No.10488
never mind the deleted comment. I see
No.10490
>>10489Well yes, I am modifying it so it would differently be different at different times.
Or did you mean that the code shouldn't be visible?
No.10491
>>10490>shouldn't be visibleYeh.
No.10492
>>10491He's messing with and trying to fix things right now, might not be a good time to judge how things are
No.10493
well, I'll consider it
No.10494
No one's noticed yet, but someone probably will today or tomorrow, that I forgot to remove a line of code that was disabling ctrl+z, ctrl+shift+z and ctrl+y.
I'm currently changing how my undo-redo works and during update forgot to remove a part. Plan to have it back by the time I finish my programming session, like 4-5 hours from now or less.
No.10495
Ver 4.23.11
Resolved undo-redo issues. Added bonus from fixing:
The TAB key will insert 4 spaces instead of jumping when a textbox(textarea) is focused.
As of this version instead of adding new things, I'll be looking to getting the software to the point where all I'm noticing is bugs and things largely undesirable. This is to make it so that I can pivot to other tasks such as server changes.
No.10496
Removing tab key might be a bit of a usability problem. I've already noticed that it gets in the way of shift+tab so I've patched that, but I don't think tab is used as much... well someone better complain about it if it's annoying
No.10503
>>10502I don't see how that addresses the bug I pointed out.
No.10504
>>10503Ok. I see what you're saying:
"when I hold down tab it doesn't create a new line in the text area. I don't want to confuse this with going down a new line because this is what tab normally does and I don't want you to be confused that this is what I mean"
No.10505
>>10504>when I hold down tab it doesn't create a new line in the text area.Yeah.
>I don't want to confuse this with going down a new line because this is what tab normally doesWhat? Tabs normally change lines?
No.10506
>>10505go into the name field and press TAB
>>10499This is normal behavior within firefox, and perhaps chrome. Beats me why they do it, it's purely visual, the actual input still has that many spaces
No.10507
>>10506>go into the name field and press TABAh, you were referring to the focus changing from one input field to another.
>This is normal behavior within firefox, and perhaps chrome.Oh.
No.10508
>>10507From what it seems I would have to basically create my own textarea, which I tried to do half a year ago and it didn't work out
No.10509
>>10507actually, it looks like it can be fixed through CSS. I assumed it was normal because other imageboards have the same issue
No.10510
kind of absurd, to be honest, that what I have to do is not the default behaviour in textareas
No.10511
Could we have a purple glow for text?
VERY important when talking about a trio of magical girls consisting of the pink one who's the heart of the group, the yellow one who's her more reserved best friend, and the spunky blue one who they meet at the beginning of the series, that halfway through the series gets joined by the purple girl who's kind of a foil to the pink one.
No.10512
>>10511The final should probably just be all web colors with a picker menu or something like that.
No.10513
>>10511I think we've got a good RGB and gold silver thing at the moment, but if it comes up I'll consider some others
No.10515
I've started the beginning stages of writing a server, which is basically me drafting out the plans to write a ton of design papers/diagrams and establish use cases and predict the future.
Essentially theoretically create the conditions for the server to exist and plan for where things could change years down the line. I probably won't start writing until a month from now, perhaps starting with some test applications I'll make for little funny side applications on kissu like a gallery page or a text only/low-res archive or something... things that I can get a bit of information about where my design ideas are going right or wrong.
No.10524
odd, php got killed by the OOM killer
No.10527
>Warning: - [05/Aug/2022:22:34:57 +0000] "POST /post.php HTTP/1.1" 502 163 "https://kissu.moe/ec/thread/7648" "Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Firefox/102.0">Your post at that that time corresponded with PHP crashing, please tell me what it was in the /b/ threadI was trying to post a picture of Hikaru from Star Twinkle Precure's cute toes.
No.10529
>>10527Can you give me a link to the image because obviously something about it consumes a ton of memory without triggering any of the warnings
No.10530
>>10529This was the picture.
No.10531
Sigh I wish it was me choking to death on Hikaru's toes instead of the PHP server...
No.10535
>>10534What's imagemagick for?
No.10536
>>10535it's a downloadable (command line) tool which modifies images. Vichan and most imageboards use it for creating thumbnails.
There are some alternatives such as Cloudflare or Amazon can generate them for you to save memory, or write your own like meguca, but for the most part it's not needed to use anything other than Imagemagick.
No.10540
>>10539Looks like I've created a solution to the issue
No.10544
>>10543The max length that's allowed is not a true length and a limitation in the function I'm using to generate it causes an error. Probably fixable, but not certain.
It's a thing with ffmpeg when the video length is past the max it can be set at. Might be a better rule I can use on the server and the application
Also you should have gotten an error message somewhere, guess I didn't set that.
Also I think it should allow you to set the slider from without the thumbnailer open.
No.10548
But, the thumbnailer is slow because the process of acquiring thumbnails from videos is not exactly fast.
No.10549
>>10191last-reply sorting seems bugged on /all/ index. Currently
>>>/sum/1292 is shown on top, meanwhile
>>>/sum/1416 had 2 saged posts in the meanwhile, and it's not shown anywhere. Also page 6 on /all/ is missing.
No.10550
>>105491416 is shown on /all/ somewhere but it's not being put on top of /all/ with last-reply sorting as you say.
Page 6 is missing from the bottom pagelist(when 15+ threads are hidden), this is not intentional, probably the leftovers of something I once fixed but didn't properly test.
No.10552
Actually I see what you mean, there's a dead spot after the catalog page ends(only 14 of 15 threads are displayed) and there is a situation where page 6 is dead so I guess I only fixed that issue in my dreams.
No.10553
meh.. the disappearance of page 6 will just have to be an idiosyncrasy with the UI software. Assuming this was due to thread hiding, fixing it isn't worth it.