Sunday, 31 July 2022

e-paper devices (part 92)

 It's been a cold blustery day here on the edge of the Alps, so in between catching up on a couple of podcasts I've been researching where we are with e-paper devices.

At the bottom of the market there's a whole slew of slate like devices, mostly from east Asian manufacturers, and some of which have been repackaged as home note pads - put it on the fridge to tell the kids you've run away to Guatemala to run a taco stand - that sort of thing.

Practically their real usefulness is as a scrap paper replacement - instead of endless scribbles on bits of scrap paper, scribble on an e-slate and wipe it when you are done.

Looking at the pile of paper scraps on my desk, perhaps not such a bad idea, although both my cats would be upset - one of their favourite games is to get scrap paper out of the bin and bring it to me to screw up into a ball and throw  down the hallway for them.

At the top end you could have an iPad with an apple pencil, a remarkable, which looks to be really well thought out, and one or two other devices which allow annotation etc. I suspect that by next year we'll see some competition in this space and the reason for Remarkable's advertising blitz is to establish market share before the competition arrives.

And the middle?

There's a number of products based around reusable notebooks, which you wite on with a special pen, scan, and then wipe clean to reuse. Some, like Rocketbook, seem really well thought out, others perhaps less so. One of the selling propositions of these reusable notebooks is that they save paper by being reusable.

As I've argued elsewhere, you can buy a hell of a lot of recycled paper notebooks for the cost of a Rocketbook (something between fifty and sixty dollars, not to mention the special erasable pens at between $10 and $12).

And of course paper is eminently recyclable (and scannable) - I've been scanning my diaries and notebooks for over ten years now, so I'd describe them as an interesting bit of technology, but really as a solution in search of a problem ...

Thursday, 28 July 2022

e-paper devices

 I don't know about you, but every time I look at Facebook I get adverts for the remarkable-2 an e-paper tablet that lets you take notes, save them as pdf's to a range of cloud services, and even does handwriting recognition on them. (Given I have the handwriting of a drunken ape, I have my doubts about the latter.)

I've never played with one of them, but I did see a colleague's yesterday and it looks like a pretty good device. Does what it says and does it well. 

It's also horrendously expensive.

If I was still going to meetings and able to claim tax offsets for work related items I might crack and jump for it, but I'm retired, don't go to meetings that much, so I can't really justify the cost at around $500.

There are other devices out there that do similar things, but they are all pretty pricy. I'm sure they're good but I can't afford them.

On the other hand at the bottom end of the market, there are a whole range of electronic memo pads that don't talk to anything useful, like the XiaoMi Mija electronic writing pad. Basically it's a slate.

It doesn't talk to anything, and you can't really store content on it. If you need to get content off it you'll need to take a picture, just as you do with a white board.

It's fairly difficult to see a use case for it, and $50 will buy you around 20 A4 recycled paper notepads from one of the big box stores, which is equivalent to something between 1500 and 2000 pages of notes.

However, the one use case I can see is for work in progress notes. 

When I dug out my old palm pilot and put it back into use, one of the use cases was for work in progress notes - these scribbled post its and marginal notes that everyone makes.

And for that I can see a use for a basic electronic scribble pad, because, even though some basic connectivity would be nice, the notes are only there to remind you to do something - they have only transient value.

My real complication is being able to write on something while wearing examination gloves - the palm pilot's hunt and peck keyboard does work well, but it's slow, and you can't do diagrams or special symbols.

So I guess I'm looking like for a poor mans remarkable ...

Tuesday, 12 July 2022

Building the Raspberry Pi x86 desktop ...


Well, in the end, I couldn’t resist trying the Raspberry Pi desktop.

There’s not a lot of information on the Raspberry Pi website about installation, so I went for it and downloaded the ISO image and used Rufus to build me a bootable installer on a USB stick. 

The download’s around 3.4GB, so I’d use an 8GB stick at a minimum.

I chose the dd option when making the bootable USB - I've always found the dd version more reliable.

First time around I chose my very elderly (like about 12 years old ⚠️) Celeron based Dell Inspiron 1545 as an install target as it was the oldest and slowest machine that I had around

Originally, it was my main machine, but as with all things it became too old, too slow. I moved it over to Xubuntu, used it a bit for a couple of projects and then stopped using it.

Despite the machine being 12 years old, installation just worked.

Booting off the USB, you first of all get presented by a slightly bizarre menu which gives you options about persistence etc, which I didn’t understand, so I picked the graphical install option, which put me into a very standard Debian 11 installer.

I didn't try anything clever and just followed the bouncing ball as far as install questions went - the whole disk for the operating system, etc.

The only oddity was when it asked me if I had the proprietary Broadcom drivers on removable media - I of course didn’t, but answering no didn’t cause a problem - it silently used the open source alternative drivers anyway. (This is only be a problem on some machines with Broadcom adapters - on machines with adapters from other manufacturers the installation program makes use of the default public domain driver for the adapter)

Installation is a two stage process and on first boot it takes you through a set of configuration options as to location, keyboard layout and so on, plus network setup and creating a default user. (Incidentally the system by default logs you in as the default user)

After configuration it was a case of rebooting into the standard environment.

The standard environment was pretty minimal, the browser was chromium, which is a plus and the default email client was claws which has never been my favourite.

However, Thunderbird installs fine, as does my favourite editor, kate. Just for fun I installed ristretto and abiword (not that there’s any need to - Libre Office is preinstalled) to prove that most things worked - which of course they would, it’s debian after all.

So, after ten or so minutes playing with it, what do I think?

It’s good, it’s got definite potential. Whether or not I keep it installed and use it, only time will tell. But it’s certainly one of the better lightweight distros I’ve come across …

[Update 17/07/2022]

I decided yesterday that I had too many old laptops - my old Dell Inspiron, J's old HP Beats audio laptop, and of course the old laptop that I bought at the start of lockdown in 2020. As it is that still leaves me with more machines than rationally anyone really needs ...

No one needs that many machines, so I wiped them prior to going to the recycler - and they now have the Raspberry Pi desktop on them as a sort of subversive present. I've taken advantage of the two stage install process to leave them ready to boot into the configuration menu, so that if anyone wants to play with them they have to configure both their own default user and wifi settings.

another 32 bit OS to run on an old netbook

 Back in the early days of the pandemic I put together a post on which linux could be run on an old 32bit pc, the problem then being that all sorts of old machines were being pressed back into service just as a lot of the major distros had dropped 32 bit support.

All of the ones I surveyed and tried then were competent but not mainstream, and that could be a problem.

Smaller linux distros  are almost always reliant on a very small team of people, all of whom have lives to lead and and are often working on a distro as a hobby.

They often produce great code and a nice distro, but then, after a few years, the team go their separate ways, and that's that. A bit like rock bands really.

So I was interested to read a review in The Register of the X86 Raspberry Pi desktop  - basically a port of the Raspberry Pi environment to a 32 bit X86 computer - otherwise a netbook.

The distro sounds good, is debian based,  and crucially is well supported by the Raspberry Pi foundation.

Unfortunately I took my old netbooks to the recycler last year when my chromebook died, otherwise I'd do  a test install on real hardware and a review - running it in a VM isn't quite the same, and I do like to eat my own dogfood.

However, that said, it sounds interesting and might be something to go for is you're still trying to get an old netbook to be useful for a data collection project ...

Wednesday, 6 July 2022

Documenting drawers


At Dow's we have an old fashioned built in cabinet behind the counter. The drawers are mostly labelled. Some of the drawers contain things - I've done a quick eyeball check and while mostly each drawer contains several of the same - eg ointment tins, aspirin packets - the drawer labels bear no relation to the contents. And of course some of the drawers are empty.

So obviously we need a standard procedure for documenting them. Up to now in the documentation methodology, we've mostly had to deal with individual items, but where the item has been a container of some sort I've usually used contains <individual item description> in the comments field, the idea of prefacing the individual item description with contains is that this would make it easier to programmatically break it down to a set of standard item descriptions at some point in the future if required.

This is quite extensible - if there are multiple and different items in the container, one simply has multiple contains statements.

This procedure was made up on the fly to solve a problem of cardboard boxes full of empty medicine bottles but it's worked quite well for a range of multi object contents description including a display of 1950's lipsticks.

So, given it works as a solution it seems sensible to use it to document the drawers - but rather than continuing as an ad hoc tweak to the methodology I've decided to document it more formally so that we (a) have a record of the procedure and (b) something to work to, because, as we know, in documentation - especially where it may end up being processed programmatically - consistency is king

So this is what we have:

Simple, but hopefully robust.


First couple of drawers in I hit a problem as to what to do when an item contains another item - such as a tube of ointment in a cardboard box

My slightly inelegant workaround is to use the string item_contains to describe the contents of an artefact. ( I originally was going to use the phrase box_contains but item_contains is more general and does not assume a particular container type. I also did not at first use an underscore in the string, but using one instead of a space makes life easier for a program to parse the description while maintaining readability for humans). So our procedure now looks like

Monday, 4 July 2022

A portable documentation success

 Six months or so ago I started using a wheeled computer bag for transporting my field documentation kit.

Despite my previous unhappy experiences with wheeled bags the use of a three compartment bag works and works well.

So what do I have in it?

Compartment 1

  • laptop (currently a 14" refurbished thinkpad)
  • box of blue examination gloves
  • field kit : pens, pencils, spare usb sticks and sd cards, and bits and pieces
  • plastic documentation wallet holding my A4 day book and a 12" ruler
Compartment 2
Compartment 3
  • notebooks ( 2 of them, both A5 - one for general notes, one for project notes and drawings)
  • paper diary
  • ipad mini
  • keys
all of this fits in - but only just. I usually have a couple of plastic rubbish bags in a side pocket for dead gloves etc, and I guess I could fit in a box of ziploc bags if required. The camera is good quality point and shoot Nikon, but any decent little camera would do.

My A4 day book consists of rough handwritten descriptions of artefacts before their final transcription to an excel spreadsheet, plus various crossings out and corrections. I'm not fanatical about these, they're basically whatever's the cheapest spiral bound A4 notebook I can pick up.

The Spirax 240 page ones are probably the nicest to use (slightly better quality paper, but in truth I usually use the Officeworks own brand 120 page ones you can pick up in a multi pack for five or six dollars).

I used to just use standard nitrile gloves that you can get from a hardware store or  big supermarket - after all they only need to act as a barrier between you and the artefact so that a 120 year old packet doesn't leak anything nasty on to you, and vice versa that you don't deposit any skin oils or grease on 120 year old packages. They don't need to be clinical grade - the ones used by plumbers when handling soil pipes are fine, and actually fit slightly better.

During the pandemic suitable gloves almost completely disappeared with manufacturing capacity being diverted to producing medical grade items. They havn't made it back reliably onto supermarket shelves so last time I was running low I bought myself half a dozen boxes from a veterinary supplies company which should last me to the end of the project.

My paper diary is of course a Moleskine planning diary, but my two A5 notebooks are simple spiral bound notebooks with rigid covers that came from one of the big box stationery stores - probably Officeworks. 

I'm not fanatical about these, as long as the paper is decent quality and they can be opened flat to scan if required, any brand will do.

We have a reasonable NBN connection at the pharmacy, otherwise I'd probably have to add my travel modem to the mix. It has an internal battery so there's no need for a power supply, but for extended periods of work I'd add a small USB charger to my bag. 

There's also a message here - over the life of the project I've (mostly) paid for all these items myself - some I had already, some were bought for other reasons - for example the ruggedised Nikon was originally bought for our Covid-aborted trip to South Africa - and the Trust did give me an old laptop, which I've since replaced with a more modern refurbished Thinkpad, which of course can be reused in a subsequent project.

Probably my participation in the project has cost me around $1500, which to put it in proportion is probably about the same as I've spent on my other recreations of cycling and walking over the same amount of time. 

I have been happy to bear the costs because I have enjoyed the project and  it has been fun, even if it might be a slightly odd idea of fun. However it's important to realise that citizen science, community history projects and the rest all involve costs for the participants, and this should be borne in mind when designing community or volunteer projects...