The Whiteboard

12:12 AM

Pretty much sums up every machine learning effort I've ever heard of. "The requirements were met, but it wasn't actually useful"

user55340

1:44 AM

@JimmyHoffa ha! my rep is palindromic... now, what would be the next highest palindrome... ;-)

user55340

2:08 AM

(and some mod ruined it by paying attention to my flags... thanks mod... ;-)

Yannis Rizos

It wasn't me. I never pay attention to anyone's flags.

Yannis Rizos

3:35 AM

Hello. I didn't downvote your question, but could you please refrain from using juvenile abbreviations and slang like "plz" and "thank u"? You are asking seasoned professionals for free help, you are not texting your buddies. — Yannis 8 mins ago

...and get off my lawn...

Jimmy Hoffa

2:24 PM

How did this get 3 upvotes?

2

Q: How might a simple IDE work? Algorithms/Pseudocode/etc

Background In a recent school software engineering module, I needed to developed a static program analyzer. Given a ficticious programming language source code, I need to analyze it, then the user can query about the program, eg. what statements uses this variable, which modifies, statements aff...

user55340

I only see two... its kind of a good question if it was narrower in scope...

Jimmy Hoffa

@MichaelT it doesn't seem NARQ to you?

Yannis Rizos

@JimmyHoffa " if it was narrower in scope" == NARQ (overly broad).

Jimmy Hoffa

He's asking how IDEs do realtime static analysis, not even giving a specific one but even then it's a question for whoever developed whichever IDE he's interested in

Yannis Rizos

That said, I can see it getting improved with a couple of helpful comments...

user55340

2:31 PM

He's asking how to design an IDE. There is a question. It is also quite broad.

Jimmy Hoffa

I suppose

user55340

Feature request: one additional character in a comment for every 100 rep.

user55340

Btw, @YannisRizos I really do appreciate the care and tending you do of MetaP.SE. Thank you.

Yannis Rizos

@MichaelT np, it's all part of the "job". I'm a bit disappointed with Mike, this kind of behaviour would have been a problem if he'd gotten elected. He's not a random new user, no idea why he's being so childish.

Jimmy Hoffa

@MichaelT Robert Harvey would have wayy too much fun with that.

user55340

2:47 PM

@YannisRizos That concerns me too. Its also the other tidbits I've seen... cleaning up the 'baan' (I don't know, maybe I was channeling some sheep while typing that...) and fixing tags.

user55340

I recall my days of tech support and all you hear are the complaints. One time I got a phone call from a customer and they said "Thank you, everything is working right. You guys probably don't hear that enough." -- and that was the entire call.

2

Jimmy Hoffa

That's awesome. I want to do that now, except that I know the guy on the other side of the phone these days is less likely a fellow like you and more likely somebody who gives bad advice and has no idea what he's talking about

user55340

@JimmyHoffa Yea... when I do those long winded comments and I see that I'm over by three characters I start going through and change two spaces to one after a period, and maybe take out an adjective... I refuse to use txtspeak. It would be nice to get that extra word or two in there.

Jimmy Hoffa

Interesting how tech support changed. Years ago tech support had actual educated technicians on the phones

user55340

@JimmyHoffa The paradox of tech support is that the guy on the other end of the phone is always an idiot. Doesn't matter which side of the phone you are on.

user55340

2:51 PM

@JimmyHoffa Some companies have actually implemented the 'shiboleet' protocol. See xkcd.com/806 (its a bit big to paste in)

Jimmy Hoffa

@MichaelT i WAS JUST TRYING TO GOOGLE THAT XKCD!

your xkcd skills are too fast for me

user55340

namesuppressed.com/support/shibboleet.shtml

Jimmy Hoffa

That's awesome

That protocol needs an RFC

I suppose tech support changed out of necessity though, user bases of technical services have increased so much, call volumes are surelt too high for technical services to maintain enough educated techs to handle every call

user55340

When I worked at SGI umpteen years ago, it was already too much. When my contract was up there, they were implementing another layer. There was the CSR (Customer Service Rep) who entered the information - outsourced (Denver IIRC), then there was the various levels of the TSE.

user55340

Denver had some TSEs to handle the easy problems, others would get routed to Mountain View, and we had the TSEs there. There were 2 levels of TSEs - the front line, and the more difficult escallations. When the TSE's couldn't handle it, it was escalated to the PSEs (Product Support) who were in depth in one aspect (this guy works on backups only).

user55340

3:00 PM

So there were potentially 2 or 3 layers before you got to the escalation layer. And that was in the late 90's.

user55340

With computers that people should know how to use. We didn't have home users calling in on Origin 2000 servers...

Jimmy Hoffa

That's actually interesting because I didn't imagine SGI to have such a large user base to warrant such? SGI mostly had private corporate clients didn't they?

user55340

Every O2 and indy/indigo that was out there... those were meant as desktop machines.

Jimmy Hoffa

Those were marketted to the public?

user55340

The last class I took in college had a professor wheeling around an indy on a cart from classroom to classroom... I showed him how to right click in netscape to save images and use xv to enlarge the image to full screen.

user55340

3:03 PM

SGI O2

The O2 is an entry-level Unix workstation introduced in 1996 by Silicon Graphics, Inc. (SGI) to replace their earlier Indy series. Like the Indy, the O2 used a single MIPS microprocessor and was intended to be used mainly for multimedia. Its larger counterpart was the SGI Octane. The O2 was SGI's last attempt at a low-end workstation. Hardware System architecture Originally known as the "Moosehead" project, the O2 architecture features a proprietary high-bandwidth Unified Memory Architecture (UMA) that connects the various system components. A PCI bus is bridged onto the UMA with one...

user55340

When I left SGI, they were getting NT to run on them.

user55340

Also consider the indigo...

user55340

SGI Indigo

The Indigo, introduced as the IRIS Indigo, was a line of workstation computers developed and manufactured by Silicon Graphics, Inc. (SGI). The Indigo was considered one of the most capable graphics workstations of its era, and was essentially peerless in the realm of hardware-accelerated three-dimensional graphics rendering. For use as a graphics workstation, the Indigo was equipped with a two-dimensional framebuffer or, for use as a 3d-graphics workstation, with the Elan graphics subsystem including one to four Geometry Engines (GEs). The Indigo was a visually pleasing design, base...

user55340

> The Indigo was considered one of the most capable graphics workstations of its era, and was essentially peerless in the realm of hardware-accelerated three-dimensional graphics rendering.

user55340

This was targeted for user desktops rather than platforms in the lab... Photoshop ran really nice on the indigo.

Jimmy Hoffa

3:10 PM

Whatever happened to SGI anyway? They didn't just dissolve did they? They were bought up while they were on their way to dissolution I would guess, I think I heard something about that in fact..

my memory is shoddy though

user55340

@JimmyHoffa SGI spun off MIPS. MIPS lives on. SGI got smaller and smaller. They sold Cray (side bit, I live ~10 miles from Cray HQ), and then got bought by Rackspace.

user55340

The 'smaller and smaller' was an attempt to switch to the intel platform, and then ultimately dumping IRIX for Linux (with an emulator for MIPS to run old apps). SGI market cap in 1995 was $7B, in 2005 it was $120M. In 2006 it said it would run out of cash by the end of the year.

user55340

vizworld.com/tag/sgi-bts/

user55340

The IP part of what SGI was (Graphics Properties Holdings, Inc) lives on as a litigation company that occasionally pokes Apple, RIM, Sony, and Samsung with law suits.

Jimmy Hoffa

3:25 PM

haha fun

Didn't know SGI owned cray

user55340

For awhile. There are a number of people I work with now who are former crayons.

user55340

(read that as cray-on s)

Jimmy Hoffa

cool

Always thought it would be cool to work on one of those things. My mom actually programmed for a Cray at NCAR

user55340

I talked with a woman who worked at UUNet in the old old days (she had a double digit employee ID number). She told me of a day when they took the shell of a cray (no actual computer) and put it in the median strip on the way to the road to the NSA. That joke apparently caused a bit of a "WTF" and a rapid inventory of their platform.

user55340

(at SGI, I once got a phone call from the NSA for tech support... it was kind of freaky when I had the actual name of the person and called asking for them and then got grilled about it... I then decided to leave calls from "Metrocom Incoprorated" located in Arlington VA well alone)

Jimmy Hoffa

3:35 PM

-3

Q: From TIFF to JPEG 2000... Can I use JPEG 2000 now, at 2013?

Some years ago, the general recomendation for record "original scanned images" was "use TIFF". But programmers need "evolution of format" for "evolution of software", so, I need to evolve my system, changing from TIFF to JP2. The question is "NOW (2013) IS THE TIME CHANGE FROM TIFF TO JP2, OR NOT...

image-manipulation

editted, good for reopen now?

I had to skim the article he linked to realize jp2 meant jpeg-2000 and why the conclusion was what it was, support for the format was poor among all software they tested resulting in poor transformations from TIFF and poor usability of the new format in consuming applications like OCR

Jimmy Hoffa

4:18 PM

Title, name, what have you, I would argue it definitely is. Data types do have specific names, this is true such that we programmers get to give them names. I can create a data type named HolyMoly, that would be capitalized as it's the name of that data type, however a variable of that data type would be a holymoly. A data type's name is capitalized as a name/title, a variable is not capitalized as it's a tangible thing. HolyMoly isn't an adjective of a data type any more than dog is an adjective of creature. Canis Lupus is the name of a type of creature, HolyMoly names a type of data. — Jimmy Hoffa 11 mins ago

Am I wrong?

Yannis Rizos

@JimmyHoffa Well... lots of things left to fix there. Editing it now.

user55340

The chart appears to reflect the relationship between java (which uses Boolean and boolean) as a type to that of C++ (where one defines BOOLEAN as a cpp macro).

user55340

> "Implement a variable using the the Boolean data type for the particular programming language that you are using."

Jimmy Hoffa

@YannisRizos I tend to try just enough changes to make a question re-openable, and no more as to avoid changing the authors intent when questions like that are often easy to misconstrue the original intent. I did almost rewrite most of it though

user55340

This is only correct if "Boolean" is not already a data type in the language. In Java, someone following that advice would implement an object rather than a primitive.

user41796

4:26 PM

@JimmyHoffa - that was one begging for an almost complete rewrite. "Would have been nice" if the OP had been more clear in what they were going for and what they had to work with.

Jimmy Hoffa

@GlenH7 yeah, to be fair he's out of his native language by a long shot, and in reality he has a real programming problem which he found great research regarding on his own already, I think the language barrier tripped him up more than anything

Yannis Rizos

@JimmyHoffa Well, I'm mostly thinking about minor fixes, like the misspelling in recommendation. Or at least that was what I first thought, I ended up rewriting it.

user41796

I must admit that I snickered when he said terabytes and "big storage" in the same sentence. But I also remind myself that I used to write for an enterprise grade backup / archive / recovery product. Big is a very subjective term

Jimmy Hoffa

@YannisRizos yeah, I saw that slippery slope which is why I just cut out where I was heh

user55340

I used to work at a storage company... terabytes were small. Peta and Exa were used when talking about our products.

user41796

4:30 PM

The engineer in me wants to know how wide his data path is as that has an impact upon his end decisions

user41796

Peta was "big" back in early 2000's. Not so much anymore

user55340

Worked there from '99 to '10.

Jimmy Hoffa

@GlenH7 probably huge, imagery archival and processing is no trifling matter. When I worked at virtual earth I did some reports on our throughput of processing imagery data and found we were generating something like an exabyte a month

user41796

Oh, the good old days. Can't say I miss those days since I get paid a lot better now. :-)

user55340

Makes me kind of frustrated now when I have to battle to get a few extra gigabytes of storage on the servers.

user41796

4:33 PM

I used to work with a graybeard (who was actually beardless, but I digress). Every time someone complained about storage, he would flip them a quarter and tell them to go buy a few more gigabytes

Jimmy Hoffa

@MichaelT yeah, scale is so relative. Last place I was at which was a horrible joke, the engineers spoke of our large databases which were ~20gb, and they spoke of the terrible performance constraints they had to work with for so much data.. mind boggling to me to hear that

user41796

back when gigabytes were actually a modest measure of "large"

user41796

"here's a quarter, go buy a real processor"

user55340

Jimmy Hoffa

@GlenH7 My thoughts exactly. They were all ex pascal programmers. The job before that we had 6 terabytes of data, largely blob data, that we had optimized to serve high volume public website with inserts updates and queries subsecond across multi-billion row tables. And they complained of how difficult it was to make their 20gb database perform... I couldn't stand that place, the people were so dense.

user41796

4:36 PM

@MichaelT - awesome, apropos dilbert there

user41796

oh wow! I just found the "ignore this user" option within chat. I have a new favorite tool.

Jimmy Hoffa

So @gnat's special power is appropriately times youtube music videos, apparently @MichaelT's special power is with webcomics... My special chat power is something along the lines of "Rant incessantly about Haskell even though we all know it only exists as a delusion in my head"

user41796

@JimmyHoffa - actually, your ranting has indirectly helped me

Jimmy Hoffa

I did say it was a special power didn't I

user55340

If you read the original bit on monads, you will learn, its all in your head.

user41796

4:39 PM

we've got a challenge within our application stack and representing values + units

user41796

screamingly great application of functional programming and monads as I understand them currently

user41796

haven't hit the "final" solution yet, but your ranting exposed some ideas that may make the solution a lot more elegant

user55340

1

A: Is it time to tighten up the quality filter at Programmers?

After analyzing the data, I'm not convinced raising the sensitivity of the current filters will do much good on Programmers: The "quality" checks were designed for sites like Stack Overflow which attract a lot of short, very poorly written questions. Of the examples you gave, only 1 would curre...

Jimmy Hoffa

Sounds plausible, though I'm not certain what you mean by representing values + units, it sounds to me like a state monad though because it's some form of product type

user55340

@JimmyHoffa Monadology from philosophy...

user55340

4:42 PM

Monadology

The Monadology (La Monadologie, 1714) is one of Gottfried Leibniz’s best known works representing his later philosophy. It is a short text which sketches in some 90 paragraphs a metaphysics of simple substances, or monads. Text During his last stay in Vienna from 1712 to September 1714, Leibniz wrote two short texts which were meant as concise expositions of his philosophy. After his death Principes de la Nature et de la Grace fondés en raison, which was intended for prince Eugene of Savoy, appeared in French in the Netherlands. Christian Wolff and collaborators published translations...

Jimmy Hoffa

@GlenH7 This guy has some great blogs, this may have a little relevant information for how to think about the type options available for a problem like you mention james-iry.blogspot.com/2011/05/…

He doesn't write at a terribly high level usually, and when he does he steps up to it bit by bit

user41796

@JimmyHoffa - thanks; I'll check out his articles

user55340

@JimmyHoffa if you wanted to troll (not that its a good thing, but just amusing to think about)... head over to Philos.SE and ask a question about philosophical monads, and then slip something about haskell in there.

Jimmy Hoffa

> Leibniz asserts "[monads] are generated, so to speak, by continual fulgurations of the Divinity", I'm not entirely clear which Divinity I should fulgurate to generate a monad, can't I just return the Divinity to generate a monad?

C. Ross

5:11 PM

mutters personal development goals

Jimmy Hoffa

@C.Ross 8k on SO and 1400 on P.SE not doing it for ya?

C. Ross

@JimmyHoffa apparently not for my employer

I wonder if I could get away with 10k on SO and 2k on P.SE as my goals for the year ...

pretty sure they don't care about my 10k on RPG and my two modships :-)

Jimmy Hoffa

@C.Ross I condone this behaviour

C. Ross

@JimmyHoffa your P.SE rep shames me

MartyIX

Hello :) Does anyone happen to know about a chat room with people who are interested in Prolog (or CSP (Constraint Satisfaction Problem), SICStus Prolog, ...)?

Jimmy Hoffa

5:16 PM

@MartyIX here or CS.SE are your best bets

user55340

The theoretical CS too...

Jimmy Hoffa

actually, I've never been to the chat room on SO, iduno why... they might have people interested as well

MartyIX

CS.SE chat rooms are empty :)

C. Ross

@JimmyHoffa lots of programming language chats there, but mostly the mainstream ones

Jimmy Hoffa

@MichaelT that's a total dead-end though, I poked around their chat once, they just babble complete nonsense once a week, and is empty all other times

user55340

5:17 PM

The trick is to make it not empty.

C. Ross

@MartyIX so get it full

chat room filling is a skill

MartyIX

@C.Ross Well, yes, but I don't think that my presence will attract many good people :)

C. Ross

@MartyIX you'd be surprised

user55340

It isn't an easy one... in some .SE sites, there seems to be a "not in chat" feeling. It is hard to pull people into chat to try to help them in a way that can serve them better than comments.

MartyIX

I can idle there, that's not a problem.

C. Ross

5:19 PM

@MichaelT yeah, takes work and time

user55340

Idle isn't enough - ask. You don't see all the people popping in and out.

C. Ross

@MartyIX just get flagged a couple times, that'll bring all the SE mods

Jimmy Hoffa

@MichaelT theoreticalCS.SE actually specifically doesn't have a chat room because they don't want a general one

or maybe it was deleted due to lack of use, I forget.

user55340

It got frozen.

user55340

chat.stackexchange.com/…

user55340

5:21 PM

chat.stackexchange.com/rooms/info/47/…

C. Ross

just lack of use

@MartyIX Are you on their meta?

MartyIX

@C.Ross No, I didn't think of that. Good idea!

Jimmy Hoffa

@C.Ross what will their meta do?

Won't have info for him about prolog etc

C. Ross

@JimmyHoffa see what the topics are, then propose they open a chat room

Jimmy Hoffa

Ah

gnat

5:52 PM

that's about 1-2 questions a week, correct? Matches my subjective perception of a frequency of shockingly bad questions. I would be interested in giving this a try for 3-4 months. What about answers, are these supposed to run through the same filter as questions? — gnat 33 mins ago

anyone interested in a stricter filter?

Yannis Rizos

6:25 PM

@gnat Well... this guy certainly isn't interested:

you're getting that because some staff somewhere was dumb enough to think that he could make an automatic script that detects quality — Sam I am 2 mins ago

gnat

@YannisRizos for SO questions, score 75 would be really interesting experiment

they gotta have some fun at MSO I think

wonder what the outcome will be

let's wait and see

user20683

@MartyIX It's a pity you're not in Atlanta, the co-chair of the CS department at Georgia State is huge into Prolog.

user41796

6:52 PM

@gnat - you've already heard my vote, but yes we could run with a stricter filter. Can't see how it would hurt much. Especially not based upon the 2nd round of data that Shog9 ran

gnat

@GlenH7 how much higher would you prefer it to be? Based on list in the first run, what value would you prefer to cut it at?

user41796

I wonder how easy it will be to game the quality score. Crap questions could / would potentially still get there if the txtspk and poor grammar are fixed. OTOH, at least I'll be able to clearly see that it's a crap question. And I may be presuming too much of those-who-ask-crap-questions.

gnat

> all questions asked on Programmers in the last year with a quality score less than 100

user41796

73 seemed pretty tight. There were only 2 marginals in there that I saw. I didn't click through and evaluate them.

user41796

I was going to resort based upon score, but that dastardly day job of mine keeps getting in the way

user41796

6:55 PM

do we know where the filter level is at now?

user41796

I'm guessing 65?

gnat

@GlenH7 okay no need to hurry, just ping me here or comment there when you're ready. Filter seems to cut at 64

there are questions with 64 in the list

none with 61, 62, 63 as far as i can tell

one question with 64 to be precise

> I'm looking for some partners.

user41796

@gnat - will do. And I'm still curious to know what the effect would be if we throttle down how quickly new users can ask questions. If the current allowed rate is 1 Q per 20 minutes then I think we should try 1 Q per hour or perhaps even per day.

gnat

funny text :)

user41796

sounds like a dating site question...

user41796

7:00 PM

I'd be fine with jumping to 73 now. 2 out of 88 exceptions is easily handled versus the 88 exceptions that had to be purged by the community

gnat

@GlenH7 throttling at 1 per hour sounds interesting

Corollary - I wonder if it would help to decrease the number of questions new users are allowed and / or to further restrict the rate at which the questions can be asked. — GlenH7 yesterday

user41796

I understood what Robert was getting at (eg. the community handled it) but I also think we could safely ease some of the community's burden too.

thorsten müller

Is there any way we can influence how the score is calculated? Like maybe punishing certain keywords like "career", "job", "book", "tutorial" etc

user41796

I was watching the results of your question to determine whether I should propose that change. It would slow down the trolls.

gnat

@GlenH7 yeah that's why I would want to experiment with higher score

user41796

7:02 PM

ohhhh, that raises an interesting question

gnat

@thorstenmüller that would be a good question to ask @Shog9

user41796

are leslar's and whomever else's questions in those low ranking ones....

gnat

@GlenH7 I doubt these are, and even if they were, bright trolls should be able to hop over any automatic filter I believe, score 73 or 83 or 203 doesn't matter

thorsten müller

because such keywords are more likely the difference between high question score and negative questions than the quality score.

user41796

As Robert is a mod on SO, I could see where he's become less reactive to dealing with high volumes of crap rolling through. And I could see the perspective of "that's just part of the gig...."

user41796

7:04 PM

@gnat - that's a good point. And our trolls are bright. There's definitely a degree of intelligence required to craft the bait that they do.

Yannis Rizos

@GlenH7 I checked for the more recent ones and they don't seem to be in the list.

user41796

@YannisRizos - thanks for doing that. I think that's a good piece of information regarding the anti-troll efforts

user41796

some days I think I should have studied more sociology

gnat

@GlenH7 as far as I can tell, Robert supports 73...

Based on that list, I think setting a threshold of 73 would be worth it. How long does that take, maybe a minute or two? — Robert Harvey 1 hour ago

Yannis Rizos

Well, I haven't spotted that particular troll in months, I'm hoping he got bored and gave up. Left some real gems behind him, like: programmers.stackexchange.com/questions/185208/… (for those with less than 10K, the url slug is pretty accurate)

user41796

7:07 PM

@YannisRizos - that particular question really annoyed the heck out of me.

user41796

ironically, he could have played his gotcha game a little longer if he hadn't dropped that question too.

user41796

That question was one where I had wished I was a mod as an instant suspension and block was warranted. IIRC, we didn't have any mods on right at that moment so it was up for a bit.

Yannis Rizos

@GlenH7 Well, the offensive flags did the trick, it was auto-deleted, no mods involved.

thorsten müller

what did he ask?

user41796

@thorstenmüller - hover your mouse pointer over Yannis' most recent link.

Yannis Rizos

7:12 PM

@thorstenmüller It's the link I posted above. You won't see the question (10K only), but the url slug alone is enough.

user20683

It's a yahoo answers question

thorsten müller

ah ok, but that's cheap. i mean there are at least three keywords that should totally disqualify the question to begin with

user41796

@YannisRizos - the community's use of offensive flags is a good point. Can that actually suspend a user as well?

Yannis Rizos

@GlenH7 No, not in the main site. Yes in chat (but for a very short time).

user41796

I see another meta request... ;-)

user20683

7:17 PM

@GlenH7 that'd be a Meta.SO request

user41796

@WorldEngineer - I'll ask on MP.SE and let it get migrated then. MSO scares me... Talk about whip-saw action there :-)

Yannis Rizos

The downvotes mean "I don't like people named Richard". Or "I don't like people with blue gravatars". Or "Hey, I haven't downvoted in a while, let's see if this still works". Or "I disagree with this. What do you mean which part? ALL OF IT!". Or "Crap, question tagged [humour], I don't even have to read it to know it's bad". Take your pick. It doesn't matter. This is Meta, and Meta means murder. — Yannis 2 hours ago

Yannis Rizos

In other news, someone thought it'd be a good idea to impersonate a SE moderator

...no idea why anyone would do that.

user41796

What I never got is why people who didn't like Richard were still down voting me

Yannis Rizos

user41796

7:26 PM

Meta means your feelings are going to get hurt

user41796

ironically, I got a down vote on SO because someone was upset that I linked to an older MS technology tutorial

user41796

never mind the 75 - 100 odd lines of text around that "oh by the way" reference. It was the link that warranted the down vote

user41796

it made me laugh if nothing else. Especially when I consider some complaints of P.SE being unfriendly

Jimmy Hoffa

@GlenH7 I would like to see a DE.SE query that counts questions on the first day a poster ever asked a question for that poster

I suspect it's very infrequent that someone posts more than 1 Q on their first day posting here

user41796

@JimmyHoffa - for those asking good or just not-crap questions, I would agree. I think it's rare to see a series of good questions at the same time on the same day

user41796

7:36 PM

I was trying to build that sort of a query, but deleted questions don't make it to DE.SE, so that would skew the results

user41796

my conjecture is that those who ask several questions on the first day tend to ask really crappy questions that we'd be safe blocking or otherwise not allowing to occur in the first place

Jimmy Hoffa

something like...
select count(1)
from post p
inner join
(
select min(createddate) [mindate], posterId from post group by posterId
) pd on pd.posterId = p.posterId and p.createddate = pd.mindate
group by pd.posterId

Yannis Rizos

btw I did something truly evil this morning. I had answered a crappy question on History, that was quickly (and rightly) closed. Then, it was brought up on Meta and someone argued that the question was on topic just because it was answered. That's it, no discussion on the merits of the question, if it brings any value to the site or not, just "it got answered" and they re-opened it. So... I deleted my answer and rendered the whole discussion moot. Lost a bit of rep, but it was worth it ;)

Jimmy Hoffa

I just made up those column names, but the semantics should work

just figure out which columns and tables associate with the meanings of my sql and I think it should give you something

user41796

@YannisRizos - props! but are you the one who opened the meta.history question too? ;-)

user41796

7:40 PM

talk about circular arguments then

Yannis Rizos

@GlenH7 No. I just posted an answer. More of a rant really.

Jimmy Hoffa

@YannisRizos nothing evil about that, I would imagine deleting the guys whole question and then start removing from the meta discussion everything that didn't make him sound like he was crazy and imagining made up things

user41796

@JimmyHoffa - it would probably work. But I think the skew from missing deleted questions in DE.SE is significant.

Jimmy Hoffa

@GlenH7 I think those deletions are a small enough portion of the whole to not skew that much...

user41796

by way of example, I know of several users who opened an account; asked three crap questions; then completely disappeared

Jimmy Hoffa

7:42 PM

But were there questions all deleted?

user41796

that could be, and I'll acknowledge my bias because I flagged their questions

user41796

sure, here's an example: programmers.stackexchange.com/users/88473/akshay-bhaiji

Jimmy Hoffa

To be fair, your example proves nothing!

user41796

he's got 3 deleted questions, all at -7 or -6 votes

Yannis Rizos

@JimmyHoffa The OP wasn't involved. In fact the OP didn't mind at all, next day he posted a question on Programmers that was a step up than his History one, he kinda got the point that he needed to do a bit more research without any drama. It was a completely unrelated user that started the Meta discussion, and one of those "let's hold everyone's hands" high reps that argued it should be on topic because it was answered.

user20683

7:44 PM

@YannisRizos I'm fine with holding people hands but you need to on solid ground not falling off a cliff.

user41796

one disadvantage of public forums is that light-hearted zings can be misinterpreted later down the line

Yannis Rizos

@GlenH7 Technically all you know is that the accounts disappeared after asking three crap questions. You don't know anything about the users themselves. Not every new account is a new user.

user41796

@YannisRizos - you're right regarding my lack of proof with that example

user41796

@YannisRizos - I would agree to that too

user20683

@GlenH7 With Gentle Ernesty: you could always start talking like an Elcor.

user41796

7:46 PM

to truly answer that would require mod powers and the ability to investigate IPs + whatever else you have access to

user41796

@WorldEngineer - I had to google that, but that's funny.

user20683

@GlenH7 I think that's kind of the commentary Bioware was going for

user41796

no complaints by me then.

user41796

@YannisRizos - at a broader level, I often wonder about some of the fringe groups of SE users. Not having access to the data to analyze and draw conclusions can be frustrating. But truth be told, I would want to analyze cross-site patterns as well.

Yannis Rizos

@WorldEngineer I think what got me really mad at the History kerfuffle was how people were quick to assume that the OP's feelings were hurt, blah blah blah. Granted, they probably didn't know that the OP had posted a followup (but better) question the very next day on a different site, but still... The guy didn't post comments complaining or anything like that, unless you have evidence to suggest otherwise, why not just assume that he's an adult that can read the close notice and the FAQ on his own?

Jimmy Hoffa

7:52 PM

Check out this late and first answer, we need more users who just show up out of the blue like this guy:

5

A: What's the first language that had the 'Unless' conditional/loop built into itself?

actually I've found a copy of the 1967 BCPL language manual with the UNLESS statement in it on section 6.7 http://www.fh-jena.de/~kleine/history/languages/Richards-BCPL-ReferenceManual.pdf this was also the first language to demonstrate the "Hello World" program BCPL became "B" at Bell Labs ...

Yannis Rizos

Anecdotally we are getting a lot of new users that don't have SO accounts. It's the first time I notice such a significant number of brand new users, and I have absolutely no idea how they find us.

user20683

Yeah

user20683

We show up on SEO pretty easily

user20683

also I pimp the hell out of the site at my school so who knows

user41796

Good to hear about the upticks

Yannis Rizos

7:56 PM

It's not so much that there's an uptick, we've been steadily getting new users. The weird part is that they don't have SO accounts.

user41796

I've heard SO can be pretty unfriendly. Jerks who down-vote just for links they don't like.

user20683

@GlenH7 note my rep on that site

Yannis Rizos

That's not entirely inaccurate in some of the larger tags. However new users wouldn't know that if they don't even have an account there.

user20683

@YannisRizos some users register separately, I've migrated people only to find they did in fact have an account I couldn't see

user41796

@WorldEngineer - you've still got more rep there than me. :-)

user20683

7:59 PM

@GlenH7 and you appear twice in my @ menu... o_O

user41796

@YannisRizos - agreed; definitely a lot more subcultures within SO. I realize I just happened to trip across someone who hates ASP web services. "oh well"

user20683

@GlenH7 send them Goma

Transcript for