I've seen terrible things

Like:
* le /g/aylords trying way too fucking hard to fit in (they already learned le e[spoiler]/prog/[spoiler]in memes, /prog/ro!)
* le same e/g/in /g/ros imitating lelcunt unironically and thinking this is the default form of protesting here
* /g/roskis trying to bring their `daily programming threads' here
* e/g/in re/g/g/it e/g/g/wings bitching about the state of the board and not leaving (like a weeaboo who went on a trip to Japan, bitches about their lack of forks and knives and doesn't go the fuck back to his third world country)

So I've set up a reasonable amount of machines to flood once again. I might have to space out the replies at some times, because I'd rather avoid the situation where all my machines get banned from 4chan.

I'm sure most of you won't accept my form of protest, but for me it seems like an efficient way of improving the board's quality. After the first flood wave, the moderators started coming here more often, which is not the optimal solutions but at least it kept the recent threads programming-related. I do have to admit I miss Nikita and some of the original /anus/ shitposting, but my purpose is not bringing more moderators, it's keeping the shitstains out of here.

Yes, this will make them engage e/g/in mode, because I remember reading on a /g/ troll bait thread about how they loved pissing us off, so I advise you to ignore their unoriginal shitposting from now on.

The setup will take me about one week, so expect the flood to start next week. I'm aiming for at least 15 days, but it's very likely to get banned, so the flooding might not be as continuous as I'd like it to.

>>118
So is it just the question of you not having better place to host it? I'm sure someone here wouldn't mind hosting it, especially considering that it's just a text board. Heck, I have a host sitting mostly idle myself.

Unless you're dead set on hosting it yourself.

>>121
It's decentralized. Everyone can run their own server. You can just grab a copy of the code (when it's done), tell it to sync against a known host, and run it yourself. Or grab a copy of the specifications and write yourself a server and run it. Whatever.

Uh, guys, I'm surely missing something, but you mention FIOC, a textboard used by 3 people, and the Internet. Why don't you just use [b][i][o][u]CLOUD ENABLED PLATFORM AS A SERVICE SOLUTIONS[u][/o][/i][/b] like Heroku, OpenShift or Stackato? They're free.

FUCKING BBCODE FAILURE

>>123
I'm pretty sure ``which is your favourite touhou and why you hate kikes'' is against their ToS...

>>125
Oh, you're right. Not many people are fine with Touhou. Poor souls.

All the preparations have been made. We will start with some Unicode spam for the one-time necrobump threads. If the spam doesn't start at the previously set time or it just stops, either:

* The power/internet went out
* I was banned
* I fucked up with the date/time of start
* The bot isn't working (even after that extensive testing)

I'll check this thread tomorrow in case any of these things happens. (I still will check it even if nothing happens.)

>>127
May the gods smile upon you.

A-and it looks like one of those posts went wild, and jumped to the front page. That was accidental and isn't supposed to start until the right time.

>>128
May HaShem keep you and bless you. ^{_{No. I'm not a Jew}}

Looks like picking threads from the front page wasn't exactly a good idea. ~10 threads were lost from the lists because mods deleted them.

Since the bot has a cooldown of about 50 seconds after making a post (attempt) and it will not find some threads, you might notice a slight delay between posts. Hopefully no more threads will be deleted.

Also, don't think I've ignored the suggestion of using advanced math topics. I need some more time for doing the formatting and it will be done on a later stage. (I'd appreciate it if some kind soul picks some articles and tries to format it with BBCode,)

hello, I am the base64 person. The prog synch is almost done. I'm fixing a few known issues right now. You are all welcome to do whatever you want with the good, be it scraping /prog/ or creating a distributed replacement. I only choose fioc because of the libraries, cross platformness, and the ease of installation. Anyone can download python for their platform and then just run the script. I would have liked to make it in lisp, but I don't want to put users through the hell of setting that up. It can be frustrating even if you are a seasoned programmer, much less a beginner.

Be careful with the spamming. If anything drastic happens as a result, regroup here as usual:

http://tablecat.ipyo.heliohost.org/lounge/read/1373068543

>>131
How many programmers does it take to screw in a light bulb?

Three: two to argue whether to do it in Lisp or in Haskell and one to do it in Java Python.

>>132
It's sad how true this is. Anyways, I hope you don't mind if I repeat this. It's perfectly accurate.

Oh wow, just as expected. My Internet went down precisely 5 seconds before starting the bot. I have access to the Internet again, as you can see.

>>134
Just when I checked this thread again. I'm excited.

That reminds me, for how long will the flood continue?

>>136
The theoretical lower bound^[1] is 15 days, the upper one is probably one month. But being realistic, it might last much less and have many hiccups.

_{[1]Abstract bullshite you would not understand anyway}

>>131
Don't worry about making it in FIOC, Actually, thank you for taking the effort to make the syncbot.

See you in tablecat's lounge in case something happens.

>>132
I loved your post.

>>103
* We'd have to see the standard you propose.
Here is a very unofficial and preliminary snapshot of the standard. Your thoughts?

============================================
Post representation

Let H160(x) denote the last 160 bits of SHA-256(x).

Let sbase64(x) denote base64url(x) (as defined in RFC 4648) with all non-encoding characters stripped (including the "=").

Each post is identified by the sbase64 representation of its H160 hash.

All strings are UTF8 unless otherwise specified.

A post consists of the following concatenated fields:

  <thread> '\n'
  <timestamp> '\n'
  <author> '\n'
  <email> '\n'
  <pk-data> '\n'
  <body>

Except for `body', the fields MUST NOT contain the '\n' character. The '\0' character MUST NOT occur anywhere inside a post.

If a post is not a thread starter, then the `thread' field references (by its sbase64'd H160 identifier) a thread starter post.

If a post is a thread starter, then the thread field consists of the following:

  <board-name> ' ' <thread-title>

where `board-name' MUST NOT contain the ' ' character.

Except for `body', the fields MUST NOT contain the '\n' character. The '\0' character MUST NOT occur anywhere inside a post.

The length in bytes of a post MUST NOT exceed 131072.

All fields above MUST NOT contain any '\0' characters.

`timestamp' is a string representing an integer representing Unix time at which the post was created. The timestamp MUST NOT be smaller than the timestamps of the posts it references within. The timestamp MUST NOT be more than 600 seconds in the future.

If the `author' field is a zero-length string, a viewer SHOULD display it as "Anonymous".

`pk-data' MUST be either a zero-length string OR the following:

  <pk-version> ":" <ed25519-public-key> ':' <ed25519-signature>

where `ed25519-public-key' and `ed25519-signature' are the 32 byte public key used to sign the post and the 64 byte signature respectively, encoded using sbase64. The value of `pk-version' MUST be "0". The signature is taken over the post itself excluding the `pk-data' field and the newline right after it, i.e.

  <thread> '\n'
  <timestamp> '\n'
  <author> '\n'
  <email> '\n'
  <body>

If the `pk-data' field is nonempty, then the signature contained therein MUST be valid.


============================================
Sync protocol

Please note that in this document, we only use "client" and "server" to distinguish between two peering users. The "client" and "server" do not play different roles. Really, the only difference between the client and the server is who gets to do requests first.

The initial handshake is as follows:

C: "DISTBB V1 SYNC" '\n'
S: "SYNC DISTBB V1" '\n'

The client may then place some requests (as described below). When it is done doing so, it MUST send the message "YOUR TURN\n" which signals to the server that it can now begin placing requests on the client. When the server is done, it should simply close the connection.

A request is either a "LIST" request, or a "GET" request. A "LIST" request is as follows:

  "LIST " <newer-than> <public-key>* '\n'

where `public-key' is

  " " <pk-version> ":" <ed25519-public-key>

and `pk-version' and `ed25519-public-key' are as described in [representation.txt].

`newer-than' is an integer representing the Unix time that the posts which will be returned must be newer than (according to their `timestamp' field).

The response to a "LIST" request is a space-separated list of

  <matches-key> <post-id>

followed by a newline. `matches-key' is "!" if the post referred to by `post-id' is signed by one of the keys listed in the "LIST" request, or the empty string otherwise. `post-id' is, as defined in [representation.txt], the sbase64 of the H160 of the post's contents.

A "GET" request is as follows:

  "GET " <post-ids> '\n'

where `post-ids' is a space-separated list of post IDs (as described above).

The response to a "GET" request is a list of posts, with a '\0' terminating each post.


[[TODO: GETALL request to speed up initial sync]]


Example session (strings have been shortened for clarity):

C: "DISTBB V1 SYNC\n"
S: "SYNC V1 DISTBB\n"
C: "LIST 1377043210 key-1 key-2 key-3\n"
S: "post-1 post-2 !post-3 !post-4 post-5\n"
C: "GET post-3 post-4\n"
S: gzip_of(post-3-contents... '\0' post-4-contents... '\0')
C: "GET post-5\n"
S: gzip_of(post-5-contents... '\0')
C: "YOUR TURN\n"
S: LIST 1377030000 key-4
C: !post-6 post-7
S: GET post-6
C: gzip_of(post-6-contents... '\0')
S closes connection.
C closes connection.

>>140
For a moment I thought you were using XML and I was about to cry.

I'll print it out and read it while I can get access to a computer with stable Internet. Sorry it it takes me hours to make a reply.

>>141
XML is like violence.

The spam seems to have been deleted.

>>143
Consider this: A pack of wild Ni/g/gers.
Savage, topleling Ni/g/gers nearing your /prog/ home. Trampling your /prog/ lawn. Raping your front page.
And you can't do shit since they're oblivious to criticism. The Ni/g/ger leader grabs your board and fucks it with his Daily Programming Thread.
The primal Ni/g/gers finally dominate your /prog/hold. They make barbaric posts and you are forced to be their slave.
Such is the downfall of /prog/rider.

Good god, it was stopped really easily. I'll try the other alternative, but it looks like mods are much more careful now. Not sure if I can keep up with this.

Wow, they deleted the spam. I'll try to be less obvious next time.

Looks like they deleted some threads too.

>>140
Everything looks good.

>>148
Nice bump.

>>148
Please be more careful next time. I already set another bot to do the flooding twice as fast, but please, please, be more careful.

>>1
You don't have multiple machines you little douche. It's not hard to create a CURL prog to auto-post, and from the intervals I have seen you are posting from 1 machine you little fa/g/get. Don't kid yourself.

Back to /g/ from where you came, you learnt some basic programming, like every new fagget does when he comes here, but they all get bored of auto spamming, just like you will.

Now go back to your shitty desktop thread board.

HAX MY ANUS

>>149-150
I'm truly sorry.

Good G-d, one bump and we let retards like >>151 in.

>>153
Pretty amazing, huh.

>>153
Age has become indirect thread murder on this board. Be careful.


(defun cum/g/uzzler_test (BackTo g)
  (if (op >= fag)
     BackTo
      (cum/g/uzzler_test (improve BackTo g) g)))

Now leave the EXPERT PROGRAMMERS in peace. Like a baby crying for attention you are disrupting the flow of our cdr, and preventing the final stages of Satori.

>>156
You aren't fooling anyone.

Suddenly I feel like stealing from threads from Tablecat's Lounge and reposting them here.

>>140
Sorry it took me so long to reply. I can't really find any flaw on your proposed standard, and it has the RFC quality MUSTs, so it can't be bad.

>>156
Please stop it already. Your abuse of ``memes'' was the biggest and most obvious giveaway.

>>158
threads from Tablecat's Lounge
The sad/funny thing is that, if the moderation paid us a visit again, these threads probably wouldn't be touched in the slightest even though they are part of the flood.