RESTfully deficient
on
Head First Rails
·
· Score: 4, Informative
I started trying to learn Rails with this book, but found the dearth of RESTful development methodology leaves this book almost completely useless. Best practices are important to learn especially when just starting out. I would recommend Simply Rails 2 as a much better starting point for the beginner. I switched over to Simply Rails 2 and it provided a much better foundation upon which to build a working knowledge of Ruby on Rails.
This is very much a positive development, though really the whole issue is eventually going to have to go before the Supreme Court. At least they seem to have been generally pretty decent in their handling of Constitutional and "IP Law" issues the past few terms.
Except for that whole Eldred v. Ashcroft fiasco described in Lessig's Free Culture.
Your ISP told you you need their "installation CD" to connect to the net?
Your ISP is treating you as ignorant prey.
Yes, my ISP (Cox) told me I needed to install Windows software to use their cable modem services. So did my previous ISP (SBC DSL). No, I didn't use it, I don't have any Windows boxes at home, and if I did I still wouldn't use it.
But most ISPs will instruct new customers to use their "Install CDs" that set things up automatically, but usually also install a bunch of useless crap and add the ISP's name and branding to IE. I don't like it, but I would guess that the average customer would use it.
This approach, though, does require a slight modification on the client side. Namely, the server has to be "trusted" within the client's certificate chain. "
Therefore, without user acceptance of the Proxy certificate, no go. Your ISP isn't sniffing your SSL traffic. Your boss MIGHT be.
That's true, I forgot to include that part. All it requires, though, is that the certificate is added to the trusted clients. I'm not sure, but can't an ISP modify its installation CDs to include this as part of a rebranded IE install? If they do, then the majority of users will never know better because they will already have a "trusted" cert from their ISP.
Here is an explaination of how Bluecoat allows businesses to create a deliberate man in the middle so it can block content on SSL encrypted sites. It's a frightening Internet we do business in.
If you use SSL at work in ways designed to elude acceptable-use filters (e.g., WebSense) or to secure applications like telephony and file-sharing, you may want to re-think that proposition.
A series of products, among them Blue Coat's SSL Proxy, provide SSL-cracking capabilities to organizations interested in shutting down SSL violations of policy.
In effect, Blue Coat's SSL Proxy breaks any SSL traffic its been configured to intercept.
When a connection request is made by the browser, it passes through the Blue Coat proxy on its way to the real SSL server. The response from the destination SSL server includes a certificate. This certificate is designed to (a) irrefutably identify the server; and (b) secure the communications between client and server. To do so, the cert wraps the server's public-key, which is tied to the domain name (or, less likely, IP address) of the server.
The real server's cert, though, is intercepted by the proxy on its way back to the browser.
Before the proxy passes the certificate through, it unwraps the public key and then re-wraps it in an "emulated certificate" (I'll go ahead and call it a spoofed cert, which I think is more accurate). This spoofed cert is then returned to the client browser. The client thinks everything is on the up-and-up and -- after it verifies the spoofed cert -- it establishes the encrypted tunnel.
The tunnel, though, is now terminated at the proxy server. The proxy itself has established a second tunnel to the real destination SSL server.
The proxy can now inspect the cleartext traffic, block the traffic, or pass it on to other devices for their use (more about this later), and otherwise fiddle with it prior to sending it down the second encrypted tunnel to the real SSL server.
Weird Al grew up in Lynwood, CA but his parents had moved to Fallbrook and that's where they were living until they passed away. This is according to the North County Times
Brought to you by the Department of Redundancy Department...
Detecting "major" quakes - those measuring a magnitude of 7 to 8.9 - which occur frequently is being investigated. NASA's planned extension of the current mission, dubbed GRACE 2, and its enhanced instrumentation should aid in that effort.
However, Han is hopeful that NASA's planned expansion of the current mission, dubbed GRACE 2, and its enhanced instrumentation, might allow the detection of "major" quakes - those measuring a magnitude 7 to 8.9 - which occur frequently.
Perhaps it was an intentional use of Chiasmus with the intention of intensifying or bringing greater attention to an important point.
And what if it turns out that sentience is required for this kind of learning? I'm really not sure if simply knowing a language's grammar structure should count as knowing the language.
The point here isn't to understand the sentence structure, but rather to understand what a person is saying and be able to respond to it appropriately.
On the question of sentience: it may be required, but if it is, so what? That's all part of the plan as it is. Self-aware systems are one of the goals that are related to this line of work. I'm not trivializing the weight of the implications of this research, nor am I assuming that it is an easy task to create such things. Goals aren't always easily reached, and chances are that we will only make it to the top of the mountain while we are aiming for the moon, but at least we'll be closer than when we started.
As far as programming languages go, I hadn't really given it much though. PHP is good though because of its great flexibility.
That project looks like a lot of fun! I would love to help out on it or at least give feedback. You can email the info to shawn [at] walkingtowel [dot] org if you want my input.
Can I assume you would argue that language is primarily learned?
While there are probably some things that are built-in to our brains that are set for language acquisition, I believe that mush of language is primarily learned. Basic traits of language are probably predestined like verbs and nouns and such. But almost ANY idea of what is built in can be disproved by counter examples in languages that have only recently been studied by Western scientists.
As for myself, I'm not really sure which way I lean. I would have to say that while there is a framework of some kind in the mind about what language is, that framework is flexible and can adapt to what it is exposed to. Also, this framework may vary slightly from person to person. These are all questions that we are nowhere close to having answers to.
Sounds to me like using ANN might be better suited to this than an algorithm, no?
You are right again. I was being sloppy with my terminology. Not a good idea for a linguist of any sort! An artificial neural network would be the way to go on something like this. Unfortunately, I don't have any practical experience with this type of research yet. I am really excited to dive in though.
won't your algorithms be (at least partly) based on sentence structure?
Not if I can accomplish what I am really after. I want a system that can learn a language the way a human does without any preconceived ideas about how a language should sound or look like.
I know this is a tall order, but it's still my goal.
...building better algorithms to mimic the learning process won't exactly tell us how it happens - it will just show us one possible way that learning can happen.
That's true. I was looking at it from a practical standpoint (ie. making a robot that speaks and understands you as well as a real person might). If we are looking at the purely scientific idea of figuring out what the brain is doing, then this could have some real potential.
I guess I was a little biased by my field of study and the focus I have been putting on machine learning rather than thinking about actual human learning.
My major is in Computational Linguistics. This sounds like a good idea as far a research goes, but the sad fact is that this will not be enough. We already know a LOT about the developmental stages in which children begin to acquire language and the relationships between the mental dictionary lookup and the rule applying mechanisms that compete with one another to produce the fastest possible production of intelligible sentences. What we don't understand is how it happens. This study will not let us know that.
What would be better is to develop algorithms that try and mimic the learning process we already have observed in native language acquisition and then continue to refine our algorithms until we have perfected that process. We will only know we have it right when you can take those same algorithms, put them to use by exposing it to a different language and have it still learn it right.
I know they have a linux client because I use it. I don't know about Mac, but I know that they provide source code to compile your own. It's not FOSS, but the source is there for compilation purposes.
I signed onto this about five days before the war began. I just hope they get it back up and running again soon. I think critical mass could be reached with all the publicity they've gotten.
Madwifi is the best. I have been able to to use any Atheros based product I've tried, and have recommended it in business settings where linux clients were connecting wirelessly in Baja California. Works beautifully.
Thank you Madwifi project!
Then again I can live with the fact that I have first to click on a link then choose to save the torrent file then go azureus select open torrent, select the torrent and finally be able to start leeching.
You would be suprised how many people would scream bloody murder if they had to do this.
Maybe this would be a problem if that were actually the case. If you have Azureus set up properly you can designate a directory in which Azureus will periodically scan for.torrent files. In addition one can have all.torrents start downloading in a specified folder.
In other words: save a.torrent to ~/torrents and within 60 seconds it has been automatically opened up in Azureus and has begun downloading to ~/torrents/incoming
Slashdot Mirror
I started trying to learn Rails with this book, but found the dearth of RESTful development methodology leaves this book almost completely useless. Best practices are important to learn especially when just starting out. I would recommend Simply Rails 2 as a much better starting point for the beginner. I switched over to Simply Rails 2 and it provided a much better foundation upon which to build a working knowledge of Ruby on Rails.
I am not alone in this assessment. Here is just two people that think the same thing.
This is very much a positive development, though really the whole issue is eventually going to have to go before the Supreme Court. At least they seem to have been generally pretty decent in their handling of Constitutional and "IP Law" issues the past few terms.
Except for that whole Eldred v. Ashcroft fiasco described in Lessig's Free Culture.
How many accidents have occurred because the driver was looking for or trying to use a control incorrectly.
Twenty-seven.
So, if the OS is not free, then will that mean that RMS won't ride in these vehicles? I see a new GNU project on the horizon!
But most ISPs will instruct new customers to use their "Install CDs" that set things up automatically, but usually also install a bunch of useless crap and add the ISP's name and branding to IE. I don't like it, but I would guess that the average customer would use it.
http://directorblue.blogspot.com/2006/07/think-yo
From the site:
I have been telling people for a few years now that when the doctor leaves the room after examining you he is googling for a cure to what ails you.
Most people I told that to thought I was joking.
Weird Al grew up in Lynwood, CA but his parents had moved to Fallbrook and that's where they were living until they passed away. This is according to the North County Times
Perhaps it was an intentional use of Chiasmus with the intention of intensifying or bringing greater attention to an important point.
I would love to see benchmarks using a USB flash drive as a swap partition. Say, 2GB flash drive with 1GB RAM. Why didn't I ever think of trying this?
Note to self: Buy big USB key.
I'm wondering, does anyone working towards self-aware systems consider implementing anything similar to Asimov's 3 Laws?
:)
Or it could be that I'm just naive.
I was thinking about them as I wrote my last response. I think you underestimate AI folks.
And what if it turns out that sentience is required for this kind of learning? I'm really not sure if simply knowing a language's grammar structure should count as knowing the language.
The point here isn't to understand the sentence structure, but rather to understand what a person is saying and be able to respond to it appropriately.
On the question of sentience: it may be required, but if it is, so what? That's all part of the plan as it is. Self-aware systems are one of the goals that are related to this line of work. I'm not trivializing the weight of the implications of this research, nor am I assuming that it is an easy task to create such things. Goals aren't always easily reached, and chances are that we will only make it to the top of the mountain while we are aiming for the moon, but at least we'll be closer than when we started.
Thanks for the information. Looks really interesting.
As far as programming languages go, I hadn't really given it much though. PHP is good though because of its great flexibility.
That project looks like a lot of fun! I would love to help out on it or at least give feedback. You can email the info to shawn [at] walkingtowel [dot] org if you want my input.
Cool stuff you've got going on!
Can I assume you would argue that language is primarily learned?
While there are probably some things that are built-in to our brains that are set for language acquisition, I believe that mush of language is primarily learned. Basic traits of language are probably predestined like verbs and nouns and such. But almost ANY idea of what is built in can be disproved by counter examples in languages that have only recently been studied by Western scientists. As for myself, I'm not really sure which way I lean. I would have to say that while there is a framework of some kind in the mind about what language is, that framework is flexible and can adapt to what it is exposed to. Also, this framework may vary slightly from person to person. These are all questions that we are nowhere close to having answers to.
Sounds to me like using ANN might be better suited to this than an algorithm, no?
You are right again. I was being sloppy with my terminology. Not a good idea for a linguist of any sort! An artificial neural network would be the way to go on something like this. Unfortunately, I don't have any practical experience with this type of research yet. I am really excited to dive in though.
won't your algorithms be (at least partly) based on sentence structure?
Not if I can accomplish what I am really after. I want a system that can learn a language the way a human does without any preconceived ideas about how a language should sound or look like.
I know this is a tall order, but it's still my goal.
...building better algorithms to mimic the learning process won't exactly tell us how it happens - it will just show us one possible way that learning can happen.
That's true. I was looking at it from a practical standpoint (ie. making a robot that speaks and understands you as well as a real person might). If we are looking at the purely scientific idea of figuring out what the brain is doing, then this could have some real potential.
I guess I was a little biased by my field of study and the focus I have been putting on machine learning rather than thinking about actual human learning.
My major is in Computational Linguistics. This sounds like a good idea as far a research goes, but the sad fact is that this will not be enough. We already know a LOT about the developmental stages in which children begin to acquire language and the relationships between the mental dictionary lookup and the rule applying mechanisms that compete with one another to produce the fastest possible production of intelligible sentences. What we don't understand is how it happens. This study will not let us know that.
What would be better is to develop algorithms that try and mimic the learning process we already have observed in native language acquisition and then continue to refine our algorithms until we have perfected that process. We will only know we have it right when you can take those same algorithms, put them to use by exposing it to a different language and have it still learn it right.
Dial-a-song's been broken for a while... I call it every weekend a few times to see if it's working again, and it just rings and rings and rings.
TMBG ROCKS!
I know they have a linux client because I use it. I don't know about Mac, but I know that they provide source code to compile your own. It's not FOSS, but the source is there for compilation purposes.
I signed onto this about five days before the war began. I just hope they get it back up and running again soon. I think critical mass could be reached with all the publicity they've gotten.
This could either be really good or really lame. We shall see.
Madwifi is the best. I have been able to to use any Atheros based product I've tried, and have recommended it in business settings where linux clients were connecting wirelessly in Baja California. Works beautifully. Thank you Madwifi project!
Then again I can live with the fact that I have first to click on a link then choose to save the torrent file then go azureus select open torrent, select the torrent and finally be able to start leeching.
.torrent files. In addition one can have all .torrents start downloading in a specified folder.
.torrent to ~/torrents and within 60 seconds it has been automatically opened up in Azureus and has begun downloading to ~/torrents/incoming
You would be suprised how many people would scream bloody murder if they had to do this.
Maybe this would be a problem if that were actually the case. If you have Azureus set up properly you can designate a directory in which Azureus will periodically scan for
In other words: save a