Sloppy

Neeva's similarity search is very useful. It takes a page and creates a query from it (oversimplified, actually a document based k-nearest-neighbors) then executes the query. This feature used to be in Google but they removed it, not used enough I guess. Neeva has very limited ways to use this.

Here's wishing for this in all search!

sounds like a good solution, can you share a few details?

I self-host file sync, calendars, contacts, photo sync, Google Workspace type services (including all Office doc types and even video meetings), as well as a blog. Here by self-host I mean run all this in a docker-compose collection on a 24 core xeon server in my closet.

Surprisingly (to some) these are easier that self-hosting email. So this is a great article than I plan to add it to my-digital-self-reliance playbook.

I also agree with the motivations and have a whole list of others. We are becoming the slaves of Big Tech. Only go there willingly, don't let the hard choice of saying "no" make the decision for you.

I was an engineering hiring manager for several small startups. I never asked for coding on a whiteboard. I asked some work style questions and then concentrated on a collaborative exercise where I laid out a problem and important applicable technology. Then I'd answer any Google-ish questions in realtime. This technique found good collaborators and also which people were good problem solvers. Never made a bad hire from this method. I only wish I had discouraged others from "code on the whiteboard" type bs.

Great article.

BTW I run a 16 core home server with 10 tb of storage accessible over Comcast home edition for the cost of electricity. At the risk of sounding old and idealistic, isn't this what the Internet was supposed to be? Isn't this where the idealists among Web3 proponents are headed?

This is an affront in several ways.

1) Apparently bait and switch is fair play now? The fact that it took years for the hook to be applied makes it no less wrong.

2) Apparently we cannot buy software anymore. I don't want 90% of upgrades but they force me to buy them as "subscriptions" now. How many do you have now that virtual monopolies are converting from buy one copy of software to "subscribe only"? Google Apps is just another in this "hook the user on free access -- then milk them forever once they are committed," business model. Office, Creative Suite, what's next Chrome?, Edge?, Firefox?, Acrobat?

Google seems now 100% devoted to evil.

My advice is:

1) create an HTTPS WEBDAV server to sync your photos to. I did this a year ago when I realized that my bill for photos would increase forever if I didn't erase the old ones as I took new ones. Switching to my own hardware is like getting a print of those film negatives... 2) host your own email 3) Get an old copy of Word and Excel (who cares about new features?)

EFF killed cert costs, may they turn to these other commodities... Once you go through this painful transition you'll only pay for Internet access and the hardware to run it.

This article defines what web3 IS then criticizes it. If web3 were only blockchain it might be forgiven but IMO it is about a lot more like IPFS, no-trust networks, peer-2-peer computing, etc.

This article therefore makes some points while missing a lot more.

This is the kind of confused weak thinking that causes new infections to rise at 150%

Of course vaccination radically reduces the ability of the vaccinated to spread Covid, 90% of them never get it. Sure of the other 10% that might get it some do and wow they can spread it. By the time you get past the IFs vaccination is astoundingly successful but the more vaccinated that are exposed to the unvaccinated the more of those 10% WILL get infected. This seems crystal clear to me and actually I'm surprised at how low the rate of "breakthrough" cases are (I think I heard < 1% but fact check).

If you start with an algorithm that accepts any type of user behavior (is multimodal) then using user-dislikes is easy.

We can easily do this with the Universal Recommender as shown in this article (reprinted from the IBM dev blog) https://actionml.com/blog/making_dislikes_predict_likes

Not sure this is true but in any case, the Universal Recommender uses a technique based on the TF-IDF algorithm of search to get long-tail recs. This de-weights popularity so that relevancy is more important.

The Universal Recommender (part of the Harness project) allows contextual and multimodal behavioral information to be used in recs. https://github.com/actionml/harness

Apache PredictionIO was designed for data scientists (I was a committer so am quite familiar with design decisions). It was built to allow new "engines" to be implemented for any arbitrary ML/AI algorithm. These came in "templates" of various flavors.

However it turned out that only the "Universal Recommender" template (that I wrote) had real demand and virtually no one was developing or using other "engines"

As designed it was hard to deploy and was not multi-tenant. I could go on about the shortcomings but suffice it to say that after I did the Universal Recommender and saw the PIO shortcomings, we (ActionML, my consulting company) decided to write a from-scratch ML server, called Harness, to solve the PIO issues and act as the host for continuing work on the Universal Recommender.

https://actionml.com/docs https://github.com/actionml/harness

So the best part of PIO (The Universal Recommender) lives on with a clean new design and modern architecture.

Timely. I just got my closet server on the internet with WebDAV behind auth and SSL. Also installed PhotoSync on my phone. I do miss the image GP browser though (a little).

Thanks, it made me (poignantly) remember Ruby and _why

The argument was, the web and its tech is too slow for phones, we need apps.

This is no longer true and apps are a pain in the rear when it comes to all the integrations you have for mobile aware sites.

How can app content be crawled, don't ask it's a major pain. How do you track usage like site clicks, ouch! How about integrations like commenting or site support chat, good luck, there might be one for your app.

Not to mention the 30% vig to Apple and Google, or the lock-in.

Many thought apps would bring money back into what sites were giving away and ok, it did -- for Apple and Google.

Can we please stop making "apps" now?

Count me signed up.

This report doesn't seem to mention burn-in directly. This is IMHO the achilles heel of the current crop of OLED TVs. This is a very hidden fact. Everyone talks about how great they are and say they're the hands-down winner among popular TVs but after having owned one for 2 years it displayed horrible burn-in. They are fantastic while they last then they are awful. It would be nice to know how mLeds, uLeds, and the new tech sizes up for burn-in.

Perhaps true, but a greater lust fills Apple with desire - the lust for control of their users. It has become increasingly hard to install software from your favorite vendor on the Mac. The advent of the Apple App Store for Mac sure looks like a play for a closed eco-system like they have with the iPhone. Moving to ARM will make the push to a closed eco-system that much easier, if not inevitable.

I'm a dev that work in OSS Machine Learning and rely on a machine that runs all this server tech. Even when I use Docker, it runs a hypervisor based VM - impossible with ARM which will need an emulated VM. This all makes ARM dev machine a non-starter. Add to this the crappy quality of the MacBook Pro (I am now 2 keyboard failures into the MBP) and I have to say goodbye to my otherwise beloved Mac.

Years back the Mac took over my area of development. I remember going to a MongoDB conference in 2011 and looking around to see nothing but Macs in the audience - I was a bit surprised to see such a huge shift a trend that has only accelerated (even with crappy MBPs). Now I wonder how many of those folks will opt-out of Macs until there is a huge demand for ARM servers.

Apple doesn't get to make this decision for me, the market for servers makes it. Until the market moves I need AMD64 + a `nix OS. Maybe a Dell laptop and Linux - oh well...

I ditched my Windows PC when Apple moved to Intel. I will ditch my MacBook when it uses ARM.

Not dogma, just practical. I'm a `nix SW Dev and require the 95% of code that works on AMD64 run on my machine. Odd to see MS now supporting more `nix. What we (many SW Devs) need is `nix + AMD64 for the foreseeable future.

Furthermore I do not trust Apple. The advent of the iPhone software lock-in echo-system shows where they want to take Macs and that is just a no starter for me.

Bye-bye Apple.

IPFS https://ipfs.io/

We developed an OSS ML Server called Harness. It does all ingest, prepare, Algorithm management, workflow bits for pugable ML "Engines". These are Algorithms + Datasets + Models and are flexible enough to do most anything. We use the build-in Universal Recommender Engine, and have built our own for other uses.

Harness exposes a framework for adding Engines and does all the routing for Engine Instance workflow and lifecycle management. It also provides a toolbox of abstractions for using the Spark ecosystem with Mongo and Elasticsearch.

It comes in a docker-compose system for vertical scaling and Kubernetes for ultimate in scaling an automation. Quite a nice general system with out of the box usefulness.

https://github.com/actionml/harness