Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?

Comment: There is a CS dumbing down going on (Score 1) 139

Donald Knuth Worried About the "Dumbing Down" of Computer Science History

Whether CS education is appropriate to all people who do computed-assisted technical work is very irrelevant to me since practical forces in real life simply solve that issue.

The problem I care about is a problem I seen in CS for real. I've met quite a few CS grads who don't know who Knuth, Lamport, Liskov, Hoare Tarjan, o Dijkstra are.

If you (the generic CS grad) do not know who they are, how the hell do you know about basic CS things like routing algorithms, pre and post conditions, data structures, you know, the very basic shit that is supposed to be the bread and butter of CS????

It is ok not to know these things and these people if you are a Computer Engineer, MIS or Network/Telecomm engineer (to a degree dependent on what your job expects from you.)

But if you are Computer Scientist, my God, this is like hiring an Electrical Engineer who doesn't know who Maxwell was. It does not inspire a lot of confidence, does it?

Comment: Re:W3C, please. (Score 2) 191

by luis_a_espinal (#48676921) Attached to: MIT Unifies Web Development In Single, Speedy New Language

Except that both languages and "application architectures" are, so as to speak, both based on usefully constraining the set of valid programs.

Sorry I don't understand what this means. If you design a data schema that can't scale no language selection, amount of clustering, sharding, money or associated BS is going to be of much help... this is just reality.

This is true, but it does not lead to what you are claiming. A data schema (in a very general sense that goes beyond relational schema or XML data schema or whatever) might or might not scale (either by poor design choice, or by design).

But that schema will depend on specific concepts and assumptions that will be best realized with a specific family of technologies (or even a single one.) It would be possible (but very hard and stupid) to try to implement a relational data schema with a document-oriented database. And it would be very painful to implement a flexible document-oriented system using a RDBMS.

A good design of each type of system would achieve most, if not all of the requirements desired for such a system, but they will make a significant number of platform and language support assumptions.

Design and architecture are not some ethereal things up in the clouds; they are meant to be rooted on very specific language and runtime constrains.

Only when machines become smart enough to do the designing will this ever change.

And since that is an undecidable problem, we know that will never occur (not without heuristics and constant human intervention, validation and verification.)

Computers can do a lot on the margins but ultimately if you want scalability and performance in a non-trivial problem space YOU will have to work for it.

Yes, YOU have to work for it... using the appropriate levels of abstraction (be them run-times, frameworks, language features or any combination thereof.) Architecture and design are about constrains, about constraining the number of ways entire systems can be put into place within limited resources. That constrain alone dictates what language and platform features make architecture and design feasible.

What does constraint validation have to do with scalability and concurrency?

I'm not sure what the OP intended by "constrain validation", so I would present my own interpretation. A constrain or set of constrains will indicate how much scalability or concurrency you need. Those constrains then become vital for describing the means for testing, validation and verification (after all, a requirement is only valid when it is testable and verifiable.)

I could architect a large-scale e-commerce site with strict fault-tolerance and consistency requirements. Great. Then, I can, in theory, implement it in assembler, or C... or with a higher-level language and platform.

Similarly, I can architect a distributed operating system. I could implemented in Java or natively compiled BASIC... or I could do it in C/C++.

Proper architectures for each problem will prescribe the constrains of the design (and the means for verification and validation), and each will be better served by specific languages and tools.

Any non-trivial architecture will have a direct dependency to a set of language features. And any set of language features will have a relation of economic feasibility to families of problem domains and related architectures.

Comment: app specific problem =/= app specific solution (Score 3, Interesting) 191

by luis_a_espinal (#48676827) Attached to: MIT Unifies Web Development In Single, Speedy New Language
I typically side with the camp that thinks concurrency and distribution (and other things like security or fault/partition tolerance) are application-specific problems because it is the set of application (or domain) specific requirements dictate how much or how little they require from each capability.

With that said, I disagree with this:

Why shouldn't a language solve the problem of concurrency and distributed applications?

Because this can only be effectively answered by the application?

An application can only effectively address such challenges when using the appropriate levels of abstraction. And by *appropriate* we mean not just appropriate in the level of high (or low) level features, but also in the amount of resources that are required to construct a system with the right synergies between application and supporting (underlying) platforms.

For instance, having an actor model supported as a language feature help application domain developers exploit (or create) the necessary abstractions for concurrency far more economically than using an actor model developed from scratch (or as an add-on framework)... at least for applications whose concurrency requirements are best served with an actor model over more low-level constructs (locks and shared resources)

Or think fault-tolerance. A language that has concepts such as a valves as actual language or run-time features is far more valuable for developing certain classes of fault tolerance systems than languages or runtimes that do not have any (a reason why most systems are not equipped with any means of throttling to cope with partial failures.)

Language does not enable non-trivial problems to scale out... application architecture enables this and concurrency is of the same coin.

Resource-efficient realization of an application architecture into a design and implementation are highly dependent on the language and run-times of choice.

Comment: Re:They're assholes. (Score 1) 327

by luis_a_espinal (#48676785) Attached to: Why Lizard Squad Took Down PSN and Xbox Live On Christmas Day

So they just gave you time to think about your game consumption, and the opportunity to think about the "silent" in silent night.

They didn't *give* shit. They *forced* it upon people without giving them a choice. Anyone who think this was benevolent or positive in any way is an idiot living ideological fallacies as if they were real, positive options. #fileitunderfuckyou

Comment: Re: Simple answer... (Score 5, Insightful) 482

by luis_a_espinal (#48633415) Attached to: Colorado Sued By Neighboring States Over Legal Pot

Hahahahahaha you think tax money goes to pay for water and highways. No. Tax money goes to pay for stuff like this, this and this.

I know that *some* (not all) taxes go to stuff like that. If you are claiming that NO TAXES go ever to public infrastructure, then you are going to have to do better than just pointing at counter examples.

I never claimed that ALL TAXES go to public infrastructure. I claim that taxes PAY for infrastructure. That claim does not says "ALL TAXES go to infrastructure" or that "infrastructure gets funded PROPERLY by ALL TAXES."

As a result, your reply, by logical necessity, is misplaced and inadequate. Unless you can prove anywhere that I've said anything that warrants your reply, you have to admit, if you are honest, that you are simply building a strawman.

Haven't you noticed that America's infrastructure is crumbling?


Now why is that?

Because its maintenance and expansion is not funded properly. This is no proof that taxes never go there. It is certainly not proof of the following statement:

And taxes are good, right? Not like that's stealing or anything.

People shouldn't expect not to be challenged when they post asinine shit like that without a context or at least some thought behind it.

Giving more tax income for the government is no better than giving a crackhead more money.

There is not one government. There is federal government, there is state and local government, and depending on the region, tribal government. Each operates differently, with different levels of efficiency and honesty (or lack thereof) when it comes to collecting taxes (and putting them to good use.)

In this specific context, this thread, taxation is being referred to state and local taxation. It is not accurate to describe taxation and public spending in such over-generalized terms. It is great from the point of rhetoric.

It has been a long time since the US government has made effective use of its money. Besides - all tax revenue is barely enough to cover the INTEREST on the deficit (even at these low low rates) - let alone the deficit. A few hundred million here or there will make zero difference to the ocean of pork.

Here you are properly elaborating a good point (finally). It still does not explain what states are to do with pot legalization, the war on drugs, state rights over their own taxation, their relation on that topic to the federal state, the nature of interstate commerce, free passage of citizens from one state to another to purchase an item and the arbiter role of federal government in such activities.

There are the goddamned subjects of this threat. Alcohol is already taxed with different sale taxes across the states, so logically legalization of pot by a state will imply its taxation by said state.

Inefficiency of (or even corruption during) taxation of an item by a government, be it local, state or federal, does not preclude a government, in particular a state government from exercising that sovereign power. If you oppose a state from taxing pot as a condition for legalization, you are going to have to do better than saying "taxation is bad or badly done."

Comment: Re: Simple answer... (Score 0) 482

by luis_a_espinal (#48633191) Attached to: Colorado Sued By Neighboring States Over Legal Pot

Yeah, because shit like highways, water and land management, law enforcement and public infrastructure is self-sustained, for free, with nothing but bunnies' farts and pixie dust magically coming out of Tinkerbell's ass.

Fuck you, you wetback cunt.

I'd spit in your face if you were in front of me, and you would like it and ask for more.

I bow to your awesome 3rd grade rhetoric.

Comment: WTF happended to "small gubmint and freedom fries" (Score 4, Interesting) 482

by luis_a_espinal (#48633153) Attached to: Colorado Sued By Neighboring States Over Legal Pot

The attorneys general of Nebraska and Oklahoma sued Colorado in the U.S. Supreme Court on Thursday, arguing state-legalized marijuana from Colorado is improperly spilling across state lines

Seriously, wtf. Oklahoma is way up there among the meth'iest states in the Union, and in Nebraska, LEO's report 1 meth lab incident per 200K people (compared to 1 incident per 376K people in Colorado.) Meth is far more dangerous than pot, I would think these two states should get their shit together before trying to drag another state to federal court.

Furthermore, Colorado is doing far better in almost all indicators than these two states. Not because of pot legalization obviously, but because of a variety of reasons (many of them social).

So, Oklahoma and Nebraska, butt off. Get your shit together. Then worry about legal consequences, if any, that you might be experiencing because Coloradoans are baking brownies the type your granny used to eat back in Woodstock (yes, either she did that there or in a barn, get over it.)

Comment: Re: Simple answer... (Score 4, Informative) 482

by luis_a_espinal (#48632913) Attached to: Colorado Sued By Neighboring States Over Legal Pot

And taxes are good, right? Not like that's stealing or anything.

And it all goes to a good cause.

Yeah, because shit like highways, water and land management, law enforcement and public infrastructure is self-sustained, for free, with nothing but bunnies' farts and pixie dust magically coming out of Tinkerbell's ass.

Comment: Re:From a C++ perspective, writing was on the wall (Score 1) 156

by luis_a_espinal (#48616839) Attached to: Dr. Dobb's 38-Year Run Comes To an End

From a C++ perspective, the only lately useful articles are from Andrew Koenig, but how the release of the articles is done has pissed me off so much I removed it from my feeds. His most recent article series, is at part 9: Abstractions for Binary Search. How about write an article that can be released in a single piece and consumed as such. Trying to consume parts of something every few weeks is an ineffective learning tool. There doesn't seem to be any more single articles. The interesting ones are broken up into multiple parts released every week or two. FUCK THAT. Give me an article that I can read, start to finish. Don't make me come back next week. I'm a developer. I'm already being torn six ways to sundown by various issues, I don't need a publication compounding that. Give me single, solitary articles that have all the content in a single page and I'm happy (it also makes the googling easier).

I am of a different opinion. I prefer to see complex topics broken down into segments. Yes, it is sometimes advantageous to have the entire enchilada. But I don't have much of a problem digesting pieces on a weekly basis (even though, like you, I'm pulled in all directions on a daily basis.)

Comment: So long... however... (Score 1) 156

by luis_a_espinal (#48616821) Attached to: Dr. Dobb's 38-Year Run Comes To an End

Dr. Dobb's 38-Year Run Comes To an End

I will miss it. I've been a fan of it since I got into CompSci back in 92. I remember fondly going through its articles. I had a subscription for it (alongside Windows Development Journal and others.) One would learn really nice stuff in these old school magazines. Hell, even catalog-like productions like "PC Shopper" would have great articles on software and hardware.

One thing, however. Couldn't Dr. Dobbs have adopted a model similar to InfoQ (which seems to be doing rather well)? I wish they had (but maybe it wouldn't have been Dr. Dobbs anymore.) Regardless, I will miss Koenig et al articles.

Comment: They Dropped The Ball (Score 3, Insightful) 440

by luis_a_espinal (#48611053) Attached to: Federal Court Nixes Weeks of Warrantless Video Surveillance

Even though police did not have a warrant,

And that deserves a Darwin award. Seriously, couldn't they have gotten one in the first place? I seriously doubt, if they had well documented reasons to believe something was up, that they wouldn't have been able to find one.

This case was in the bag (or would have been in the bag), but authorities dropped the ball. I've been on jury duty, and I've seen this before. Cops drop the technical ball, and we in jury duty have to say "not guilty" even though we know deep in our guts that the guy on the stand did it.

It is annoying, but this is how the law is meant to operate in a civilized country. This just stresses the point that authorities need to do their shit better, all the time.

This login session: $13.76, but for you $11.88.