The Whiteboard - 2013-11-07

« first day (1161 days earlier) ← previous day next day → last day (3837 days later) »

5:08 AM

@MichaelT most of the "bad stuff" back log that has been stored in my favorites list is done (was quite a relief to find that out:)

5:57 AM

4

Q: Working with CPU cycles in Gameboy Advance

I am working on an GBA emulator and stuck at implementing CPU cycles. I just know the basic knowledge about it, each instruction of ARM and THUMB mode as each different set of cycles for each instructions. Currently I am simply saying every ARM instructions cost 4 cycles and THUMB instructions c...

game-development cpu process

4 hours later…

9:48 AM

4

Q: How to handle a project manager who is not getting satisfied with his own design/plan?

I am working in a start-up where we have a team that consists of 2 developers, 2 testers, and 1 tech lead, that's me. Database design, UI design, and the workflow are decided and managed by our project manager. Additionally, we have two managing directors, but they don't interfere in the technic...

professionalism software-industry project-management

^^^ Is there a good Programmers question hidden in there?

10:22 AM

@YannisRizos I for one can promise to abstain of close voting if it gets migrated :)

3 hours later…

1:16 PM

Just stumbled upon this little gem...

0

A: Multiple Interfaces in Java - Good or bad

If interface B needs more than which are available in A then it is good,

...this comma in the end drives me nuts, what depths of knowledge (or ignorance) does it hide? These guys at SO have funny habits (as far as I can tell, this is typical draft FGITW "answer", posted two minutes after the question was asked and abandoned by the author after the question has been closed and migrated). Another thing that drives me nuts is how it managed to get an upvote?

2 hours later…

user55340

2:58 PM

@YannisRizos Its got more than a few comments in it that look like its still trying to be specified.

3:30 PM

@jozefg the applicative type construction is perfect for what they're looking for, though you might think about combining validation and type construction with the Either applicative so if validation fails, construction fails, and thus they get the error. I hear what you're saying about separating parsing and validation, but what they were explicitly looking for was validation only so the intermediary structure yields no real value to them

Though even that they probably wouldn't be happy with

what they actually want is to take a BNF grammar and generate a validator based on it

For which a parser generator is likely out there though I don't even pay attention to those enough to know which ones work directly from a BNF grammar

Parser generators are stinky, I avoid them

@jozefg but LISP is an acronym, it stands for Lost In Superfluous Parentheses

user55340

4:09 PM

@JimmyHoffa acronymfinder.com/LISP.html

user55340

And from another site... Large and Incredibly Slow Programs

user55340

4:22 PM

1

Q: What is the abbreviation of JAVA language?

Any one know, What is the abbreviation of JAVA language?

5:10 PM

@MichaelT Just a vile ass

2 hours later…

7:12 PM

@JimmyHoffa Hey, I wrote useful code :O bitbucket.org/jozefg/reified-records The version on Hackage is currently out of date though

@jozefg Have you used Leksah? Just out of curiosity

user55340

Hmm... found an instance of an SE employee following up with a comment in a question after the question was deleted (on Meta.P.SE - I have the question favorited so that I could refer back to it if the user brought it up again). Just one of those "interesting" things.

7:27 PM

@jozefg ok that's pretty far over my head O_O I understand what you're going for but you're using far more haskell libraries than I've any familiarity with. Nice small bit of code though for doing something of that complexity

7:38 PM

About all I gathered from it is that you have the ability to retrieve the names of all the record fields for any given record type, and what looks like a bit of type info as well which I'm guessing is a string description of the lexicographical information

and no QQ or TH needed

Though arguably either of those would have made it easier and more robust (but both of those probably have existing available libraries, yours for folks who don't want QQ or TH which is most people is much cleaner)

7:51 PM

wisdomgroup.com/blog/10000-hours-of-practice

2

Huh, great link to keep on hand. More eloquently gets to the heart than norvig's ten years to learn article which is a great "shutup and do it" piece, but less informative on the details.

@JimmyHoffa that 10,000 hours thing gets repeated a lot and I try to be conscious of that every time I'm around people who are way better than me at something...

user55340

In other bits... glance at the list of languages and groupings used in

user55340

Relational operator

In computer science, a relational operator is a programming language construct or operator that tests or defines some kind of relation between two entities. These include numerical equality (e.g., 5 = 5) and inequalities (e.g., 4 ≥ 3). In programming languages that include a distinct boolean data type in their type system, like Java, these operators return true or false, depending on whether the conditional relationship between the two operands holds or not. In other languages such as C, relational operators return the integers 0 or 1. Some programming languages make a...

user20683

8:09 PM

@JimmyHoffa

user20683

amazon.com/Real-World-OCaml-Functional-programming-ebook/dp/…

@MichaelT Take that perl! Real FP camels have 2 humps !!

(though they both look terribly photoshopped...)

user55340

@JimmyHoffa They're from classic drawings / depictions.

user20683

They've also recently changed the cover style

user55340

> Edie Freedman designed the cover of this book using a 19th-century engraving from the Dover Pictorial Archive. Emma Colby produced the cover layout with Quark™XPress 4.1 using Adobe's ITC Garamond font.

8:14 PM

0

Q: Design Patterns - Why the need for interfaces?

OK. I am learning design patterns. Every time I see someone code an example of a design pattern they use interfaces. Here is an example: http://visualstudiomagazine.com/Articles/2013/06/18/the-facade-pattern-in-net.aspx?Page=1 Can someone explain to me why was the interfaces needed in this exam...

c# design-patterns

I don't even know where to start...

Maybe someone should edit that question to basically ask for a description and detail of what interfaces are, their mechanics, and why/when someone might use them

the question is an X/Y where that's what the guy needs to know but he's asking something else

I could write a lot of an answer but I have work to do and am still fighting this cold for my mental facilities and not totally winning

user20683

I can do it

It might not need an edit, I'm not sure. It might be a good question as it is, he just won't get the answer he needs the way he's asking it

user20683

It needs an edit sheerly because he didn't finish the last sentence.

@WorldEngineer but that part just adds to the mystique

user55340

Its like gnat's trailing comma.

user55340

8:18 PM

7 hours ago, by gnat

0

A: Multiple Interfaces in Java - Good or bad

If interface B needs more than which are available in A then it is good,

user20683

@JimmyHoffa or it's like an interface. It provides a method creating the beginning of a sentence but leaves the creation of the end of that sentence as well as the general implementation up to the individual programmer.

user20683

Also Martok has been elected to office in New York:

user20683

polhudson.lohudblogs.com/2013/11/06/…

user20683

Har Har

8:28 PM

<-- proud like a cat when he shits on the rug.

@jozefg what do you think about my comment of combining parse and validate?

8:46 PM

@JimmyHoffa About once, I didn't like it that much

@JimmyHoffa I was thinking, should that really be grouped in the form Thingy -> Either Error Thingy and then doing some more monadic composition. It'd be awkard to handle errors in validation areas

user55340

@JimmyHoffa Watch sad cat diaries.

user55340

@JimmyHoffa Well reified-records 0.2.0.0 is now on Hackage :)

@jozefg Nice!

@jozefg Yeah, I don't really like it much either, though the guy seems to be asking for it.

@jozefg You think? I should have thought the composition of validatePiece . parsePiece should have been able to be plopped into each applicative spot

shouldn't have been tricky to handle but dunno

You think in those terms much more clearly than I do so you would know better than I how tricky it would be to chain the eithers into an applicative for type construction validation

9:14 PM

Well Either Errror a -> Either Error b can't be composed with applicatives can it? We'd need Either Error (a -> b) which is pretty useless in terms of validation. Since we're dependeing on the result of a previous computation for our new side effects we do need monads

@jozefg no no that's not what i was suggesting

I was saying the validation is on success id, on failure mempty

Either Error a -> Either Error a

it just decorates the parser which is input -> Either Error a

and than you want to compose it with the piece for validation leaving Parser a

Oh yes that is more pleasant :)

I'll edit when I'm near a compiler and can typecheck thing

user55340

but i always get errors. -- all our crystal balls are in the shop for repairs, so we can't magically read your mind to see what your errors are. — Amal Murali 4 hours ago

@jozefg yeah, I can't mentally typecheck what I'm talking about, but conceptually it should work heh (validation during type construction was explicitly what I learned applicatives to do)

Haha fair enough

user55340

9:21 PM

0

Q: When there's no TCO, when to worry about blowing the stack?

Every single time there's a discussion about a new programming language targetting the JVM, there are inevitably people saying things like: "The JVM doesn't support tail-call optimization, so I predict lots of exploding stacks" There are thousands of variations on that theme. Now I know that s...

recursion stack stackoverflow tail-call

user55340

My first read of "TCO" was "Total Cost of Ownership" which made no sense.

applicatives make such a nice API for decorating type construction such as: createPerson age height stupid = Person <$> validateAge age <*> validateHeight height <*> isStupid stupid, which in real-world type applications validation at construction to get specific instruction messages what you did wrong trying to construct something is a requirement, the applicative interface makes it very nice and easy to do

or I might be remembering that syntax wrong, but that's acceptable I'm not so sharp, and my mind is clouded with cough syrup

@jozefg

The O in TCO is "optimization", as a rule optimizations aren't necessary, and when they are it's usually because you did something wrong. That's what @ratchetfreak is saying above. But you're asking a question that isn't relevant, how important is it? Well optimizations are important insofar as your software will run slowly with none of them. If you rely on any single optimization completely you are in trouble, if you however write completely unoptimized, non-optimal code you are then also in trouble. — Jimmy Hoffa 24 secs ago

People totally misunderstand the premature optimization quotes so much, yes optimizations aren't necessary, but non-optimal code with no optimizations will be crappy. Performance is a fine-touch art, like all of engineering it's about tradeoffs.

@JimmyHoffa Yep! What do you think of my answer btw

9:40 PM

@jozefg +1 makes sense, answers his question, though you might try and find an article or two out there that talks about optimizations being trade-offs and how YAGNI towards optimizations is a bad idea just to link because they might balance the guys perspective a bit

@JimmyHoffa Agreed, to google!

user55340

Its not a concern, because you (the programmer) could (should?) just write the code as a loop rather than recursive. Use recursion to frame the proper answer to the problem... but that doesn't mean its the correct implementation of the problem.

user55340

> However, there is a decent reason why we don't yet have TCO. The JVM gives us stack traces. With TCO we systematically eliminate stackframes that we know are "doomed", but the JVM might actually want these later for a stacktrace! Between this and the fact that Java has loops, TCO looks like a bit more trouble than it's worth to the JVM engineers.

user55340

Which is what @Jozefg said.

10:15 PM

@MichaelT what's that "encryption in plain text" word again?

Steganography () is the art and science of encoding hidden messages in such a way that no one, apart from the sender and intended recipient, suspects the existence of the message. It is a form of security through obscurity. The word steganography is of Greek origin and means "concealed writing." It combines the Greek words steganos (στεγανός), meaning "covered or protected," and graphei (γραφή) meaning "writing." The first recorded use of the term was in 1499 by Johannes Trithemius in his Steganographia, a treatise on cryptography and steganography, disguised as a book on magic. Generally...

that's it

user55340

A classic form of that was a etching that had the blades of grass as morse code.

user55340

@JimmyHoffa books.google.com/…

user20683

So apparently the Curiosity Rover's chipset is allowed to Bluescreen at most once in 15 years

11:28 PM

on the point of recursion

its something that you should use very sparingly in production environments

its something that you should never use in limited memory scenarios

and that includes web browsers

user20683

@MattD unless you use tail recursion or are using a language like Haskell or APL.

i've seen recursive JS hit the function invocation limit of various browsers

@WorldEngineer yeah, i basically consider tail recursion to be iterative

user20683

@MattD it effectively is

user20683

@morons I've edited your blog post for grammar and the like. Will post at Midnight tonight UTC.

user20683

that's in about 1.5 hours

11:36 PM

@WorldEngineer its just entertaining when people find out about recursion and then EVERYTHING WILL BE RECURSIVE FOREVER!

@WorldEngineer and then ask you why their stuff fails after a long time....

user20683

@MattD yeah...

« first day (1161 days earlier) ← previous day next day → last day (3837 days later) »

Transcript for

Nov '137

The Whiteboard

General Discussion for softwareengineering.stackexchange.com P...

coffee-day horror-stories monads my-code-is-compiling scotch the-other-monitor

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile