating CS programs like job training: if you want Java for a resume, go to a Sun training class. In SF there are "learn IT in 3 weeks!" posters in bathrooms. Blah.
Don't forget Squeak which is open-source Smalltalk. A steeper learning curve than VAST or VW, but in a lot of ways much more interesting.
Sometimes people just explode. Natural Causes.
Reading philg's piece [quickly] it seems one reason that aD's hand was forced to take VC money is because they were experiencing 1000% growth -- which presented some serious organizational issues. (There's also the need for a pile of money, but let's play with this one)
I'm curious if this constraint could have been relaxed: could aD have "spun off" aD1, aD2, etc.. and kept the informal management style which made them as effective.