[ home ] [ GET / ref ] [ test ]

/GET/ - Hangout

Random
Name?

This field is optional. You can choose any name you want, or you can post anonymously by leaving this field empty.

Tripcodes are a way to identify yourself between posts without having to register with the site. To use a tripcode, enter your name as ‹name›#‹key›.You can choose anything you want as the key—it is private and will never be shown to other posters or stored on the server. For example:

Rarity#bestpony → Rarity!.4PK7yxdII

If you want a tripcode containing specific words, you can download a program designed to search for tripcodes, such as Tripcode Explorer.

Email
?

Entering an e-mail is optional.

There are also code words you can enter here which perform certain actions when you submit your post.

  • sage — lets you post without bumping a thread.
  • nonoko — uses the original post behavior to redirect to the board index.

These can be used at the same time as an e-mail address by typing ‹email›#‹action›.

You can also use Skype names in place of an e-mail. The notation is the same as a link to a username on skype itself, which is skype:‹username›

Subject
Comment?
Giving emphasis
[b] Bold [/b] Ctrl + B
[i] Italic [/i] Ctrl + I
[u] Underlined [/u] Ctrl + U
[s] Strikethrough [/s] Ctrl + R
Hiding text
[?] Spoiler text [/?] Ctrl + S
[h] Hide block of text [/h] Ctrl + H
Special
[rcv] Royal Canterlot voice [/rcv] Ctrl + K
[shy] Fluttershy voice [/shy]
[cs] Comic Sans [/cs]
[tt] Monospaced [/tt]
[d20], [4d6] — Dice rolls
URLs and linking
Link to a post on the current board
>>1234
Link to another board
>>>/pony/
Link to a post on another board
>>>/pony/1234
Hypertext links
[url=https://www.ponychan.net/] Ponychan [/url]
File
Flag
Options
Password?

This field is for editing and deletions.

File: 1706196587924.png (117.68 KB, 453x331, 139637488159.png)

⛵Country code: blank.gif, country type: blank, valid: 98377

Watcher README.md:

>Redis is recommended but not required (though MySQL is required), and the watcher should be able to function without a running Redis server. If you do not have Redis, remove the redis section from config.json

>removes Redis

./tinyboard-watcher
10:15AM FTL main.go:40 > Unable to connect to Redis error="no Redis connection configured"

suggestions?

Zeke Roa's phone  98378

That was written before I was involved.
If you check out the branch client-core-typescript, does the original watcher require redis?

⛵ 98379

File: 1706201920044.png (302.45 KB, 1200x1023, 5761450.png)

>>98378
I'm currently using client-core-typescript from /master, but it's non-functional. The watcher is there, and it lists threads, but all threads are marked as "Thread not found"
I'm not sure if this has anything to do with Redis, but I didn't get any errors while building it.

Zeke Roa's phone  98380

>>98379
As it is now, the watcher rewrite requires Redis. I meant the watcher directory from the client-core-typescript branch, not the master branch. Cct has the latest version of the old watcher, which is written in pseudo JavaScript/Typescript (Flow). It's old and a bit creaky, but that watcher should still work, possibly without requiring Redis, though I didn't make it so I could be wrong

⛵ 98381

>>98380
Would I be better off just using Redis? I'd rather not, but if that's what works best then I will. I'm going to take a break from this for a while since it's late and I've been working on it for hours. but a couple of questions:
1) If I use the CCT version, would it be just an overwrite of the js directory? or just part of it? possible conflicts?
2) If I use Redis, what happens to the current watcher? Do I have to disable it somehow?
Thanks for your help

Zeke Roa's phone  98383

>>98381
You may as well just use the CCT version, at least of the watcher. CCT has its own issues with the frontend, but the CC and CCT frontends are effectively unsupported, as while new-frontend isn't finished yet (I've overextended myself with projects a bit, and have IRL stuff as well), it's the version that is being worked on
If you use Redis, you'll need to restart the watcher if you want to keep the current version you're using, but it should work

⛵ 98385

File: 1706235258269.gif (928.97 KB, 700x394, 579056.gif)

>>98383
Alright, I'll give it a shot; thanks. Just so you know, the repositories are kind of confusing, especially since using 'git' virtually always downloads a completely different version than the one you're looking at, which is frankly idiotic on the part of bitbucket. I've gotten used to it since then, but still.
It's cool that you're still working on it, but I'm kind of surprised. Wasn't the notion that you were distancing yourself from the project the main justification (or rather, pathetic excuse that they used) when arguing why the site had to go down? Seems they really were full of shit all along.

Zeke Roa  98392

>>98385
I'm not sure how you're cloning it, but the default branch is set to new-frontend, since that's the one that's currently in development and has a lot of work done on it as opposed to a small branch with a just few commits to be merged. When it's complete, it'll be merged back into master, and I'll mark that one as the default branch.
If you run git clone --recurse-submodules https://bitbucket.org/ponychan/ponychan-tinyboard, it should get the new-frontend branch, as well as the frontend and watcher submodules

⛵ 98572

File: 1706845260315.gif (1.03 MB, 500x281, 1468888565333.gif)

>>98392 client-core-typescript also requires Redis. so I gave in and just installed Redis and went back to the watcher linked from /master/. TCP was already listening on 6379, so I changed Redis to 16379 (which might conflict with the cluster-bus, but shouldn't since it should theoretically remap, but hasn't). So since I have no experience with any of this now...

1) Does Redis need to be reachable by the client directly?
2) In the watcher config, what is listen port 4000 in reference to?
2a) Does port 4000* need to be accessible to the client?
3) Does this require a pre-existing Redis database, or will it create one on its own?
4) Does Redis "cache" refer to the cluster bus?
4a) Is the port the cluster-announce-port or the cluster-anounce-bus-port, or something completely different?
4b) Does this also have to be accessible to the client directly?
5) Do I ignore "recaptcha"?
6) Does "templates" refer to templates from $WATCHER/templates, which is {filters_dashboard.html, layout.html, script_page.html}, or something else?
This post was edited by its author on .

⛵ 98573

File: 1706857261364.gif (1.23 MB, 322x253, 1395292818783.gif)

>>98572 ok scratch most of that, but now it's too late to edit my post. I found that Redis was conflicting with itself, so I have it running on the default port 6379. The Watcher is "running" but isn't working. I have no idea why at this moment since I'm not sure whether things are mapped properly. so the remaining questions are:
1) Does Redis need to be reachable by the client directly? I assume no?
2) In the watcher config, what is listen port 4000 in reference to?
2a) Does port 4000* need to be accessible to the client?
3) Does this require a pre-existing Redis database, or will it create one on its own?
4) Does Redis "cache" refer to the cluster bus?
4a) Is the port the cluster-announce-port or the cluster-anounce-bus-port, or something completely different?
4b) Does this also have to be accessible to the client directly?
5) Do I ignore "recaptcha"?
6) Does "templates" refer to templates from $WATCHER/templates, which is {filters_dashboard.html, layout.html, script_page.html}, or something else?
7) ./tinyboard-watcher is finally running with no logged errors. Yet it isn't working. How would I check to see if it's making changes and/or watching anything?
INFO keyspace on DBs 1 and 2 both return nothing.
This post was edited by its author on .

Zeke Roa  98585

>>98573
1. No, Redis is just used server-side. Depending on your setup, it might be best to only have it use a socket.
2. 4000 is the port that it listens on for requests to /watcher. Nginx passes requests to /watcher to the watcher service, whether the original JS one or the Go one.
2a. No, only server side.
7. in Redis, run
SELECT 1
KEYS *
to get a list of the keys that TinyBoard is writing to, and
SELECT 2
KEYS *
to get a list of the keys that the watcher is writing to.

The DB #1 is referred to as cache in the watcher config because that's the Redis DB that TinyBoard uses for caching. 2 is only used by the watcher AFAIK, though I think the watcher sometimes checks on 1, like the all_boards key
This post was edited by its author on .

⛵ 98591

File: 1706949569322.gif (845.71 KB, 700x394, nod.gif)

>>98585 I'm not using nginx. Could that be the problem? Is there some sort of nginx built-in function, or would it be running a script I can peruse?

>(empty array)
in both cases

ah ok, that's cool.

Thanks again

Zeke Roa  98593

>>98591
I don't think so. As long as you can see something (aside from an error page from your server) when you go to /watcher in a browser, that part is fine.

It could be that Redis isn't configured properly in TinyBoard. I believe when you do a rebuild or add a board, it updates the all_boards key in Redis.

⛵ 98597

File: 1707003683352.gif (547.74 KB, 576x324, 575943.gif)

>>98593 tried rebuilding first, still empty arrays.
added a new board, still empty arrays.
How would I test if it's configured properly? I know the pass/auth is correct. port is correct. db's are 1 and 2. host is localhost. the prefix in the config is "watcher_" which I haven't touched.
watcher log indicates normal start of server when running ./tinyboard-watcher and gives no errors.
This post was edited by its author on .

Zeke Roa  98618

>>98597
Are you seeing any Redis-related errors in the TinyBoard error log?

⛵ 98675

File: 1707203235702.gif (1.92 MB, 400x345, 1395501986649.gif)

>>98618 not that I have seen. the redis-server.log is empty. The Apache logs (where all the board errors show up) say nothing about Redis.

Zeke Roa  98697

>>98675
No, the TinyBoard log. If you have it enabled, it should be in /var/log/tinyboard

⛵ 98717

File: 1707290136231.gif (928.97 KB, 700x394, 579056.gif)

>>98697 There's nothing there.
I see. All of the $config['*log'] entries were commented out in config.php. Uncommenting them so far hasn't enabled logging. I'm looking for whatever may be holding things up.
needs www-data: permissions instead of root:. Logging is up and running. will see what I can find.
No Redis errors in the tinyboard logs upon running ./tinyboard-watcher.
I have noticed that checking developer tools on the site itself, there's the error:
Failed to refresh watched threads SyntaxError: JSON.parse: unexpected end of data at line 1 column 1 of the JSON dat

On an unrelated note, I'm trying to fix the embed Youtube video functionality. I've tracked down the functions and confirmed that the iframe is being generated properly. However, it seems to be appended to the page being generated as escaped text; i.e. it doesn't actually recognize the <iframe></iframe> as anything other than text. I think this has something to do with
$this->embed = embed_html($this->embed);
but I'm not sure what to make of this as I don't understand the syntax. Any suggestions?
This post was edited by its author on .

Zeke Roa  98773

>>98717
>I have noticed that checking developer tools on the site itself, there's the error:
>Failed to refresh watched threads SyntaxError: JSON.parse: unexpected end of data at line 1 column 1 of the JSON dat
What is the url (or at least the path) where you're getting that?
>On an unrelated note, I'm trying to fix the embed Youtube video functionality. I've tracked down the functions and confirmed that the iframe is being generated properly. However, it seems to be appended to the page being generated as escaped text; i.e. it doesn't actually recognize the <iframe></iframe> as anything other than text. I think this has something to do with
I don't think I ever touched Youtube embeds, aside from having it replace the URL text with the video title, so I don't know what's going on there.

Anonymous Comrade 98950

yo zeke

not sure if it's something the site wants, but it's possible to build a script to automatically scrape archives (web.archive.org) and re-insert (as in, retaining post number) old threads, especially for slower boards like /ref/.
note: around 10-2023, Cloudflare blocked images and threads being scraped, so they have to be from before then.
This post was edited by its author on .

Zeke Roa  98957

File: 1707973931815.gif (39.53 KB, 369x237, typing.gif)

>>98950
That isn't really necessary now, I have my own archival script backing up database and upload data so if the whole site were to get nuked, we'd be fine, if a couple days behind.

Delete Post [ ]
Edit Post
Posts on this board may be edited for 2 hours after being made.
[Return] [Catalog] [Top]
[ home ] [ GET / ref ] [ test ]