indianmouse

Might be. But as an Indie developer and maintaining app is just not the app developer and maintenance. Upgrades and the development requirements around it are costly and needs constant upgrades! I'm not justifying, but trying my best to bring the reality. The time spent on development, hardware costs, licensing costs, design costs etc, are to be factored.

If people are working on a full-time job and trying to create something in the free (!!) time (it still needs to be valued for the time and effort!) and they could chose to or not to charge for the app. It's more personal and a general view.

IMHO, any and all subscription fees are justified based on the thoughts of the developer and how he values his effort, time and development complexity.

Not stating anything about it even if it is not appealing or convincing. It is purely a personal choice and the dynamics of value, usage, benefits, cost and the problem it solves from the buyer's side will also kick into the thought process and decision making.

I'm just saying only one side of the story and want to record my thoughts. Peace out!

True. I have installed PMOS on a Xiaomi Redmi Note 4 and it is has great support for the hardware.

Amazed to see Linux booting natively on a palm sized device with full hardware support and computer capabilities was out of the world!

Given the fact that the device is more than a decade old, it still runs the mainline 6.x Linux kernel and the UI is good! (desktop manager which I've installed it is based on POSH Shell).

Looks cool and since calls, SMS, Wi-Fi and data works fine to a good extent, the only gripe was that camera is unusable in it's current form. Have not tried the android subsystem yet I don't think it'll work due to device limitations.

Have installed Tailscale and primarily using it for remote SSH administration of all my devices through Tailscale. The device build is still solid (which I don't see in the newer devices! due to planned obsolescence) and browser based netbanking access works fine.

Caveats of a small device, keyboard and user experience matters though I'm not complaining because I can pretty much see what happens on the device and I'm not feeding any unwanted signals to anyone.

Constantly on the lookout for pure Linux based devices. Unfortunately, yet to my hands on Librem based device or on a Fairphone!

I hope the world moves towards a Linux based device (being a Linux user for close to 3+ decades (since the early Sco Unix times with core systems programming, I always dream of it! BTW, being a hardcore C programmer who has worked across almost all OSes and CPU architectures!) and such wish and dream will fully be fulfilled only on Linux or on *nix with firm belief!

Hope someone takes a serious move...

Exactly! I like the comment!

Anything connected to network can become a national security threat. IoT devices, internet connected smart appliances and all. Computers that run some OS is least of the worry because of the various layers of protection that can be added to them. But it is not the case with the IoT devices or SMART(!) devices where, if the firmware or any app ecosystem gets hacked / cracked / modified or sideloaded without any user intervention, could turn into a large zombie C2C botnet that can cause havoc.

Proxying is least of the worries!

When manufacturers bring out the smart features that require constant data collection, monitoring (obviously in the name of service quality, troubleshooting, firmware / app updates and subscription of services (!!)), it is oblivious to the fact that security and privacy at a personal / state / national level is never considered (for various reasons such as the design, profit interests and overheads / compliance perspectives and what not!) to be part of the ecosystem. Bad actors with sufficient knowledge and tools could exploit and profit from it.

It has become more of a planned obsolescence / profit / greed based industrial / corporate culture in rolling out unwanted stuff / monitoring and controlling as a feature that gets exploited to the core.

Honesntly, I don't have an answer for that, but have some thoughts that I wanted to share.

I do not want a cleaning robot or a water purifier or a printer or a TV or a refrigerator or an oven or a smart light bulb or a smart lock or even my car to have undisclosed, unwarranted connectivity to internet for whatever reasons.

I do not care if it is the manufacture or the service provider or whomsoever it may be!

I want all of the network connectivity options to be explained with all the controlling options and boundaries and data collection that happens on any of the connected medium to be as much transparent with the option of preventing or restricting what one does not want to go out of the device.

Will anyone do that?

It ill never happen! Sadly!

This problem will balloon further without adequate controls and killing the networking at a ground level would only be the only solution!

But, in the case of a connected device having an inbuilt connectivity option (like the m2m based options) in a smart vehicle, even that is not possible!

Could not get any usable output from it for any real usecase.

Can't see any real value in using 1 bit llms except and apart for the sake of running it for demo or feel good factor of running a higher param model locally and in limited hardware!

Since use as it is very limited, could someone throw some light on any actual use case?

How to trust the heavily obfuscated source code? Did not want to waste tokens on doing the review, but if one cannot trust the code (when something gets published as OSS), it becomes a difficult choice to adoption / usage. I can't even run in on my device. Probably could try a containerized version, but short of time.

Very nice implementation.

Really appreciate it.

Solves a lot of issues with the other sites. Though I haven't used it much or need it, it could be useful to students who need it and one PDF of it solves the issue.

Just keep it running. Have you considered open sourcing it at some point?

Thanks. It looks good. Will test it out.

Is there a way to export or import accounts into the setup? I'm looking for moving my linux / windows configured connections on to the mobile and currently running with Termius, but I need multi device syncs and keep switching devices which proving to be a costly running affair. Also, I've some devices running on postmarketOS on which I could setup things easily over SSH. But if there is a sync mechanism, it'll save some time.

Fantastic read and article.

As an ardent follower of IOCCC since ages (where I learnt a trick or two during the initial C days around late 90s) this is seriously cool!

Many may not even appreciate the power of C! Becoz hardly it is getting taught and used nowadays. People who know knows it better. The laziness and wrappers have taken over the programming world and now AI is taking it to another level.

In near future, I guess people might even forget to code! Where to optimize and obfuscate then? Skilled ones would need to be preserved with awe!

Need to document them (not as a training material for AI for sure!), but somewhere else where it could be learnt by the one who needs it! Privately!

Having it on the internet is definitely prone to stealing / misusing it in the name of fair use policy.

Exactly! Can't even imagine how this runs?

Smirks and shruggs...

Extremely sorry for this unrelated rant. Got triggered by the keyword "Modi".

Quoting Modi is a joke which one cannot even remotely relate w.r.t to AI and don't even feel sorry for saying that since that man blabbers on every stage about non-sensical / non-existent stuff! Watching his videos is the best timepass one can have!

Having said that, the infra and mindset is definitely not there in India to even remotely to innovate or compete in AI race!

Academia is a huge BS where every other person is a backstabber!

A lot of talent is there for sure, but all wants to work for some company or another since there is absolutely zero support for entrepreneurs . No real innovation.

All copy cats as they have proven with mobile and robotics. Just copying or masking Chinese products with the local brand names and reselling. That's all they are good at and that's the irony.

So far, nothing has come from that country which is a real innovation or ground breaking. The day it happens probably one can consider that they are good.

But otherwise, they are good at selling / reselling and scamming the world and nothing else! They cannot produce anything or whatever they produce is taken control by a handful of big corporates from the western region. That's a narcistic corporate monopoly!

Extremely bad tax structure, endless corruption and useless and unqualified ministers occupying worst portfolios, people are really struggling to survive!

Where will they innovate or compete in the global AI race?

Everyone at every level just want to scam and make money to surive that's all!

Period.

Exactly! Though it is sarcastic, it is the way in which everything is moving. No end to it and it'll get worse by day.

But the site has brilliantly captured the thoughts and the little nuances behind agentic coding. It is sure good for all the LLM providers, but on a slightly serious note, it just burns cash which could have been avoided all together.

All said, it's just too good and satirically correct with the prevailing attitude!

Nothing to complain or comment on about the thought process or content. Just don't get into an opinion forming on what is written, but just take a step back and retrospect, it is all on the wall!

Nice work IMHO!

Any alternative to bedrock?

I'm ok to pay for GPUs and currently trying vast.ai, I just need the firepower and I'll use it for my org's usage and the way I want to using opensource models.

I don't want to policing or get locked down or throttled based on my usage or volume.

One flat fee and pricing.

The reason is that I currently don't have the necessary infra available (though working towards acquiring GPUs), I want the dev to continue without any bottlenecks.

May be trying creating one and see how much effort and time is required to clone such a functionality to a proper working state! Something for personal use can be created in about 5-10 days, but even then the skill that is required and the amount of tokens to burn, hosting and security etc, will easily kill. This is exactly the thought process of many, but it will surely kill many opensource contributors. I've stopped committing anything to any open source repos as a personal choice. I do not want to train a LLM which will eventually create more slop and headaches since for me, time is the only important factor which holds the maximum value! Nothing else!