« first day (1837 days earlier)      last day (3304 days later) » 

11:00
@Katenkyo It looks like I'm starting to fix it... I'm rebooting, see if this works.
Well.
This is weird.
My desktop changed.
But it's working. \o/
(but it uninstalled mono. :I)
Okay, good.
@VoteToClose weird... But at least, it's working !
Looks like I've made a new breed of Linux. O.o
And re-installing mono shouldn't be hard
I did this all remotely. :D
11:02
SSH connection held like a charm.
Good job ! ;)
I probably broke something, though.
I'll need to check the USB, HDMI, and SD ports. :\
@VoteToClose So, what did you finally do for your "kill httpd" problem?
3
A: "Aliasing" of Commands with Arguments

Karoly HorvathYou want to intercept killall, so: killall() { if [ "$1" = "sshd" ]; then echo "some warning message, and perhaps service sshd restart" else command killall "$@" fi }

much more elegant than what I proposed :D
11:17
@Katenkyo I did it again. This is what happens when you have to write your own bootloader every single time ... in nano...
reluctantly starts installing ubuntu. ugh...
@mınxomaτ do you install it from scratch each time?
Maybe have a script/image saved somewhere that will prepare it to a stable version
@Katenkyo Yeah, the platform is kind of unusual.
(maybe up to the point you screwed)
I didn't do anything. It turns out the latest dist upgrade (arch4ARM) is flawed. It is either linux-armv7h or systemd-armv7 that bricks my system.
almost done with my irc relay
11:19
Internet relay chat relay?
@mınxomaτ aww
@ӍѲꝆΛҐӍΛПҒЦꝆ <- this username breaks, though
never upgrade if you don't have to ^^'
a faux irc server, that connects to chat.se, and relays messages
@Katenkyo Every update in arch is a full system upgrade. There is no versioning like in ubuntu-based distros.
11:35
0
Q: Sort by Multiplying

randomraYour should write a program or function that given a list of positive integers multiplies each element with the smallest positive integer possible to create a strictly increasing list. For example if the input is 5 4 12 1 3 the multiplications will be 5*1=5 4*2=8 12*1=12 1*13=13 3*5=15 and...

0
Q: Output the van der Corput sequence

BobThe van der Corput sequence is one of the simplest example of low-discrepancy sequence. Its n-th term is just 0.(n written in base 10 and mirrored), so its first terms are : 0.1, 0.2,0.3,0.4, 0.5,0.6,0.7, 0.8,0.9, 0.01, 0.11,0.21,0.31, 0.41,0.51,0.61, 0.71,0.81,0.91, 0.02, 0.12,0.22,0.32, 0.42,0...

12:02
1
Q: Recursive 2x2 determinant

Sp3000The determinant of a 2 by 2 matrix a b c d is given by ad - bc. Given a matrix of digits with dimensions 2n by 2n, n ≥ 1, output the result obtained by recursively computing the determinant of each 2 by 2 sub-block until we reach a single number. For example, given the input 3 1 4 1 5 9 2 6...

@mınxomaτ I would be quite upset if it messed up my entire system each time !
Aaaand I accidentally killed the sshd again. :I
@VoteToClose ...
And you blame your mates for doing it :)
Oh, no, I share my server with no one.
Yet.
So, how did you kill your httpd this time?
12:15
sshd*
I was testing the fix for the killall, and I used sudo (which means I need to put the .bashrc update in root, which I forgot).
(to much used to do systemd httpd restart, sorry :D)
oh, at least you thought about testing it !
Better learning it that way than the hard way !
Yeah, well, I did it remotely. :I
So I basically was am an idiot. xD
:I
Oct 29 '15 at 2:17, by quartata
I'm an idiot
^ 23 stars
@Fatalize Yay for the (first?) declarative golfing language Brachylog! Are you the one who develops it?
12:43
@randomra Yes I am. I doubt it's the first one, even if I started it in September 2015
(On the same note does anyone know what the hell you're supposed to do so that Esolang actually updates its category list?)
I wonder, why this challenge is bad?
@ChristianIrwan I think it is because the only way to make an answer not "crackable" by robbers, it to write one sooooooo long it will be impossible to analyze it within days
Plus, you ask for a binary answer from robbers, so it might not really be inspiring for them
12:59
@ChristianIrwan It's not a bad idea (though as Katenkyo says, it could have many flaws in practice), but I think you're getting down votes because the post is confusingly written and has bad grammar.
@Calvin'sHobbies I don't mind the grammar (mainly because mine is far from good), but yes, it's confusing. In fact, I can't understand the scoring :/
confusing: Easy to understand, even to noncoder, Easy to write, No informal proof.
@Katenkyo Binary answer? I ask proof.
@ChristianIrwan It is still a binary answer -> Yes it halts, no it loops
@ChristianIrwan Also, for example, someone could write a program that looks for the first even number that is not the sum of two primes - i.e. a program that disproves the Goldbach conjecture, an unsolved problem in mathematics. Literally no one could say if it will stop or not. (Or if they could they'd be world famous)
13:07
Even when you have to prove it, it doesn't change that fact, and it offers poor possibility to the robber
@Calvin'sHobbies could it work with calculating most (supposed) infinite decimal number (like PI), because we don't really know if they are infinite or note?
Well, pi's decimal is infinite, pi is transcendental, but yes, the idea could be applied to other unsolved math problems
Or just very complex problems like the 33=a²+b²+c², where you would run into all sorts of physical and compiler design constraints.
Actually, the cops have to provide proof, but not shown and after 6 days, the cops reveals his proof.
And I meant all proof is formal proof.
@ChristianIrwan That's unclear. You say the cop "can" reveal the proof, not "must"
@Calvin'sHobbies Fixed
13:20
But maybe with that changed (and other confusing stuff) it could work
@ChristianIrwan so one of the only actual trick that could do the job in a cops-and-robber style would be a program looking for the xth prime, which is the value y and y>MAX_SIGNED_INT, so it loops forever (because of overflow) while it appears to be ending.
other than that, as we can't obfuscate our submissions, it become only a math challenge, which belongs to puzzle.se ^^
hum... anyone see a reason why a working mouse doesn't work on a freshly formatted windows?
nevermind, solved
So algorithm challenge doesn't belong here?
@ChristianIrwan You ask for a formal proof, and also ban the only languages in which a formal proof would be remotely feasible.
13:39
@feersum I mean the submission is just algorithm to be decided if it is halting or not.
Oh? The question in no way involves proofs?
@ChristianIrwan providing formal proofs of an algorythm is pure mathematics. I think PPCG is for challenges around the whole programming and IT things
In facts, you could ban every languages existing, and only allow pseudo-code, or plain literal algorythm, it wouldn't change your challenge and the task of the robbers
That's the problem I, personnally, have with it
So, this challenge is off topic.
@ChristianIrwan Try doing a meta post if you think it's a reasonable doubt, my opinion isn't absolute. Else, you could still work on it and do it anyway, as Calvin said, it's not a bad challenge in itself
13:57
Vim is cool
14
A: Do all challenges have to be answered with code?

Nathan MerrillWe should allow algorithm-only answers. We currently accept code that is impossible to verify. Furthermore, we rarely actually run the code to verify anyways. Hence, the fact that it is impossible to test should be taken into consideration by the OP, but it should not prevent the OP from askin...

@Quill until you want to exit it
@Katenkyo @ChristianIrwan ^^
@MartinBüttner u ruined it
@Optimizer my pleasure
13:58
well, I did too.
@Optimizer ?
you really haven't heard of the story of the developer who was unable to exit vim? oh wait... that story has not ended yet..
@MartinBüttner oh thanks :)
@Optimizer the camera pans and it's revealed he speaks his melancholic words to a mirror, however the virtual image shows an earlier version of himself, less tormented by the key command horrors that we call Vim.
When can we expect that story to hit the box office?
@Optimizer happened to me the first time I used vim. I killed the process and never used it again...
14:09
0
A: Sandbox for Proposed Challenges

Stewie GriffinCreate a football field. code-golf, ascii-art Football is the sport where players kick the ball, not carry it. Some confused individuals might call this soccer. A football team has one goalkeeper, and 10 players out on the pitch. There are many formations used in football, that dictates wher...

@Quill when linux crashes due to int overflow in time epoch milliseconds
2036, wasn't it?
i think
don't remember
@Katenkyo typical story of an emacs enthusiast ...
@Quill 2038
ah, thanks
14:11
-4
Q: String that hashes to hello world in any commonly used hashing algorithm (e.g md5, sha-1)

curiosIs there a string that hashes to hello world in any commonly used hashing algorithm (e.g. HAVAL, MD2, MD4, MD5, MD6, RIPEMD, RIPEMD-128, RIPEMD-256, RIPEMD-160, RIPEMD-320, SHA-0, SHA-1, SHA-256, SHA-512, SHA-384, SHA-224, Tiger, Whirlpool, SHA-3 etc)? It's ok if the Hello World is contained in...

@Optimizer not really, either nano or geany/gedit
We're now on 9.3 questions/day
That is damn high
@Adnan not for normal SE site.
That is true
But it's the highest I've ever experienced on this site
I, for one, like vim
14:22
vim is pretty. I might try the vim plugins for Atom
I've been using VsVim in Visual Studio for about a year now, it's pretty neat =)
I personnaly prefer things with less... functionnalities. I only want my text editors to have syntax highlighting, and why not a project view for the most complex ones...
@Adnan As long as no one does anything foolish like confirm graduation, that number should keep going up and up...
@Katenkyo Right, vim would be pretty overkill in that case =P
@Roujo isn't vim overkill for everything? :p
@trichoplax do you really think that the graduation will make this number go down?
14:34
@trichoplax In fact, I'm probably going to write a challenge right now
@Katenkyo Meh. I like being able to move/copy blocks of code with just a few keystrokes, plus macros are pretty cool as well.
@Roujo more often than not, if you have to copy blocks of code, it means you did something bad ^^
Using all of vim's functions all the time might be overkill, but I don't see much of a downside in just using vim. Worst case, you just go into insert mode and type and that's it. Best case, you use a couple of feature and you're better off.
@Katenkyo Sure, but who doesn't? Sometimes I realize that I should have placed the code elsewhere =P
Or I'm refactoring code, or I'm reviewing my own while debugging, etc.
@Roujo My bad, I only thought about copy/pasting, skipped the move part ^^
@Katenkyo Oh, right. Yeah, just repeating the same line over and over with minor difference is kind of a code smell. =)
14:40
And that's why I think a simple text editor is sufficient to do stuff (even without a mouse ^^). I know that vim can be useful, but I love simplicity, as it's the best tool of a developper ^^
@Katenkyo No I'm joking because there is a general impression that graduation is the reason for having a question push, whereas in fact I think our gradual growth is now inevitable.
@Roujo at my previous workplace, there was some really smelly code. Nested if blocks repeated 5/6 times, each time with the same conditions, just some instructions changed...
@trichoplax I agree with that
I will always be amazed by what I see on this site and its community
@Katenkyo Maybe they've never heard of switch
@TimmyD don't even know if this language had the switch statement
Ok, so my baseline didn't go very well. Some basic responses, but nothing I'd call good. Cleaned up the tags in the conversation, increased the net size, and starting again.
14:49
Pshaw ... if you can even call that a language ...
It was in something like PL/1
@Geobits Can you make one from just a few days of chat so we see what it might be like? (Can it work like that?)
@Calvin'sHobbies I'm not sure. I might try that at some point, but I think (in general) the more the better. The papers I was reading were using datasets with >=22M lines, whereas I have about 500K right now.
Here's one more
We'll work on chatting more and in shorter lines
14:51
>_>
I ' m it ' s a the the the the the the the the
@Geobits I keenly await the day the bot appropriately responds to something with a text face
@Geobits What's the name of the program/library again?
Tensorflow. Specifically the seq2seq model.
I really should have a stack of GPUs for this, but oh well :/
15:16
2
Q: Preparing a multiple choice test

AdnanIntroduction Note: this is not some kind of method to encourage cheating. As Cᴏɴᴏʀ O'Bʀɪᴇɴ already said, studying is the best solution for passing a test :3. Consider the following answers to the multiple choice test: ABCBCAAB Here is a table which indicates whether the answers match: A...

0
A: Sandbox for Proposed Challenges

Christian IrwanGuess my number code-golfnumber Your task is to create program that try to guess the number. Your program will take input N, after that the game begins. Your may guess any integer. the integer then is outputted. Then you read input. If it is "+", then the guess is too high. Else if the output ...

15:50
Reading model parameters from train/translate.ckpt-1100
> hi
I
> Are you still doing that?
Yeah
> Why not talk like a person?
Yeah
> You clearly need more training.
I
> Yea, see, that's what I mean.
I
> Dammit, say something different.
:
> Why would you just type a colon?
Yeah
> Do you answer every question with "Yeah"?
Yeah
> Do you like pickled fish and chocolate?
_UNK
_UNK is my favorite chocolate.
@Geobits That last response was a little unexpected.
Yea, he ran past the end of his vocabulary I think. That's the 'unknown' token.
You'd think it would force it to choose a word it knows instead of vomiting garbage.
3
It's only been training for a bit over an hour, though, so I wasn't expecting much.
16:09
_UNK
^ poisoning well for future chatbots
Anonymous
9.4 q/day! Now that we've gotten confirmation that we're graduating, we're actually going to meet the vague requirement :P
> confirmation
Anonymous
I think Grace Note's comment is pretty much confirmation. It's not official confirmation (i.e. a meta post by her), but it's a hell of a lot more than we've had for the past year+ (that one comment by Jon notwithstanding).
@Downgoat How does the byte counter work in your userscript? It's super cool.
16:18
@Mego :O
So I guess you guys don't count my comments as "confirmation"? Fine, I see how it is.
3
Anonymous
@Geobits Until you have a blue diamond of power on Meta.SE, your comments about our graduation don't count :P
Anonymous
On a related note, Geobits for community manager 2016
2
@MartinBüttner I'm thinking about it but not coming up with anything yet. I hope I can figure it out without a hint, but I probably will need the hint...
16:22
@DigitalTrauma let me know when you do (either way) :)
Anonymous
Our graduation date should be our 6-year anniversary of the A51 proposal
Six seasons and a graduation?
2
Anonymous
Hey
Wow, I had no idea we had 2k+ "avid users".
16:26
Anyone have any comments about this one? I'm wondering if it's a good challenge, and if it's ready for posting.
5k visitors per day??
@trichoplax Yeah.
Anonymous
@StewieGriffin 2 hours in the sandbox != ready for posting
@Mego depends
Anonymous
Peter hasn't even had a chance to point out all the things you're doing wrong in your life the challenge
16:27
@Mego, 0 hours in the sandbox == ready for posting
In many cases :)
Anonymous
@Calvin'sHobbies You are the exception :P
While we're talking about Peter
1
A: Sandbox for Proposed Challenges

quartataChallenge about advacado Inspired by this. No avocados were harmed in the making of this challenge. Hello I have challenge I need help juic an avocado so I need program to tell me how long to juic avocad for Observe this ASCII art avocado: ###### # # # #### # # # p# # ## ...

Is this a dupe of this:
24
Q: Are you in the biggest room?

turbulencetooIntroduction You have recently accepted a job offer at a Pretty Good Software Company. You're pretty content with the size of your office, but do you have the biggest office? Its kinda hard to tell from just eyeballing your co-workers' offices when you stop by. The only way to figure this out is...

@Mego, Give me a minute, I have 30 challenges to delete... =P
Related. It's sufficiently borderline that I wouldn't dupe-hammer it, but I would cast the fifth close vote if others vote it as a dupe. — Peter Taylor 18 hours ago
Anonymous
@quartata How on earth is it a dupe? I'm not seeing it...
Anonymous
16:30
Oh counting enclosed spaces
Yeah.
Anonymous
Yeah ok I'd call it a dupe too
@Mego I almost never use the sandbox. ¯\_(ツ)_/¯
@Doorknob would you dupe-hammer
@Mego They graduated with those flibbin stats?!
Anonymous
16:32
@Doorknob You and Calvin have been around for an eternity a while, and have gotten good at posting challenges that are well-specified. Us peasants mortals normies newer members should still use the sandbox because we don't have as much experience.
Anonymous
@Calvin'sHobbies I know right
Anonymous
It launched 5 years ago with those stats, back when it was graduate or die after 90 days
When will people stop paying attention to those stats... :/
@Mego I dunno, I've generally always been anti-sandbox, and while my stance on it now isn't quite the same as it was two years ago, I still think it shouldn't be considered as necessary/required as it is now.
Anonymous
@MartinBüttner When the CMs stop acting like they're meaningful
16:33
except they're not
like, at all
for the past year or so
9.4 questions per day: Just OK
@Mego Suppose they do announce that we are set to graduate. What happens next? Is it instant? (I know the design takes a while.) I've never paid attention to other site graduations to know how it works
@Calvin'sHobbies It is instant.
But the design comes later.
The "beta" label goes away immediately.
Anonymous
7.9 questions a day means you all are getting close. I'm guessing (at this rate) you'll get an election later in the year and a design sometime later. Not much else to say, really. — Jon Ericson ♦ Jan 15 at 4:11
How much later?
16:34
@Mego every single CM I've ever talked to has said that the A51 stats are absolutely meaningless
@Mego quoting meta.stackexchange.com/q/257614/201409 (the source regarding graduation criteria at the moment) "This is not 100% automated; CMs will still be manually checking on how individual communities are doing, and some sites might still graduate "earlier" or "later" than their question activity alone would suggest; but it's an effective rule of thumb."
And the swag comes after the design. rubs hands together
@Calvin'sHobbies Depends. CR waited a year :P
@Mego that's one single stat
and an unimportant one at that
(I was about to link what Martin linked)
Anonymous
@Doorknob It's a CM quoting our A51 stat as if it was meaningful
16:35
@Mego stats plural?
@Calvin'sHobbies Worldbuilding is currently going through community guided site design, as an example
Hmm I wonder who'll run in the mod elections
@trichoplax ooh, ty
@Calvin'sHobbies Sorry - pressed enter before paste...
@trichoplax nice maybe we'll have that too
16:36
of course; everyone does
The responsiveness to feedback is most encouraging
Anonymous
More specifically, it's a CM quoting an A51 stat as if it were meaningful, 6 months after the meta post which states 10 q/day was the threshold, even though that number doesn't mean anything, as sites have graduated with less, and it's not a good measure of the quality of the questions posted to the site.
@Mego it's not a threshold; it's an indicator
Anonymous
A threshold for being considered for graduation
16:38
@Calvin'sHobbies you misunderstand - design is done primarily by the SE team. After the design is 90% finished, it gets posted on meta for some feedback, and maybe it gets some minor changes as a direct result.
@Mego it's not a threshold. We're at around 9 last time I checked and the comment on the most recent mother meta post said we're way past "ready"
4 mins ago, by Martin Büttner
@Mego quoting http://meta.stackexchange.com/q/257614/201409 (the source regarding graduation criteria at the moment) "This is not 100% automated; CMs will still be manually checking on how individual communities are doing, and some sites might still graduate "earlier" or "later" than their question activity alone would suggest; but it's an effective rule of thumb."
@Geobits What are all the types of swag?
@Calvin'sHobbies depends. Different for each site. Usually a t-shirt for everyone, and then some other stuff.
I'm going to get smacked for this, but I'm not entirely sure if we should graduate right away. I feel like we need to resolve popularity contests and all that first.
@Calvin'sHobbies I've seen (and have) mugs, shirts and stickers at the least.
@Calvin'sHobbies No offence to all the users who poured lots of time into that, but I kinda hope they will...
16:39
@Doorknob wait t-shirts
@MartinBüttner :(
T-shirts and stickers are usually for every site, yeah
At least I'd like to see a completely different take on the design by SE, and then we can still suggest to incorporate some of those ideas into it.
@quartata it's a valid point
Anonymous
@Doorknob My point was, the CMs clearly still consider the A51 stats (at least, the q/day stat) as being meaningful, even though it's not a good indicator of site health at all (since those 10 q/day could be very low quality Qs and still get counted)
Anonymous
16:40
@quartata I agree with this. Hence why I made the popcon meta post, which a grand total of 2 people answered :P
@Mego I'm with Mego. Obviously the stat has some meaning else it wouldn't be there
@MartinBüttner honestly a professional design is bound to be significantly better than one made by programmers. Programmers are notoriously bad at design and art ;)
@Doorknob hey I like our design
Anonymous
@Doorknob The real reason we ban art contests is because we're insecure about being bad at art
Although some parts are too bright
@Mego hahaha
Sure the stat has some meaning. There's clearly a difference between a site with 2 q/d and one with 30. But the difference between 6.4 and 10.1 is not very important.
@Mego @Calvin'sHobbies the only stat on the A51 page that carries any semblance of weight is questions per day
I'm not really looking forward to the rep requirement changes, though.
16:42
A51 has been the SE-site-equivalent of a run-down, beat-up shack for years now
@PhiNotPi I am
@Doorknob I don't know if it varies from site to site, but for worldbuilding the main design image for the background is directly inspired by the top voted hand drawn user suggestion on meta.
bad close/reopen votes is a far too common sight already
We've got plenty of 10k+ers now, so the rep changes shouldn't be too bad.
Anonymous
16:43
@Doorknob And it doesn't mean anything really, because of the point I've been making: quantity != quality. I would like to see the CMs start recognizing that and stop quoting it as if it was a meaningful indicator of site health.
@Mego it's a rule of thumb
Can you really put toilet paper in your printer?
I really also want to get to 10k before we graduate so I hope we don't graduate for a couple months
Rep rush!
@DigitalTrauma can you? Well... ;)
16:44
@Doorknob css doesn't count? ;)
@Geobits That's surprisingly bad.
It was a fresh data set, only trained for an hour. Not unexpected really.
@Geobits Are you sure you're not trying to translate PPCG chat to scifi chat or something?
Scifi has chat?
... while we're on the topic, I do think this is a new peak
16:48
@Geobits hmm, figured it'd be more active: chat.stackexchange.com/rooms/198/mos-eisley wait, all the inactive people were blending into the bg
@Doorknob Can we easily access that graph of questions per day over time again?
@Doorknob I don't think we've cracked 9.4 before yeah
@Calvin'sHobbies Mos Eisley is pretty active in the evening
@Calvin'sHobbies There are chat themes??
They have 480k messages to our 518k or so total.
16:49
Also the design userscript totally breaks their CSS lmao
5k privilege though iirc
@trichoplax Yeah.
Unique one for every graduated site
How much rep will site stats need when we graduate? Isn't like 40k?
I do think that those rep requirements are way too high
16:50
:/
So 5 people?
30k for stats? I thought it was the 20k "trusted" level?
(almost 6 :D)
@Geobits nope
@Geobits It's a new privilege
16:51
@Calvin'sHobbies Yep and 3 of those will most likely be mods
@Calvin'sHobbies plus Alex :P
Damn, I need some more rep.
@trichoplax
I'm going to miss being able to protect questions.
16:51
@Doorknob Thank you!
@Calvin'sHobbies oh, I was wrong; thought it was 30k for some reason
I'm not so sure I want PPCG to graduate lmao
Time to think up a new KotH. Those are usually good for rep for me.
What's the threshold for close voting after graduation?
16:52
3k
@Rainbolt 3,000
So 9 people privleged to the max
@Geobits Good. No more challenges from me then.
But mod tools are 10,000 as opposed to 2,000 which really sucks
@Rainbolt Yea, CV is the one really important one IMO.
Well, beside the 125 for downvotes, obviously :D
16:54
@Geobits Mod tools are actually really important though IMO
@Geobits Tron
Obviously =P
Delete votes for one
And review queues and all that
I'm going to seriously miss that
Review queues are 2k/3k on graduated
Why are delete votes important?
16:54
@Calvin'sHobbies I think the racers have been done at least. I think Doorknob did that. "King of the walls" or similar.
@Doorknob oh?
I thought they were a part of mod tools.
@Geobits I... don't recall this
@quartata Suggested edit review comes with the edit priv; close vote/reopen review comes with the close/reopen vote priv
@Doorknob Huh.
LQP is lumped in with suggested edits IIRC
Nope, wasn't Doorknob. I remember it because I won: codegolf.stackexchange.com/questions/26293/king-of-the-walls
16:55
First posts and late answers are like 500
I'll still be missing my beloved "new answers to old questions"
@quartata You mean late answers?
@Geobits But 10 upvotes? We could do better than that
@Geobits oh heh I participated in that
Well sure. That was two years ago :P
16:56
@Calvin'sHobbies It had 15 ;)
@Doorknob No, the "New Answers to Questions More Than 30 Days Old" section of the mod tools.
That's where I find all the invalid answers
You can kinda sorta simulate that with search
Except I guess that counts edits and stuff
@Adnan While adding 05AB1E to TIO, I've discovered that the latest version of the interpreter only understands UTF-8. That means that, according to this, you cannot use CP-1252 to compute your byte count. That's easily fixed by adding a flag to decode the source code as CP-1252 though.

« first day (1837 days earlier)      last day (3304 days later) »