EuphoricPenguin22

Hello,

In my previous post, I outlined a detailed plan for the likely shutdown of NCL (NormalCity.life). The hope was that the post might encourage donations, alleviating the ever-growing financial burden of running NCL. The hope was that this and a few other issues would've sorted themselves out by December. Unfortunately, we are not in a better position today than when I wrote that first post. It's not an announcement I take lightly, but NCL is shutting down for good in January.

You can look back on that post for the specific plan we have in place for the shutdown, but I wanted to spend a little more time going into some of the reasons why NCL is ceasing operations. I suppose the best place to start is why NCL exists at all. In June of this year, Reddit had a significant upheaval concerning API pricing. If you recall, a sitewide protest resulted in much of the site being inaccessible for a significant period. People were upset with how the website was being run, and people were searching for alternatives to Reddit that still filled a similar hole in their social media diet. Lemmy was perhaps the best-equipped alternative, although many of the instances that I came across before NCL were home to vile and divisive content, often moderated on extremist political lines, and generally were one-trick ponies in terms of what local communities were allowed under the rules. It was clear to me that I was in a position to create an instance focused on tackling these problem areas, and so NCL was born. NCL became a home for federated alternatives to the subreddits I moderated at the time, as well as several other communities I created based on my own interests. NCL was founded on the idea that the divisive political content typical of other instances should be minimized without resorting to outright censorship. It's a tricky thing to get right, so we ended up essentially banning communities that were created with explicit political intent but left well-meaning users with as much freedom to speak their minds as possible. If you're interested, there are some older posts I made on this topic that clarify our stance on moderation. We also were interested in gaining a general user base and wanted to avoid locking users into too much of a box if they wanted to create their own communities here. In fact, NCL originally intended to be purely general-purpose in this regard but more recently switched its focus to STEM-related topics. In general, we wanted to be a place that offered more freedom, less drama, and something more dependable.

So, did we achieve what we set out to do? In some regards, I think we did. NCL gained a respectable, albeit small, base of local users. We never reached the level of growth needed to become fully supported by donations, but we did receive a modest amount of user donations early on. Aside from the communities that I've created here, we haven't had much activity in terms of community creation. Still, our [email protected] community is probably our biggest success story regarding activity and participation. We offered someplace with lenient moderation; we tried to stay as federated and functional as possible, and I believe we managed to set something of an example in terms of what a Lemmy instance could be. We didn't reach our glorious aspirations to be self-funded, but we ticked the main boxes we wanted to for the short time we've been online.

I encourage you to read the last post on our shutdown, which will be the playbook for the coming weeks. The short version is that we're shutting down on January 15th, 2024. This provides a two-week grace period after the end of 2023, outlined in that initial post. Everything here will be deleted along with the VPS where NCL is hosted, although a few key posts that I value were/will be archived before that time. I can attempt to recover data if you request it at [email protected], but I'm really not making promises regarding what I'll be able to gather before the shutdown. Chances are, you're in a better position than I am to recover your own data, and some tools are available to make moving key bits of information to a new account on another instance easy. I'll try to address that gamut of concerns on January 1st with another post, so look out for that.

Aside from financial concerns, which were the main reason for the shutdown (it's now $32 a month to run this site), legal and technical concerns were also part of the decision. Ensuring continual compliance with the multitude of global and state-level privacy laws is nigh impossible, and more recent issues with our automated update system Ansible meant that hours would need to be spent fiddling with configs to get that working. Combine that with an already busy schedule, waning interest in the concept of Lemmy as a whole (on the part of me and others), and other considerations; it's simply the right time to say goodbye. So, once again, NCL will definitely be shutting down on January 15th, which provides a two-week period after New Year's Eve for people to get their things in order and figure out their next steps. I'll make a post before that period with more information.

I hope everyone has a great December, and I will see you at the start of our grace period to help you all out as much as I can.

Sincerely,

EuphoricPenguin22

Hello, regular users and passers-by,

NormalCity.life first came online some three months ago as a response to the Reddit protests that were taking place at that time. I wasn't really sure whether or not NCL would be a success, but I thought it would be worth the risk. We've had our ups and downs, especially with several initial configuration headaches and issues with spam accounts. We're still here, though, and plan to be online at least until the end of this year. That's really what I want to talk about, though, as server costs are still steadily increasing.

I host NCL in the cloud because it's faster and more secure than a honkey-tonk network configuration here at home. Of course, that comes with a cost. Initially, it was palatable for the small user base we had at that point. Now, for whatever combination of reasons, we've maxed out our storage again. If you noticed issues with the site today, that was the reason. I had to upgrade the plan to a higher tier, which means it costs more to run NCL month-to-month. Now, it's not astronomical ($28-30 per month), but it's quickly getting to the point where I cannot support it long-term without some help. This server is run in the interest of those who use it and without the sale of user data or profit-seeking activities. So here's what the future of NCL looks like:

• NCL will pledge to keep its servers online until January 1st, 2024. After that date, if it is unsustainable to operate the server, a two-week notice will be given to users via a post in this community. An additional post will contain as much information as I can reasonably provide to help with account migration and user data management. Users will have the opportunity to request user data; however, certain types of data retrieval may be unfeasible (remember that this is a one-man operation on the backend, after all). After that period, assuming the server is still unsustainable to operate, the server will be taken offline, and all user data will be permanently deleted. NCL promises to never sell or otherwise directly profit from your data. We will only make copies of your data as requested by you via the email associated with your account. We will only offer data retrieval before or during the two-week window; after the sunset period elapses, all data will be unretrievable.

• If you would like to help fund this server, a recurring donation to our tip jar would greatly help in offsetting the costs of operating NormalCity.life. If we can reach a more sustainable month-to-month operating cost by the end of the year, NCL will continue to operate as usual. I hate to set a deadline for something like this, but I have a lot of personal costs that I have to balance with the operating costs of this server. I can reasonably promise three or four more months, but if costs keep increasing due to resource usage like they are, that is all. A recurring donation, even $1-5 per month, would make a big difference, and one or two of those donations would likely put us in a place where we can continue to operate into 2024.

I don't like making posts like this, but I wanted to formalize this in writing so you all know what to expect. I've been thinking about the cost of operating NCL long-term for some time now, and this recent issue reminded me to make this post. I've been pretty happy with Lemmy overall, if only for the constant supply of Star Trek memes at [email protected]. This is my main account, and this server is my main way of accessing Lemmy. When I said a while ago that this was my ideal Lemmy instance, I wasn't mincing words. I've run it to treat others the way I want to be treated on a social media platform, and I hope that's still obvious, especially now. Your support is appreciated, and I sincerely hope that we can move forward together in a more sustainable fashion.

All the best,

EuphoricPenguin

Send all requests, suggestions, and concerns to: [email protected]

EDIT: I'll add this at the bottom here, but we've moved to registration applications until further notice. I'll do my best to let people in, but this is better than letting spammers slip through without a care in the world. I'm periodically trying to get NCL updated to the more recent version of Lemmy, but I've hit a few snags with the Ansible install. In any case, we're proceeding with the above plans unless the funding situation improves for this instance.

The engine mount finally cracked due to improper installation, but otherwise, I've been pretty happy with this one. I'm currently working on Rig V2, which will be on a new bike with a Phantom 85 V3. I gave this bike away for free to someone else.

Just for fun, here's a laundry list of some of the issues in this particular photo:

• Missing filter: no fuel filter, although this was addressed not long after this picture was taken.

• Carb not level: carb is at an angle due to clearance issues, although a different intake manifold and mods to the intake on the carb itself would have probably solved this. It still runs fine enough but might have been underperforming due to this.

• Mediocre fenders: ModPD fenders are probably the bomb on a stock bike, but they're pretty flimsy and don't like to stay attached to this rattle machine. Still, plastic fenders like this are far safer than the full metal alternatives; these are far less likely to get wrapped around a wheel.

• Bad engine mount: my down tube was fairly large, but I used a stock mounting bracket with longer bolts. This ended up being a major problem, as the bolts cross-threaded and eventually cracked away from the engine. There exists a proper large down tube mount, which I should've used from the start. That's mostly why I retired this bike, but I'm hoping the person I gave it to can find a way to fix that aspect of it.

A promotional card from the 1989 DRK Next Stage! press conference. A prototype rally car made from officially-licensed parts was exhibited on a private course. This was several years before the final version was made available for sale.

Some are from when I toured around Western PA a few weeks ago, and the more recent ones are from Myrtle Beach.

cross-posted from: https://normalcity.life/post/253737

The philosophy of motorized bikes is generally to create a cheap mode of transportation that isn't really filled in well by other options. Cars are expensive, take up a lot of space, and aren't always the best option when road conditions get sketchy. Electric bikes and scooters, while falling in cost, often are still far more expensive than building a gas-powered bike yourself. Mopeds, dirt bikes, and mini bikes are often more expensive as well. Plus, most of the cheap options don't strike the same balance between road-friendly tires, lightweight construction, and the ability to add shocks and other vibration-dampening. Motorized bike kits and bicycles, in general, also allow for far more customization, as essentially, any part on the bike can be easily replaced, added to, or upgraded when desired. For example, the bike we will look at could easily be upgraded with disc brakes, a mountain bike fork for added front shocks, and a spring seat for additional dampening. Motorized bikes do often need servicing, but they can prove to be a reliable source of fun or transportation if used properly.

The Bike

The 700c Kent Rideland bicycle.

AliExpress Link

This is the first time I've personally bought a new bike with the intention of using it in a motorized bike build, and I'm really happy with most of what this bike offers. For starters, this bike is a hybrid v-style frame with regular diameter down tubes, so this should allow an engine to be mounted with stock hardware. Second, this bike employs dual v-style brakes, which are superior to coaster brakes in several ways, and leave room on the rear wheel for mounting the motor chain sprocket. It's a single-speed bike, greatly reducing complexity and decreasing the likelihood of the pedal chain falling off. Lastly, this bike is cheap and comes at just over $100 on AliExpress. You can order this bike everywhere, though, and multiple sellers on eBay, Walmart.com, and more have listings up for this bike at various prices. It also looks cool as heck and has a super eye-catching color scheme.

The Kit

One of the many motorized kits available online.

AliExpress Link

With the cheapest options, there isn't really a single kit that's worth getting over another. Despite claiming to be 80cc/100cc, most of these motors are closer to 40cc/60cc. They don't make a ton of power, but enough that they can get up to 20-30 MPH on flat ground using the 36T sprocket. Most run well enough and are fairly reliable when broken in properly. The option I linked above is listed for around $100 and includes everything to mount the motor and add appropriate controls. Again, these are listed everywhere and can be found easily for around $100 (or sometimes less).

There are many things you'll learn in the process of completing a build like this, and the end result is something you can actually ride and potentially use as transportation. It's a $200 DIY moped, and it can be done with a brand-new bike and motor for that price.

Resources

Assembly Overview

Basic Assembly Video

The philosophy of motorized bikes is generally to create a cheap mode of transportation that isn't really filled in well by other options. Cars are expensive, take up a lot of space, and aren't always the best option when road conditions get sketchy. Electric bikes and scooters, while falling in cost, often are still far more expensive than building a gas-powered bike yourself. Mopeds, dirt bikes, and mini bikes are often more expensive as well. Plus, most of the cheap options don't strike the same balance between road-friendly tires, lightweight construction, and the ability to add shocks and other vibration-dampening. Motorized bike kits and bicycles, in general, also allow for far more customization, as essentially, any part on the bike can be easily replaced, added to, or upgraded when desired. For example, the bike we will look at could easily be upgraded with disc brakes, a mountain bike fork for added front shocks, and a spring seat for additional dampening. Motorized bikes do often need servicing, but they can prove to be a reliable source of fun or transportation if used properly.

The Bike

The 700c Kent Rideland bicycle.

AliExpress Link

This is the first time I've personally bought a new bike with the intention of using it in a motorized bike build, and I'm really happy with most of what this bike offers. For starters, this bike is a hybrid v-style frame with regular diameter down tubes, so this should allow an engine to be mounted with stock hardware. Second, this bike employs dual v-style brakes, which are superior to coaster brakes in several ways, and leave room on the rear wheel for mounting the motor chain sprocket. It's a single-speed bike, greatly reducing complexity and decreasing the likelihood of the pedal chain falling off. Lastly, this bike is cheap and comes at just over $100 on AliExpress. You can order this bike everywhere, though, and multiple sellers on eBay, Walmart.com, and more have listings up for this bike at various prices. It also looks cool as heck and has a super eye-catching color scheme.

The Kit

One of the many motorized kits available online.

AliExpress Link

With the cheapest options, there isn't really a single kit that's worth getting over another. Despite claiming to be 80cc/100cc, most of these motors are closer to 40cc/60cc. They don't make a ton of power, but enough that they can get up to 20-30 MPH on flat ground using the 36T sprocket. Most run well enough and are fairly reliable when broken in properly. The option I linked above is listed for around $100 and includes everything to mount the motor and add appropriate controls. Again, these are listed everywhere and can be found easily for around $100 (or sometimes less).

There are many things you'll learn in the process of completing a build like this, and the end result is something you can actually ride and potentially use as transportation. It's a $200 DIY moped, and it can be done with a brand-new bike and motor for that price.

Resources

Assembly Overview

Basic Assembly Video

Prompt:

A modern smartphone picture of a man riding a motorcycle in front of a row of brightly-colored buildings.

Settings:

Rendered using various steps and CFG values, Euler a for the sampler, no manual VAE override (default VAE), and no refiner model. All images were generated at 1024*1024. This is using the 1.0 version of SDXL.

Summary:

Subjectively, 50-200 steps look best, with higher step counts generally adding more detail. A CFG of 7-10 is generally best, as going over will tend to overbake, as we've seen in earlier SD models. Prompting and the refiner model aside, it seems like the fundamental settings you're used to using will probably still hold true for SDXL. Granted, prompting is a bit easier for photorealistic outputs now, and the refiner model might allow you to use fewer steps for the initial generation with the base model.

Mine is probably the crane or the jumping frog at this point, but I'm probably overlooking a ton of other cool traditional models. Anyone else have any they like to fold?

Entire Art Portfolio

Since I was revisiting some of my attempts to use AI in my work, here's perhaps my very first attempt at combining vaporware with text-to-image generation. It's super messy and not really all that spectacular, but it's pretty cool to think that we've come this far in only two years. I mean, this (VQGAN+CLIP) was already impressive back in 2021, and it looks absolutely weird and incoherent next to even the first official public release of Stable Diffusion (from August 2022). This piece is one of those that I tend to ignore in my portfolio, as it's not all that great on its own, bit in with context, it helps to highlight the march of technological progress and changes to how I approach art in general.