pexavc

joined 2 years ago
MODERATOR OF
[–] pexavc 1 points 1 year ago* (last edited 1 year ago) (1 children)

Edit: tbh, I see how the question is framed as a general question too.

I think overall, therapy is a great solution. I wish it was easier to start the process in the states. Surprisingly I haven't had any experience truly understanding anti-depressants or being close to those that took them and were open talking about them. I wish I knew more about their effects on how they help with self-assurance.

[–] pexavc 1 points 1 year ago (1 children)

Yeah, I also tend to log everything that is supporting my argument. But, it definitely feels like it could be worse "gathering evidence" sometimes. Especially if the topic is around self-improvement rather than something objective.

[–] pexavc 2 points 1 year ago* (last edited 1 year ago)

that’s the time to reëxamine what you just said

yeah this is definitely hard, I feel like sometimes it is hard to see what caused it. or overthinking on what is it that provoked. and then focusing on probably the wrong causation and then basing everything afterwards on that.

Genuinely asking, Do you usually ask for clarification even on that or not? I feel it would make it worse, "What did I do, to make you say that just now to me?" I would normally think it comes off as arrogant.

someone disagrees, try to politely ask why. Most people are willing to explain where you fucked up a long as you don’t get defensive.

Yeah, the problem is sometimes, the experiences don't match up properly and the explanation will still not fit the reason for disagreement. Navigating past that, is really difficult. Because I feel it then becomes a battle of egos. Because all points on the table, self-included, will not fit the solution. But, I guess maybe in these situations a third party is necessary?

[–] pexavc 2 points 1 year ago* (last edited 1 year ago)

can manage is to find fault with a self-serving characterization of a falsely dichotomous opposing position. So they need to be able to assign me to one or the other team

Oh wow, this is kind of what I have experienced. The tougher part for me, it was someone that wasn't a stranger. It made me self-doubt intensely. And I resorted to doing the same, without thinking that I changed myself completely at the moment. Pointing out flaws rather than bringing it back to the main "issue". (I never am one to "confront", so it felt like a new frontier).

Cutting ties with these types, has probably been the biggest mental improvement I have had. And a huge boost in most other aspects of my life. But, I still have these self-doubt questions. But, this time around trying to discover those answers via the suggestions/similar strategies listed in this thread, I feel is much healthier moving forward.

And to go all the way back, it could be said that the exact problem is that they have unfounded confidence.

And it’s sort of ironic really, because they’re generally driven by a psychological need to be right, and clinging desperately to one fixed position pretty much guarantees that right is the one thing they will not be.

This is all spot on to be honest

And to go all the way back, it could be said that the exact problem is that they have unfounded confidence.

I definitely have unfounded confidence as well, but am one to internalize all the causation or experiences that aggravate it. Leaning on those I view have "resolved" those issues I see in myself.

And it’s sort of ironic really, because they’re generally driven by a psychological need to be right, and clinging desperately to one fixed position pretty much guarantees that right is the one thing they will not be.

Which is why when I see traits like this, I tend to mirror thinking its the correct approach. Instead of realizing the flaws of absolutism.

[–] pexavc 2 points 1 year ago

This is great. It's a nice little addition to a list of gratitudes.

[–] pexavc 3 points 1 year ago* (last edited 1 year ago)

It definitely makes sense. and that's what makes it more complicated. because it is also hard to relay context to get second opinions. sometimes what's left is personal reflection, but practicing how to remove self-bias and not re-adjusting past memories to fit an argument is very difficult.

Thank you for listing those categories, it's nice seeing them out in print to properly remember.

[–] pexavc 1 points 1 year ago

That’s really cool

[–] pexavc 49 points 1 year ago (7 children)

Yeah noticed it too. For some of them. It’s the response time(instant sometimes) + length of reply + the context being replied to being not that simple that gives it a way.

[–] pexavc 7 points 1 year ago* (last edited 1 year ago)

StarCraft 2

Mostly for building in-game awareness. Helped in a lot of games, even FPSs. Just always being proactive when facing some kind of meta.

[–] pexavc 1 points 1 year ago

What’s the difference between a federal agency managing this. Potentially susceptible to private lobbying vs. keeping it unregulated?

2
submitted 1 year ago* (last edited 1 year ago) by pexavc to c/swiftui
 

This project is really amazing for rendering HLSStreams and video in general. Especially on SwiftUI.

It exposes the metal textures prior to rendering which you can also tweak if you wanted to add some layer of effects in between.

And honestly overall having the ability to use ffmpeg easily as a swift package is a huge plus.

This is their FFmpeg kit: https://github.com/kingslay/FFmpegKit/tree/main

Here’s the SwiftUI implementation example: https://github.com/kingslay/KSPlayer/tree/develop/Demo/SwiftUI

 

Curious to start a discussion around the usage of NavigationView or the new NavigationStack in a complex app structure.

How people have implemented these in their own Apps or if they even use these, or have they come up with their own custom solutions?

I’ll start:

I experimented with a custom solution (not using navigationview or stack at all) using ZStacks on the topmost level of the application essentially. I’ve seen significant improvement in performance and cleaner code execution allowing me to route views inside nested sub views simply referencing an EnvironmentValue. But, the downside is losing the out of the box, didAppear and didDissappear lifecycle events.

I see NavigationStack is following suit with storing a stack of paths to route. But, sadly doesn’t support iOS 15. Using UIKit’s navigation controller is kind of a pain when wanting to support all 3 platforms too.

But, I wonder if others have done some analysis on trade offs regarding Navigation. Since it’s kind of the cornerstone of an Application as a whole.

5
submitted 1 year ago* (last edited 1 year ago) by pexavc to c/loom
 

Instead of being strict on Friday updates like this. I will leave it as between Friday and Sunday.

Some features (RSS Feeds/custom feed styles) and improvements will not appear until build v1.1.0b (2023.9.4) which should be sent out by tonight PST.

Loom: https://github.com/neatia/Loom

LemmyKit: https://github.com/pexavc/LemmyKit

Granite: https://github.com/pexavc/Granite

Testflight: https://testflight.apple.com/join/owwIagmV


The 2 notable updates this week is content filtering and integrating FederationKit, which will aggregate not only Lemmy, but RSS Feeds, Mastodon, and more.

RSS and Mastodon's current integration is just for bookmarks and the main feed at the moment.

I have also added a new option to change the appearance of the feed to a more classic reader feel rather than social app.

Aggregation Sources (Aside from Lemmy)
Image RSS feeds can be pulled when setting the instance URL to a valid endpoint. Pagination is not currently supported, but will be in the future. Looking at supporting Atom/JSON streams as well.
Image Mastodon feeds can be pulled when setting the instance URL to a valid mastodon instance. Pagination is not currently supported, but will be in the future. Interactions are disabled for the time being.
Safety
Image On-device CoreML models + keyword filtering to remove posts from the feed. NSFW Extended will remove posts that are detected as NSFW even if they are not flagged as so. Requires an image for inference, which can be generated even without a thumbnail_url property. (As long as a post url itself is available).
Styling
Image Intended to allow full user customization in the future. Currently supports a "social app feel" and a "reddit-like reader".

Note:

This week was lots of performance optimizations on the front-end as well as major underlying API changes for communicating with federated networks.

I have not kept iPad and macOS variants up to date (or at least have not tested them thoroughly. In theory they should be at least close to upto date as its the same codebase).

But since there are more testers on iPhone at the moment. I am looking at mobile as the flagship and goal to achieve very high stability and performance up to all iOS 15.4+ devices. I am hoping to bring the other form factors up to speed by the next feature summary.

 

Also the uploads are always using 3rd party with multiple options setting the posts as URL posts. Built around the aspect of not using a pictrs backed lemmy

Some advantages of seeing so far:

  1. Frugality on storage
  2. Can secure content that is shared on the instance from the core members
3
submitted 1 year ago* (last edited 1 year ago) by pexavc to c/loom
 

Felt this was important enough to share a mid-week update.

The incoming build will have a new option in settings under a section called Filtering.

The option is NSFW Extended

This will run an offline CoreML model that will run through all thumbnails and present media that is SFW. This is based on the data set used here https://github.com/bhky/opennsfw2

I will always find offline solutions first. The goal is for data to never leave the device when incorporating such solutions.

I will stay vigilant on better and more up-to date solutions. Since this is a fairly new integration, I will also spend time optimizing, but the speed of inferencing is negligible and almost instant, so experience does not seem to be affected.

I have made a new open-source package called ModerationKit, where I will incorporate solutions of any kind to prevent harmful media from entering a user's feed.

2 Notes to be aware of:

  1. There is an issue with config restoration per fresh app launch that will be resolved later this week. You may need to re-enable this filter each open.

  2. Computer Vision is never a 100% guarantee.

78
submitted 1 year ago* (last edited 1 year ago) by pexavc to c/[email protected]
 

Other samples:

Android: https://github.com/nipunru/nsfw-detector-android

Flutter (BSD-3): https://github.com/ahsanalidev/flutter_nsfw

Keras MIT https://github.com/bhky/opennsfw2

I feel it's a good idea for those building native clients for Lemmy implement projects like these to run offline inferences on feed content for the time-being. To cover content that are not marked NSFW and should be.

What does everyone think, about enforcing further censorship, especially in open-source clients, on the client side as long as it pertains to this type of content?

Edit:

There's also this, but it takes a bit more effort to implement properly. And provides a hash that can be used for reporting needs. https://github.com/AsuharietYgvar/AppleNeuralHash2ONNX .

Python package MIT: https://pypi.org/project/opennsfw-standalone/

 

The solutions I will employ will be on-device, and just be able to provide additional warnings/censorship for posts not marked appropriately and obviously for posts that are incredibly harmful. This is not the type of content I want to be building something to be near. And will do my best to find solutions in preventing such.

35
submitted 1 year ago* (last edited 1 year ago) by pexavc to c/lemmyapps
 

I'd like to share the TestFlight in this community with a possible listing on other resources for Lemmy federated clients that are being actively supported.

Testflight: https://testflight.apple.com/join/owwIagmV

Feel free to use the https://lemmy.world/c/loom community to post concerns, suggestions, bugs, and I will tend to them promptly!

macOS clients are always released via Github as notarized archives. Or, you could build it locally on your own.

Known Issues:

  • Scrolling stutter
  • Registration is not functional
  • macOS (intel) optimizations
    • I do not have an M-chip, so may be missing more macOS related bugs that need to be fixed

List of some major features are below:

cross-posted from: https://lemmy.world/post/3804045

v1.1.0b is in Review. Will be available by end of day.

1.0.4b was skipped. Since lots of changes were made in Granite, decided to mark v1.1.0 to denote.

Loom: https://github.com/neatia/Loom

LemmyKit: https://github.com/pexavc/LemmyKit

Granite: https://github.com/pexavc/Granite

Testflight: https://testflight.apple.com/join/owwIagmV

Looms
Image Create "Looms" of any group of communities from any instance. Merging their content into a singular feed.
Explore
Image Visit linked instances when connected to a lemmy server. View small snippets of each and their ping response time.
Travelling Modals
Image Writing modals can travel with your viewing context. Allowing you to browse content in any stack prior to publishing.
Bookmark Syncing
Image All your saved posts and comments can be switched between accounts in a dedicated component.
Sharing
Image Share posts or comments as images. With the intention of supporting QR codes and Engravings in the future.
Search
Image Either search all, a specific subcategory with sort support. View expanded contents within the view, interacting with content as normal.
Profile
Image Dynamic threads, despite viewing context. Swipe to reply to comments or tap the more button, to modify, remove, block, etc.
Image Deleting and restoring with toast reactions.
Image Switch accounts, view their profiles, and their scores.
Embedded Webview
Image Custom JS insertion supported, customizing how webpages are viewed via direct user input is a possibility.
Light Mode
Image Light mode and Dark mode supported. Refer to the color group in Assets to define preferences.
view more: ‹ prev next ›