For me, having invested immensely more in Clojure than other languages (I think that's an anomaly), it's faster for me to write Clojure than anything else going forward.
I've been having a lot of fun using the same knowledge on multiple platforms, be it JVM-Clojure or JS with Clojurescript.
I've also used python and ruby (and PHP), but I'm finding that node.js is a lot of fun, and npm is a coherent package manager.
If you're just starting out, I recommend doing Javascript + Node for max bang for your time-buck.
You'll get your backend and frontend at the same time, and it's quite a fast runtime. Javascript itself has some thorny edges, but it's a relatively simple language compared to the others you're considering.
The core is pretty compatible, but certain features are present in one, absent in the other, and vice-versa.
You generally wouldn't care to write a full program in one and expect it to work in the other, but you can share code, there's projects like CLJX that facilitate this.
This isn't true. Clojurescript uses the Google Closure optimizing javascript compiler for things like minification, including dead code elimination. Source maps support makes debugging this kind of code still possible.
I'm 4 years experienced, I still remember when I had 0 years experience.
The difference between then and now is a few things:
1. I actually, objectively, learn faster, I have more experience, more techniques at my disposal, and fast paths in my brain to do so.
2. The discomfort of not knowing something is much more pronounced relative to my usual competency.
When I was just starting out, the relative cost of #2 was the same for a number of topics.
Once you are an expert in one, working through feeling dumb all the time I imagine is much more frustrating.
What worked for me and the best advice I could possibly give is to identify the smartest programming writers and speakers that are doing what you want to do, read their books, watch their talks, and embody their viewpoints of code.
In the course of a few years of this you will gain a foundation whose benefits become more apparent over time.
Dealing with the perceived discomfort from #2, I'm not sure as I'm super-crabby and still young, but the most up-to-date and fresh experienced developers I've had the pleasure to work with embody a childlike curiosity.
#2 can be mitigated
#1 is a little more biological
I think #2 and related perceived tradeoffs dominate the learning equilibrium.
yea, I don't really miss the 'one right way to do anything' attitude. I wouldn't really say python's a paradigm, it's sort of just a highly-opinionated and dynamic OO language and community. Ruby tries to be friendlier, but I don't use that either.
The whitespace thing.. I guess it can be inconvenient. I just always use spaces. I have felt the pain (in java) of badly formatted and mixed-formatted code being weird in different editors, but I use a lisp full-time now, which means I'm willing to trudge through syntax unfamiliarity for the sake of some benefits. I prefer the regularity, though you can't control how other people format their code or what editors they use.
If it's code that I'm working on, I make sure it's indented properly.
This is a road fraught with peril, likely because the people that implemented the original functions are the same ones that made crappy language choices that prevent new functions from being used effectively, so it's possible to both paper over issues, conveniently in the short term, and simultaneously creating more issues in the long-term.
Hate for python whitespace is a sure sign of a superficial appreciation for its tradeoffs. It doesn't say anything bad about the language, it just shows you haven't actually used it for anything and are willing to make egregious statements based on unfamiliarity and ignorance. Why learn a language if it's not a bit unfamiliar in the first place?
(disclaimer, haven't used python since 2010)
Maybe you're a good coder, but this would not be advice I'd give anyone except the most elite folks, who don't need it. I personally prefer to spend extra up-front time to learn a system and would advise anyone I work with to do the same. How do you know when things might interact when you don't know the premises of the system? You would have to rely on someone else to provide negative feedback. The approach seems a little flippant, but the details of how you implement it might make it work.
I've always been a little ADD and impatient with having to do things systematically, unfortunately I found this is the only approach that works for me. If I don't do this, I end up staring at the screen for long periods of time and not getting anywhere.
So, I wrote up a worksheet to help myself in these situations, here are the steps I identified as helpful to deal with massive levels of complexity in unfamiliar code:
1. Establish a clear goal and sub-goals.
2. Use the goals to determine the scope of your reading.
3. Allocate quite a bit of time in large chunks.
4. Identify key layers of abstraction.
5. Enumerate classes (functions, namespaces) of interest.
6. Systematically, read through each class superficially.
7. Pick 8 classes to focus on.
8. Do a deep dive.
9. List/sketch inputs and outputs in terms of function names, types referenced.
10. Look at relevant tests for usage as needed.
11. Check off each class once looked through.
12. Measure the complexity of a component by how many checks are required for full understanding.
13. Iterate until goals achieved.
I didn't learn how to learn until late into my master's studies... I was near the top of the class in one of the best public high schools in Georgia. I would have very much more enjoyed a school like this.
Ugh... It's not MY argument, it's the reality of history. A software translation layer (VM) will be less efficient than microcode, which is what intel processors use anyway to decode x86 into a RISC ISA. You really don't know what you're talking about here.
that's totally silly, a non-x86 cpu has no benefit for running windows. You can't run x86 software on it, and that's why anyone uses windows in the first place. Windows without backwards compatibility will not sell.
I don't get this at all. The router menu is only a convenience. It would stand up to reason to expect that if you have a use for dyndns, you have some sort of box behind your router that stays up continuously. With this in mind, it should be fairly simple to update any DNS provider's records with an api from that box behind your router. If they don't support that, then bitch at them.
In the case that you're just accessing your router, not quite sure what the point would be.
Killing off the offspring of a lesser-fit being because they fail at parenting is just as much natural selection as killing off the lesser-fit being before they procreate in the first place.
actually, if he ends up having 2 kids, it's only 1/2 as much natural selection.
This is simple to answer. It's a form of smugness and self-promotion, to publicly wish a famous stranger well on an online forum makes yourself look well-aligned and better compared to other folks. If not that, then it's likely a comment in anticipation of poor taste of future posts (based on past slashdot posts).
Slashdot Mirror
For me, having invested immensely more in Clojure than other languages (I think that's an anomaly), it's faster for me to write Clojure than anything else going forward. I've been having a lot of fun using the same knowledge on multiple platforms, be it JVM-Clojure or JS with Clojurescript. I've also used python and ruby (and PHP), but I'm finding that node.js is a lot of fun, and npm is a coherent package manager. If you're just starting out, I recommend doing Javascript + Node for max bang for your time-buck. You'll get your backend and frontend at the same time, and it's quite a fast runtime. Javascript itself has some thorny edges, but it's a relatively simple language compared to the others you're considering.
The core is pretty compatible, but certain features are present in one, absent in the other, and vice-versa. You generally wouldn't care to write a full program in one and expect it to work in the other, but you can share code, there's projects like CLJX that facilitate this.
This isn't true. Clojurescript uses the Google Closure optimizing javascript compiler for things like minification, including dead code elimination. Source maps support makes debugging this kind of code still possible.
The difference between then and now is a few things:
When I was just starting out, the relative cost of #2 was the same for a number of topics.
Once you are an expert in one, working through feeling dumb all the time I imagine is much more frustrating.
What worked for me and the best advice I could possibly give is to identify the smartest programming writers and speakers that are doing what you want to do, read their books, watch their talks, and embody their viewpoints of code.
In the course of a few years of this you will gain a foundation whose benefits become more apparent over time.
Dealing with the perceived discomfort from #2, I'm not sure as I'm super-crabby and still young, but the most up-to-date and fresh experienced developers I've had the pleasure to work with embody a childlike curiosity.
#2 can be mitigated
#1 is a little more biological
I think #2 and related perceived tradeoffs dominate the learning equilibrium.
I've watched him do it 2-feet-away from me with Clojure, does that count?
as is Linux, but a successful, socially relevant one.
People have the right to ask him to stop being abusive, but he has the right to do what he wants.
Take your EE background and rigor, and get a career in software. You'll have an edge over the hipsters and will find it easy to rise.
me:
BS Biomedical Engineering from Georgia Tech.
MS Electrical and Computer Engineering from Georgia Tech
Some of the most talented programmers I work with are EE or physics guys.
yea, I don't really miss the 'one right way to do anything' attitude. I wouldn't really say python's a paradigm, it's sort of just a highly-opinionated and dynamic OO language and community. Ruby tries to be friendlier, but I don't use that either. The whitespace thing.. I guess it can be inconvenient. I just always use spaces. I have felt the pain (in java) of badly formatted and mixed-formatted code being weird in different editors, but I use a lisp full-time now, which means I'm willing to trudge through syntax unfamiliarity for the sake of some benefits. I prefer the regularity, though you can't control how other people format their code or what editors they use. If it's code that I'm working on, I make sure it's indented properly.
You can redefine the functions w/ your own.
This is a road fraught with peril, likely because the people that implemented the original functions are the same ones that made crappy language choices that prevent new functions from being used effectively, so it's possible to both paper over issues, conveniently in the short term, and simultaneously creating more issues in the long-term.
Hate for python whitespace is a sure sign of a superficial appreciation for its tradeoffs. It doesn't say anything bad about the language, it just shows you haven't actually used it for anything and are willing to make egregious statements based on unfamiliarity and ignorance. Why learn a language if it's not a bit unfamiliar in the first place? (disclaimer, haven't used python since 2010)
Maybe you're a good coder, but this would not be advice I'd give anyone except the most elite folks, who don't need it. I personally prefer to spend extra up-front time to learn a system and would advise anyone I work with to do the same. How do you know when things might interact when you don't know the premises of the system? You would have to rely on someone else to provide negative feedback. The approach seems a little flippant, but the details of how you implement it might make it work.
I've always been a little ADD and impatient with having to do things systematically, unfortunately I found this is the only approach that works for me. If I don't do this, I end up staring at the screen for long periods of time and not getting anywhere. So, I wrote up a worksheet to help myself in these situations, here are the steps I identified as helpful to deal with massive levels of complexity in unfamiliar code:
1. Establish a clear goal and sub-goals.
2. Use the goals to determine the scope of your reading.
3. Allocate quite a bit of time in large chunks.
4. Identify key layers of abstraction.
5. Enumerate classes (functions, namespaces) of interest.
6. Systematically, read through each class superficially.
7. Pick 8 classes to focus on.
8. Do a deep dive.
9. List/sketch inputs and outputs in terms of function names, types referenced.
10. Look at relevant tests for usage as needed.
11. Check off each class once looked through.
12. Measure the complexity of a component by how many checks are required for full understanding.
13. Iterate until goals achieved.
isn't a language just a turing machine API?
I didn't learn how to learn until late into my master's studies... I was near the top of the class in one of the best public high schools in Georgia. I would have very much more enjoyed a school like this.
Ugh... It's not MY argument, it's the reality of history. A software translation layer (VM) will be less efficient than microcode, which is what intel processors use anyway to decode x86 into a RISC ISA. You really don't know what you're talking about here.
that's totally silly, a non-x86 cpu has no benefit for running windows. You can't run x86 software on it, and that's why anyone uses windows in the first place. Windows without backwards compatibility will not sell.
I don't get this at all. The router menu is only a convenience. It would stand up to reason to expect that if you have a use for dyndns, you have some sort of box behind your router that stays up continuously. With this in mind, it should be fairly simple to update any DNS provider's records with an api from that box behind your router. If they don't support that, then bitch at them. In the case that you're just accessing your router, not quite sure what the point would be.
Killing off the offspring of a lesser-fit being because they fail at parenting is just as much natural selection as killing off the lesser-fit being before they procreate in the first place.
actually, if he ends up having 2 kids, it's only 1/2 as much natural selection.
Carmack seems to agree with MS here: "I agree with Microsoft’s assessment that WebGL is a severe security risk. The gfx driver culture is not the culture of security." http://twitter.com/#!/ID_AA_Carmack/status/81732190949486592
This is simple to answer. It's a form of smugness and self-promotion, to publicly wish a famous stranger well on an online forum makes yourself look well-aligned and better compared to other folks. If not that, then it's likely a comment in anticipation of poor taste of future posts (based on past slashdot posts).
This has implications for exercise, no? Would you get stronger if you wear this all day, growing more bone and muscles than necessary?
Arnold Schwarzenegger gets the honor of being the top male that's categorized a female with 100% confidence, followed by Colin Firth http://rtw.ml.cmu.edu/rtw/kbbrowser/pred:female