Computer Science

9:54 AM

@Elton Hey Elton

10:55 AM

@FrankW just introduced mathematical-foundations. I don't know whether it's a good tag; what qualifies as foundation(al) is probably too subjective. Should we have a meta discussion?

If we decide to keep it, some retagging is probably in order.

For instance, most asymptotics questions are, arguably, about mathematical foundations.

FrankW

11:16 AM

I was mostly thinking about the "what kind/level of math do I need for X" type of question, that comes up from time to time.

Given that most of these questions are closed as duplicate nowadays, the main use of the tag will be that potential askers or (more likely) close voters can find these questions more easily.

Having all asymptotics questions tagged with the new tag would mostly subvert that use, though.

Raphael

12:50 PM

@FrankW I'd rather close all these questions; they are pretty pointless (since either easy to look up or very subjective).

So you want to use it as a meta tag?

I'd prefer study-recommendation or something then, but then we'd really have to have the discussion about meta tags and whether such questions are at all appropriated.

FrankW

@Raphael I don't want to use it as a meta tag.

Arguably, quite a lot of the questions here could fit under the tag -- much of asymptotics, graph-theory, discrete-mathematics, probability-theory, ...

Raphael

@FrankW Then you'd have to tag all questions about mathematical foundations, according to the tag description. (It does not seem to say what you want?)

@FrankW My point exactly. I don't think a tag for that is useful.

FrankW

The situation I saw with that question was that none of those tags really fit, since it wasn't abaout a specific area of maths.

Using all or an arbitrary subset of the tags didn't feel right.

Raphael

@FrankW Imho, that's a indicator of "not a good question for cs.SE".

FrankW

My initial gut feeling was that there could be other types of questions with the same problem, but thinking more about it I'm not able to come up with one.

Raphael

1:04 PM

We have too many of similar questions already. We probably should close most; I think many remained open with a "maybe they'll improve it" reasoning.

FrankW

And I agree, that these questions are at the border of on-topicity and the number of variations that could come up and not be duplicates is quite small.

Gilles

1:35 PM

@FrankW do we need a tag for that? This kind of question sounds like it should be tagged x

Raphael

@Gilles x? Is that meta-slang for "burn it with fire"?

Gilles

@Raphael no, "what kind/level of math do I need for X" → tag X

the tag formatting forces letters to lowercase, which here was indeed confusing

Raphael

@Gilles Ah, I see. I did not catch the reference to Frank's X above. :)

FrankW

@Gilles We don't need one. Initially, I thuoght it would be useful. I am not as convinced anymore. -> I wouldn't mind if it is deleted again.

Raphael

What do you think of this type of question in general, Gilles?

To me, it tastes like book recommendations or software recommendations, a type of question usually outlawed on SE.

@FrankW If it's not used, it'll go away on its own.

Gilles

1:51 PM

depends: “what's the best book” → no. “What math topics are prerequisites” → ok. “What can I use as study material” → depends

Raphael

While we are talking tags, before I forget: I suggest we blacklist [tag:java] and [tag:c++] -- I have not seen it used on an ontopic question in ages, and blacklisting may discourage folks to post their questions in the first page.

I remember I proposed this once but we decided not to do it because there may be ontopic questions (e.g. language design). I still think that such questions should not need the PL's name as tag. The only set of questions I have in mind that is valid are a couple of Prolog questions; they could do with only [tag:logical-programming] and without [tag:prolog], I gu…

Gilles

and I still disagree, there are on-topic questions with these tags

Raphael

@Gilles Regarding item #2: we see hugely differing opinions among teaching members of the site. You can give any CS course with any amount of math prior knowledge between "arithmetic with fractions" and "BSc in Maths".

Gilles

preventing people from using the tags will just cause them to pick different tags

Raphael

@Gilles Why are these tags elementary? I'd think that any question that is OT on Computer Science should apply to multiple languages and/or whole paradigms. Anything else is too specific and unlikely to have scientific reasons, thus OT. (Maybe I lack fantasy; can you link one?)

@Gilles That may happen, true.

Gilles

1:56 PM

2

Q: Java, strong typing, covariance and contra-variance

While studying for a test in my OOP course, I came upon this question which had an answer I didn't really understand. The question is as follows (translated): The programming language "Sava" is similar to Java in every way except in the following: It's structurally typed When overriding a met...

3

Q: Is the type inference here really complicated?

There's a question on SO asking why in Java the right type doesn't get picked in a concrete case. I know that Java can't do it in such "complicated" cases, but I'm asking myself WHY? The (for simplicity slightly modified) line failing to compile is Map<String, Number> m = ImmutableMap.builder()...

programming-languages typing java type-inference

Raphael

@Gilles Good question, but does not need the Java tag (imho). It should have the OOP tag, though.

Gilles

2

Q: Contracts for Java Bytecode

Introduction For a paper I need contracts, which are also referred to as Design by Contract (DbC)1, and conceptually go back to Hoare[2]. For my work I need to apply contracts to Java bytecode. The question is only about using contracts on Java bytecode, nothing more. If you ask, why do you want...

programming-languages formal-methods object-oriented java

7

Q: Inheritance, and dynamic access to members/attributes and methods in Java-like languages

I have a question about inheritance in Java-like OO programming languages. It came up in my compiler class, when I explained how to compile methods and their invocation. I was using Java as example source language to compile. Now consider this Java program. class A { public int x = 0; void...

programming-languages semantics typing object-oriented java

and I think that's the lot, I've retagged or closed the others

now C++ we can probably kill

it's the difference between a language designed with a small injection of science, and a language designed by throwing everything at it and applying glue if it didn't stick

Raphael

@Gilles Here, too, the tag seems unnecessary. The OP even talks about other languages.

@Gilles For the sake of consistency, we should then kill all of them.

Gilles

@Raphael but it's contrasting Java with other languages

@Raphael I'm going through that list now

Raphael

@Gilles I appreciate any question about language design and concrete examples probably enrich such questions. I just think that having specific PLs as tags encourages people to think that questions about Java, C++, ... are fine. We mean "questions about principles of programming languages, illustrated here at the example of [Java]".

@Gilles So programming-languages and object-oriented, plus semantics, type-theory or other more specific tags as appropriate.

Gilles

2:07 PM

3

Q: How could I implement, in C++, Mathematica's capability of rationalizing machine reals?

If I have a variable x bound to a machine precision real in Mathematica, I can use y = FromDigits[RealDigits[x]] then y is bound to a completely equivalent rational number which has infinite precision. What are the basic principles behind FromDigits and RealDigits How could I impleme...

c++

^^^^ What do we do about this question?

It should have been migrated to SO, not here

it's too old to remigrate

@Raphael in these cases, java is an appropriate more specific tag

I'll edit that question to be about doing without arbitrary-precision reals

2

Q: Finding amortized time complexity

So I wrote a function push_back for a vector class, and I'm now trying to figure out the amortized time complexity. I'm pretty new to the theoretical side of programming, so if someone could walk me through it that would be amazing. This is my function: void Vector<T>::push_back(const T &e) { ...

time-complexity amortized-analysis c++

^^^^ I've closed this one. If someone knows how to edit it to not require knowledge of C++, please edit so we can reopen.

Raphael

@Gilles There's a fine, general question there, I think, using Mathematica and C++ only as examples. (If the specifics of Mathematica are relevant, the question should have moved to Mathematica.)

Apparently not, since it came from there. :D (need to read more before I type :>)

The question is really about number representation, not C++.

I must have missed that one back then. :/

@Gilles I think your closing reason is wrong. It's a language-independent question, but it's a dump, too.

(cf the comments)

They never came back, so I don't think we need to spend time on that one.

Anyway, this is one thing we can and probably should involve the larger community in by ways of a Meta post, shouldn't we?

Gilles

@Raphael sure

didn't we go through this already? or only in chat?

I'm looking at c now. 3 questions, 2 deserve the tag

and one deserves a better answer

Raphael

6

Q: Why are there tags for programming languages?

One of the defining factors about this SE is that it's NOT about programming. I recently edited a question which, I was surprised to see, the user had tagged as "C++", and had asked for an implementation. I'm wondering, is there a specific reason we have tags for programming languages? It seems ...

discussion tags

I would not call this a definitive treatment of the issue. ;)

Back then, two moderators basically saying "it's fine" probably killed the discussion before it began.

Gilles

2:24 PM

I'm surprised that we don't have an ml or ocaml tag, but I think we only have my two seeding questions and I didn't use either of these tags. Should I?

1

Q: Is procedural programming weaker than OOP in practice?

I have studied the book "C++ Demystified". Now I have started to read "Object-Oriented Programming in Turbo C++ first edition (1st edition)" by Robert Lafore. I do not have any knowledge of programmin which is beyond these books. This book might be outdated because it's 20 years old. I do have th...

reference-request books object-oriented c c++

^^^^ just been edited, what do you think of the edit?

Raphael

@Gilles Imho, no. I think we have a number of functional-programming questions and I'd prefer them to be general(ly tagged).

@Gilles I think it's a bad/misguided question (opinionated). There may me good answers along the line of "There is no better or worse here, both apply well to different settings."

I don't think it needs the C(++) tags.

Gilles

2:44 PM

@Raphael functional-programming is anecdotal on these questions. FP isn't about typing: Scheme is functional but not typed

@Raphael that's what I think too, but should we reopen the question in the hope that someone answers that (as opposed to the predictable religious answers)?

(origin unknown)

Raphael

@Gilles XD

@Gilles Imho, we can and should demand that the OP formulates a resonably scoped, factually answerable question. Such as: "In which scenarios is OOP strictly better than PP, and in which the other way round?"

Gilles

ah, it was from sec.se and isn't deleted yet

3

A: Where should someone start if they want to become a 1337 H4x0r? {serious.(true)}

First, install several subdomains to anonymize your DNS and wrap your IP with a cat5 SATA. Next, load up your ISP with the MapReduce -n flag and nettrace your target to obtain their octagonal megahertz address. Plug this address into your latest hardware compiler (probably either ELF or ASM). No...

vzn

3:12 PM

2 hours ago, by Raphael

@FrankW Imho, that's a indicator of "not a good question for cs.SE".

sigh why are mods so fixated on tags? do you guys actually understand the concept of a "folksonomy"? do you think that all valid tags are already "discovered/used"?

think the tags need some room to "breathe" & discover new uses. killing em off instantly as they are created, seems just micromanaging

esp if they have some basic relationship to CS...

its more up to users to invent/apply correct tags

Gilles

@vzn tags that are the product of one person's imagination are useless

they're only useful if most people agree on the meaning

which for PL tags is the case

vzn

ppl "agree on meaning" by using the tags.

that "agreement" cant happen if they are killed off immed before any reuse.

it seems there is little harm with allowing some openness here unless there is something obviously "off" ie basically non cs related.

alas have pointed this out many times but seems losing battle :(

defn not only se where there is noticeable mod fixation on tags...

vzn

3:44 PM

mods are quite rightly highly focused on scope but it would seem that, despite superficial appearances, tags are not nec a great way to manage scope...

also esp if a se is in beta, it is presumably still "discovering" tags. sure tags tend to eventually "crystallize" but later in the se lifetime.

major retagging is great for consistency but not really supported that well by the se interface (because all retagged questions, even by mods, go to queue front)

3 hours ago, by Raphael

We have too many of similar questions already. We probably should close most; I think many remained open with a "maybe they'll improve it" reasoning.

this seems like an expressed inclination to close (old?) questions that are borderline that dont have answers.

3 hours ago, by FrankW

My initial gut feeling was that there could be other types of questions with the same problem, but thinking more about it I'm not able to come up with one.

its not up to a user to figure out subsequent uses of a tag. only the current use.

tags have to be introduced by user its part of their nature

←earned new tag privilege but never uses it due to se-wide rigidity/constriction on that

2 hours ago, by Raphael

@Gilles I appreciate any question about language design and concrete examples probably enrich such questions. I just think that having specific PLs as tags encourages people to think that questions about Java, C++, ... are fine. We mean "questions about principles of programming languages, illustrated here at the example of [Java]".

seems that issue can be handled with good tag descriptions that clarify correct/incorrect uses!

FrankW

@vzn It could, if new users would typically read such descriptions. Experience tells, they don't.

vzn

3:59 PM

so then are we "dumbing down" the site so it is more amenable to the inattn of new users?

Gilles

@vzn no: tags are a construction of the community, they need to be the result of an agreement, or at least (and this is the common case) to have a reasonable expectation of reaching an agreement

vzn

there are many other issues where new users are just dinged if they dont follow conventions correctly

Gilles

@vzn The site is not the playground of one user. We're a community. Please keep this in mind.

vzn

the "agreement" is the labelling that goes on with initial posts. its very simple.

?!? and who treats the site as a playground? what a straw man

think that relabelling incorrectly labelled questions is somewhat rare. that is the activity that should increase if ppl think labels are incorrectly applied.

it teaches the user also.

if there is a lot of relabelling over certain incorrectly applied tags then thats the indication maybe they are not helpful.

Raphael

@Gilles Amen.

vzn

4:03 PM

the basic concept of folksonomies is not so much that they are (top-down) "arranged" or "organized" its more that they (bottom-up) emerge....

eg twitter hashtags.

Raphael

Iirc, we have had quite a number of discussions about tagging issues, few of which resulted in actions taken.

FrankW

@vzn BTW: What exactly are you fighting against? The number of removed tags today is exactly 0 AFAICS.

vzn

FW did you notice? was supporting your position before you immed caved

tired of the rather high attn to tags. they are not such a big deal.

the ratio of discussion to use seems unbalanced sometimes.

re "playground", se is not intended as a playground for mods, either, although se seems to encourage that pov at times...

"quite a number of discussions few of which resulted in actions taken" hmmm

look how simple introduction of a single new tag has generated literally dozens of lines of comments & sig mod attn/focus. myself incl, ugh

overreaction?

alas not at all uncommon on se which has many hammers to smash flies

Juho

@vzn Well, at least it's better to deal with things already before they become a problem :-)

vzn

right, and some things left alone never become problems

Raphael

4:17 PM

@FrankW We have three tags users can't enter: cs, homework and untagged.

Juho

sure, but I also appreciate the experience some people have with tags. So I appreciate their opinion on what tags are bad

vzn

"bad tags" = "ghettoized"

Raphael

Blacklisting a tag is a very invasive action and we don't take such lightly.

vzn

"burn with fire" lol... sounds like device for pyromaniacs

Raphael

In all but the most clear settings (say, some users would keep using tags that violate norms in every society) that means community involvement.

Juho

4:19 PM

I personally don't use tags much. The volume of new questions is so small it's easy to check all that seem interesting and worth reading more closely

Maybe they are more useful on sites with a lot of traffic, like SO

vzn

the "community involvement" is ideally generally focused use of the tags, and retagging.

exactly they make more sense on high volume sites & low volume site its not as big a deal

is there any se policy statements on tagging?

have seen very little.

so there is so much opinion/attitude on them vs actual official policy.

Juho

even so, I see nothing wrong with making sure we have good and meaningful tags before we reach that high volume (if ever)

vzn

its almost like theres so much freedom there it leaves ppl confused & grasping for more structure.

its a tabula rasa

"good/meaningful" is determined by significant/widespread use. the tags that are "good/meaningful" get the most use. tags that are not as important have lower use.

this happens naturally as users strive independently to look for the best tags.

btw se is attempting to find AI-like auto-propose-tagging methods... its under development...

Juho

users don't always know how to tag their questions appropriately. also, some tags (or tags that we used to have) add next to no information. I don't think a tag that can apply to pretty much any question is good nor meaningful

vzn

←falling into trap/la brea tar pit himself with every sentence typed on this subj, feeling like a struggling dinosaur aieee

Raphael

4:26 PM

For reference, I explained here why I think tags and proper tagging are important for the site.

One big thing is that it might be possible for most to filter by hand now, but you will want old questions to be tagged properly in the future, once volume and backlist grow.

5

A: A guide to moderating Computer Science Stack Exchange yourself — editing

Tagging Tagging is an important feature of Stack Exchange for several reasons. Tags help searching. Tags are used to select "Related" questions for the site bar. Tags help disambiguate titles in question lists. Tags allow users to filter what they see. The last point is probably the most rel...

vzn

←favorite tag p-vs-np

Raphael

@Juho Do you have any in mind?

vzn

an interesting use of using tags. they can be cited as urls....

let me dig that up

Juho

@Raphael I'm recalling the soft-question tag at least...

vzn

just put it on Programming Puzzles & Code Golf

fyi further inquiry there are P vs NP tags on cs.se & tcs.se. see also how not to solve P=?NP cs.se and math monster/many P vs NP refs/stackexchange questions — vzn 2 days ago

FrankW

4:30 PM

@vzn From my POV you were not supporting my position. I suggested a new tag and explained (part of) my reasoning; people explained, why they felt it wasn't as useful as I initially thought; in parallel I did some research and noted that my other reasons were not as valid as I initially thought. At this point my position was "I neither see a big advantage nor a diadvantage with it."

vzn

just feel like theres so much overthinking of this by mods. its a job for users. ie mods should be able to relax somewhat about the whole tagging mechanism. just an idea!

maybe they dont want to relax :p

←likes tags. likes freedom too

FrankW

Possibly useful bit of information: This discussion started after I asked Raphael (outside of chat) for feedback on the new tag.

vzn

the site quality is correlated with overall users effort. if mods are always attempting to rescue quality of the site by fixing user actions/choice (eg wrt tags etc)... hmm.... better to educate the users somehow via communication mechanisms....

hey. just found this link re science (anyone remember that?)

Billionaires With Big Ideas Are Privatizing American Science NYT

this is showing up in CS eg Claymath awards & Simons institute etc

...

FW skimmed chat session. you basically looked around for a tag you wanted, didnt find it, proposed a new one. you had that privilege. then chged your mind quickly as mod R. pushed against it.

there is no right answer here but from my pov it looked like a user having some initiative, basically following the right idea, & then backtracking

ah! so a new tag maybe being born goes away not with a bang but a whimper

then many ensuing dozens of lines of analysis paralysis with yours truly caught up in it

← slowly going crazy 1 line at a time on se & chat rooms am soon gonna have to look into 12 step programs

← believes in delegation/empowerment

← maybe will create new fake tag that looks real and attempt to create long involved chat discussion on subj... j/k

ps re tag descriptions, how many major tags dont have tag descriptions? isnt that a/the basic tag communication mechanism? hmmm

FrankW

5:01 PM

@vzn Why don't you go write some? Might be more productive than filling chat with rants about the mods suppressing everyone.

2

vzn

5:14 PM

this is not writing?

thanks for your great suggestion/advice on engaging with stuff thats productive

yeah youll really fit in around here!

FrankW

@vzn It is not writing tag descriptions.

vzn

FW you showed some brief )( signs of independent thinking... guess made mistake on encouraging/supporting that!

as usual!

FrankW

Why? Because I came to a similar conclusion as most other people?

vzn

whatever

FrankW

"If one person calls you a horse, ignore them. If many people call you a horse, check if you have a saddle on your back."

vzn

5:21 PM

"& if it seems there are many horses being ridden, then you are on se"

...

bigger fish to fry on here: R. himself notes there seems to be low voting....

have noticed that somewhat

R. also hints that higher rep users might not be voting much. see that somewhat

so yeah we've certainly beat a dead horse re tagging. but voting is a big deal on se. now theres something significant to think about...

Raphael

7:25 PM

@Juho Ah, okay. That one we don't use. I thought you had some in mind we actively use.

Juho

Oh yeah no. I don't recall any of the current ones being bad

Raphael

@Juho Good, good. What a relief... my mind went all "What? OMG what does he have in mind?" :D

Juho

haha :-)

Raphael

8:56 PM

5

Q: How to prove that a language is context-free?

There are many techniques to prove that a language is not context-free, but how do I prove that a language is context-free? What techniques are there to prove this? Obviously, one way is to exhibit a context-free grammar for the language. Are there any systematic techniques to find a context-f...

formal-languages context-free formal-grammars proof-techniques reference-question

This question could use a well-written answer that exhibits some general techniques of grammar/automaton design, probably with links to examples on the site.

Raphael

9:16 PM

@Gilles Regarding our earlier discussion: a question of the form "Is the syntax of C completely specified by a CFG?" (you edited a similar one) does not need the c tag. From a CS view point, C is just one programming/formal language, why should it have its own tag? It would be tagged just fine with, say, programming-languages, formal-grammars, context-free and syntax.

Especially since good answers probably answer the same questions for many other PLs, too.

vzn

11:12 PM

@Realz hey this was written partly with you in mind :)

0

Q: boolean circuits/expression induction via visualization

this is a proposal, hopefully first in a series (depending on reception), that will explore boolean circuits, empirical research, visualization, reverse engineering, and advanced (T)CS, and possibly even have some real research value. it is different that other proposals on this site in distinct ...

code-challenge graphical-output combinatorics number-theory

hope something turns up

vzn

11:46 PM

ugh closed in 30m... they have no clue :(

FGITW

Transcript for