Massassi Forums Logo

This is the static archive of the Massassi Forums. The forums are closed indefinitely. Thanks for all the memories!

You can also download Super Old Archived Message Boards from when Massassi first started.

"View" counts are as of the day the forums were archived, and will no longer increase.

☖ Forums → Discussion Forum → Computer Science and Math and Stuff
1234567891011121314151617181920212223242526272829303132333435
Computer Science and Math and Stuff
2018-05-31, 9:49 PM #961
Originally posted by Jon`C:
It's a language that lets you monkey patch the global namespace, of course it has a built-in eval function.


I'm trying to figure out if this is a knock at Lua or a spurious assumption that by pointing this fact out in passing I meant to express surprise at this state of affairs.
2018-05-31, 10:00 PM #962
Originally posted by Reid:
From everything I've personally seen in ESO and from WoW: if you're going to support scripting, be damn well sure you understand how the language you're implementing works.


Definitely this.

If you're going to have people copying and pasting untrusted code, there's no way the developers shouldn't have audited their use of Lua.
2018-06-01, 1:43 PM #963
Maybe I'm wrong and weird but I hate languages without types. I know what kind of type I want to deal with, and how to properly handle these types in the sorts of applications I write. I don't see the benefit of not having types.
2018-06-01, 2:10 PM #964
Strong static typing supremacy
2018-06-01, 2:36 PM #965
I've honestly never heard a good argument for weak or dynamic typing. The TL;DR of it is "less typing" - er, keyboard typing, I mean. But then you'd have to write many assertions and unit tests in order to enforce your assumptions about the data you're passing around.

I wonder how many Python projects use systems hungarian?
2018-06-01, 3:38 PM #966
You should always use types if they help. And the "less (keyboard) typing" argument applies just as well to languages that can infer types. I'd rather write Scala than Python or Clojure for example, and I'd always try to avoid joining a large project that didn't use some kind of type system... even it gives me a headache. Better than a headache from unchecked run-time errors or having to write a gazillion tests. Also, can't you usually subvert the type system if you want to?

Maybe the right place to not use types is when you don't know yet what you want your types to be. For example, in Scheme or Racket, you'd probably extend the language by creating a DSL rather than a type. But why not use the best of both worlds and use Scala or Typed Racket? Given a choice I'd rather use any of these than Python, except for trivial and one-off programs.

Rather than procedures (Scheme) or types (Haskell) being fundamental, there's also languages which take Actors as being fundamental, such as Erlang (or Go). In these languages, it may not matter how good your type system would have been, since you the concurrency primitives could be subverting it anyway. For example, channels in Go have been called the new "goto", and in the Scala Actor framework Akka, the type system might not map very well to Akka abstractions either. Tl;dr: programmers can subvert type systems, but languages can too, at which point the existing type system becomes an aberration.

One approach is to compromise between static and dynamic type systems, e.g., gradual typing, such as what Dialyzer offers you in Erlang / Elixer:

Quote:
Static typing versus dynamic typing is an age-old debate amongst computer scientists and programmers, and the fact that we still argue about it suggests that there is no single right answer for all circumstances. But what if we could have the best of both worlds by combining the safety guarantees of static type systems and the freedom and flexibility of dynamic type systems? In this talk, I will present an introduction to an optimistic, gradual type system as implemented by the Dialyzer tool for Erlang and Elixir. I will highlight the differences and trade-offs between static and dynamic typing, and present optimistic, gradual typing as a good compromise.


http://www.codemash.org/session_old/dialyzer-optimistic-type-checking-for-erlang-and-elixir/
2018-06-01, 3:59 PM #967
Lol! The ghost of Vladimir Arnold is at it again, pissing off some Reids over at Hacker News... as of now, every single parent post is a complaint about the phrase, "mathematics is a part of physics".

https://news.ycombinator.com/item?id=17209444
2018-06-01, 4:02 PM #968
based
2018-06-01, 9:15 PM #969
I don't know where to put this, but it looks tasty.

2018-06-01, 11:18 PM #970
Originally posted by Jon`C:
I've honestly never heard a good argument for weak or dynamic typing. The TL;DR of it is "less typing" - er, keyboard typing, I mean. But then you'd have to write many assertions and unit tests in order to enforce your assumptions about the data you're passing around.

I wonder how many Python projects use systems hungarian?

This, I guess it's "less typing" to write, but I find myself often struggling with whatever dynamic typing is doing and find myself unsure of how their type handler is going to deal with my code.

I was going to say I didn't think there was a good reason for it, but you said it for me.
2018-06-02, 12:54 AM #971
Originally posted by Jon`C:
I've been experimenting with a lossless binary texture compression algorithm.


I started experimenting with ways to make this compression "two-dimensional".

I've got nothin'.

I arbitrarily chose the following test function: if at most 1 channel is less than 0.5, the pixel is off. Otherwise, the pixel is on. It's probably not a surprise to more analysis-minded people, but this simple test function can produce some surprisingly complicated shapes. Each of the following are generated from a 2x2 random source texture.

Code: 
                                
                  #             
             ########           
       ################         
#########################       
###########################     
#############################   
############################### 
################################
################################
################################
############################### 
#############################   
###########################     
###########################     

----------------------------------------------------------

         ####                   
#       ####                    
###    #####                    
##### ######                    
###########                     
###########                     
##########                      
############                    
#############                   
##############                  
###############                 
################                
#################               
###############                 
#############                   
###########                     

----------------------------------------------------------
                                
                                
                                
###                             
#####                           
#######                         
#########            #          
##########           #####      
###########         ##########  
############       #############
#############    ###############
#############  #################
################################
################################
################################
################################

----------------------------------------------------------
                                
                                
                                
                                
                                
                                
                                
                #               
                ############    
                ################
##              ################
####            ################
######          ################
#######         ################
#########       ################
##########      ################

----------------------------------------------------------

                    ############
                    ############
                     ###########
                      ##########
                        ########
                            ####
                                
                                
########                     ###
###########                #####
#############            #######
##############          ########
##############         #########
###############       ##########
###############      ###########
###############     ############


Judging by my random sample, this test function looks like a win for a general purpose decoder. The challenge then is encoding source images into these particularly chosen texels. ...and I have absolutely no idea how to do that.
2018-06-02, 3:40 PM #972
The inverse problem is complicated?
2018-06-02, 9:17 PM #973
Originally posted by Reid:
The inverse problem is complicated?


The one-dimensional inverse problem is crazy easy.
2018-06-03, 6:04 PM #974
In light of Microsoft / LinkedIn agreeing to buy Github, I thought Jon`C might appreciate the following `shopped screen-cap:

2018-06-07, 3:34 PM #975
I think I realized why people like dynamically typed programming languages: because the way people think about the world is the way they think using natural language, which is "weakly typed".

Humans invented mathematical notation and rigor to avoid the pathological complexity that comes with trying to make natural language precise; failing to take advantage of the corresponding analog of mathematical rigor in programming languages is going to leave you with gazillions of tests to write, and your code will start to look a little bit like a tax form, as you clumsily try to catch a ballooning number of unanticipated cases.
2018-06-07, 3:46 PM #976
That said, there are those who argue that type systems can be taken too far.
2018-06-07, 4:16 PM #977
“Types are bad for mathematics”
“Mathematics is a part of physics”

your citations need to get their story straight
2018-06-07, 4:31 PM #978
For what it's worth, Lamport was Lambasted for that paper at the time, and in that little blurb he says that he didn't know much about types. But I still think he's a genius / Turing award winner, etc., and at the very least thought provoking.
2018-06-07, 4:38 PM #979
I was tempted to combine those two quotes into "Types are a part of physics", but by the admission of the author of the paper I was about to link to in support of that little aphorism, by "type", he is actually referring to what physicists call units, which makes the statement trivially true.

I think the Pascal language tried (poorly?) to argue this analogy in the reverse direction (though I could be wrong about that).
2018-06-07, 5:43 PM #980
Does this mean that Alphabet will be shutting down their driverless-car program?

[quote=Sundar Pichai]We will continue to develop and apply strong safety and security practices to avoid unintended results that create risks of harm. We will design our AI systems to be appropriately cautious, and seek to develop them in accordance with best practices in AI safety research. In appropriate cases, we will test AI technologies in constrained environments and monitor their operation after deployment.[/quote]

https://blog.google/topics/ai/ai-principles/
2018-06-07, 6:14 PM #981
Originally posted by Reverend Jones:
Does this mean that Alphabet will be shutting down their driverless-car program?


you got my hopes up
I had a blog. It sucked.
2018-06-07, 6:18 PM #982
Although if they did, would handing the maket over to Uber make the roads safer overall? Unless Google then decided to lobby against driverless cars altogether, citing poor safety.

Of course if they did that, it would just be sour grapes that Uber got away with Waymo's IP, wouldn't it? :P
2018-06-07, 6:26 PM #983
I still can’t believe how many simpleminded technophiliac doofs have fallen for Google’s PR about self driving car safety.
2018-06-07, 6:45 PM #984
Every time I see some ducking normie or teenager talking about self driving cars, they talk about how much safer we’ll all be. As though the current ****ty, sparking prototypes are anything close to as safe as Google et al claim they are.

67% of the companies with self driving cars on the road have been killing people and lying about it. They are hundreds of times more dangerous than human driven cars, maybe thousands.

33% of them drive like a ****in glaucomic grandma and are too scared to put them on streets anywhere other than San Francisco Bay Area, the city known both for offering the same physical sensation whether you’re inside or outside, and an insufferable tolerance for bad Google products.

Mark my words, if these tech companies get their way self driving cars will be made safe the same way we made roads safe for normal cars. By making them unsafe for everyone else.
2018-06-07, 7:32 PM #985
Move somewhere you can walk to work crossing only pedestrian bridges.
2018-06-07, 7:34 PM #986
I've seen enough footage of self driving cars doing weird ****ing **** to be a bit scared of them.

All they're trying to achieve too is to get rid of labor.
2018-06-07, 7:37 PM #987
Originally posted by Jon`C:
I still can’t believe how many simpleminded technophiliac doofs have fallen for Google’s PR about self driving car safety.


Can you blame them? The rest of the world is a hellscape, might as well pretend someone's the good guy.
2018-06-07, 7:42 PM #988
Originally posted by Reid:
Can you blame them? The rest of the world is a hellscape, might as well pretend someone's the good guy.


Yes, I can blame them. It’s called pattern recognition. Corporations aren’t your friends.
2018-06-07, 7:44 PM #989
But hey, we live in a timeline where people are befriending waffles on the Internet, sharing waffle jokes with each other like it’s some in joke and chortling about all of the shade one hamburger restaurant throws on another. So maybe corporations are our closest friends after all.
2018-06-07, 8:08 PM #990
Originally posted by Jon`C:
Yes, I can blame them. It’s called pattern recognition. Corporations aren’t your friends.


Well, if you don't have any other friends...
2018-06-07, 8:20 PM #991
I was going to post some stuff about how dangerous Tesla's marketing is regarding their "autopilot", but it's talked about in the Some More News video already. 14m36s
2018-06-08, 2:53 PM #992
I hurt a friendship today. A friend from undergrad told me they were going to study at a school with a DoD grant, and I told them I highly disapprove of accepting military funding. They didn't take it well, and we kinda argued about it.

:/
2018-06-08, 3:09 PM #993
"Military funding" is taxpayer funding. Do you refuse to pay taxes? Do you refuse to use any services or technology originally funded or supported by the military? Have you ever heard of these guys? https://en.wikipedia.org/wiki/United_States_Army_Corps_of_Engineers

Did you know Massassi was founded on "military funding?" I was in the Air Force when this site was created and I used part of my meager salary to pay hosting fees, buy the computer I used to play the game, write tutorials, update the site, learn how program, etc. Are you going to boycott Massassi? :(
2018-06-08, 3:13 PM #994
So Massassi is part of the military industrial complex. I knew it! This is just like the end of Ender's Game! All that time I spent with JED I was actually bombing Kosovo, wasn't I??
former entrepreneur
2018-06-08, 3:17 PM #995
Originally posted by Reid:
I hurt a friendship today. A friend from undergrad told me they were going to study at a school with a DoD grant, and I told them I highly disapprove of accepting military funding. They didn't take it well, and we kinda argued about it.

:/


Sounds like a learning experience.
former entrepreneur
2018-06-08, 3:23 PM #996
Originally posted by Eversor:
So Massassi is part of the military industrial complex. I knew it! This is just like the end of Ender's Game! All that time I spent with JED I was actually bombing Kosovo, wasn't I??


Well, if you dig through old forum/news posts you might find references to me being shipped off during this "conflict." I thought I was going closer to the front lines, they even made me do chem gear training; I was so freaked out. But when I landed in London they redirected me to an RAF base there and I spent 2 months working on computers there, supporting the people who were supporting the planes/pilots that were doing the bombing in Yugoslavia. It's hard to remember whether I actually updated that site during that time or what; I can't remember. I remember having very limited access to phones for overseas calls but can't remember the internet situation very well.

I do remember frying a power supply because I forgot to switch the voltage selector. Oops. Oh, and I got to drive a sweet Land Rover with a spare tire mounted on the hood. And a crazy Mercedes military van. And I got to eat really crappy food.
2018-06-08, 5:17 PM #997
Originally posted by Reid:
I hurt a friendship today. A friend from undergrad told me they were going to study at a school with a DoD grant, and I told them I highly disapprove of accepting military funding. They didn't take it well, and we kinda argued about it.

:/


Unless your friend is literally working on killbots, what does it matter? A lot of the things the military needs research for are logistical. They are just as helpful for civilians but often nobody has the foresight or capital to invest in them.
2018-06-08, 11:20 PM #998
http://blog.counter-strike.net/index.php/expresstrade/

When cease & desist letters are now sent through Facebook..
2018-06-08, 11:36 PM #999
Originally posted by Brian:
"Military funding" is taxpayer funding. Do you refuse to pay taxes? Do you refuse to use any services or technology originally funded or supported by the military? Have you ever heard of these guys? https://en.wikipedia.org/wiki/United_States_Army_Corps_of_Engineers

Did you know Massassi was founded on "military funding?" I was in the Air Force when this site was created and I used part of my meager salary to pay hosting fees, buy the computer I used to play the game, write tutorials, update the site, learn how program, etc. Are you going to boycott Massassi? :(


I don't hold any judgment for people who enlist.

Originally posted by Jon`C:
Unless your friend is literally working on killbots, what does it matter? A lot of the things the military needs research for are logistical. They are just as helpful for civilians but often nobody has the foresight or capital to invest in them.


Their internship is not killbot related but is not logistical. It's intelligence-based.
2018-06-09, 3:47 AM #1000
Quote:
I don't hold any judgment for people who enlist.


Why not?
1234567891011121314151617181920212223242526272829303132333435

↑ Up to the top!