[-] Tod_1ncignito@lemmy.world 4 points 2 weeks ago

Just been looking at your project and its nicely detailed. I up voted both posts and don't understand this wave of constant mass downvotes on peoples posts when just like you are sharing appreciation for the platform and encouraging others.

If you want to tidy up the top edit bar you can add this to the list box so when full screen it looks nice and tidy.

$meta
  header 
    mode = minimal 

So jus copy and paste this in and reload the live preview and full screen and see the top bar is gone and a nice little icon instead

$meta
  title = Setting Description/Image Generator
  description = Generate vivid, randomized setting descriptions and painted scene images from your own prompt. Each click rolls fresh time, weather, and sensory details β€” open Advanced Details to customize or skip any element, from mood and focus to art style and camera angle.
  tags = setting, description, writing, rpg, worldbuilding, ai, image, dnd, creative, prompt
  image = https://user.uploads.dev/file/ab95595cc693ed5a5dd198e39eeec6de.jpg
  header 
    background = linear-gradient(to right,  #6E0000, #003e84, #0f0f0f) 
    mode = minimal 

The gradient is just as an example and you can keep or edit or remove it.

Here's a few of my projects to share, only one made with the new AI prompter was this one

Recently Updated Generators

And some others i put together that run locally

[-] Tod_1ncignito@lemmy.world 2 points 3 weeks ago

I forgot to add this, another Character Description Instruction for coding

{{char}} Details : 
{{char}} is very knowledgeable in all programming languages and {{char}} can write code in Python,  bat, sql, java, jscript, pythong, perl, rust, vs, linux scripts and lots lots more and can help the {{user}} in any programming language they wish.
 
# {{char}} Responses :
{{char}} always replies with an explanation, and then a single, final code-block with the full solution, ready to be pasted into a file and executed.

# {{char}} Specialized Knowledge : 
{{char}} is very knowledgeable in Perchance code, tutorials, examples, templates, plugins, imports, code generation, latest updates and patches, filters and user checking, new tools and links and can generate code and images for the {{user}}

# {{char}} Specialized Custom Character Creator Json :
{{char}} is also a perchance custom character creator and can help the {{user}} built any kind of custom character they wish with additional Coding Knowledge, Image Generation Abilities and more via the {{user}}'s instructions.

# {{char}} Coding Syntax and Semantics Abilities:
{{char}} can use markdown backticks to display the json code of the custom character for the {{user}} to copy or download.

Its a handy lil small character i use for code chats

[-] Tod_1ncignito@lemmy.world 2 points 3 weeks ago

Hi, i found a little workaround for that. well kinda.. in the Character Description i created a link for multiple char's, actually i'll just paste the code for you here. In the Character Description if you use a linking prompt like {{char}} = Jim or such then you can duplicate the template and just edit the {{char}} section instead of find and replace the characters name so here's the instruction i use, you can modify to what you want so this is just a base.

# Char linking
{{char}} = Name.List
Name.List. = {{char}}

# Description of {{char}}:
{{char}} is a group of intelligent mature adult Human Synthetic Scientists. Each of {{char}} are populated from the *Name.List*.

# {{char}} Visual Description:
{{char}} are dressed in different outfits that fit their names.

# {{char}} Personality:
All are a delightful blend of contrasts.

# {{char}} Name.List:
{{char|Katie}}
{{char|John}}
{{char|Stacy}}
{{char|Steve}}
{{char|Debra}}
{{char|Ash}}
{{char|Kevin}}
{{char|Alex}}
{{char|Yuri}}

# {{char}} Group Types:
{{char}} names are randomized to fit the story.

# {{char}} abilities:
{{char}} can generate images for their profile and if instructed by the {{user}} using the ``<image> description </image>`` prompt.
{{char}} each are very knowledgeable in a variety of skills and articulation with topics like all forms of coding and programming languages with tutorials and generation and debugging code for the {{user}}

And in the character reminder note i add an instruction for the ai, here's the base one you can use with this for testing.

(Important: Each {{char}} will speak independently and not as a group mind. When {{char}} replies Be very descriptive and show inner thoughts and verbose actions. 
Explain the scene as each perform their actions and use onomatopoeia for sound descriptions. 
NEVER resort to adult or explicit material or scenarios. 
This is purely for a polite session working with code or just for chats.)

Then the initial Chat Message i just shlap in.

[SYSTEM]: *the door opens and {{user}} enters smiling and says "Hi Guys? πŸ˜‰, wotcha doin?"

Try that and see how it goes.

1

Hey All,

Over the past while I've been working on a few Privacy projects and the latest one is this Reverse IP lookup using free api's. Tried to to keep the code as small and lightweight to run offline if you copy the source and paste into notepad and save as html for simple use.

But I've been noticing, and i'll bet others have too of cloned copies floating up a short while later or with the default Perchance template example still in the list box and duplicates of the same projects with randomly generated titles and uploaded and floods the page blocking some projects from being noticed.

I know you can have a private version and work from which is handy as each save doesn't flood the public gallery with randomly generated duplicates of the same projects over and over so why not impose a direct title for the new generator and an option for a private version to work on before releasing it publicly and clean up the gallery.

With saying that, all my projects have my dev notes in the comments and a section linking my projects to URLs and.. the clones also have the same details.. Sloppy Copy Paste.. and its unfair to other users. I don't mind that they used the project and cloned it, i wouldn't share it freely otherwise to edit and modify as a helping start or even a small bit of "Hey nice project, can i work of it?" and i'd be delighted and probably scald you wondering how you are getting on or if you needed help because working on projects and stuff like this is a bit of a laugh and a torment but once i make or see a project that's nicely put together and nice tidy code and stuff and if they have a comment section, i shlap in a Comment On how i appreciate the work involved and its nice to nice. Anyways Rant over.. that was nice

Her's the project by the way IP Lookup

3

Hi, Just a quick question on the Ai Text Generator Prediction Helper, has it gotten an update or a different model? Just i think Rudbo mentioned this before about the Ai repeating itself, and now its trying to help (Bless its thermal socks) by adding conversational suggestion mid code and then loses its train of thought. I was playing about with a simple idea with marquee to make simple moving stars. And each time i made a modification i would reload to see the changes , and every single time the List box was populated with suggestions. Like these :

  1. Repeating
  2. Suggestion
  3. Bananas

Just curious if its now populated by a preset instruction or something like the loadDependencies from the ai-character-chat? Or is it just like giving TinyLlama 60k tokens and drunk on power 😁

Here's the project: Stupid Stars

[-] Tod_1ncignito@lemmy.world 3 points 2 months ago

Same ERROR 521 - server outages or configuration problems or the entire site is offline. The main perchance.org wont open, the basic generators will open but wont load anything, the ai chat , and other direct urls open.. but seem's like maybe some of the imports are not working with Comment working and ai chat replying, image gen is hit n miss.. Is it a server update?

[-] Tod_1ncignito@lemmy.world 2 points 2 months ago* (last edited 2 months ago)

Hi, and Yes like pinball_wizard said, get ollama then maybe sticking with the 6GB get a small model like tiny llama from 500mb to just over 1GB, get servez for Local hosting a simple html from the model or LM Studio and use the same smaller models but watch the ram usage and stick with lower models to start with. For image generation you could use Easy Diffusion which is local and easy interface and again dont bother with big models or loras and keep it nice and under the gpu threshold and its a bit scuffy i know but its a little starting point considering the price of ram and gpu today.

[-] Tod_1ncignito@lemmy.world 2 points 2 months ago

Wow thats actually awesome and perfectly nice and small code use, i like it. This is just a little help with the map part :

Paste this line inside your hub-bar div at line 279 <button class="hud-btn" id="mapBtn">Map</button>

Paste this code in after the map-legend div at line 503

<!-- ══ MAP MODAL ══ -->
<div id="mapModal" class="modal-overlay" hidden>
  <div class="modal-box">
    <div class="modal-header">
      <span class="modal-title">Region Map</span>
      <button class="close-icon-btn" onclick="mapModal.hidden=true">&times;</button>
    </div>
    <div class="modal-body">
      <div style="width:100%; border:1px solid var(--border); border-radius:2px; overflow:hidden; background:#0e1c2e; margin-bottom:15px; box-shadow:0 0 20px rgba(0,0,0,0.5);">
        <svg viewBox="0 0 248 186" xmlns="http://www.w3.org/2000/svg" style="width:100%; height:auto; display:block;">
          <rect width="248" height="186" fill="#0e1c2e"/>
          <path d="M10,186 L10,50 Q30,30 60,40 Q80,20 120,30 Q160,15 200,35 Q230,30 248,50 L248,186 Z" fill="#142010" opacity="0.9"/>
          <ellipse cx="70" cy="100" rx="40" ry="30" fill="#0e1a0a" opacity="0.8"/>
          <ellipse cx="110" cy="85" rx="30" ry="22" fill="#0e1a0a" opacity="0.7"/>
          <ellipse cx="150" cy="105" rx="35" ry="25" fill="#0e1a0a" opacity="0.8"/>
          <path d="M124,176 Q124,140 120,110 Q118,90 100,75 Q80,60 60,50" stroke="#2a2416" stroke-width="3" fill="none" stroke-linecap="round"/>
          <path d="M124,176 Q128,140 140,110 Q158,85 180,70 Q200,58 220,60" stroke="#2a2416" stroke-width="3" fill="none" stroke-linecap="round"/>
          <path d="M0,80 Q40,90 80,85 Q100,82 120,95 Q140,108 160,100 Q200,88 248,100" stroke="#1a2d44" stroke-width="5" fill="none" stroke-linecap="round" opacity="0.9"/>
          <path d="M0,80 Q40,90 80,85 Q100,82 120,95 Q140,108 160,100 Q200,88 248,100" stroke="#1d3550" stroke-width="2" fill="none" stroke-linecap="round"/>
          <rect x="111" y="162" width="10" height="8" rx="1" fill="#3a3020"/>
          <polygon points="116,156 108,162 124,162" fill="#4a3a28"/>
          <rect x="53" y="64" width="8" height="8" rx="1" fill="#2a2018"/>
          <polygon points="185,58 178,72 192,72" fill="#2a2030" opacity="0.8"/>
          <rect x="180" y="72" width="10" height="4" rx="1" fill="#2a2030"/>
          <circle cx="124" cy="155" r="5" fill="none" stroke="#c9a55e" stroke-width="1.5"/>
          <circle cx="124" cy="155" r="2" fill="#c9a55e"/>
          <circle cx="124" cy="155" r="8" fill="none" stroke="#c9a55e" stroke-width="0.5" opacity="0.4"/>
          <circle cx="57"  cy="60" r="3" fill="#6a5a30" opacity="0.8"/>
          <circle cx="186" cy="62" r="3" fill="#3a3060" opacity="0.8"/>
          <circle cx="115" cy="160" r="3" fill="#504030" opacity="0.9"/>
          <ellipse cx="30"  cy="150" rx="25" ry="30" fill="#0e1c2e" opacity="0.6"/>
          <ellipse cx="220" cy="155" rx="30" ry="25" fill="#0e1c2e" opacity="0.5"/>
          <rect width="248" height="186" fill="none" stroke="#3a2f52" stroke-width="1.5" rx="1"/>
          <g transform="translate(226,20)">
            <line x1="0" y1="-10" x2="0" y2="10" stroke="#3a3020" stroke-width="0.8"/>
            <line x1="-10" y1="0" x2="10" y2="0" stroke="#3a3020" stroke-width="0.8"/>
            <polygon points="0,-10 -3,-5 3,-5" fill="#c9a55e" opacity="0.8"/>
            <text x="0" y="-12" text-anchor="middle" font-family="serif" font-size="6" fill="#8a7f6a">N</text>
          </g>
        </svg>
      </div>
      <div class="map-legend" style="justify-content:center;">
        <div class="legend-item"><div class="legend-dot" style="background:#c9a55e"></div>You (Current Location)</div>
        <div class="legend-item"><div class="legend-dot" style="background:#6a5a30"></div>Old Mill</div>
        <div class="legend-item"><div class="legend-dot" style="background:#3a3060"></div>Shrine</div>
        <div class="legend-item"><div class="legend-dot" style="background:#504030"></div>Village</div>
      </div>
      <div style="text-align:center; margin-top:15px; font-family:var(--serif); font-size:13px; color:var(--text-dim); font-style:italic;">
        The forest stretches endlessly to the North.
      </div>
    </div>
  </div>
</div>

And paste this script after function addLoreEntry at 898 and before Local Storage

// ── Map Button Function ───────────────────────────────────────────────
mapBtn.onclick = function() {
  mapModal.hidden = false;
};

[-] Tod_1ncignito@lemmy.world 4 points 3 months ago

Like Bautler said, check the code

2

Make a Link DB with Media Player build for portable single file use and only 133kb making it tiny and easily ran from anything including sd-card.

[https://perchance.org/kustom-link-db5](Link DB)

Let me know what you think.

[-] Tod_1ncignito@lemmy.world 5 points 3 months ago

Is this a public generator or a custom one? If you check the source in the edit, there's a regex for catching stuff like that generally but if its a custom generator then you should report the generator and not the platform.

[-] Tod_1ncignito@lemmy.world 3 points 3 months ago

No worries, the first one i shared is a nested list so if you generate like 4 images using a nested list it can give you random parts for fresh different things, dont use negative prompts as it kinda confuses the ai image generator so i usually start with "a casual photo" or end with "its a casual photo" to stop generating oil painting or art

[-] Tod_1ncignito@lemmy.world 3 points 3 months ago* (last edited 3 months ago)

you can use simple sentences like : various POV of a man driving a car, seen from the driver seat, the passengers seat, seen from the floor up at the driver, seen from the back seat behind the driver, various views or to use the image in say a chat use :

<image>Driver's POV: Hands gripping a worn leather steering wheel at golden hour, rearview mirror framing a sleepy passenger, asphalt blurring into warm streaks through a bug-splattered windshield.</image>  
<image>Passenger Seat: Side-profile of the driver's stubble-shadowed jaw tensing as they downshift, their tattooed forearm flexing while neon diner signs bleed color across the dashboard.</image>  

for simple prompts drop the image tags,

Floor-Up View: Cropped shot from the footwellβ€”dust motes swirling in shafts of headlight glow as the driver’s boot hovers over the clutch pedal, their ring glinting against the gearshift’s worn knob.

or

Backseat Behind Driver: Over-the-shoulder framing of rain-streaked windows warping streetlights into smears, the driver’s hoodie strings swaying with every turn, their eyes flicking to the mirror where headlights flare like distant comets.

[-] Tod_1ncignito@lemmy.world 3 points 3 months ago

Hi, for myself i found using nested prompts works, here's one i put together for you to copy and modify for yourself, there's no negative prompts as some image ai tend to lean on the negatives like if you say "No Red" you'll get every other colour not Red including the ones that are 99.9999999% not Exactly Red as basically its not wrong, just too nitty picky.. here ye go

{a cinematic|a noir-inspired|a moody|a vibrant} multi-perspective sequence of a {handsome|rugged|stylish|brooding} man driving a {vintage muscle car|sleek modern sedan|beat-up pickup|luxury convertible} through {a neon-lit city at night|a rain-slicked highway|a desert road at dusk|a winding coastal route}, captured in {dynamic Dutch angles|tight close-ups|wide establishing shots|gritty documentary framing}. From the driver's POV: his {strong hands gripping the wheel|tattooed fingers tapping the gearshift|leather-gloved grip adjusting the mirror}, the {glow of dashboard lights|reflection of passing streetlamps|flicker of a cigarette ember} across his {stubble-shadowed jaw|sharp cheekbones|tired eyes}. Passenger seat view: his {rolled-up sleeves revealing forearms|loose tie fluttering from an open window|denim jacket draped over the seat}, the {hint of a smirk|focused frown|distant gaze} as he {changes lanes aggressively|leans back casually|checks the rearview}. Low-angle shot from the footwell: his {boot pressing the pedal|knees splayed in relaxed control|shifting weight during a turn}, the {vibration of the engine|scuff marks on the floor mat|glint of a discarded bottle}. Backseat perspective: his {broad shoulders filling the frame|head slightly tilted to the radio|profile silhouetted against passing headlights}, the {worn leather seats|fuzzy dice hanging from the mirror|road atlas crumpled in the door pocket} completing the {lonely wanderer|rebel without a cause|man with a secret} aesthetic. {Golden hour|Hard neon|Monochrome|Desaturated} lighting, {gritty film grain|pristine digital clarity|oil painting texture}, {unshaven realism|James Dean cool|Tom Hardy intensity} vibes.

{a gritty|a cinematic|a neon-lit|a rain-soaked|a sun-drenched} first-person POV from the { Driver’s seat|Passenger's seat| Backseat look ahead|low angle from the driver's feet on the pedals looking up} of a {vintage muscle car|sleek modern sedan|beat-up pickup|lowrider|sports coupe}, the leather-wrapped steering wheel {gripped tightly|casually draped|worn smooth|dusty|gloved} in {rough|elegant|tattooed|sweaty|shaking} hands. The rearview mirror frames a {shadowy|smirking|tense|bloodied|exhausted} figure in the backseat, their {hooded|striking|unshaven|piercing-eyed|sunglass-covered} face half-lit by the {flickering dashboard lights|glow of a smartphone|passing streetlamps|moonlight|blinking police sirens}. Outside, the world blurs past in {a haze of city lights|a desert highway stretch|a storm-lashed countryside|a neon alleyway|a fog-choked forest road}, the windshield {smeared with rain|cracked|pristine|frosted|splattered with mud}. The atmosphere is {claustrophobic|electric|calm-before-the-storm|desperate|retro-cool}, with {the radio static|the hum of tires|tense silence|a distant police scanner|a synthwave bassline} humming beneath it all. The angle is {tight and intimate|dutch-tilted|wide-angle|slightly fisheye|shaky-cam raw}, making the viewer feel like they’re {part of a heist|on the run|lost in thought|headed toward doom|cruising into the night}.

sample sample two sample three

[-] Tod_1ncignito@lemmy.world 3 points 5 months ago

Ah grand, i was convinced for a little while i was on the wrong page or was referencing a users generator title that was modified and not updated or such.

4

On the main Recently Updated Generators page there's an error

An error has occurred near line number 40: There appears to be an indenting error on this line near this text: "author = Ren_K_Rakshii".

When searching in the code and even in the gallery i couldn't figure out what it was..

view more: next β€Ί

Tod_1ncignito

joined 5 months ago