I'd remove the top header re: "this is subjective...". We can have subjective questions and its scope defined by the community.

@yoda Because it's a best practices / code review type of subjective question (okay, not as bad as I put it here ...)

@acl asked me about it, and suggested I split it into pieces (based on the requirements section)

We've been very good about that on Stack Overflow, and I don't foresee any craziness immediately

that's why I edited it

@Szabolcs posted here.

my existing code does almost everything I want now. So it's not about "how to do this specific thing"

But to come up with this patterns, it was not at all trivial for me.

If I had seen only a couple of other palettes, made by others, it would have helped I guess

I'm saying your question is perfectly fine. If you're worried about subjectivity, we can try to avoid "best" in the title so as to not give new users an opportunity to mimic with poorer questions

@yoda true. Best should be defined by a specific context.

It is reasonably scoped, focused and specific, and it's the kind of subjective questions we want

I say avoid "best" because tomorrow someone will write "Hey guys, I just figured out pattern matching. What's the best pattern to do X?" or "What's the best pattern you've ever seen?"

I don't see a problem with the question, either, for what it's worth.

@yoda allo. and sorry q:

@yoda exactly. There "best" isn't well defined.

OK. Edited a bit.

It's even more dramatic when you enter @acl. A still icon doesn't do much

it's like watching the Undertaker do a tombstone move in slowmotion

in black&white.

Indeed :-)

I'm off for an hour or two

oh I didn't notice it, let me try

here I am(?)

brilliant animation :)

Just looking at the user list, 6 people above 400. A couple near 500, and if they would get off they're lazy butts, we could have people with edit capability. :) Overall: 23 people with over 200 rep.

@rcollyer a good number of us have rep capped, which is why no one's moving since yesterday

szabolcs missed out on a lot of rep due to that, and I'm guessing Leonid and Vitaliy too

@yoda as I said, lazy bums. ;)

Leonid's score should be 800ish

Well, 9 mortarboards have been awarded, so far. And I can see, Leonid getting Legendary and Epic before too long.

Of course, the problem is compounded by the number of Suffrages and Vox Populi that have been awarded, too.

@yoda Rep cap isn't fun. I wish it was a bit higher for beta.

@MikeBantegui especially private beta where there is a limited number of people.

If I knew about this beta earlier on, I would've had a bit more rep.

I started maybe an hour before the end of the first day IIRC.

@MikeBantegui It started only 3 hours before end of UTC

So you didn't miss much

Short first day.. Odd choice of time to launch

well, UTC can suck at times (like now) and be beneficial at times... for example, I can not turn up for two days (almost) and still have consecutive days counted

of course, there's no "benefit" to that, but I'd say there's no real "benefit" to rep either :)

anyway, don't bother suggesting on meta... will be shot down before you can finish typing it :)

@yoda: Well there is a real benefit when you're starting a private beta and stuff needs editing :)

@rcollyer: Your link in your proposed tag synonyms isn't working.

My Locator is doing weird side-jumps ... can anyone please test? Code is coming:

What weird jumps are you having? Works fine under 8.0.4.0

Not always following the mouse ...

I have to restart my FE I guess

Hm. seems to work fine for me. Does it stop following it or something?

No, it jumps perpendicular to the direction of the drag sometimes

Wait

Can't seem to repro, sorry to hear that.

Okay, some more info. Try this now:

p = {3, 3};
Graphics[Circle[{5, 5}, 5], Epilog -> Dynamic@Locator[Dynamic[p]],
 Axes -> True, GridLines -> {{3}, {3}}]

The Locator is almost "snapper" to the grid lines. I don't know why. The movement is not completely smooth around the grid lines.

I am simply unable to position it at {3.25, 4}

It either gets snapper to {3,4} or to {3.5,4}. It's very weird!

If you could try this for me again, I'd be very thankful!

I don't want to post a question which will resolve as "this is a bug" ... Though I'm still hoping that this isn't.

@Szabolcs I see the same thing. The Locator seems to jump when I get close to the original position.

Hi everyone. I'm going to ask a question about CJK encoding and rendering in the Front End. So before posting it I would like to ask: is that an appropriate question during the private beta period?

@Silvia I would think so. And it is a good question because it would have been off-topic on StackOverflow. (PS Welcome! It is nice to see people who weren't particularly active on SO using the new site).

Thanks @Verbeia . I'll do it:)

Ergh, so frustrated with Wikipedia right now.

MathJax trouble on the site? I see this:

hmm.. a quick question – I have an expression that roughly looks like $\sum_N \sum_N (-1)^N (Nx)^N e^{-Nx} det(A_{N\times N})$

Would it be terribly incorrect to say that the computational complexity is roughly O(N^4)?

I reckon two double sums (O(N^2)) and a det (O(N^2.3ish)) together give something O(N^4ish)

Sound right?

I don't need a precise answer... just a ballpark figure

@rcollyer working on it. :)

@Silvia I admit I have no idea about the answer, though. Maybe someone else does.

@Verbeia I think the default support of CJK characters in Mathematica is not very good, especially for Chinese. There are two Chinese Simplified codepage file: CP936.m and MacintoshChineseSimplified.m, neither of them covered the most commonly used range..

@Szabolcs Happens sometimes, even on math.SE. Just Shift-Refresh. Post on meta if it's persistent.

@yoda Well, you did say "-ish", so I think you're in the ballpark.

@JM I posted on scicomp again, but I can't find an appropriate existing tag. Maybe you can help?

@Szabolcs Wait...

so, nobody can vote? seems I chose the perfect time to start answering questions!

I can vote!

(or is it just Verbeia and myself who've run out of votes?)

oh!

well, go and find nice answers and questions and upvote them, then!

How many votes do we have per day?

@Szabolcs Thirty to forty, depending on whether you voted more questions than answers.

you are also warned 5 votes before the limit that you now have 5, then 4, etc votes left

The thirty is when you're voting mostly for answers. The specifics are somewhere in meta.SO ...

I votes 33 times, just got Suffrage

Well, I was spent in 34 votes.

@JM try voting on more questions

@Verbeia On the next day. :)

(which is in a few hours.)

so @Verbeia, I have responded to your criticism of my transformation function by making it user-defined; so now if it does the wrong thing, it's the user's fault :)

@acl I saw - nice! I spent some time trying to work out a suitable parametric function to do what's needed, along the lines of Erf but I couldn't think of one off the top of my head. Vitaliy's answer nailed it, though.

@all check out our numbers on the Area 51 beta page compared with some other recently started betas. Wow, let's keep it up!

graph is indeed a tad ambiguous...

@verbeia I was doing exactly the same, but then I saw your answer

(I'm about to retag to graph-theory...)

I had been thinking of Erf[x] before but was looking for a similar looking analytic function to make it faster

Tanh[]?

like y=x^3 but then mirrored in the line x==y

As we keep saying in math.SE, there are way too many sigmoidal functions...

yes, graph is ambiguous! but graph-theory sounds like maths, while the question is often about Graph[] ... we need to agree on a tag to use for questions about handling graphs

@acl @Verbeia Which post are you talking about?

@JM yep

@Szabolcs graph-plot?

@Szabolcs Apparently this.

graph-plot is only about plotting, but one question is about importing/exporting graphs

or ArcTan[x]

@JM correct mathematica.stackexchange.com/q/207/8

how about graphs-networks? It would include everything graph-related as well as everything Graph[]-related

@SjoerdCdeVries Yeah, arctangent works splendidly too.

@JM economists never use trigonometric functions, so they just don't spring to mind for me anymore...

@Sjoerd ArcTan is the canonical kink-like function for most physicists (that I know)

@Verbeia Heh. Somebody with an engineering bent would think the error function is a bit of a sledgehammer if all you want is sigmoidal behavior... :)

but I think it's better to allow for the curve to be defined interactively (like I did in my last edit)

@JM @rcollyer @acl @SjoerdCdeVries What do you think about using graphs-networks for everything graph-related and Graph[]-related, may it be about graph theory (maths) or about graph exchange formats? graph is best avoided completely because of all the misunderstandings. I have been following graph on SO, and it's an awful mix of plotting, graph theory & algorithms, and using the Facebook graph API. I don't want that here!

and graph-theory sounds like maths

although from past experience I am guessing I've wasted my time in terms of upvotes :)

@rcollyer hard to say, it depends on how people will interpret it and that depends on their backgrounds. here we have more varied backgrounds than before, so it's hard to tell

@Verbeia how about a Fermi-function (1/(1+exp(x))?

@Szabolcs I meant

@Szabolcs can you forbid a tag from getting entered?

@SjoerdCdeVries Diamond users can blacklist tags if need be.

(e.g. they can ensure mathematica will never be created here...)

@JM Then I can see some merit in this proposal

@Szabolcs I honestly have no opinion on this one. I encounter graph-theory so rarely, accept for Feynman diagrams for which I don't know their graph-theoretic implications.

@SjoerdCdeVries Right now no because now everyone can create tags. But in two weeks that'll not be the case any more, and people who've been here long enough will know what we agreed on. All we need is an agreement, I just want to avoid a mess with these

@SjoerdCdeVries On SciComp, which launched recently, I can't create tags anymore, becuse I don't have enough rep

@acl how so?

An hour and a quarter before a new day...

(Bah, if I woke up only a tad earlier, I could've gotten user number 42...)

Here's the question:

Hi @MikaelÖhman - Mathematica is in private beta at the moment, but if you are interested, please follow the proposal on Area 51 and you will be notified when it goes public (hopefully next week)

Can non-committers join this room, even though they can't see the site yet?

@Szabolcs yes. A couple of people have followed in the last day so they will be notified of the public beta starting

@JM thanks :)

@Szabolcs this room is public. All they need is 20 rep anywhere on SE to talk

@yoda what do you think about the graph tagging question on meta?

boy, I see everyone is busy forgoing rep today for interest tomorrow... seeding all those answers for upvotes when people have slowed down the frenzy

@Szabolcs I agree with all your points. I'll be busy intermittently over the next few days, so won't write an answer, but I gave it an upvote

Also to all: Generally, meta upvotes == agree and downvotes == disagree. Also, generally upvotes==haha, clever and downvotes==you suck.

have others noticed that sometimes when you post a comment -- not here, a comment to questions asked or to answers -- starting with e.g. @Name the @Name gets removed? It seems to be totally random.

So downvoting to disagree with an unpopular or silly suggestion is good

@MikeHoneychurch No, it's not random. It's done when someone @replies to the post owner and no one else (other than you) has commented on the post

It's a highly highly contentious topic and math.se would probably love to talk to you about it over a cup of coffee pitcher of beer.

but it's here to stay. So yes, they'll be stripped off

not sure if I comprehend that. ...and I have had my breakfast and coffee. let me consider it ...(think music)

The reason it's done is because the post owner is always notified, no matter what. So calling out someone when they're the only other person is sorta meaningless

@yoda, ok. got it. :)

@yoda *snicker*

meta.math.se is a treasure trove.

@JM Actually blacklisting a tag requires the intervention of a developer - mods don't have that power

Not that I know of, anyway

also, easier said than done :)

@DavidZaslavsky Ah. I guess I misremembered; there was a time where I saw a mod nuke a tag, but it was a reeeally long time ago.

needing to completely burninate a tag isn't as common as one might think

@JM Mods can merge one tag into another, so it's possible that's what you're remembering

Maybe...

I updated the post about tagging graph-related things