« first day (1478 days earlier)      last day (3435 days later) » 
00:00 - 23:0023:00 - 00:00

12:21 AM
When you have a running shirt, then running is not an adjective because the shirt is not running. On the other hand, if after you finish running you find that now you have a dripping shirt, then dripping is indeed an adjective because the shirt is dripping. Applying this logic to your question will reveal its answer.
 
 
2 hours later…
2:15 AM
 
3:14 AM
 
 
7 hours later…
9:50 AM
@matt Boo! It's Fri!
 
10:21 AM
It is!
 
11:20 AM
I am glad to see we have people in this chat who passionately support a party for which the Westminster cathedral is a mosque.
 
11:41 AM
 
in The Bridge, 6 mins ago, by spyder
LMAO http://developerexcuses.com/
 
> I haven't been able to reproduce that
That's not an excuse. It's actually 100% true 100% of the time.
 
I've just put that on two of my tickets
> I must not have understood what you were asking for
 
That's so American it hurts.
 
I think that accounts for at least 30% of my bugs
 
11:52 AM
No normal person outside of the US would put the blame on themselves.
 
yeah it'd be more like "that's not what you asked for"
 
Yup.
 
> We spent three months debugging it because we only had one month to build it
I wonder if we'd had 3 months to build it then we'd take 9 months to debug it
 
12:15 PM
If you build it, they will debug.
 
12:45 PM
@MattЭллен There's never enough time to do it right, but there's always enough time to do it over.
 
Jez
GAHHH
i hate TFS so much
stupid checkout mechanism. so when i go to my "pending changes" window, i see... a bunch of files with no pending changes, because I changed them back.
 
user116848
1:44 PM
yello folks!
 
2:17 PM
How many years can a laptop last if you take care of it?
 
3:33 PM
I find that I don’t know what data structure to use to represent text with foo-ML tags that overlap, like "<p>this <i>is tough <b>to</i> say</b> and harder to represent</p>".
 
@JasperLoy There are working laptops that are decades old.
 
The <i> part is "is tough to", but the <b> part is "to say", and those overlap.
 
@tchrist That's not valid in some *ML dialects. which ones are you using?
 
So I can’t just use a simple array of tag names and what they hold.
@Mr.ShinyandNew安宇 This is the special-purpose SGML that the OED2 source is in.
 
@tchrist ah. too bad
 
3:35 PM
I wanted to use a standard parser on it, and realized that due to the evil of the format, I couldn’t.
Let me find you a real example.
 
Have you tried a generic sgml parser?
Also you could try a tag soup HTML parser which is meant to deal with garbage and pound it into shape
 
</ET><p><S0><S2><#>I</#> <DEF>To beat down, demolish, destroy.</p><p></DEF>
Notice how the <p> container ends in the middle of the <DEF> container.
I haven’t tried anything yet. Just wondering about it.
This isn’t HTML.
The only nice property is that every tag is exactly described by <\S+>, and usually by <\w+>.
 
@tchrist It's not HTML, but HTML started as a kind of SGML. it might be possible to whip it into shape.
I guess I have to ask what you want to do with a parsed representation
 
Well, I already have something of an OED-SGML-to-HTML converter, which works by responding to <TAG> and </TAG> events. This is mostly good enough. But I lose all the meaningful markup that way, which I would like to preserve for better searching.
I have an index built on all the headwords and other terms from each entry. But searching on any other field requires pretty uncouth regexes.
I guess I just need to make a front-end that makes that easier.
I was hoping to be able store better indexes though.
Here’s an example of one entry, (semi-)pretty-printed.
<E>
  <ST>ali</ST>
  <HG>
    <HL>
      <LF>extraordinaire</LF>
      <SF>extraordinaire</SF>
      <MF>extraordinaire</MF>
    </HL>
    <MPR>ekstr<i>o&hook.&mac.</i>&revr.dine&mac.<su>&schwa.</su>&sd.&revr., &trli.ekstra&breve.&sylab.<i>o&hook.&mac.</i>&revr.d-</MPR>
    <IPR>
      <IPH>&ope.kstr&revc.&lm.&revr.d&shti.&sm.n&ope.&schwa.(r)</IPH>,
      <IPH>&trli.&ope.kstra&revc.rdin&ope.r</IPH>
    </IPR>,
    <PS>a.</PS>
  </HG>
  <ET>a.
    <L>Fr.</L>
    <CF>extraordinaire</CF> extraordinary, outstanding; used in colloq.
I’m sure you can see why the to-HTML conversion loses almost all the useful information.
It makes for something pretty to look at.
But the original SGML is much richer than that.
One thing that is hard to represent is that they use an <r>...</r> sequence to force something into Regular/Roman fonting, overriding any surrounding thingies that would be in italic, bold, smallcaps, or some combo thereof.
I don’t know how to map that to HTML without keeping track of what other effects were turne on.
Whereas in troff it is trivial to make that <r> into a \fR to switch the fontset back to regular and then when done because you see a </r> to emit a \fP to pop the font-stack.
But I don’t know how to do that in HMTL.
For example, in "<SF>fixed <r>or</r> mineral alkalis</SF>", the SF tag for "short form" (well, or "shown form") forces the style into a certain fonting due to the necessary style class span it turns into. But the embedded <r> exempts that part from it.
So suppose SF created bold and italic, or bold and smcaps. I don’t know how to tell HTML not to do that for the <r> portion.
At least, not without keeping track of stuff I don’t really want to worry about keeping track of.
 
4:20 PM
in HTML + CSS, you would probably have to specify SPAN.sf { font-weight bold;} SPAN.r { font-weight: normal }
That'd be my first stab at it, anyway, but yeah, parsing that mess to make working HTML will be annoying.
I'd start by parsing it and finding out what tags aren't nested
Of those, I'd determine what they're used for
and, best-case, I'd just insert more of them to nest properly
oh, but now I see that that doesn't work easily for your example
It's hard to say what that example should even become
<p><S0><S2><#>I</#> <DEF>To beat down, demolish, destroy. </DEF></S2></S0> </p><p> <S0><S2> </DEF>
maybe?
It would really depend on what the <p>, <S0>, <S2>, <#>, and <DEF> all mean
 
5:04 PM
Well, yes.
Although I do know those.
 
def's a regular html tag, I think
 
No no no, don’t think about that.
 
oh, sorry, I'm coming in without context :D
 
Every single SGML tag turns into something else in HTML.
In fact, I think only <p> <i> <b> map to themselves. Plus those ones’ close tags.
Here’s a short list of what certain tags map to:
<#>      Numbered Sense (render tag contents in bold with a FULL STOP and space following)
<A>      Author (in <Q>uotation)
<BL>     Bold List (like HL but used internally; render as non-italic bold smallcaps)
<CF>     Cited Form (this is a mention not a use: hence italicize)
<D>      Date of Quotation (in <Q>uotation)
<DAT>    Date Alternate (in <Q>uotation; used for alternate publication date variants)
<DEF>    Definition
<E>      Entry (this is the outermost tag; each entry is an <E>...</E>)
I have all the others. That’s just an example selection.
And the entities as well, whose mappings are in another table.
Note that these entities are all of the regex form "&\w+\." though; they end in dots.
They only print the IPH, which is real IPA. The stuff in MPR is from the OED1 or NED or v1 OED or whatever you care to call it.
So those are <IPH>ɛkstrɔː˞dɪˈnɛə(r)</IPH>, <IPH>∥ ɛkstraɔrdinɛr</IPH>
I also wish I had a font with ligatures for all the IPA vowels with the rhotic-coloring ˞, not just one with precombined forms for ɚ and ɝ.
I know of no font that has a ligature to merge "ɔ" with "˞" as the IPA designers intended.
 
posted on November 28, 2014 by sgdi

Science is making me worried Many new things have been hurried Implications of which Are making me twitch I hope the result isn’t bloody

 
5:21 PM
They really should have used "<IPH><ST>ali</ST>&ope.kstra&revc.rdin&ope.r</IPH>" instead of "<IPH>&trli.&ope.kstra&revc.rdin&ope.r</IPH>" there, and in places, they do. But this has been hacked on by a thousand hands, and there is no consistency.
There are 125 ways of specifying a language tag that is some form of Latin.
I counted.
:)
 
5:33 PM
that seems like a lot
 
6:30 PM
"Hemingway would rather have died than get old"
@MattЭллен More is less. Which is then again more. Win-win.
 
I'd prefer win-win-win, but that'll do
 
Well, that would be 'more', innit?
 
I'd never settle for less
 
6:51 PM
I could settle for less. Not much less. I don't want to be too forgiving. Unless you've done something unforgiveable. That would be unforgiveable. That's sort of how things work.
 
you don't want to be too forgiving unless I've done something unforgivable. So you'd forgive unforgivable things?
 
I’m saving up all my dippy tongue questions for the Hat Dash.
 
7:35 PM
hi all
I have a question!
 
7:49 PM
A stone's throw is defined by the NIST to be thirty-five feet. In metric, it's thirty meters. The difference is just that stones are bigger in Europe. — Mitch 16 secs ago
@MattЭллен Totally. I'd forgive all sorts of things that are unforgiveable. But I'm not the Pope so it's not really enforceable. Or binding. Or actionable. What I'm saying is forgiving is a load of horse shit.
 
@abforce that's good!
@Mitch especially if your follow regular advice which is to forgive and then to forget. if you forget you've forgiven then what was the point?
@tchrist sounds like a good plan
 
@MattЭллен You may forget, but you'll still wonder anxiously why you have a nagging feeling for punching that bastard in the face.
 
well, obviously you can't go around without needing to be forgiven for something!
 
8:05 PM
What? covers up quickly Was I 'showing'?
 
I wasn't looking, so you got away with it
 
8:25 PM
@tchrist I'm not really sure there will be a nice way to fix all this stuff up. Your input is essentially unstructured garbage.
 
Jez
woohoo! kebab time
dances to kebab
 
I'd be inclined to try to categorize the various structural failings and try creating fixes for each category. But possibly there is no useful set of categories of failures.
 
@Jez kebab winks back at you
 
8:43 PM
@MattЭллен I think forget and forgive are different. But they will each mean different things to different people.
 
@JasperLoy ah
 
@MattЭллен I am very troubled these few days by issues similar to what I mentioned in my secret email. It is difficult to try to change the world. Maybe I should forget about it.
 
The important thing is to feel well. focus on that first.
 
Yes, I should save myself first before saving others.
I have lost almost everything in life. I hope many miracles will happen in the next two thirds of my life.
The bus that runs till 4 am on weekends now runs till 2 am only. This means that I must return earlier if I go out, or wait for the train at 6 am.
 
9:09 PM
@JasperLoy That sucks.
Can't you cycle?
 
I can't, and it's too far to cycle too, lol.
I can't cycle at all and I can't swim properly, though I can swim improperly, lol.
I took some swimming lessons in the past, but I never felt I could do it the way I am meant to.
 
@JasperLoy how far is it?
 
@Mr.ShinyandNew安宇 Well, I only know that it would take about 4 hours to walk home, lol. I tried it once or twice.
 
@JasperLoy That's not THAT far. You could probably cycle it.
If you cycled it often you'd build up strength and endurance and probably get it down to under an hour
 
@tchrist thanks!
 
Welcome.
 
10:08 PM
This guy needs guidance about what SE isn’t.
 
Hey @Cerberus, does Dutch have a calque for "idea/invention (floating) in the air"? Perhaps even more to the point, does Latin or Greek?
Same @tchrist re:Spanish or Portuguese.
 
@RegDwigнt Sure. “Proyecto Cartele es el fruto de una idea de Machi Mendieta, Gastón ... De pronto fue claro que la idea estaba en el aire, esperando que alguien le diera forma, ...’
 
Thank you. I was guessing as much.
 
It’s just what one says.
Better than Zeitgeist. :)
 
So something somehow somewhere in the back of my head does tell me this comes from some Socrates guy or something. But I wouldn't know why it does that.
 
10:11 PM
Well, Platonic ideas are kinda the opposite.
 
Yeah I've always been more into Plasubdominant ideas, myself.
 
Are we doing chord progressions?
 
Are we ever not?
 
Depends on the interval.
 
Also, this is not exactly a progression. This is more like Status–Quo-level stuff. Which is the opposite of, well, really any progress.
There's my en dash.
 
10:14 PM
Tshure.
 
Interval Kilmer.
 
Status aeternus.
 
Interval Kilmer is sort of the opposite of aeternus. Changes size over time.
 
And yes, status was masculine, just 4th declension not 2nd. Hence aeternus.
 
I would not know. Even if I knew who Declension was.
 
10:16 PM
Friend of Cart & Morgan.
 
Fun fact: the plural of Status in German is Status, and out of 82 Germans exactly 0 know that.
 
Of course it is.
 
Well you go tell them.
Stati? Staten? Statuten?
 
But why does the Wiktionary have it not?
 
Either because reasons, or more likely still, because no reasons.
Also, I just misspelled 82 millions as 82.
Now that's what I call a typo.
 
10:18 PM
I seem to recall German doing something odd with declensions of Jesus.
 
Yeah they use Latin ones except nobody knows that Latin even existed, so basically it's a free-for-all.
Die Leiden Jesu.
I suppose most folks would just say "des Jesus" these days.
 
Really?
It’s the genitive Jesu I always found weird.
 
Jesus Christus (von griechisch Ἰησοῦς Χριστός Iēsous Christos, iɛːˈsuːs kʰrisˈtos, Jesus, der Gesalbte) ist nach dem Neuen Testament (NT) der von Gott zur Erlösung aller Menschen gesandte Messias und Sohn Gottes. Mit seinem Namen drückten die Urchristen ihren Glauben aus und bezogen die Heilsverheißungen des Alten Testaments (AT) auf die historische Person Jesus von Nazaret. Kirchliche Lehren zu Jesus Christus behandelt der Artikel Christologie. == Die urchristlichen Quellen == Das Neue Testament überliefert die Botschaft von Jesus Christus in verschiedenen Literaturformen für verschiedene Zwecke…
> Jesu Geburt
 
> Jesu Erscheinung
> einige Worte Jesu
> Jesu Wirken und Schicksal
Bored now, just Ctrl+F for "jesu " yourself.
@tchrist And yeah WTF Jesum? Nobody ever at any time at all.
 
10:22 PM
Well, that would be using the Latin accusative.
Which like -m.
Or likes.
 
M in German is exclusively dative.
 
I know.
So, I have been doing almost nothing but translation stuff at the new job.
 
From to?
Indian to English?
 
Software.
 
Software does not translate to English well.
It doesn't even have a plural, for crying out loud.
 
10:25 PM
Replacing a hell of a lot of print "You should be able to understand this" calls to print localize("You should be able to understand this") ones.
 
What a joy.
 
But also being the second-string Spanish speaker.
We have a native speaker.
But he isn’t always around.
 
No siempre está.
 
And he typos a lot.
 
That's the way of the native speakers, everywhere.
 
10:26 PM
So whichever one of us does the translation, the other one has to approve it.
 
Well that should work actually.
Two cooks a broth do not spoil.
 
I keep telling the programming team to stop throwing single words for translation, since they’re doomed to combine wrong.
 
Haha. Can they spell one-oh-one?
 
Well, you can have a "context" argument, but we haven’t used enough of those to make translation to something with cases work right.
Like into German or Russian.
At least, not for single words. Which are usually GUI elements.
Things like gender and case count as "context".
I beg for nothing but full phrases.
 
A basic human right.
 
10:29 PM
But they keep wanting to do things like printf "%s %s", $customer, $department" and it drives me mad. Since the order will be wrong.
 
Ugh.
101 again.
Haven't they ever run into a name or address?
 
That has to be something like print localize_x("{client} {dept}", client => $customer, dept => $department) so that I can rearrange that into the other order and insert appropriate "de" etc pieces.
Names are a problem.
 
There are like five ways to format "221b, Baker Street", too.
Only one at a time being correct.
 
I keep entering "Gran Vía, 25" for "25 Main Street" and "Cañon City" for the city just to see what happens.
 
You should enter some Chinese addresses. Or even just Japanese.
 
10:33 PM
We actually have those in our unit tests.
Not that I can flipping read it.
Then again, neither can your local postal worker.
 
Or actually, even in Chile. Was it Chile? There are no street names, just things like "second house after the homeless guy with the dog".
 
Haven’t heard that one.
 
Like, fifty years after the guy has died, mind you. Not to mention the dog.
@tchrist lemme google. That's actually a specific example for a reason. I heard just that specific example somewhere.
 
The problem with name is you are supposed to be able to look up a record by the start of the last name. But what is that?
We have Spanish-speaking customers who are constantly entering three or four names.
 
The last name is anything that comes last. Or first. Or in the middle. DUH.
 
10:35 PM
Fernando María Sánchez Miret.
His last name is Sánchez Miret, but how the hell would anybody else know that?
The only solution is to have separate entry boxes for forenames and surnames.
If you think you need to tell the one from the other.
Because otherwise, you cannot.
 
Most recently we had an Indian customer with four names. Each 12+ letters long. I think two of those were surnames, and two forenames. But he wasn't too sure himself even after we flat-out asked him.
 
hahah
 
So basically he told us to call him Joe or something.
 
The entry document you fill out when entering the United States has a place for "Nombres" and a place for "Apellidos" on the Spanish-language version.
But the English-language version doesn’t admit that there can be more than one of each.
 
What does it have for Saudis?
 
10:38 PM
How does international postal mail ever get where it should go?
 
BTW my favorite field on that document is "do you plan to kill the President".
 
It’s a great filter.
 
@tchrist how does national anything in Japan ever get anywhere?
 
No, seriously. How does international mail whose source and destination use different scripts ever get anywhere?
 
Experience.
 
10:40 PM
If you write my address in Cyrillic, our post will not love you.
Let alone in Arabic or Kana.
 
You could write anything on a letter to Germany. Anything at all. They will deliver it to that exact person. Within two days, too.
I does not knows how them does it.
But I've done it myself in point of fact.
 
In small towns, you can skip the address and just give the town.
 
I only knew the city and the street, and I knew the street wrong.
 
They will probably find you.
 
Twas Munich. Not the smallest town. Even less so in the context of Germany.
 
10:42 PM
Now, a wrong street in a big town is a problem.
That seems . . . creepy? Efficient? Magical?
 
Two days. I is telling you. I must have had awoken their sportliches Interesse.
 
heh
 
I can offer no other explanation.
 
’Twas a challenge.
 
Other than really some guys being so much into it, they are doing nothing else all day long, and for free cuz they love it.
 
10:43 PM
I regularly get mail in my postbox that isn’t for me.
Sometimes I politely walk up or down the street to put it where it goes if I know.
The problem is that they will deliver to the street address no matter the name.
Which is the opposite of what happened to you.
And hand-written addresses confuse things like 1, 4, 7, 9.
So if they wrote 675 but crossed the 7 enough to look like 645, I get it.
Even though it is not my name.
 
@tchrist Yeah that's the weirdest thing. If you write, "Tom in München", they will find you. But if you tell them "Tom Soundso, Sonnenblumenallee 42, 82111 München", they will sometimes leave it on the doorstep of Sonnenblumenallee 44.
 
@RegDwigнt Umm I'm not sure I understand what you mean.
 
I imagine the people it 615 get the mail for the people in 675 a lot, too.
 
In what language do you have such a word?
 
Then again that probably goes to show that it's just a specific guy doing the hard stuff, and the normal postmen doing the screwups.
 
10:46 PM
@Cerberus He wants to know if the phrase is used.
If a notion is floating around in the collective mind, can it be said to be in the air?
 
Ah.
 
@Cerberus not word. The whole expression. English: idea/invention floating in the air. German: die Idee liegt in der Luft. Russian: идея витает в воздухе. Spanish: whatever Tchrist said.
 
I see.
 
37 mins ago, by tchrist
@RegDwigнt Sure. “Proyecto Cartele es el fruto de una idea de Machi Mendieta, Gastón ... De pronto fue claro que la idea estaba en el aire, esperando que alguien le diera forma, ...’
 
The metaphor is certainly clear and alive in Dutch.
 
10:47 PM
Está en el aire.
O frig, why am I even scrolling up if he's faster.
 
So, did Latin talk about ideas being in the air?
Or Greek?
 
@Cerberus haha, then I wonder why you even asked.
 
But I'm not sure I would say there is a fixed expression exactly like that. You can say something "hangs in the air"; it could be a technological development, or an idea, the Zeigeist, an imminent war.
 
We’ve already ge-Zeitgeisted.
 
Ge- is for the past participle.
Better.
 
10:49 PM
@Cerberus good enough for my needs.
Teh kvestchon is hier:
1
Q: What is the word when people come up with the same idea independently

benSuppose Darwin and Wallace independently come up with a similar idea. It's like the idea has entered the social consciousness at that time. What is the word for this called? Kind of the tipping point where everyone catches on and starts doing similar stuff independently.

Also, I am stopping to be drunk. Probably because I have stopped drinking. This is inacceptable.
AFK
 
Drink.
 
@RegDwigнt An expression that is currently en vogue in Dutch is "een proefballonnetje oplaten", to release a trial balloon, to "fly a kite". It is said about e.g. when a minister has an informal proposal for a new law and lets people know.
 
Bah. Of course it has to be an etje. Who are you, Brazilians?
Everything is an inho.
 
Hehe.
You can say "een proefballon oplaten", but the diminutive reinforces its state of an unimportant thing.
 
@RegDwigнt Small, cute, friendly.
 
10:54 PM
@RegDwigнt So in this example, I think a Dutchman would say "het hing in de lucht" ("es hing in der(?) Luft").
 
katje
 
Somehow, most Dutchmen prefer poesje.
 
spits
 
Even though, officially, a poes is a female kat.
But some would say katje.
 
Really?
 
10:55 PM
Yes: technically, poes is female, kater male.
 
We can use it for either gender of kitty. "Hey puss, don’t be climbing on the table while we’re eating.”
 
Well, we are used to using it sloppily.
 
It’s a bit old, though.
 
In fact, I prefer the sloppy style.
I can all pussies poezen.
 
Normally now it’s just kitty.
 
10:56 PM
Right.
 
Where kittykitten
 
And did you know that many people in Dutch use kitten for young kitties, even though we have the perfectly normal equivalents poesje and katje?
 
Is that plural?
 
It is an ugly word in Dutch.
No! It is English!
Singular.
 
Wow.
 
10:58 PM
It's interesting how importing certain words from English is nice and appropriate, while importing others is hideous and stupid.
 
I am looking at you, German, with your Handy.
 
Kitten is German for to kit.
 
I think kittentje bothers me more.
 
As in, to cement. Cementate. Seal.
 
10:59 PM
@tchrist Haha, nobody says that. It would be hugely pleonastic. Of course that has never stopped anyone before, but...no.
@RegDwigнt Ah, yes, it is the same in Dutch.
 
00:00 - 23:0023:00 - 00:00

« first day (1478 days earlier)      last day (3435 days later) »