I spent a bunch of time researching this and figuring out "how you get the most work out of programmers."
What it comes down to is that there are no shortcuts. Treating your programmers well is the best way to get the most work out of them. That doesn't mean pampering them, but 10 hour days are just going to hammer them if kept on for more than two weeks.
He was told to get back into his car after the officer had just punched him in the face.
I can't even imagine where this timeline of events was created. Probably in your head. It doesn't appear in any credible news source.
Ah crap. The command was to lie down on the ground, not to get back in the car.
As for the source, it's direct from Dr. Watts himself:
"So what it came down to, ultimately, was those moments after I was repeatedly struck in the face by Beaudry (an event not in dispute, incidentally). After Beaudry had finished whaling on me in the car, and stepped outside, and ordered me out of the vehicle; after I’d complied with that, and was standing motionless beside the car, and Beaudry told me to get on the ground — I just stood there, saying “What is the problem?”, just before Beaudry maced me.
And that, said the Prosecutor in her final remarks — that, right there, was failure to comply. That was enough to convict."
He was told to get back into his car after the officer had just punched him in the face.
It doesn't surprise me that he'd be confused and disoriented, or that he'd be slow to comply. Try punching someone in the face some time. It hurts.
The really sad bit is that under these laws, you could not only punch someone in the face, you could pepper spray them, kick them in the nuts while they were down, and then tell them you wanted them to stand up and empty out your pockets. Don't do it because you're screaming and in pain, or trying to run away? You're committing a crime.
I've been accumulating books for years. I think the best ones are the down to earth programming ones -- Code Complete, The Practice of Programming, Programming Pearls, and the like.
I have a photo on Flickr that I like to look at when I feel stupid.
There are a few more books at work, but most of these have served me well at various times. Apart from the BEEP protocol book, that was a total waste.
The Christian Science Monitor has a decent news website, csmonitor.com. They actually do a pretty good job of reporting events (although their analysis reflects their bent, of course).
A couple of other people have suggested going to an open source project. I think you should as well. One of the best things that an OSS project will do is show you the true complexity of a working application, as opposed to the theoretical concepts presented in an education.
You will also learn far more good practices from a well known / respected OSS project than you will in your average organization. I learnt a ton just reading through the Mozilla source code and following how they broke out interfaces and practiced defensive programming.
Finally, you will have the chance to work with people who need you for something. Programming is very much a team exercise, and you can learn about teamwork as well as programming through the activity. Apprenticing yourself to a senior mentor is a great way to piggyback through your mentor's experience, as long as you have the necessary humility to have your work ripped to shreds...
Finally, don't sell yourself short. Alan Cox has a story about a random guy who hung out on the Linux IRC channel until someone told him to make himself useful. He ended up writing most of the IPv6 stack.
> I am a professional programmer, I make a 6 figure salary, and I suck at math.
Ditto. I can't even split the bill in restaurants.
In practice, the only time I've had to worry about math is in determining which algorithm to use and some awareness of how primes work in hashing. I've had to do some permutation and combinatorial stuff, but nothing Knuth couldn't destroy before breakfast.
There are a number of utilities that allow you more control over Windows paging. notably at Sysinternals which has PageDefrag and CacheSet.
There are also a couple of registry tweaks you can try: DisableExecutivePaging and LargeSystemCache. Arstechnica has a guide at http://arstechnica.com/guides/tweaks/memory-1.ars which is useful. No hard data on whether this actually made any difference or not, of course.
I use Perforce, which works great for me. You can get a two person license and it comes with a GUI client that works much better for me than any of the Subversion ones. It also has the useful feature of keeping track of all changes I make to the repository without keeping any hidden directories around (and you'll appreciate this the first time you move a subversion directory from one place to another).
Subversion is probably a reasonable second, although Perforce wins out on features and simplicity.
I would say so. You can highlight a section of code and tell Eclipse to "Extract Method" and it will take all the variables external to that section of code and make them parameters to the method it creates. You can move methods up and down the class hierarchy. Implement interfaces automatically.
Again, I haven't found anything that IntelliJ does that Eclipse doesn't do, but I'm officially behind the current package now and therefore don't know what I'm talking about.:-)
I didn't like the Java editor. There are many things that IntelliJ does differently, but there are very few things I could say that IntelliJ does that Eclipse doesn't in the Java editor. At least for 4.5.4. I have no experience with 5.0.
However, the JSP and XML support in IntelliJ freaking rocks. Live templates combined with the IntelliJ JSP editor is enough that I switch out of Eclipse to IntelliJ whenever I have to edit JSP, even though I have WTP installed. I've been told that JDeveloper and Netbeans also have JSP editor support, but haven't looked at them closely.
Since you haven't mentioned a specific programming language, I'll assume you're conversant in C, but not C++ or Java. In this case, you'd probably be good learning Java: you'd be learning a new style of programming, have a "safe" environment to fail (since everything in is in a JVM) and you can leverage the cross-platform GUI to run on Mac and Linux.
Eventually, you'll want to move to Eclipse, but you can start off with BlueJ, which will help you learn the basics before you try anything more advanced.
You should be using the final version, not the beta.
Having said that, I don't doubt it will get fixed sharpish. The writer of ShadowPlan is VERY on the ball, if a little insane in his business practices.
You are still missing my point. I'm not saying that it's impossible, or even hard to write transactionally aware code. I'm saying that Prevayler doesn't provide that functionality for you, and it is a mistake to say that it does.
I bitch and rant because for every aware practitioner like yourself who examines the technology carefully and can fill in the gaps, there are ten or more people who take it at face value, stick some code in, go live without testing, and then wonder what they did wrong.
You say "if you take an approach like Prevayler you can still have reliability and transactional integrity."
Prevayler does not provide ACID transactions. Saying "you don't need rollback or undo" is all well and good, but you can't call it a transaction. It's not atomic, and it's not consistent.
So if you use Prevayler, do you have "reliability" and "transactional integrity"? No. It's something you have to code yourself. Among other things, you have to make sure no code in a transaction ever throws a RuntimeException: http://www.prevayler.org/wiki.jsp?topic=RollbackIs Needless. So your statement is incorrect.
Prevayler is not a reliable system. It does not have any transactional integrity. You can't run "entire systems" from it because you are limited to the maximum heap size of the JVM.
I thought it sounded like a scam just reading the documentation, and I've never seen a technical explanation of the "black magic" that enables it to magically outperform Oracle or even MySQL.
This is completely confirmed on reading Mike Spille's blog. Spille's a solid, no-bullshit kind of guy, and his technical calls are right on every time I've had independent knowledge, so I trust him on this. Check out the link:
"The key thing to realize here is that the "transactional" capability that Prevayler is giving you here is serialized execution of Transactions, and that it won't write to the log if the transaction can't be pulled off. And of course, on success it will write to the transactional log. That's it. There's no rollback/undo, no locking, no thought whatsoever for concurrency in any form."
Okay, I know logically there must be a reason why Itaniums exist (someone buys them), but the mind really boggles at the idea of ~1800 Itaniums on a cluster. At what point is it more cost effective to buy an Itanium than a gaggle of Celerons for CPU power? Is there really that much of a compelling advantage to buying a $2000 chip?
Slashdot Mirror
Seconded. It takes a while to get used to some of the fine details of Play, but it's very much worth it.
I spent a bunch of time researching this and figuring out "how you get the most work out of programmers."
What it comes down to is that there are no shortcuts. Treating your programmers well is the best way to get the most work out of them. That doesn't mean pampering them, but 10 hour days are just going to hammer them if kept on for more than two weeks.
Here are the blog posts:
http://tersesystems.com/2007/08/16/getting-work-out-of-programmers-part-1
http://tersesystems.com/2007/08/20/getting-work-out-of-programmers-part-2
Good luck.
The entire 240/ block is reserved. Is there something wrong with those IP addresses?
He was told to get back into his car after the officer had just punched him in the face.
I can't even imagine where this timeline of events was created. Probably in your head. It doesn't appear in any credible news source.
Ah crap. The command was to lie down on the ground, not to get back in the car.
As for the source, it's direct from Dr. Watts himself:
"So what it came down to, ultimately, was those moments after I was repeatedly struck in the face by Beaudry (an event not in dispute, incidentally). After Beaudry had finished whaling on me in the car, and stepped outside, and ordered me out of the vehicle; after I’d complied with that, and was standing motionless beside the car, and Beaudry told me to get on the ground — I just stood there, saying “What is the problem?”, just before Beaudry maced me.
And that, said the Prosecutor in her final remarks — that, right there, was failure to comply. That was enough to convict."
He was told to get back into his car after the officer had just punched him in the face.
It doesn't surprise me that he'd be confused and disoriented, or that he'd be slow to comply. Try punching someone in the face some time. It hurts.
The really sad bit is that under these laws, you could not only punch someone in the face, you could pepper spray them, kick them in the nuts while they were down, and then tell them you wanted them to stand up and empty out your pockets. Don't do it because you're screaming and in pain, or trying to run away? You're committing a crime.
There is a subversion plugin to JIRA which is free and will automatically scan for bug references. It's pretty good.
You can change the size of the photo. At the original size (5 megapixel camera thingy) it's very readable.
I've been accumulating books for years. I think the best ones are the down to earth programming ones -- Code Complete, The Practice of Programming, Programming Pearls, and the like.
I have a photo on Flickr that I like to look at when I feel stupid.
There are a few more books at work, but most of these have served me well at various times. Apart from the BEEP protocol book, that was a total waste.
The Christian Science Monitor has a decent news website, csmonitor.com. They actually do a pretty good job of reporting events (although their analysis reflects their bent, of course).
Also see Mike Spille's criticism.
A couple of other people have suggested going to an open source project. I think you should as well. One of the best things that an OSS project will do is show you the true complexity of a working application, as opposed to the theoretical concepts presented in an education.
You will also learn far more good practices from a well known / respected OSS project than you will in your average organization. I learnt a ton just reading through the Mozilla source code and following how they broke out interfaces and practiced defensive programming.
Finally, you will have the chance to work with people who need you for something. Programming is very much a team exercise, and you can learn about teamwork as well as programming through the activity. Apprenticing yourself to a senior mentor is a great way to piggyback through your mentor's experience, as long as you have the necessary humility to have your work ripped to shreds...
Finally, don't sell yourself short. Alan Cox has a story about a random guy who hung out on the Linux IRC channel until someone told him to make himself useful. He ended up writing most of the IPv6 stack.
> I am a professional programmer, I make a 6 figure salary, and I suck at math.
Ditto. I can't even split the bill in restaurants.
In practice, the only time I've had to worry about math is in determining which algorithm to use and some awareness of how primes work in hashing. I've had to do some permutation and combinatorial stuff, but nothing Knuth couldn't destroy before breakfast.
Seconded. Alienbrain is the single best tool you can get for this kind of work.
There are a number of utilities that allow you more control over Windows paging. notably at Sysinternals which has PageDefrag and CacheSet.
There are also a couple of registry tweaks you can try: DisableExecutivePaging and LargeSystemCache. Arstechnica has a guide at http://arstechnica.com/guides/tweaks/memory-1.ars which is useful. No hard data on whether this actually made any difference or not, of course.
I use Perforce, which works great for me. You can get a two person license and it comes with a GUI client that works much better for me than any of the Subversion ones. It also has the useful feature of keeping track of all changes I make to the repository without keeping any hidden directories around (and you'll appreciate this the first time you move a subversion directory from one place to another).
Subversion is probably a reasonable second, although Perforce wins out on features and simplicity.
I would say so. You can highlight a section of code and tell Eclipse to "Extract Method" and it will take all the variables external to that section of code and make them parameters to the method it creates. You can move methods up and down the class hierarchy. Implement interfaces automatically.
:-)
Again, I haven't found anything that IntelliJ does that Eclipse doesn't do, but I'm officially behind the current package now and therefore don't know what I'm talking about.
However, the JSP and XML support in IntelliJ freaking rocks. Live templates combined with the IntelliJ JSP editor is enough that I switch out of Eclipse to IntelliJ whenever I have to edit JSP, even though I have WTP installed. I've been told that JDeveloper and Netbeans also have JSP editor support, but haven't looked at them closely.
Eventually, you'll want to move to Eclipse, but you can start off with BlueJ, which will help you learn the basics before you try anything more advanced.
Have you tried other Windows search solutions? I've tried a few and I prefer Copernic to Google desktop by a fair margin.
You should be using the final version, not the beta.
Having said that, I don't doubt it will get fixed sharpish. The writer of ShadowPlan is VERY on the ball, if a little insane in his business practices.
Will.
You are still missing my point. I'm not saying that it's impossible, or even hard to write transactionally aware code. I'm saying that Prevayler doesn't provide that functionality for you, and it is a mistake to say that it does.
I bitch and rant because for every aware practitioner like yourself who examines the technology carefully and can fill in the gaps, there are ten or more people who take it at face value, stick some code in, go live without testing, and then wonder what they did wrong.
You say "if you take an approach like Prevayler you can still have reliability and transactional integrity."
s Needless. So your statement is incorrect.
Prevayler does not provide ACID transactions. Saying "you don't need rollback or undo" is all well and good, but you can't call it a transaction. It's not atomic, and it's not consistent.
So if you use Prevayler, do you have "reliability" and "transactional integrity"? No. It's something you have to code yourself. Among other things, you have to make sure no code in a transaction ever throws a RuntimeException: http://www.prevayler.org/wiki.jsp?topic=RollbackI
I thought it sounded like a scam just reading the documentation, and I've never seen a technical explanation of the "black magic" that enables it to magically outperform Oracle or even MySQL.
This is completely confirmed on reading Mike Spille's blog. Spille's a solid, no-bullshit kind of guy, and his technical calls are right on every time I've had independent knowledge, so I trust him on this. Check out the link:
http://www.pyrasun.com/mike/mt/archives/2004/12/25 /15.02.00/index.html
"The key thing to realize here is that the "transactional" capability that Prevayler is giving you here is serialized execution of Transactions, and that it won't write to the log if the transaction can't be pulled off. And of course, on success it will write to the transactional log. That's it. There's no rollback/undo, no locking, no thought whatsoever for concurrency in any form."
Yay. I feel so safe.
Will.Reporter interviews JWZ, invited to explore own sexuality.
Okay, I know logically there must be a reason why Itaniums exist (someone buys them), but the mind really boggles at the idea of ~1800 Itaniums on a cluster. At what point is it more cost effective to buy an Itanium than a gaggle of Celerons for CPU power? Is there really that much of a compelling advantage to buying a $2000 chip?