A week with Anytype

 

Anytype looks like a local first, privacy-focused alternative to Notion. I have used Notion in the past and generally liked it, though not for its web-based philosophy and price structure. Anytype avoids those Notion disadvantages by being secure, local and free - though with paid options if needed. There are many similarities between Anytype and Notion, so from a superficial perspective, the comparison with Notion is fair. 

 

I first tried Anytype shortly after it was released as a beta product in 2021. Then, it was too early in its development to be a serious prospect, and I put it to one side. Now, four years later, that has changed, and Anytype, though still in beta (v0.47.5), is maturing and getting fairly feature-packed. Perhaps I could use it to build a PKM (Personal Knowledge Management) system that is better suited to my specific needs than Obsidian? 

 

Managing expectations: - Anytype isn't really an application; it's a tool to build an application. It's only an "everything app" because, like putty, you can shape it into whatever you can imagine. So, when you can't find an app that does exactly what you want, then you may be able to create it in Anytype

 

First impressions

First impressions are not very good. Not because Anytype is bad per se, but because it is not easy to learn.

• There is currently a lack of well-made tutorials on YouTube. Those that do exist tend to be of the 'do this, click here, do that' variety, lacking in any explanation of 'why'.
• There are examples available (in the gallery), but again, they lack any explanation for the design choices made. They may get you started, but they are not good learning tools.
• There is application documentation on the web, but it is low-level technical and lacking in any high-level "if you want to do this sort of thing, you might approach it in this way" style of information.

 

To my mind, learning Anytype is summed up best by analogy: Imagine signing up for driving lessons and when you turn up on the first day, you are shown a garage full of car parts and told, "Just assemble these and then we can go for a drive". Anytype is a bit like that, and after the first three days, I still felt as though I was learning what the parts were and how they best fit together to achieve my objective of a PKM system.

 

I guess that many potential Anytype users might give up during this 'assembly' process and that Anytype need to address this problem before adding many more 'parts' to the garage.

Second impressions

I'm a persistent old sod and won't be easily deterred, so I pressed on and continued building. By the end of the first week, I had learned what object types possessed what properties, that what I had assumed could be "collections" (thank you template) worked better as "queries", and so many other detailed things. The car still isn't built yet, but it is beginning to look more like a car, and each time I add a new part, there is a sense of achievement.

 

The issue is that I am conflicted about continuing. I really shouldn't be investing this amount of time in building a customised PKM system that exactly suits my needs. But, at the same time, it is an intellectual challenge that produces a certain sense of satisfaction when a new part of the puzzle comes together.

 

Perhaps that's the thing, To use another analogy, Anytype is like a bottomless box of Lego bricks, you can make almost anything you like. There are no plans, so there's going to be a lot of trial and error, lots of searching for a brick that's just the right shape (if it exists). The payoff at the end is that you could have something you'll be proud of creating. If that's your thing, then Anytype may be for you. But I suspect that most people should move on - nothing to see here, folks.

 

As for now, and to keep the analogy train rolling, I'm continuing to build the plane while flying it. I wrote this article in Anytype; it belongs to my "Writing & Blog" area, which is part of a modified PARA* template provided by Anytype. Maybe learning while doing is not the best approach, but for now, the plane's tank still has loads of fun-juice, so let's keep on flying and hope the fun-tank doesn't run dry before I land this thing.

 

In the future?

I've been looking (with a certain amount of amusement) at one of the latest approaches to software development - A.I. vibe-coding. You tell an AI what you want to achieve (the vibe), and the AI goes away and writes the code to make an application to do just that (sort of). 

 

It seems to me that Anytype could well benefit from such an AI front end. It would allow users to describe the application they wanted, and an AI could simply use Anytype's objects to create that application "automagically", as my developer friends used to say. I don't know what an AI-built Anytype app would be like, but it wouldn't take a week to build.

 

You can find Anytype at: https://anytype.io

*PARA - Projects, Areas, Resources, Archive. An approach to knowledge management by Tiago Forte.

Luminar 3 (a.k.a. "Luminar with Libraries") came out yesterday. L3's predecessor, Luminar 2018, is a capable piece of software on the Windows platform, but Luminar 3 is not.

I had only been waiting a week or so for this release so my disappointment, though heavy, would not have been as soul destroying as those that have waited for over a year. L3 for Windows should never have been released, it's not even ready for beta testing. </opinion>

Installation was a nightmare. It took four attempts to get it installed in a form that would even start up again after it had first been started and closed. And, in case you're wondering, yes I have a reasonably powerful machine (i7) with plenty of  RAM (16Gb) a modern graphics card (2Gb) and disk space to spare - all running an up to date version of Windows 10. I only got L3 working eventually, by limiting the catalogue to about 2,000 images rather than the 60,000 in my library.

When it did start working, it quickly became obvious that L3 for Windows was nowhere near ready for prime-time. </more opinion> The fact that some features were missing is, perhaps, forgivable (software can be updated). The fact that some necessary functions don't work at all is not. There is, for example, no folder refresh - once L3 for Windows has built the catalogue for a particular folder, it cannot be refreshed or updated without closing and reopening the application. That means that no other applications can put files in that folder and have Luminar recognise them during that work session. Not even Luminar its self!

Another deal breaker is that Luminar has no ability to send files to other applications. Any DAM (Digital Asset Manager) must be able to do this and the implementation of this feature can only be described as trivial. For example, my current DAM (ACDSee) allows me to configure any external editors I wish, and the folders update with changed and added files on the fly. This is all base DAM functionality.

There are heaps of missing features that need to be added and some of the existing functions can only be described as "flakey". Did I mention S-L-O-W, very, very, slow? It all adds up to a broken package that isn't fit for purpose. The combination of ACDSee and the previous Luminar 2018 is still a way better overall solution than Luminar 3 on Windows. Shame on you Skylum.

Slumming it (photographically speaking)

I've lost count of the amount of money I have spent on photographic gear over the last fifty or so years, but I'm guessing that it comes to tens of thousands of dollars in today's money. There is a fairly high correlation between the cost of camera gear and its quality so, if you want the quality, you generally have to pay the price. But this year I have been 'slumming it', carrying some modest 'point and shoot' cameras that lurk nearer the bottom end of the cost and quality spectrum.

The thing is, there have been some significant advances in software this year that allow even photos from cheaper cameras to end up looking and feeling as if they have come from more expensive gear. Take these five photos; photographically they are nothing special - all of them came from a single one-hour walk 'around the block'. But the final image quality is similar to the quality I would get from my full frame DSLR, rather than a JPG from a point and shoot; which is what they actually are.

This is a very freeing development: To get good quality results, I don't need to carry a heavy rucksack of expensive gear around - just a tiny camera that fits in my shirt pocket. The expensive DSLR? It still has it's special uses; just not so many and not so frequently. Anyway, I'm finding that slumming it is fun.

Topaz AI Gigapixel - tested

When Topaz came out with their AI Gigapixel (AIG), image resizing product, I was sceptical. Firstly, I had tried many resizing products in the past and had settled on one that worked quite well (within the accepted constraints of resizing). Could AIG, do better than what I was already using? I doubted it.

My first attempt was not good. AIG crashed on first use and clearly didn't want to work with my graphics card. I walked away disappointed but not altogether surprised. I told myself that, anyway, it probably wouldn't have been any better than what I was already using.

I must be a glutton for punishment because when the email arrived announcing an improved 1.1 version, I downloaded it and tried again. This version included a CPU mode, which bypassed the GPU on my graphics card. That should, at least, allow it to run.

It did run, and the first output looked surprisingly good. So I devised a stricter test - I took a 24MP DSLR image and downsized it, in my normal editing program, to 6MP (half the width and height). I then put the 6MP image through AI Gigapixel at 200% to bring it back to its original size. I could then compare the original 24MP image with the enlarged 24MP AIG image and see how different they were.

The AIG image was sharper than the original. AIG obviously adds some sharpening during the resizing process. To compensate, I added some sharpening to the original file so that they were more directly comparable. What I was really looking for was a loss of detail and fat 'jaggy' edges in the enlargement. At 100% I could not see any difference between the two files. Definitely, not the result I was expecting.

Screenshot - Original 24MP file on left - 6MP file enlarged to 24MP on right (click to see full size) 

The screen-shot above is taken at 200% magnification and contains both in focus and out of focus items. The original image is on the left and the AIG resized image is on the right. I am still struggling to see any significant difference between the two images. Not quite believing what I was seeing, I repeated the same test on other images all with a similar result.

As a final test, I took an image from a different camera - a compact with a 1/2.3", 16Mp sized sensor. I enlarged the image by 200% in my image editor, using the bicubic sampling method and then enlarged the same image in AIG and compared the results. The screenshot below, is at 200% magnification and the results are quite surprising - the AIG image has much finer detail (particularly at the edges), texture seems to have been replicated rather than enlarged, and image grain has been suppressed to a far greater extent, leading to a much cleaner image that would be difficult to pick as an enlargement.

Screenshot - Bicubic enlargement on left, AIG enlargement on right (click to see full size)

All my tests with AIG have been at 200% enlargement because that is what I would normally use. But AIG can enlarge up to 600% (though the results may not be so acceptable). The shrink-enlarge experiment indicates that AIG could be an effective substitute for a camera sensor with four times as many pixels (16MP becomes 64MP) or for a longer focal length lens. AIG produces significantly better results when compared with either a standard bicubic enlargement or previous industry-leading enlargers.

US$99 does seem a lot for a single task piece of software. But when compared with the cost of a camera with a larger sensor or a lens with a longer reach, AI Gigapixel may well be a reasonably priced alternative.