Fix the terrible generation algorithm to be more efficient so that threads can bypass 4k without causing the board to crash. Maybe instead of rewriting the whole thing try appending new posts? Who knows.

change the text on the settings button
i don't think it's really new anymore

Can we get some bbcode-like memes? Like I cound post [topkek] and have all the post be topkek-y?

Or have [doge] and have le doge faic!

Or have [walruse] and make a walruse noise for people who turn that on in settings

or like have [meme] and have it generate a random meme!

>>159883
That might possibly be an inherent issue with Kusaba X (I could be wrong though). But regardless, maybe the better solution would be to implement a posting cap. I think the bigger concern is not that kusaba can't handle a thread over 4000 posts long, but rather, the browser being unable to open them.

>Fix the terrible generation algorithm to be more efficient so that threads can bypass 4k without causing the board to crash
What do you mean?
And appending the posts while keeping the thread/board structure intact would be far more trouble than it's worth, and if anything, would slow things down.

>>159886
What do you mean?

>>159887
Maybe.

>>159890

>>159892
Oh, that's there if you don't have Ponychan-X. But I'll fix that to make it less ugly, and maybe have Ponchan-X on by default

>>159890
Speaking of full threads only. Kusaba X has a terribly inefficient algorithm whereby it has to regenerate entire threads. All threads that reach the 4500~ mark cannot be regenerated anymore because it takes too long and PHP times out or otherwise crashes (but instantly reboots), and replies can only be seen from the board page.

And for the love of God don't install dog memes or "memeicons" like that please.

Last edited at Wed, Aug 6th, 2014 20:51

>>159898
Oh, I see what you mean. On my imageboard software I'm working on, gochan, I'm going to have it use separate pages instead of just a +50. Like <postnumber>p1/p2/p3.html

>>159899
That could work. But please be mindful, if you're going to do that, in stuff like Gala threads you are going to duplicate the header and footers by more than twentifold (space, space, space!). Splitting them up like that might be a good initiative, but consider including only the posts in those files and use the page buttons to give JavaScript instructions on which file to load. Will also look smoother that way - just ajax the contents and on success replace the content with the data.

Last edited at Wed, Aug 6th, 2014 21:01

>>159898
>"memeicons"
good idea! We could have customized favicons for our favorite memes!

Really though, yeah, don't do that.

Please add .webm support so that every visitor from 4chan can finally shut up about us not having it.

The "old" list: http://www.ponychan.net/chan/predictionsandprophecies/res/13055.html#13090

I don't know if the suggestions are still relevant but you could give the list a look.

>>159902
That shouldn't be too difficult. 4chan's limit is 2 MB, right?

>>159900
Well yeah. But it might/will probably be more trouble than it's worth.

>>159906
I'll add some of those

>>159888
How'd I miss this?

But no, nobody expects you to open a 4k post-thread, and a post cap is the worst thing we can do. The solution listed a few posts above is the answer to people not being able to view previous posts, but the problem is that even if you're in the +50 version of a 4k post thread, the thread will no longer able to generate as the website is trying to generate 4,000 posts at once.

Either the splitting up of posts inherently (as to physically not being able to open more than 50 posts without going to a new page) or reducing excess SQL queries and/or data is the answer to the problem. I wouldn't be surprised if the stupid thing is sending 4,000 queries instead of building one. To debug this, I strongly recommend timing the process and saving the time passed at several breakpoints and outputting to some file. This will allow him to find where most of the processing is being wasted.

The short-term solution though would be to build the +50 page before the full one, I will footnote. Until OP figures out a workaround, this will save threads from being killed.

Last edited at Thu, Aug 7th, 2014 16:54

How about.. fixing the youtube margins so that it actually looks nice?
Really really, it just needs a tiny bit of CSS..

http://www.ponychan.net/chan/test/res/36068.html#36297

Last edited at Thu, Aug 7th, 2014 18:26

Regarding video embedding: Based on some /test/ing I did two or three months ago, it was no longer possible to embed Blip videos, at least back then. And also, didn't Google Videos close down a year or two ago and shouldn't be necessary anymore?

This is however not the most urgent of suggestions I've had over the years, as the only use I'd have for Blip.tv embedding would be to link to videos from the That Guy With The Glasses site.

And while we're at it, would Dailymotion be a site worth adding embedding support for, even if just for those who embed FiM episodes?

Personally I'd suggest letting people embed soundclouds more than webms

>>159913
Added your bit. It might take a bit because of CloudFlare's caching, but if you do a hard refresh, it should fix it.
Really though, that Youtube video margin (or lack thereof) has been there a while, I don't know why it wasn't fixed before

>>159919
I've never used Blip, and barely even know what it is. And I think so, I'll just remove that.
I'll see if I can add DailyMotion

>>159920
Will do. Unlike on 4chan, people here have just been using GIFs, which makes them look silly when they only use GIFs and then complain that using webm embedding is too difficult.

Alright, Blip embedding works, but it's a pain in the ass to get the embed ID, and you can't even play it here anyway, it just links you to the video on their site. Should we even bother having it?

>>159922
I'd have no use for it, but I feel like I'm one of like three people having input here. I'd say not to turn it off until you get more input somehow (blotter?).

>>159923
Eh, leaving it there doesn't hurt anyone, and I doubt anyone will use it anyway. Like I said, getting the video id is annoying, and people will complain about it more than they did with webm embedding (which is relatively simple)

>>159926
I saw that one, but they aren't the same, and it seems to increase the webm margin

>>159891

>>159927
That's weeeird.
Is http://www.ponychan.net/chan/test/res/36068.html#36328 any better?

>>159930
>>159927
Oh I think I see, for you it looks like the interface was resizing to the resolution of the gif I used. But the actual size of the video container didn't change and just stayed at 200.. so it makes it look like it has a huge margin.
I don't know how to fix it though. I can't even get any browser to do what yours is doing.

>>159930
Nope. Webms still have a crazy margin. I'll look into it later today

>find out the problem was caused by me using Aurora (unstable Firefox build) instead of the main channel

>>159964
Haha!
Please add the CSS.

also
>Aurora
You really are Zamoonda, aren't you

>>159964
As soon as I get my VM set up, I will. I got the network fixed, so that shouldn't take too long

>>159966
>You really are Zamoonda, aren't you
I've been spotted!

Webm compatibility please.

  >>160132
>implying

>>160136
Now we just need someone to paste the CSS in img_global so that it has margins and doesn't look awful.

>>160136
Oh when did that happen?
huh..

>>160138
happened about a week ago as far as I remember. There was as sticky on /oat/ and everything.

>>160137
It's a caching issue. Do a shift+refresh. What browser are you using?

In testing the creation of a mobile version of the site, I have managed to create a CSS that resizes the site in a very pleasing way. I have tested it and have come up with some interesting results.

Through tons of trial and error, I was able to resize and reorganize quite a few things to fit a mobile screen.

It's done by placing the following CSS at the end of the main CSS file:

@media all and (max-width: 999px) {
@viewport{
zoom: 1.0;
width: extend-to-zoom;
}
table {
display: inline-block;
table-layout: fixed;
width: 100%;
}
.postblock {
width: 90px;
}
textarea {
resize: none;
width: 280px;
}
.navbar {
overflow: auto;
text-overflow: hidden;
white-space: initial;
}
#verytopbar {
position: relative !important;
}
#mainmenu {
position: relative;
float: right;
}
#bodywrap1 {
margin-top: 0px !important;
}
}

And placing the following meta element near the beginning of the html:

<meta name="viewport" content="width=device-width,initial-scale=1.0">

You can test it to see what works.
The biggest issue is the form for posting, but other than that, everything seems to scale nicely.

Last edited at Thu, Aug 14th, 2014 09:28

Also, there is a bug with the edit button.

If you go to edit a post with a BBCode "code" tag in it, it will rip the code tag from where it was and place it directly beside the "endcode" tag.

>I learned this while editing the above post

Last edited at Thu, Aug 14th, 2014 09:33

Also, can we make the derpy permanent? Something like a chan pet or something?

>>160169
>

>>160170
Is that why it doesn't seem to be displaying correctly?

>>160172
You can't do it with the CSS test because it replaces the theme completely.

That and the HTML meta element won't be counted by the browser until the site itself reloads.

It'll need to be tested by going directly to a live site with the code already applied.

Last edited at Thu, Aug 14th, 2014 10:13

>>160173
Oh, alright
and what does "<meta name="viewport" content="width=device-width,initial-scale=1.0">" do?

>>160171
I'll add that as an option turned off by default. I'll also make the settings part of the settings menu instead of its own page.

>>160174
This tells the mobile browser to force the correct zoom and resize according to the size of the screen.

Without that, many mobile browsers will still display a zoomed out and full scaled version of the site, despite the site's ability to be resized.

>>160175
Nice, that's perfect then.

I suppose to go along with the "no longer new settings area" bit, why not turn Ponychan X on by default, delete the settings options that don't actually work, and then condense all the settings pages into just one thing?

Also, possibly don't have them on separate pages. I think PchanX actually just works in a pop up window.

>>160169
http://imgur.com/uAQiSzB,MPyZQC1
What it looks like on my phone, kind of. It keeps making a really big empty spot to the right.

It shouldn't be too hard to add some extra CSS to this to make it looks really nice!
i'd have fun with it tonight, but i'm busy and tired..

Last edited at Thu, Aug 14th, 2014 17:53

>>160180
An empty spot to the right? Then there is something that isn't collapsing as it should.

I'll have to see what is sticking out and experiment a little more. Is it a comment box, the top menu, the navigation bar at the bottom or what? I'm willing to bet it's most likely the form at the top for posting a reply or a new thread.

If it's just an empty space on the right from top to bottom, then I'll have to see it and mess around til I find it.

It's still a WIP, and hopefully I'll find a good way to make it work.

Last edited at Thu, Aug 14th, 2014 18:08

I have a habit of not reading stickies. I should've posted this thread into here I suppose. >>160190

>>160192
Addendum: Make it a dropdown menu like the reverse imagesearch feature that 4chan has, call it 'Share' or something and have it automatically add a shortened url to your clipboard.
Addendum to the Addendum: Add reverse imagesearch, possibly through a tiny dropdown menu.

>>160199
Drink with me, Fen.

>>160202
No alcohol, too tired and too late, maybe next time.

>>160199
Ponychan-X already has a reverse image search

>>160480
It's a ☲ for me, and i have never noticed that until now, god damn.

Also

/r/ OP's being unable to delete their thread once it reaches 50 replies or a length of time has gone by, like an hour or something.

>>160490
>OP's being unable to delete their thread once it reaches 50 replies
Yes.
>or a length of time has gone by, like an hour or something.
But what if the thread's not doing well (like two replies after that hour) and they're like 'fuck it' and nuke it to make way for something else/better

>>160490
Isn't it safer to just ask for a mod to undelete a thread rather than prohibiting people from deleting it in the first place? Just add an option for password salting in the thread undeletion. Or just add a scramble password capability all in all.

>>160490
>OP's being unable to delete their thread once it reaches 50 replies or a length of time has gone by
Really? They can't?

>>160498

No, people can delete them fine, he's saying he wants them not to.

>>160492
>and they're like 'fuck it' and nuke it to make way for something else/better
The thread could possibly pick up later on in the day, and if not then it would just get pushed off the board as if it were bump locked.

>>160493
Because mods might not be around 24/7, and an OP deleting their thread because they didnt like what has happening is an occurrence I've seen a few times.
The thread getting deleted also screws with the feel of the thread, and usually kills it even if it's undeleted afterwards.
I just feel as if it'd be better if an investment of people's time like that wasn't taken away just because of butthurt or some other reason.
It usually happens past 100+ posts, but i said 50 because it fits with the +50 link.

>>160500
I've only ever deleted one thread, and that was a duplicate, but still preventing a thread owner from being able to get rid of their thread doesn't seem right to me. I actually don't think I've ever been part of a thread that was deleted by OP that didn't need to get deleted in the first place, badly. At least the users can then have some semblance of quality control since this site sorely needed it. If my thread derails into user flirting and circlejerking because a few users stepped in you bet I'm deleting it.

>>160500
Fair enough - some just might stop caring for their thread after a while, then again if they stop caring then they probably won't care if someone finds it on page 8

>>160501
You might as well report it for deletion/sage/lock/cleanup, which in my opinion would be a better way of handling it

>>160501
Would you delete a 200+ /meta/ or /pony/ thread because some guy's derailed it right at the end?
Because thats the kind of thing i'm talking about, i guess when you put it like that though i can see your point.

Seems to be more a case of moderators not doing their jobs though, in that instance.

Could go either way i suppose.

So then is the general consensus that if a thread has been up for an hour or more, it can't be deleted unless a mod deletes it? Are there any other circumstances that should bar it from being deleted by the OP?

>>160509

I think the times that a user really needs to delete their own thread are few and far between, personally. The most obvious and plentiful case is "Oops, double post." In which case they'll probably notice there's two fairly quickly and delete one. But if the thread is active with people discussing stuff I think it goes against culture a bit for the OP to be able to shut down discussion. If the OP has a problem with it they should be able to request a deletion so the mods can look at it and see why, but with how rare that situation comes up and how it can interfere with stuff on some boards like Fennec was talkin' about, I think it would even be safe to limit that time to 15 minutes or less.

I dunno, though, just my opinion.

>>160509
Since this seems to might actually become a thing, I'm going to have to be forced to continue this argument, aren't I?

>>160503
You mixed and matched a bit. A thread in /meta/ that gets deleted doesn't mean the, pardon the term, temp "community" sort of that resides in it will vanish, the thread only resets. Take for example the thread asking for the deletion of /meta/ - it was heavily derailed with stupidity and the OP simply deleted it and restarted over. Was this a bad thing?

If an /oat/ thread gets to 200+ posts and the OP decides to delete it, why would the OP decide to a delete a thread for any reason other than a derail? I have never seen any instance where this was not the case. The ending of shitty derails that go against the topic of the OP is something that the mods should be enforcing already; the only reason an OP would delete it is if the mods refuse to follow the rules they themselves have written down.

It is much more practical to ask for a mod rather than make an automated system. I remember a time when we had at least one mod available at every hour of the day, if a person has to wait more than two hours for a mod to respond you are clearly low on moderators.

If a system has to be made, make it 200 posts with no time limit. I can see that by the 200th post the thread must have surely spiraled to the point where the OP's opinion doesn't matter, but anything before that / time limit is silly.

>>160512
So then they can't delete it after an hour or if there have been at least two posts in the last 15 minutes

>>160513

I agree with you that thread deletion due to derails should be able to be handled by mods. It's an easy thing to handle. That's why I think it's okay to leave that up to the mods.

Basically, there's two potential cases for deleting a thread. One in which it needed to be deleted because it had gone to shit, in which case the mods should be able to respond to and delete the thread. Another in which the OP didn't like something or other about the thread, but it's actually still totally on topic and the OP deleting it would just upset the people talking, though there is something to be said for making a new thread in that case.

Neither case actually seems all that common, though, so the real question kind of comes down to a discussion of ownership. Does the OP have the right to do what they want with their thread, even if it disrupts other posters, or do other people take ownership from the OP by posting in it? Both have problems, and both of those problems would have to be fixed by mods. From my point of view, it's easier to delete a thread a bit late, than restore a thread after it's been deleted (as was noted, the interruption in discussion can kill it outright), so I'd lean towards leaving thread deletion up to the mods.

>>160514

I think that would solve the problems Fen's describing while not causing any new ones, yes.

Did sound like some people think an OP should still just always be able to delete stuff, though? So maybe hear them out, first.

>>160515
Alright then



To update, should an OP be allowed to delete a thread if it was made an hour ago and no posts have been made in the last 15 minutes?

>>160516
We are heavily derailing this thread. This is a policy change so to speak, not a feature. If somebody wants to revoke OPs' rights over their thread - which since time immemorial, OPs owned the thread and had deletion privileges over their own topics - it should be a separate /meta/ thread for discussion.

Last edited at Sat, Aug 16th, 2014 22:28

>>160517
Not really. It's a feature/code change that is directly affected by a policy change. It's a slight tangent, but not enough to say it's completely off topic. I'd like to keep this on topic of features/bugfixes after this, but this tangent isn't bad.

I THINK THAT NOT BEING ABLE TO DELETE YOUR THREADS IS A DUMB CHANGE AND I DON'T LIKE IT

"..I wonder if I can still delete a few /pic/ threads"

If someone's evil enough to think like that then byebye 1000-post-Chrysalis-thread without anyone noticing because the place is dead

>>160523
>tfw we will never reform /pic/
Forever inferior to derpibooru it is.

>>160523
well yeah, if it has more than 500 posts, it won't be able to

>>160526
I could possibly implement some sort of tags system

>>160526
>>160527

What I imagine for /pic/ is more of a premium gallery type deal. 'cause really, otherwise we won't be able to compete with derpibooru on any level whatsoever. With a superior tagging system for a more limited picture set, we might be able to draw in people who don't want to sift through literally every picture of Fluttershy ever, instead only grabbing fan art or screencaps or something like that. It'd require a bit more refinement to put into practice, but I can see /pic/ having use as a database more than an actual board.

>>160532
That makes sense.



also, youtube video data coming soon

Last edited at Sun, Aug 17th, 2014 15:40

>>160527
>>160532
I don't really think tags are enough honestly. Derpibooru also has tags.

Let me start of by explain just how awful /pic/'s format is right now. First of all, it's an immense pain to save every image in a thread, if you so wish, unless you have something like downthemall, which I'm not sure every browser has. Even then, there is no way to get rid of duplicate pics unless you manually do so yourself, leaving you wish a really messy folder that you have to organize yourself. This is especially prevalent in those threads that have thousands of pics in them. It's just a complete mess. The /pic/ directory is a good idea, but there is still no picture duplication detection, so there's going to be a lot of overlap in the threads.

So if we want to do bare minimum improvement to /pic/ we need
1. Picture duplication detection across the board (which I know MLPchan's /pic/ had when it existed, so it's possible)
2. Tags
3. Some sort of way organize the pics that's not a thread (which may be weird for ponychan, but a thread is definitely a terrible way to organize a picture board.

Even then though, there is still no reason to use it over derpibooru, which should be our main concern.
My idea for it, which I haven't received comment on, so I don't know if it's possible, would be something like 4chan's old /rs/ board, in that it would be a way to fileshare your folders. To my knowledge, there is no other site like this for ponies, and the advantage it would have over derpibooru is that you wouldn't have to individually save every picture. Again I'm not sure how feasible this is, but it would be wicked cool if we could get it together.

Also as a side note, ponychan has been freezing for me lately. As in, I can scroll through the page, but I can't click on anything. This is only with ponychan and after a few minutes it goes away. Just thought I would mention it since it's thread appropriate.

>>160541

You're definitely right about the need for a mass downloader of sorts, but I know those exist. We just need to make it available at the top of a /pic/ sticky.

Same with duplication detection, definitely important, and if /pic/ is gonna serve more as a gallery than a board (which it should, because if people wanted a generic Rarity thread we've got /oat/ for that).

Tags should be easy, because Derpibooru has a horrendous tagging system with hardly any rhyme or reason to it and a bunch of worthless extra joke tags appended to make things extra hard to search through.

I think threads actually do a potentially decent job of sorting things, though. Each thread would basically be like a search term. Like..."Fluttershy" or "Humanized Fluttershy" and stuff like that. In the same sticky that we put the downloader (So, the directory) we could organize the threads in an easy to follow fashion, so they know exactly what "tags" are available and don't have to follow weird search terms and dig through stupid junk.

With that I think we'd actually fill a niche that Derpibooru doesn't. An image gallery that's more easily sortable and focused, without having to filter out porn or gore or just plain shitty art.

As an additional opportunity, and this might be more difficult, but as a potential idea we could set up a series of torrents for each thread, too. Update each torrent weekly and anyone who has the torrent will automatically snatch any new files added to the list. Could be cool stuff.

>>160541
>>160542
Saving pics from /pic/ is easier/fast than from derpibooru: alt-click the thumbnail and that's it

And dupe-detection is a thing in ponychanX

>>160542
>Derpibooru has a horrendous tagging system with hardly any rhyme or reason to it

What? Derpibooru's tagging system is perfectly fine. I don't see what's wrong with it.

You have the image rating, Safe/Suggestive/Questionable/Explicit. And then you have things like character tags, tagging every character in the image. Then...etc, you just tag whatever's in the image. Simple as that.

Like
>this image
has tags of "alfred hitchcock, artist:pixelkitties, big macintosh, crossover, goldie delicious, littlest pet shop, peter new, safe, sunil nevla, vertigo (film), voice actor joke"

What's so "no rhyme and reason" about that?

Alright, for all the coders on the site, I've started the Coders' thread.

My hope is that, if we can wrangle in enough coders, we can collaborate to create projects for features or fun little gags for the site, among other things.

>>160542
I think the one problem is that /pic/s lacking the sheer amount of images that derpibooru has. Though you do make a good point that theoretically it's possible to not have to sift through shitty art, that's not a guarantee unless a new rule on /pic/ is made where you can't upload shitty art, but then you have to define what shitty art is, and that's tricky.

Knowing what tags you have with the directory is a good thing. The problem is we have a very limited amount of tags. For instance, if someone is looking for fluttershy+clown pictures, unless someone has somehow made a thread about that, they're going to have a tough time using /pic/, while it would be quite easy on derpibooru.

I'm pretty sure we have this thing called "The Great and Powerful Archiver", but it doesn't work for macs, and it's a bit buggy from what I've heard.

Torrents would be interesting, but as good as that would be, it's just a fact that not everyone torrents, so you would get a lot less people downloading.
>>160543
1. That's not that much faster than derpibooru
2. Derpibooru has a lot more images than /pic/
3. I would like to know how this dup detection work on pchanX and how long it's been available for, because when I downloaded threads from /pic/ a while ago, there wasn't a duplication detection of any kind. I was using downthemmall though, so maybe it doesn't work with that or something?

Last edited at Mon, Aug 18th, 2014 11:41

>>160547
>that's not a guarantee unless a new rule on /pic/ is made where you can't upload shitty art, but then you have to define what shitty art is, and that's tricky.

Nah, it doesn't need any new rules. Derpibooru aims to be a collection of nearly all pony art ever made. Here, people just upload their favorites. It's a natural thing, doesn't require extra rules.

And if someone's looking for something as specific as Fluttershy+Clown pictures then...that'll probably have to go back to Derpibooru.

The thing is, we're not actually going to beat Derpibooru, and we shouldn't try. We should try to offer something different from Derpibooru. Not equivalent, not something that's going to compete with Derpibooru, just something useful enough that /pic/ isn't entirely obsolete.

>>160544

Lots of tags are undefined duplicates, there's a collection of weird in-joke gag tags clogging stuff up, even just comparing it to other boorus, the tagging system feels incredibly chaotic to me, like they just let the users make up tags on the spot and run with it.

Even in your example, did we really needs tags for "Vertigo" and "Alfred Hitchcock"? And while the artist tags are nice, it's a bit of extra clutter for people who might not care and just want to nab some Big Macintosh pictures.

It's true that it's not really a huge problem unless you're trying to look up porn, Derpibooru's porn tags are the worst, but like I was saying to Shuckle, we really can't actually compete with Derpibooru, all we can hope to do at this point is attempt to fill a niche that they are not. I feel that /pic/'s niche could be nice complete ready to mass download folders of various major tags, like one for each pony.

>>160548
>like they just let the users make up tags on the spot and run with it.

Uh...that's exactly what happens. Tags are user-defined, and anyone can edit the tags. (although a log of edits is kept to prevent abuse) But that doesn't mean anyone can just spam any tags they want. The userbase is pretty good about keeping tags neat and orderly. And fucking with tags is against the rules, so abuse is kept to a minimum.

>it's a bit of extra clutter for people who might not care and just want to nab some Big Macintosh pictures.

Then just search "big macintosh" ? That's really what the tags are used for: searching. Artist tags let you search for all art by a specific artist.

>>160548
>The thing is, we're not actually going to beat Derpibooru, and we shouldn't try. We should try to offer something different from Derpibooru. Not equivalent, not something that's going to compete with Derpibooru, just something useful enough that /pic/ isn't entirely obsolete.
Oh I completely agree, we can't outright beat derpibooru, but we can give people a reason to go here instead of there in certain instances. The one advantage that I think we can have is the ability to massive download images without individually clicking on and saving them. The two ways I think we can do that are
1. The /rs/-like idea I had (which again, I don't know if it's possible)
2. Have a built-in archiver like downthemall on ponychan where people could download the threads pictures.
The second one is probably the most realistic, but not as cool or innovative as the first.

/arch/'s:
"Return"
"Entire Thread"
"Last 50 Posts"

Overlap the OP pics ever since it was made read-only.

>>160547
1./2. If your point is that db is better than /pic/ for pics, OF COURSE it's better, all I said was that alt-clicking is faster than opening a new tab and clicking the downloadbutton.

3. >>157446
>Thu, Jun 5
It searches for the file you're about to post and tells you that it's already in the thread

>The /rs/-like idea I had (which again, I don't know if it's possible)
It already happened a long time ago, people just forgot about it, or aren't contributing to it anymore.
And that is the problem with /pic/, contribution, no one is there to post anything and no matter how much code you change or try to make it better there just aren't any people left

Last edited at Mon, Aug 18th, 2014 14:21

>>160560

We'd have to find and assign some people to maintain /pic/ as a gallery database rather than an ordinary board.

>>160560
Oh neat, well that's one issue solved. I guess I hadn't heard of it because it's so recent.
>It already happened a long time ago, people just forgot about it, or aren't contributing to it anymore.
How long ago? I never heard about or saw any of this.
>And that is the problem with /pic/, contribution, no one is there to post anything and no matter how much code you change or try to make it better there just aren't any people left
Well I do think part of that is because of the existence of derpibooru. Also the fact that /pic/ relies on people wanting new pictures, which means normally new users or new ponies, in order to get traffic, so unless we get new users, we aren't going to get new traffic. Therefore the hiatus will definitely hurt pic. I personally think you need to give posters a reason to post somewhere, instead of not improving something with the logic of "well nobody is there, so we shouldn't improve it".

>>160562
Good luck. really, i mean it
>>160565
>>/pic/922818

By all means, go for it.

>>160566
Oh right, that thread. Yeah I contributed to that when I was capping things and such. That's not exactly what I meant by the whole /rs/ thing. What I mean is a searchable database of downloadable folders with tags, sort of like derpibooru, but with zips, and on an imageboard.
I think that one of the reasons nobody used that is that it's limited to one thread and it just looks so unappealing. Though you may be onto something where maybe on /pic/ people just don't want to upload their folders, they want to post pictures. To which case the second option I suggested definitely sounds like the best.

>>160560
>It already happened a long time ago, people just forgot about it, or aren't contributing to it anymore.
I'm also curious about what you're referring to right here as well.
>>160566
Ideally, a directory of any sort should probably be automated to ensure the users that it is indeed up to date and reliable.
>>160548
>We should try to offer something different from Derpibooru
My suggestion is that instead of trying to align /pic/'s purpose to serve similar features to that of a booru that's already clearly doing well on its own, that instead we exploit the benefit of the fact that every last image that the user reposts from /pic/ is already located on the same server where it will be reuploaded? From what I've heard both here and on MLPchan is that image thumbnail processing is by far the largest chunk of server CPU usage, and the sheer fact that it was compromised with some client-end image thumbnailing via draft-spec html5 APIs [www.w3.org] on MLPchan is pretty telling if you ask me. I say we go a step even farther and make it possible for the user to not have to reupload any images at all.

I haven't thought of many suggestions of how exactly this could be done, other than maybe an image search in pX's quick reply that autocompletes (typing in "rain" would be enough of a query long enough to understand that the user is looking for images of pinkie pie). Then with a mix of randomness in results combined with some emote word-matching with contents in the post message to narrow the results if the user is expression viewerfagging (which most of us do, face it), then take the emote words that evaluated as true and match them in the appropriate /pic/ thread by filename, then return a list of at least 50% of the image results and the remaining <50% a list of random results of the same thread.

In theory if all goes well, not only would this effort give /pic/ a defined purpose as well as a reason for people to post there, but it would save the user time from doing the tedious work of reuploading images as well as saving massive amounts of cpu processing (the most expensive element of web hosting I should add) as well as the bandwidth it takes to both download and upload an image, effectively making the site faster than ever.

And since the catalog page would be requested by the client frequently, that page will likely be cached by the browser. And since most of us expression viewerfag, the search results will just return the same threads of the same pony, and thus that thread will likely be cached as well.

I'm just saiyan.

Last edited at Mon, Aug 18th, 2014 19:37

>>160593
>I say we go a step even farther and make it possible for the user to not have to reupload any images at all.
The only issue I see with this is the fact that it scans a thumbnail against its database everytime...and if you want to post a similar but different pic, it might get caught because of the inherent fuzziness of thumbnails compared to actual pictures.

I'm not really familiar with anything on thumbnail generating algorithms and cpu usage, but to me it seems like just as efficient to scan the image, symlink the image thumbnail, but upload the real image also and just not generate already covered thumbnails. So in reality the thumbnail is just a symlink or embed to the /pic/ pic and the image is what the user uploaded originally

>>160594
If the image has been changed in any way it will generate an entirely different result. No "too similar", either it is an exact copy down to the last pixel or it isn't.

Ontopic though, it's much more ideal and less process-heavy if the file is simply checked only against posts in the thread. Plus, this could also be a new option in /pic/ threads, i.e. whether to allow duplicate images. (The talk only with images comes to mind as a possible thread that will be harmed by a global and permanent no-duplicates.)

>>160595
Alright, I was thinking anon was referring to the reverse image search image matching algorithms being used for finding the thumbnail like derpibooru and google have that have a fuzziness setting.

Then yeah, I don't have a problem with this at all. That's in no way a bad thing to me.

Last edited at Mon, Aug 18th, 2014 20:01

>>160596
>You
Different anon.

>>160597
fixed.

Last edited at Mon, Aug 18th, 2014 20:03

>>160593
So what I understand from that is, you would search a query, and it would search pretty much every image that's been posted on ponychan, and then you would be able to post that image? And /pic/ would be a place to store those images? So it would give people a reason to post on /pic/ while redefining it's purpose? That sounds really awesome!
Only real problem I see with that people would actually have to upload images to /pic/ for it to work. That can be a pain if you have a large folder. So it's really dependent on others being willing to make it work, which, no offense to ponychan, I don't really have confidence in it's ability to make things happen as a community anymore.
Also it wouldn't really draw any outside traffic, which is part of my reasoning for how to re-purpose /pic/. Should I make a new thread for this? I feel as though this is derailing this thread.

>>160599
Does that not create an incentive for /pic/ to be alive again?

>>160599
Theoretically everything except the image search would *edit: search images in /pic/ but from other boards*, yes. Now the only thing I feel might be an issue now is that Zeke is probably overwhelmed with requests by now. So if it were to happen, it might take some time before it does.

Last edited at Mon, Aug 18th, 2014 20:59

>>160600
Oh it certainly does, and that's great, but it doesn't draw outside traffic, though I seem to be the only one actually concerned about that. Then again, /pic/ probably isn't going to be a calling card for getting traffic back on ponychan.
>>160601
So wait, /pic/ would be the thing that searches? What does
>search images in /pic/ but from other boards
Mean? Would it search all the boards of Ponychan through /pic/? So like /pic/ would be a search engine? I'm confused.

>>160603
What I meant was that all boards could utilize a search engine to get specific images from /pic/, the board that organizes them best.

>>160604
Oh ok, that makes sense. Then yeah that sounds good! We need to get someone on that.

Crap, my IP got reset after the power outage, so my watched threads list was kill

>>160559
I've noticed that as well

>>160688
Is Mithent's threadwatcher no longer a thing? If so, it's another thing you could work on.

>>160690
Might be too much for Zeke on top of everything else he has to do. If anyone has the source code and assuming the thread watcher actually needs updates, I'd be willing to work on it. Can't imagine it's run on some obscure language or, I don't know, Visual Basic. So I should know the language.

>>160691
Yeah, this is probably the way to go. Working on the thread watcher really doesn't require special site privileges because it was always unofficial in the first place. I'm sure Mithent would provide the source code if he's around, assuming he hasn't provided it somewhere already.

>>160690
No, I thought it's still a thing.

>>160691
By source code do you mean the site's source code? It's all in PHP

>>160694
I think he's referring to the desktop thread notifier that Mithent wrote three years ago hosted here:

http://www.ponychan.net/chan/PonychanNotifier.zip

>>160695
Oh heah that's pretty much dead. I think that was written in a .NET language, maybe Visual Basic but I don't remember. I started working on a Java based one for cross-platform compatibility, but I lost interest

>>160692
I released the source for it before, but I no longer have the hosting that was on. I can dig it out if desired but not for a few days. It's in C# (Visual Basic, gasp!) but it was very much My First C# Project, and I'm not very proud of the code now. As far as I'm aware, the server API hasn't changed, though, so it should still be functional?

>>160698
We have an API? I always assumed all of Ponychan's addons and interactions were just scraping! Is there documentation somewhere?

Last edited at Fri, Aug 22nd, 2014 17:46

If you want to see the title of an embedded Youtube video, just hover your mouse over the thumbnail.

>>160726
just pointing out, this only works if the video is already playing
not before you click on it or anything

>>160731

But that's how it was before! You sure he didn't change it? I guess I'll find out before too long here...

Would be nice to be able to tell what a video is before I play it, though, so if it wasn't just done already it would be a great thing.

>>160731

>>160806
I was shocked for a second upon seeing my own test post in an image. "Oh god did I do something wrong"

>>160808
Nah, that was just closest to my screen, so I used that.

I actually wanted to have the title show above the video, but I was having trouble getting it to fit right, and then I realized that with long titles, it would look like shit anyway.

>>160806
well at the time i made the post that wasn't yet working

Its minor, but autosage seems to be broken. I've seen several autosaged threads on several boards getting sporadically bumped by people who kept posting in them.

>>160971

Oh, yeah, I've seen that, too. I think I heard something about people "figuring out the trick"?

>>160976
I was claiming that but I'm full of shit.

>>160971
>>160976
I'll look into that. It could possibly have something to do with deletes or something, but I'm not entirely sure.

>>160971
This is a recent issue right?

>>160983

I have only seen it just recently with one or two threads, so I would call it recent. And not really universal, either.

i thought a mod was doing that to be annoying..

>>160999

Probably Orange trying to siphon nude selfies out of people.

Could you make the youtube embeds larger or something?
Not being able to control the volume is somewhat annoying.

>>161021
Alright, that shouldn't take too long. I just decided to have it have the same size as the images, without thinking.

>>161021
By the way, how did you make that GIF?

>>161042

Maybe you could make it so they're small prior to clicking on them but then larger after you click on them?

Random thought.

>>161042
Thank you. That happens.

>>161043
http://screentogif.codeplex.com/
I think there are better recorders out there but it did what I wanted it to do.
Or maybe I haven't fiddled enough with the settings.

How about giving /all/ the option to manually add hidden boards?

>>161044
>>161045
To make things cleaner, I'll do like 4chan. Set them to the full size of the screen, or if they're larger than the screen, keep them inside the screen. When it's open, there will be a close button to turn it back into the image.

>>161154
Adding that to the list. I haven't looked into /all/ much, but that might not be too difficult.

Last edited at Sun, Sep 7th, 2014 14:39

A second spoiler-(joke)-tag to cover up images so you can use the real spoiler-tag for real spoilers.

The mods would need to start enforcing the spoiler-rules though.

>>161197

I'm fairly certain that nowhere on the site has spoiler rules anymore.

>>161198
Damn, really?
If that's the case then remove the spoilertag and text

>>161200

Just because it isn't enforceable doesn't mean people won't politely spoiler things for people. Even non-pony stuff like a recent movie might have the cool ending spoilered by the guy raving about how great it was.

Also, there might still be rules for spoiling previews of things? Though I know that as soon as an episode has aired, all spoilers are allowed. There isn't a "grace period" anymore.

>>161201
>Even non-pony stuff
Yea that's why I'm not too mad about Doctor Who stuff I see here, good thing that hiding works.

>as soon as an episode has aired, all spoilers are allowed. There isn't a "grace period" anymore.
I knew that, but with EG2 and idunnoS5 coming up...
some people might care enough to avoid teasers and trailers and god-knows-what

>>161203

Well I have definitely not seen any of the EG2 stuff spoilered. That's not only left in plain sight, but also rather popular and tossed about everywhere. You could suggest that this be changed, but I think that takes a different thread.

>>161204
>Well I have definitely not seen any of the EG2 stuff spoilered
...
...
I'd write more and make a thread about the spoilerstuff but things came up and I have to go, maybe later

>>161205

I think we may have gotten "spoilered" (hidden from sight with warnings attached) mixed up with "spoiled" (the opposite of that, where you can see it).

But yeah, I should really get to bed myself.

Ponychan should bring back the /hi/ board

>>161207
What I meant was that I haven't seen any EG2-threads at all, I did have a couple of ponychanbreaks so I might've simply missed them.

I decided against making a thread for the current spoiler-rule-situation, I'll just leave if some s5-news show up.

>>161197
>>161198
>>161200
we do on /pony/ and /oat/ for new episodes

>>161222
I vaguely remember that, but I don't remember what it was about

>>161275

It was literally just an introduction board, as far as I can recall. It was where new people went to say "Hi, I'm new." Instead of going to /oat/ or wherever they happen to fall.

>Add a Celestia theme - 0%
^mfw I just noticed
I can't wait!

>>161274
Ok, so, what's it gonna be?

>>161346
What do you mean? The new thread spoiling policy obviously won't be applicable until S5 airs

>>161481
So every leak/trailer-threads about s5 or eg2 are acceptable without a spoiler ...until it airs? What?

Either way, my point is adding a fun-spoiler - yay or neigh?

Last edited at Thu, Sep 18th, 2014 09:00

>>161505
Oh yeah, forgot about that...but that isn't really an issue of technical concern, that's just an issue of spoiler rule enforcement.

>>161512
>that's an issue of spoiler rule enforcement.
Yes, you guys might want to talk about that.

But even if you don't change anything there, some people want to hide their potential spoilers of any show and some others just like to have a little fun, so consider implementing it regardless of rules.

>>161513
Can you show me cases of people ignoring the spoiler rule?

I believe what zeke is talking about is the fact that new episodes cant be spoiled in their entirety until the episode starts and we actually know what happens, while you are talking about bits and pieces of content being spoiled.

Regardless, the spoiler rule should be getting enforced for pony content, while a non-pony spoiler is something to consider.
I could toss up a thread on /oat/ and/or /pony/ and see what thumbnails people can come up with if we decide to add that.
Not sure how to implement an alternative spoiler through text other than have it be a colour different than black though, like white for instance.

And now i want a rainbow spoiler, Zeke pls.

Pretty sure the S5 and EQG2 threads are supposed to be spoilered.

Marimo probably forgot because we haven't had the need to spoiler anything in a long time.

I tried to report it to notify a mod to spoiler it, but it looks like someone else already reported it for something else. (most likely Marimo reporting it to get the thread stickied)

Last edited at Fri, Sep 26th, 2014 18:21

>>161925
/pony/ is mostly fine from what I remember

/oat/ had some EG2-character threads, which are probably dead by now

>>/pic/952452
>>/pic/952487
unless picdumps are good to go, the /oat/-ones were similar except with a little more 'mah waifu' in it


Aside from all that, I didn't bring it up because I saw oh-so-much unreleased content, but to get some answers about what's good and what's not (..and to suggest the fun-spoiler)

>you are talking about bits and pieces of content
Putting all those pieces together, I have a pretty good idea what's going to happen in Rainbow Rocks, without watching a second of trailer-material.
Just sayin'

>>161928
Threads should be spoilered now.

On a related topic;
Zeke, is it possible for you to tweak the "Spoiler thread" option so that the subject line is still revealed?


>>161947
*A lot of bits and pieces.

And those should have been spoilered, sorry about that.

>>161970
Sure, I'll check that out

Alright, subject thread should be revealed when you load a page. With the way CloudFlare handles caching, you may need to do a hard refresh (shift + click reload)

I don't know if anyone even still uses rolls for any serious purpose, but the little blue border denoting a legitimate roll is removed after editing a post. Warrants a look, anyway.

On a lighter note, really love the new editing feature. Good for my sleepless nights when I hit [reply] before reading my post.

The youtube videos are still too small to change the sound with me with any ease.

I think this would be a pretty easy fix

If anyone was having issues with videos embedded in OPs not playing go here in your browser, hold shift and click on the refresh button. Then reload the page with the faulty OP embed, and it should work

>>162049
That's probably an issue with the underlying php, since the dice don't have a real bbcode that can be reversed

>>162050
I've been meaning to work on that but I've been really busy, it's still on my immediate list

Last edited at Sun, Oct 5th, 2014 00:14

In the "settings" area, the User tab won't save on my phone. It works fine on my PC, and the Preferences tab works on my phone...it's weird. (Didn't try the General tab since there's nothing on there I want changed.)

>>162128
I'll need you to try the general tab before I can figure out what's going on.

>>162164
Right. It seems to work, I can turn on watched threads and the WT menu appears.

But User is weirder than I thought. This image is what I typed in...

>>162190
...but when I hit Save it reverts to this. And then when I go to post, the full name is there, not just "the". The email is the same. I'll try in a moment which delete password works...

Result: the gibberish password is the correct one. Also, I can't clear the email field from that page, only when posting. It stayed as gaben otherwise.

>>162190
>>162191
>>162193
Well I just looked at it myself, and for the Name box, it only says Zeke, so it must be broken (unless it works for someone else)

As stated in the new features section in the OP,
>Make Youtube embeds increase in size when the thumbnail is clicked for better viewing, and add close button to return to thumb - 100%

  Here's an example

>>162544
>>162545
Awesome!

Small bugs:
It doesn't seem to be working for videos in the OP.
The [close] leaves the brackets behind when closed.

>>162551
>It doesn't seem to be working for videos in the OP.
>pic
>The [close] leaves the brackets behind when closed.
fixed

>>162711
After clearing my cache and cookies I made a thread on /test/.
The OP video doesn't change size and can't be closed.
(This goes both "inside" and "outside" the thread.)

The video in the reply works great and the brackets are removed when it's closed.

>>162718
Just tested on it again on a PC and it's working as you say it is.
Still not working on my phone though. Not that it's super important ofc.

>>162719
probably a caching issue, phones can be funky with that sometimes
Go to http://ponychan.net/chan/lib/javascript/kusaba.js on your phone and refresh a couple times.

>>162720
Cleared the cache and did as you said. It is still the same for me on the phone.
I suppose it's something at my end. It's working on your phone, right?

>>162722
I meant to post this, it's not working on my phone either

>>162748
Ah, good. I mean. Not good good. But at least it's not just me.

Can we disable the vocaroo embed for now? It causes a flash player runtime error dialog to be opened on every page load with one. At least if you have the content debugging version of the Flash player plugin installed.

>>162859
Update your flash player

Though yes, I'll do what I did with Youtube embeds having it load an image. It may or may not have a HTML player

>>159881
please fix oat?

>>163182
You can't fix what isn't broken

zeke, can you make a feature that makes you not a faggot?

>>163315
2edgy

>>163315
see >>163211

Is anyone else getting a bug where being on page 0 stops you from clicking on page 1 on the bottom of the screen?

I'm not sure if this is the correct place to mention this, but I got banned by the spambot for posting either a word with too many consecutive letters or for posting a shortened youtube link (the kind you get if you choose the share option when viewing a video on youtube). I think it might have been the latter, since I have a hard time imagening that typing a word with too many single letters in it (like knooooow, fx) would cause the spam filter to go crazy.

>>164973
The shortened youtube link was added to the spam filter yesterday. It's since banned at least three people on accident and been removed.

>>164974
Cool beans. Thanks for handling that so swiftly.

So as of this post, Pchan keeps defaulting back to a black and white theme, even if I try changing it. Does this have to do with the merger?

>>165296
nevermind. It was just Admiral giving /oat/ a Dash of fun

The quick reply box thingy doesn't pop up anymore when i click on a post's number to respond to it, any idea what could cause this?

Oh hey it's working again, apparantly just needed to restart my browser.

I'm using Chrome, why do the mod tags show up as wingdings?

>>165345
I'm guessing you're using the Derpy theme.
That's the most likely explanation at least.

>>165387
oh hey, you were right. I swapped it to Rarity and now it's fine. thank you.

>>165397
Happy to help.

Shit, I have been rather absent from this thread

>>165515
In my defense, my IP has changed a few times without my knowledge, and my cookies have been weird.

The edit button messes up some of the BBCode.

The most prominent example to me is when I put the 'code' open tag and the 'code' close tag on different lines in order to cover more of the text. When I click on edit, it will pull the open tag down to the beginning of the same line where the close tag is.

So this:

[-code-]Here are;
three lines;
of code;[-/code-]

...becomes this:

Here are;
three lines;
[-code-]of code;[-/code-]

>>165745
Posts are stored as HTML, so it's actually an HTML to BBcode "disassembler" that I wrote. I guess that's a case where it doesn't quite work properly.

I wouldn't expect it to be fixed, though - I can't fix it, and I think Tinyboard stores the original BBcode.

>>165765
At this point, since the merger is so soon, I'm only going to take care of major site-breaking issues.

I'd like to suggest a blacklist on at least /anon/, because I currently can't seem to hide threads, and looking at "MILKY WAY THREAD" and "PONY SCAT THREAD" every god damn time is fucking disturbing.

>>165893
There is a little box in the upper-left corner of each post, that looks like [-]. Click it to hide that post; if that post is the opening post of a thread, it will hide the entire thread.

Hey code ponies,
As exciting as it is to have my page scroll out, can you add a catch to automatically insert a carriage return or something if a string is over a certain number of characters? As we're already parsing the text for wordfilters, I wouldn't think this would be too painful as an array.