/qa/ - Questions and Answers

>>6233
also i think that the dice placeholder should be automatically always on top in case people try to fudge the roll somehow

>>6234
4chan's /i/ uses https://github.com/desuwa/tegaki it has:
drawing size from 1x1 to 9999x9999(or at least that is what 4chan allows)
replay animation
50% 100% 200% 400% 800% 1600% zooms
3 pages of colors(no more, not even custom input)
brush size from 1 to 64 pixels
opacity from 0 to 100
add layer, remove layer, merge layers, move layer up, move layer down
25 layer limit, numbers start at 1, not 0(you might want to change that)
new image button that asks width and height of the new pic
open image button that allows you to open images and draw on them
export button that allows you to save the image directly without posting and then saving the result
undo button(also doable with ctrl+z), redo button(also doable with ctrl+y)
close button, finish button
pressure control of size and/or opacity option
preserve alpha option
pencil, pen, airbrush, bucket, tone, pipette, blur, eraser
indicator of it recording the replay

you currently have https://github.com/websanova/wPaint which has:
undo button, redo button, clear button
rectangle and elipse tools(tegaki does not have this)
line and pencil
text imput(tegaki CERTAINLY does not have this)
eraser, bucket
bucket color picker, stroke with selector from 1 to 10, color picker for everything else
if you are to tweak one of them, id go with tweaking tegaki by adding the few tools from wPaint that tegaki lacks.

testing a few things
[1d1-99999 = -99998]
[-1d1]

>>6259
>[diceroll]
intardesting...
[diceroll]what im saying by that post is that you should add negative dices, see exotic weapon part of pic, although even is there where no system that utilized negative rolls, having it would be a really neat feature of the system making it capable to go to a minimum of -1999998... i think[/diceroll]
oh and [1d0 = 1] [diceroll]0d1=0{}[/diceroll] [diceroll]0d0=0{}[/diceroll] [1d1+0 = 1] [1d1-0 = 1] [1d1 = 1] [1d1-1 = 0] [1d1-2 = -1] testing some more

>>6260
huh, the [diceroll][/diceroll] can't be inputted manually, neat anticheat
and by my newer tests it seems your system isn't happy displaying 0 as a result of 0 dices and also not that euphoric to show negative values, which in my opinion ought to be inputted as [-1d1] and shown as *dice pic*[-1d1 = -1]... perhaps in blue(000088)?
mainly due to the positive=red, negative=blue that magnets have, and the following quite likely will be read as dumb, but could you make it so that either/and a [0d*any number here*(with perhaps a + or -999999)] or a result of 0 has the color 008800?
im mainly saying this because if + is 880000 and - is 000088 that leaves 008800 unused, and that slightly bothers me, but its completely fine if you don't... and to sound greedy on top of dumb(although im quite certain that i have gone WELL beyond the line of what is palatable to request with the following), how about a completely random dice? as in random +/- dice amount, dice digit and +/-0 to 999999 with the rainbow shift of >>6180 as it's color? as for the syntax, how about [rdr&r](which could stand for RainbowDashRoll) and the obvious variants of [rd1&1], [1dr&1], [1d1&r], [rdr&1], [rd1&r], [1dr&r])?, the "&" is to randomize the addition or subtraction since [rdr+r] or [rdr-r] might be wanted, and to be fancy the 6 variants that aren't [rdr&r] could have color shifts based on the other mane 6(+spike?),
and the other 47 variants(iv'e counted them and put them out in a diagram, im sorry) could have color shifts based on other ponies(perhaps 7 of them to the filly gang+little league????)
here is the full unique random dice combination diagram, again sorry for letting my stupidity flow
[rdr&r], [rd1&1], [1dr&1], [1d1&r], [rdr&1], [rd1&r], [1dr&r]
[rdr+r], [rd1+1], [1dr+1], [1d1+r], [rdr+1], [rd1+r], [1dr+r]
[rdr-r], [rd1-1], [1dr-1], [1d1-r], [rdr-1], [rd1-r], [1dr-r]
[+rdr&r], [+rd1&1], , , [+rdr&1], [+rd1&r]
[-rdr&r], [-rd1&1], [-1dr&1], [-1d1&r], [-rdr&1], [-rd1&r], [-1dr&r]
[+rdr+r], [+rd1+1], , , [+rdr+1], [+rd1+r]
[+rdr-r], [+rd1-1], , , [+rdr-1], [+rd1-r]
[-rdr+r], [-rd1+1], [-1dr+1], [-1d1+r], [-rdr+1], [-rd1+r], [-1dr+r]
[-rdr-r], [-rd1-1], [-1dr-1], [-1d1-r], [-rdr-1], [-rd1-r], [-1dr-r]
the 9 "missing" are because it's as unnecessary to mark a regular number of dices as positive as it is to add the random dice, if i'm honest, but i might as well request.

>>6260
>>6261
Yep I see there is a bug in the dice system when it comes to displaying negative results, and that a dice roll of zero dices should not be performed. I'll look into the other portions regarding dicerolls you mention, but have to read it more thoroughly.
Thanks for doing the testing and finding the bug. I will fix and update the code so negative dice rolls also shows up.

>>6262
>and that a dice roll of zero dices should not be performed
i think it should, but it should always spit zero... or perhaps to give it utility, it could spit out some random phrases based on taking the 0 as a one(so only one phrase) and taking the dn+/-N as variables(which could be "r", [0dr&r], etc etc...) and -0 could be sad/gloomy/negative phrases, this also jives well with my request of making it (008800)green, since you know:
>green
this could in turn be taken as a more useful(?) version of the fortune option of 4chan's /s4s/, although me mentioning that might have just made the idea look to you as a "definitely hard pass, if anything perhaps(P.E.R.H.A.P.S) the always 0, but no random phrases, we want zero to none spammadge, thank you very much"
or just to be silly: 0d/-0d(regardless of the other parameters on the roll)= 0 and -0 respectively, and 00d/-00d is the phrase thing(which i might add, both the green color and 0/00 thing would put them in line with an american roulette in terms of aesthetic)
ALSO: i saw that over at 8chan.moe they don't have an oekaki tool, and so after making the request here(>>6231 its me) i made an as poorly written as >>6231 request on there(https://8chan.moe/site/res/134.html#3194)
then later i wrote https://8chan.moe/site/res/134.html#q3213 a more thorough version of >>6235, you might want to check it
for among the things i requested over there is for the moe devmod to get in contact with you so that you can work together on the oekaki tool instead of ether of you having to do it by themselves, that being said he will be occupied with the 2.6 update for a while, so you should talk to him and see when he is available
oh and according to both the 8moe devmod and wolfram alpha 1d0 should give 0 as the answer(https://www.wolframalpha.com/input/?i=1+0+face+dice+roll+chance), currently does [1d0 = 1] that being said wolfram doesn't give me anything for 0d(says the number of dice should be a positive integer), i'd still say the result should be zero though because you are rolling zero sided nothing(which only different from a zero sided something by the fact that is nothing, it is still 0 sided)
[16d0 = 16] [1d1+1 = 2] [2d1 = 2] and a few tests i did on moe but not here
...and a few more things that just came to my mind:
could you add the shift JIS art font and the [aa][/aa] selector?(either with the Mona or MS PGothic(ＭＳ Ｐゴシック) fonts(https://en.wikipedia.org/wiki/Mona_(font))), and remember that the line height should be 1.1 on the css for the [aa] rule(at least that is how 8moe has it configured)
a selector [wolf][/wolf] that applies the wolfentsein font to the text with the color 4D5D53(Feldgrau) to be the counterpart of the [doom][/doom] text they have over at 8moe?, and i must say that i would not be opposed to doomtext here as well(i don't lean one way or another for the [moe][/moe] sparkle effect they have there, i mean i don't dislike it but i also wont insist on it being added here)
another selector that changes the text color to 6a1de2 AND shrinks it by the same amount as red text(what you call heading on the help menu) grows it, so reg=1.0em, redtext=1.1em and tinytext=0.9em?(its something 8kun has that here doesn't and since i suggested it on 8kun during the loli fiasco as small text, i see no problem requesting this of you) and id say [t][/t] for tiny text is a good selector
testing (((echoes))) and ///slope\\\(like ((())) but for chinese), and i must say, since you can change the color of the text here, how about also changing the color of its background?(like what those selectors do, at least on 8moe)
testing LaTeX support \LaTeX $$\LaTeX$$ [$]\LaTeX[/$] (i personally think that [$]...[/$] should be the way to go to implement it)
and i think you should link to here trough the ponerpics header(like you do from here to there), although i freely admit that this might be my dumbest request yet
thanks again for your time.

>>6260
Think I got all the bugs with dice roll fixed.
I also added the option to do multiple rolls (up to three rolls).
[2d10-1d6 = (10-4) = 6]

>>6299
neat

I'm not sure if this is really a bug, but Archived threads clutter up the search results. Is there a "Don't show archived threads" option when searching?

>>6315
I will look into adding this option to the search.

>>>/mlpol/322722 →
Just a bug (or a feature) so you know. So it still opens up the post and the good news it doesn't add a post number next to the post in question. The other part of the oddness is that the post retains the colorized underline.
[color #00f[u>>6314[/u[/color
>>6314

>>6325
Thanks for letting me know. I'll look into the error with the post number not being added to "quoted by". It should have been added, but I guess I have made a small oversight in the regex used for adding "quoted by" to posts.

>quoted by
*cited by

This is more of a question regarding of NPFchan. Would mind telling my why did store all the files in the archive table as mediumtext? Because of this, there's no way I can't properly access the uri of the files in that string.
I'm trying to modifying the archive class and how it handles, but without success so far without messing up with previous archived threads.
I really liked how you handled your archive like it was a catalog, but I can't fetch the OP's thumb or even the json of the thread because of the json_scramble. Can you give me a light in the right direction?

>>6331
Been a long time since I worked with NPFChan. The scrambled json can be turned off in setting and is not needed at all (unless you want it, but it was never fully hashed out and polished feature). The way NPFChan saved the archive was that it just saved the html files generated as I remember and copied the accompanying files (not sure what it kept in the database if anything except the thread info). It didn't do any alterations of the files (except fix the links in the html to the new locations) from what I remember. But to be fully frank I can't remember how I coded it all for the archive - part of the problem remembering is that there was a two tier archive system set up (that evolved over time). If you turn off the scrambling of json, and easiest next step is perhaps to creating a script that reads the json files one by one (in the archive) get the thread id from it and rename or save a copy with the correct thread id.

Bug when -------------------- the output disregards the previous portions. For rough timestamp. 325157

Can the Overboard get a clickable button at the bottom of the screen that loads the next page? Sometimes it doesn't register that I've scrolled down enough or completely.

Why cant we upload .webp images?

>>6434
>Can the Overboard get a clickable button at the bottom of the screen that loads the next page?
I will look into adding this.

>>6447
>Why cant we upload .webp images?
I haven't added support for it, but I will look into adding support for the webp format. Only verified fileformats are accepted after analysing the filecontent. Reason I did it this way was to prevent every imaginable file format to safeguard users and to be able to handle Exif stripping as best as possible. So I will have to look into how exif is embedded in webp and how to strip it. And as always can only do my best to do this but can't guarantee that all exifdata is removed from every possible way it is embedded in various file formats.

>I have been a bit busy and been slacking a bit in the coding as I am doing rewrite (cleanup) and at the same time maintaining the current so working on two codesets at the same time.

>>6448
What if the site converted uploaded .webps into jpgs before hosting them on the site?

>>6449
It is an option that I can look into. I don't like to change the uploaded files since uploader probably intended it to be in the format they wanted and might loose quality be converting etc. Only time I change file extension is in cases the uploaded file has wrong extension like changing .png to a .jpg or visa versa as an example. But it is an possibility, but I would have to add an extra step in the upload process and convert image so I have to look at the code and how it is done. As code is now it is likely easier for me to look into potential exif data in webp and see if Taglib and/or Exiftool supports it (and if there is any custom cleaning needed as addition).

How do I shut this goddamn snow off? Its lagging out my mobile

>>6451
To manually turn it off you can go to Options and User CSS and add

.container-extra-4 {
display: none;
}

but snow should be turned off soon as the celebrations comes to an end.

>The extra field might be used for other things in the future so it is not just for the snow. It was just an convenient field I added it to when I was scrambling to make some festivities for Christmas.

>>6452
Thanks. Goddamn was a little harsh, I was frustrated because I could only post so much before the text would take 3 sec per character added, and would lag out if I had too many queued characters

>>6451
>>6453
I confirm, the snow may cause high CPU, in my case a desktop computer.

>>6453
No worries, it is understandable.

>>6453
>>6454
Yes it does raise the CPU usage by a bit more than I was hoping, but I was only intending to run it for a few days (will end today around 9pm CST). I see the snow at 4chan is using comparable CPU usage so it is the nature of CSS and animation.

Refined the Custom CSS to reduce the CPU usage a bit to lower the usage a bit more - seams like the animation runs even if not displayed so ending it after one cycle will stop that.


.container-extra-4 {
display: none;
animation-iteration-count: 1 !important;
}


>>6455
Thank you, the new script seems to work nicely

>>6455
Im not sure if its a bug or not, but the Oekaki tool isnt on the quick reply window.

>>6583
It's not on the 4chan /i/ window either, iirc. Probably not a bug.

>>6583
It is not on the Quick Reply since it was troublesome to get it in there in current form (would have had to add big drawing area to the quick reply). In the new code I plan on changing Oekaki to the same one used on /i/ and I will try to get it on the Quick Reply and also add option to draw on images added. But I'm not at that point in the code yet.

No biggie, but moving threads resets the posting time/date of all posts to the time it was moved.

>>6624
Yes, the move function in current code is not too good.
Moving of threads/posts is one of the areas I will totally redo and fix in the rewrite/cleanup I'm doing. Another aspect I will fix regarding Moving threads is that the post links gets updated to the correct new post links (so I don't have to manually fix them after move).
Moving threads never worked that well in the old code either (NPFChan), so moving threads/posts is something that always have been used very rarely.

We had small incident of downtime. Hopefully it didn't cause too much inconvenience.

>>6627
It scared me the hell out.

>>6629
Not sure what the reason was. Server stopped responding.
The server might just have had a Derp moment, but all looks good now.

Fixed small bug regarding posting links and getting false positive "Spam detected" alert.

>>6631
That's good news. It's annoying getting flood detected every few minutes when I'm dumping images.

Updated a major part of the code hoping I fixed the bug that caused server to not respond. If you notice anything that isn't working as it should please let me know.

>>6658
>I fixed the bug that caused server to not respond
Yup, I got that problem. Last time was 10 minutes ago.

>>6659
Last one you experienced might have been when i took down server shortly to activate the new code. But hopefully the change I did to the code will make the site more responsive and not have the occasional slow to respond as happened from time to time.

Bug: There is no "Whole words only" option in the search menu.

Search for "ADL" and you see countless posts with the letters adl somewhere in them. Posts where someone used a word like badly, sadly, adler, and so on.

>>6679
I'd recommend add a space before or after. Pro tip " ADL " or " ADL" or "ADL ".

I have seen an OP post where there were links provided to lolicon websites. The post was since deleted, but the post itself may have taken a while to be deleted from the catalog. Is there any reason for this?

>>6684
Kiss loli 😭😭😭😭😭😭

>6684
Do you even understand how servers operate nigger?
>6685
Surrounded by retards and boomers.

>>6686
I do, but Johnny Law may not.

>6687
>I do, but Johnny Law may not.
So what's the point? Are you also a lawyer giving legal advice?
The 'under the hood' stuff is better left in the dark as proprietary knowledge for the time being. Maybe not, in any case not my call.
>I do,
Then you ahould understand the trials and tribulations associated with any content hosting from third parties.
So there may be multitudes of any kind of reasons and different solutions to problems. The ugly gritty details are by nature just handled. Some technical, some biological, some in human fallibility, some in human intuition, some by mechanical check sums, or any other possibility. Maybe it's communication time, maybe it's time taken to type and double check that really is the post who knows.
Maybe it's something else completely.
You can speculate, but that doesn't solve anything. Nor is speculation how you quickly and efficiently and robustly solve these kinds of happenstances.
If you do undertand you also may understand what kind of responsibilities you have and don't have, and the current available actions at hand.
In any case it's only worth a theoretical programming-lite thread in and of itself for stand alone projects.
Or even a theoretical law thread.
Wait.
>may have taken
So you don't know. That looks awfully alot like legal bait.

>>6693
>So what's the point? Are you also a lawyer giving legal advice?
>The 'under the hood' stuff is better left in the dark as proprietary knowledge for the time being. Maybe not, in any case not my call.
I’m not a lawyer giving legal advice, but there may be a case where server owners and/or administrators are strictly liable for forbidden content hosted on their servers. This means that no matter the explanation, they are responsible for it, no excuses.
>inb4 “That also sounds like legal bait!”
I have taken a course in business law, but I don’t have a law license nor do I plan on getting one anytime soon. Ask a real lawyer if you want the kind of stuff you are suggesting I’m giving.
>Then you ahould understand the trials and tribulations associated with any content hosting from third parties.
>So there may be multitudes of any kind of reasons and different solutions to problems.
And none of them may matter in a lawsuit or prosecution such as the one that I had previously suggested could occur. In fact, it could only mean that far worse problems could be caused as a result of this “virtual bureaucracy”, such as the continued spread of viruses or worms long after they would have been allegedly been deleted.

Can you add a "Sort by newest" and "sort by oldest" option to the Search function?