Saturday, 17 October 2020

Old machines and education

 Way back in May, I wrote about how (a) old machines had vanished off of the second hand market, and (b) why taking an older, less suitable machine, and sticking linux on it was a bad idea.

Not that it won't revive the machine and make it useful, but that it comes with a support cost.

Basically, being a fully paid up geek, and someone who has played with multiple operating systems for years, I can cope with using just about anything. 

That's fine for me, but it's unrealistic to expect a teacher, with no experience of linux, to cope with student using open source products to do their work, or be able to fully support the student.

Basically it would be sink or swim.

To work, online learning needs a predictable environment that gives a degree of standardization. There's no reason why you couldn't standardize on linux, but you need to plan it properly.

Windows and OS X both offer predictable environments and ones where one can assume the presence of certain browsers - edge and safari respectively and the presence of some standard applications. The joy of linux means you can't do that - while there are a lot of components in common, various distributions are different enough to complicate things, and as I showed some time back, if you are using an older revived machine, you may be using a less than mainstream distribution.

So my heart sank when I saw an article in the Register about taking an early 2007 vintage Macbook (one of the early intel based machines - the article doesn't make that entirely clear) and sticking Elementary OS on it.

Actually, I'm lying when I said my heart sank - I actually thought it would be a fun thing to do, well except that no one's selling polycarbonate Macs on ebay in Australia for fifty bucks - more like a $150, and that's too much for a fun project.

But reviving a machine is only the start of it - if you provide it to someone you need to provide some support, and if you have multiple linux's which do you support?

Consider the start menu. In Xubuntu it's at the top left. Gnome or KDE distributions usually have it at the bottom left, and OpenBox based distros like Bunsen Labs prefer you to right click on the desktop.

Nothing wrong with any of them, but a nightmare to support.

So while I'm all in favour of reviving old machines by running linux on them - basically there are social and environmental positives in doing so, I'm well aware of the support costs involved and why, to succeed, any project needs to have a carefully thought out end user support plan ...

Using an Android tablet for a serious purpose

Recently, I came across a New York Times article on how to get various older computers to be useful that, while it agreed that iPads could be useful, was quite dismissive of using android tablets for real work.

(I've lost the URL, otherwise I would link to the article)(see https://www.nytimes.com/wirecutter/lists/getting-work-done-on-an-ipad/)

As always the problem is what you define as real work. 

Way back in 2012 I spent around a $130 on a Shenzen special seven inch tablet packaged with a keyboard - and I used the device for around three years to take notes in meetings and so on - it was both incredibly useful, and terribly nonstandard, with a USB micro B connector for the keyboard and a separate 3mm jack for charging, but it worked, and I could get a day out of it.

I still have the device - the battery failed and was not replaceable - it really ought to go to the e-waste disposal people


but the damned thing was so useful I went out and replaced it with another cheap tablet - this time an Alcatel Pixi 7 I got from Telstra's disposal shop on eBay for round about eighty bucks to which I added a case with an inbuilt bluetooth keyboard:


Even though nowadays I use the refurbished  iPad Mini I bought a couple of years ago as a carry round device, the Alcatel was a good machine, and you could type reasonably well on it, and save your notes to Dropbox for further processing - as with the Shenzen special I tended to use Markdown to create semi structured text that you could feed through PanDoc to provide something a little more corporate when required, but Polaris Office  also worked well with the advantage of being able to sync to OneDrive or Google drive.

And there I left it, or rather I did until my recently acquired Huawei Mediapad.

Having retired, I no longer have to go to meetings, or at least, not very often, and an A4 pad is usually good enough for notes. Occasionally, I did some research or writing work in a public library, and when I did I used to first use the Alcatel, and latterly the iPad Mini when I didn't want to cart around a full size computer - a netbook or my MacBook would have done, but there's the question of battery life - a surprising number of public libraries provide desks, but nowhere to plug in your device.

Now you might recall that the MediaPad came bundled with Microsoft's Office tools for Android. I had no intention of using as a laptop substitute, but out of curiosity I invested the princely sum of eighteen bucks (including delivery) in a no name bluetooth keyboard


and it was surprisingly good. They keyboard was a little bouncy, but good enough to type quickly on and Microsoft's word tool for android was equally responsive, and allowed documents to be saved OneDrive using one's Office 365 credentials.

As an experience it was provocative - a tablet could be used for real work with a standard application. While I've no intention of using it as a poor man's surface the whole thing worked so well one could imagine doing so ...

Wednesday, 30 September 2020

Of bookscanning and image sizes

 J, my life companion, is an accomplished pastel artist, and wanted to put some of her artwork into a competition.

Pre-Covid, this would have meant selecting a picture or two, getting them framed, driving somewhere, and watching someone from the exhibition team put them on the wall.

This year, of course, everything is different. Pictures are photographed, and the images uploaded to the exhibition website, where they are loaded into some gallery software.

Now, what was interesting about this process is that the exhibition organisers said to use a digital SLR for the images, not a mobile phone because of the image quality.

Now, J's artworks are normally something between A4 and A3 in size (that's because that's the sizes specialist paper for pastel work comes in), and for archival purposes she takes a picture with her iPhone, which has an 8 Megapixel camera, and archives them in iCloud, using what I'll call iPhoto (it's actually called Photos these days).

Apart from iPhoto's tendency to produce smaller than expected jpegs on export this works well as a process


Internally, Photos uses the newer High Efficiency Image File Format  rather than one of the other more standard formats to achieve an efficient use of resources using lossless compression.

As always, we can argue about compression, image formats and archiving, but using HEIF is no more at risk of introducing compression artefacts than anything else, and may even be better as it is claimed to be lossless.

Professionally though, most people use cameras for archiving work rather than mobile phones.

We've all seen pictures of archivists using digital SLR's mounted vertically on a stand to take images of old photographs, and obviously when you don't know the exact size of the image and want a high quality image this makes sense. 

But the question is what is good enough?

Well my little experiment using a photoscanning app on a phone has convinced me that a phone produces a good enough image, even if the OCR's result of the text would need a little work:


and there was report in Nature this morning (which I retweeted) about a group of scientists using the Covid hiatus to scan old lab  notebooks


now the interesting thing is that most of the work was done using mobile phone cameras and a phone scanning app - in other words the scientists concerned found the images perfectly adequate.

At the same time if one searches for book scanner Google shopping or Amazon, one gets results similar to this


delving into the specifications one finds that they all use a camera with a fixed image size - the cheaper ones tend to be designed to image only a set page size, usually A4, the more sophisticated 'bendy' ones can be adjusted to scan a page to a maximum paper size - usually A4 or A3. All, or almost all, use either an 8Megapixel or 5Megapixel camera - assuming the better or pricier devices using an 8MP camera, the cheaper fixed image size devices a 5MP camera.

I don't know this, but I'd guess that the scanners are using mobile phone camera assemblies. An 8MP image of an A4 page would give you roughly 300 dots per inch, which is pretty sharp and as sharp as many high quality printed images. (If you are planning to OCR the text, you actually don't want a supersharp image of old typeset pages as these can introduce artefacts that confuse the OCR software.)

So, where does that leave us?

For J's artwork, for a sub A4 image is probably good enough at 8MP and for book scanning it's certainly good enough for OCR.

If your image is bigger, yes there's probably an advantage in using a higher quality camera, but for most purposes 8MP is good enough ...



Sunday, 27 September 2020

Huawei mediapad

 It's no secret that I like messing around with old documents.

Normally, when working with digitised content, I use an old 2008 vintage iMac - long unsupported but still with an excellent screen - to display the item, and I'll  type the notes into a laptop.

I could, I guess, have a single machine with dual screens, but at the moment this works for me. What this solution is not, is portable, which can be a pain when working somewhere like a library (which I havn't done for six months because Covid.)

Now the little note taking ipad I bought myself a year or two ago has become useful as a carry around device - but the screen is a little too small to work with when looking at old documents. 

Given that I normally work off of a laptop, I decided that a standard format tablet would probably be the thing to go for.

An iPad Air would have filled the bill, but not at the price Apple charge for a new one, and decent refurbished items have disappeared off the market.

So that meant Android.

Now if you go to any of the big box stores you have a choice of Lenovo or Samsung, and the items with decent quality displays are reasonably pricy. 

So I read some reviews and overseas people seemed to rate the Huawei mediapad - decent screen, good battery life etc. There's two models and the better specc'd 64GB model isn't currently available in Australia - except that for some reason Amazon will sell you one from Amazon UK via their market place.



so that's what I did.

It only took a couple of weeks to get here. Gratifyingly it was not crammed full bloatware, giving you a fairly vanilla machine to work with. The only problem was of course that it came with one of these bizarre UK claw chargers:


which wasn't a problem as, like most people, I've oodles of spare micro-USB chargers. I've also got an array of international sockets in my workshop dating from the days when I used to play with kit from overseas


Setup is standard Android, and the tablet comes bundled with the Office 365 tools for Android. There's not a lot in the way of unnecessary bundled apps, but the device comes with  Huawei's own app store as well as Google Play. The whole setup experience is pretty vanilla.

In use the device is responsive and the bundled Microsoft swiftkey virtual keyboard is one of the nicest I've come across.  Screen quality is as good as promised, and the device is light and sits nicely in the hand.

Definitely a business class machine despite its low price.

Huawei include their own mail client and 5GB of their own cloud storage but there's no compunction to use them, or their own app store - you can just as easily use your own preferred mail client and cloud provider, and delete their apps off of the device should you prefer.

Due to Huawei being banned from 5G networks in Australia and the recent reported hacks of university computer data, there's obviously going to be some questions around security.

Personally, if like me, you are a private individual, your data is probably no more at risk with Huawei than with any other cloud provider. 

If, however I still worked for a university or a government body, paranoia might kick in and I might think twice about buying such a device, but equally, you can be too paranoid - after all Telstra, no less, sold me a Huawei 4G broadband modem a couple of years ago ...

Equally, if you want to be careful, you can simply avoid installing applications like online banking on the device, or simply access them via the web.

It's a shame that the Mediapad is not better known in Australia. It's a good well made, well priced device that does what it says.






Saturday, 12 September 2020

What should happen when an online journal dies

 

Over the last few days I’ve tweeted links to a research paper and two news articles, one from The Register, the other from Nature, on the phenomenon of disappearing open access journals.

I must say I’m not surprised.

While I have never worked on an open access journal, I have built a number of data repository solutions for both higher education and government, and was once even on the management committee for the long gone UK Higher Education National Software Archive.

And if there’s one problem with every solution I’ve built, it’s sustainablity.

While the systems are comparatively cheap and simple to deploy – you can build an Omeka instance in an afternoon, and building a non customised Dspace install is similarly quick, production based systems need hardening, security and customisation, all of which requires a small of software engineers – usually about two, and a part time manager to manage the install and deployment of the solution – and because the only metric we have is money, we can say that if deployment takes a year it will cost around $300,000.

Pre-cloud, and pre-virtualisation, the cost of hardware and storage was a significant consideration – nowadays, less so, so let’s stick with the $300,000 annual cost but assume we manage to deploy and get signed off in less that twelve months, and that we are using a virtualised server and cloud based storage. Sure there are hosting fees and storage costs, but you don’t need to worry about redundancy, backups, and maintenance costs for the hardware – a lot of these costs are simply abstracted into your monthly hosting and cloud storage bill.

After you’ve got your solution deployed, there’s probably less work for your deployment team, but they still need to have a role patching your repository or journal system, adding features, and so on.

So while you may not need so much of your repository or journal system team’s time you’ll still need a reasonable bit of it, so let’s stick our fingers in the air and say that the ongoing costs of running a solution is around $200,000 a year.

Remember that’s the cost of keeping it running. It doesn’t cover any of the costs, in the case of a journal solution, associated with managing the publication workflow – getting the submitted paper in, out to the peer reviewers, back from the peer reviewers, updated, revised, returned to the reviewers etc.

It’s quite a lot of work and require employing at least a couple of staff and a journal manager. Obviously, you can reduce your costs by running a preprint server as opposed to an open journal solution. Typically, though, preprint servers do not charge a submission fee, and trust that anyone submitting a preprint cares enough about their academic reputation not to publish rubbish.

Many open access journals work by charging a fee for you to publish your research – for example PLOS One charges a one off fee of US$1350. In the case of PLOS One, a well known journal with high impact scores, they almost certainly have a submission rate that allows them to cover their operating costs.

For smaller journals, and ones dealing with a highly specialist area, it may be difficult to charge a fee sufficient to cover their costs, or indeed achieve a submission rate that generates a sufficient level of income.

Inevitably, that will mean that the cost of running the journal is subsidised in some way by a learned society or by an academic institution, sometimes for reasons of prestige.

Now times are hard in academia. Government funding is grudging to say the least, and in these Covid times, student fees don’t provide the income they once did.

And departmental managers then look at the $200,000 or so it’s costing them to host a journal and not unnaturally think ‘we could get three, even four, postdocs for that, and they might do something significant’.

And so the journal ceases publication.

But of course it doesn’t end there.

To keep the already published content available, you need to keep the server running and patched, which means employing someone with suitable skills. In the old days you could trust that some libraries would keep the old issues on the shelf. With electronic journals it's a little more tricky.

So not surprisingly, sometimes the host ends up killing the whole thing and the content simply goes. Specialist dark archives such as CLOCKSS sometimes ensure that the content survives, but CLOCKSS is no resourced to cover everything, so smaller journals might simply be missed and disappear down through the cracks.

People who start small specialist journals sometimes fail to understand that starting a specialist journal is a bit like owning a cat – when you take on a cat you agree to cover its costs, feed it, take it to the vet, and in return you get affection, companionship and the occasional dead rodent – but the point is that you agree, implicitly, to pay for the animal for the fifteen or seventeen years of its life, and if circumstances change you get the animal rehomed so it can continue to scratch furniture for the rest of its natural life – in other words you have a tacit sustainability plan.

Small online journals need to have such a sustainability plan to cover what happens when the host institution can no longer afford to cover the costs of the journal, including alternative hosting arrangements …



Wednesday, 12 August 2020

Using a phone photoscanning app to capture old documents

When you work with old documents and photographs, not to mention ephemera such as labels and packaging, sooner or later you need to make a copy.

Being an utter geek, I've both a flat bed scanner and a film and slide scanner at home, which means I can scan most things, and at the Dow's Pharmacy documentation project I use a little Nikon camera to photograph artefacts. The only thing I can't so is  photograph or scan books, but I do have half a thought to make myself a DIY book scanner. Or perhaps not - basic book scanners are remarkably cheap these days:



But sometimes, when you are out, you come across a photograph or document that you want a copy of. You can't take it home to scan, so what to do?

The obvious solution is to take a photo with your phone, but you then end up fiddling about afterwards tweaking and straightening the image.

But over the last few years, various scanning applications have appeared, initially with the aim of allowing you to easily and accurately scan invoices and receipts, but clearly you could use them to scan anything.

So how good are they?

Well I did some experiments using the Google Photoscan app on my iPhone (it's an iPhone 8, so the camera is nothing remarkable in terms of capability and resolution).

First of all, I scanned a page from an old notebook, and certainly I got a nicely lined up and legible image:


Now we have a copy of the Compleat English Gardener which belonged to J's great^n grandfather who was a market gardener in Barnard Castle in the 1810's. I took a photograph of page 9 and also scanned it with the scanning app.



(top: iphone photo app :: bottom photoscan app)


The photographic image is more realistic than the scanned image, but the scanned image is perhaps more legible, and has got rid of the shadow of my chair. What it does show is that photographs can be perfectly usable, but you might want to use the scanning app to guarantee an accurate image.

I then decided to compare how it handled photos. I've a picture on my pinboard that was taken of us by a photographer friend of ours after Christmas lunch in 2003. Despite showing how depressingly young we looked compared to now, it's a nice test example as it was taken on a reasonable quality Olympus SLR on standard Kodachrome and processed in an automated film lab as opposed to any clever stuff. Despite being taken by a professional press photographer, it was taken on his everyday camera and treated as an ordinary fun picture.


(top: scanning app :: bottom: flat bed scanner)

The scanned image is certainly better, but again the image taken with the scanning app is perfectly usable though a little bleached out - better lighting may help here)

So what do I think? - for documents it's certainly more accurate out of the box and gives more consistent results than simply photographing them. It also does a reasonable job of photographing images, but not to the quality obtainable from a proper flatbed scanner. 

Undubtedly, dedicated equipment will give better results, but where this is not possible, using a photo scanning app may give better and more consistent results than simply using the camera app. As always your mileage may vary and more recent phones with higher resolution cameras may give better results. 


Saturday, 1 August 2020

Using a cheap fitness tracker ...

When I joined the herd and bought myself an iPhone, I discovered that I'd become reliant on using the Samsung health app on my old Galaxy to track my bike rides and cross trainer sessions, and what's more Apple's equivalent app was not nearly as sophisticated.

So I bought myself a $30 no name Chinese fitness tracker that let me track work outs, bush walks, bike rides and so on, as well as measuring heart rate and blood pressure. The tracker came with a nice little app that interfaced with Apple health, which was pretty cool.



So, for the last few months I've used it purely as a fitness tracker for workouts and bike rides. It's certainly not the jazziest, but it does the job, and the data it records is not too different from what I got out of my old phone, so we'll say it's accurate enough - a reasonable indication rather than clinical grade data.

And then, a couple of weeks ago my watch broke. Actually the story's more complicated:

About a year ago my old trusty Seiko died, like never to go again died. So I took to wearing a minimalist black plastic Swatch I'd bought for travel and rough stuff as my everyday watch. And as always happens with plastic watches, the strap died long before the watch did.

The simple solution was to get a replacement strap and the poky tool to get the pins out, but this of course all took time, and in the interim I used a $10 unbranded army style watch I'd bought off ebay a couple of months ago.

Well $10 watches are fine - some I've bought for everyday use have lasted two or three years. This one wasn't one of them. It stopped and the battery wasn't easily replaceable.

So this left me without a watch. Now I live in rural Victoria and the nearest town with any decent watch shops is in NSW and inaccessible due to the COVID-19 border closure. The local pharmacy has some overpriced shiny no name watches, and that's about it.

So I held my nose and bought a watch from Amazon. Stupidly, I failed to notice that they were shipping the watch from Canada, so instead of being without a watch for two or three days, it was potentially two or three weeks.

Now, I still like a watch to know the time, so I decided to start wearing my fitness tracker all the time.

And that was quite interesting - turned out that I was more active than I thought I was, and that my resting heart rate and blood pressure were lower than I thought. The latter being of interest as I've always had blood pressure at the high end of normal, and doctors periodically like to give me the healthy lifestyle speech as a consequence, and don't want to hear when I tell them it's probably genetic, and that when I regularly ran 10km it was still the high end of normal.

But I digress. What was it like in daily use?

Surprisingly good - battery life was good, running for three or four days without having to be charged - which was a blessing because the charger clip - basically a large USB crocodile clip with two prongs that need to be positioned just so, was incredibly fiddly to use.

Basically it told the time. Features like vibrating when you got a new mail message were less useful - the screen really was too small to read, and of course you have to be in bluetooth range of your phone - like you can  hear your phone ping.

I suppose if you had your phone on silent in your bag when you were in a meeting, or else somewhere noisy like a train station it would be a useful feature, but for me, less so. I didn't test how well it's claimed ability to start and stop your music play list or remotely operate your phone's camera worked.

The heart rate and blood pressure monitors use the standard flashing light technique, which meant that you had to have the band pretty tight to get a decent reading.

Other than that, it did the job - and did it well. My major gripes were (a) it was IP27 rated, and not fully waterproof - something to bear in mind nowadays when we're all scrubbing our hands so much, and (b) the device was designed to be worn on the inside of the wrist, rather than the outside.

However, these aren't major criticisms. I came away quietly impressed how well a $30 fitness tracker worked in daily use. In fact it did the job so well that it's not clear to me why you would use one of the more expensive named brand  models unless you were engaged on a serious training regime...