Saturday, March 25, 2017

Smart-phone Screencast to Airline Seat-back Screens?

If the ban on electronic devices larger than a cellphone in aircraft cabins continues, it may be worth airlines adding smart-phone interfaces to their seat-back screens. Many airlines already provide a USB socket which can be used to charge a phone, or access data on a flash drive. However the interface offered by the in-flight entertainment system for accessing the flash drive data tends to be slow and cumbersome. An alternative would be to screen-cast from the smart-phone to the seat-back display, via the USB cable (which would power the phone at the same time). The passenger would use their phone touch screen as the interface, and have access to the phone's data and applications, but see the results on the much bigger aircraft screen.

Thursday, March 23, 2017

Limit Electronic Devices on Airlines to 300 mm and 200 g?

The US and UK have banned electronic devices larger than a cellphone from aircraft cabins on flights from some countries. But what is “larger than a cellphone”?

Perhaps airline check-in counters need a gauge, like the one for carry-on luggage, or weight alone would be sufficient.

It is not clear what the threat from laptops and tablets is. There could be good reasons for authorities are not wanting to make this clear, but I don't know what they are.

The Guardian article mentions the risk of bombs and of lithium batteries. The bigger and heavier the device is, the bigger the bomb or battery it can hold. So it would make sense to limit the size of devices, but to do that an actual size needs to be specified, not just “larger than a cellphone”.

Large phones ("phablets"):
  • iPhone 7 Plus, with a 5.5 in screen is 158.2 mm x 77.9 mm x 7.3 mm and weighs 188 g.
  • Samsung Galaxy S7 Edge with a 5.5 in screen is 150.9 mm x 72.6 mm x 7.7 mm and weighs 177 g.
Are not that much smaller than a small tablet:
  • Samsung Galaxy Tab 7.0 with a 7.0 in display is 190.09 mm x 120.45 mm x 11.98 mm and weighs 380 g.
A reasonable limit for devices might be: height, width and depth added (the way baggage sizes are set), must be no more than 300 mm and the weight no more than 200 g. That would allow the phablets, but not tablets.

Just to check, my chunky old Lenovo A588T Android Flip Phone is 120.6 mm + 62.5mm + 16.8mm = 199.9 mm at 178.4 g,. That would pass, provided I don't open the flip, which makes the phone 115 mm longer. ;-)

Tuesday, March 21, 2017

Virtual Power Station Cheaper Than Snowy Hydro 2.0?

Tomorrow I am giving a guest lecture to Systems Engineering students (COMP3530) at the Australian National University on "Environmental Concerns". As usual I will run them through "How Green is My Computer?", an exercise in estimating the carbon emissions from a computer. However, to make it topical, I thought I would see if computers can provide a cheaper alternative to the PM's Snowy Hydro 2.0 Proposal.


Smoothing the peaks ...
Blakers and Fulton,
The Conversation,
CC BY SA 4
In February 2017, a team of ANU researcher, lead by Professor Andrew Blakers released a report proposing that "100% renewable electricity in Australia" was feasible. This would use pumped hydro energy storage: excess power from wind turbines and solar panels would be used to pump water into a high reservoir. At times when there was insufficient wind and sun, the water would run down to a lower reservoir and drive a turbine. Blakers and Fulton described how this could be implemented at the existing Snowy Hydro Scheme, in 2014.


Murray 1 Snowy Hydro Station,
by Martin Kraft/Wikimedia
Commons
, CC BY-SA
Malcolm Turnbull, Prime Minister of Australia then announced the "Snowy Mountains Scheme 2.0." (16 March 2017), to add 2000 megawatts of pumped hydro storage, at a cost of AU$2B ("Malcolm Turnbull plans to upgrade Snowy Hydro to version 2.0" (Video), ABC TV, 16 March 2917).

Dr Lachlan Blackhall
However, are there cheaper, quicker alternatives for low pollution energy, using computer and telecommunications? Companies including Canberra's Reposit Power (Founded by ANU graduate Dr Lachlan Blackhall) aggregate the power from home solar systems (and batteries) and sell this back to the grid.

However, there is another underused source of on-demand power available in people's homes: the air-conditioner. The air-conditioner is a large part of the energy supply problem but could be part of the solution. A home ducted home air-conditioner uses about 5KW of power. A medium room air-conditioner uses  2.3KW. What if we could pay householders to halve their air-conditioning energy use at times of high energy demand?

Smart meters and some air conditioners have a Demand Response Enabling Device (DRED) option. This allows the electricity supplier to remotely switch the air-conditioner to a lower power mode at times of peak demand. The supplier offers a cash payment up front and a lower electricity charge to householders who take up the option. However, this is not very popular.

If we could use smart-phones, and sharing economy techniques to provide more of an incentive to consumers, would this be cost effective?

A quick back-of the-envelope calculation:
Cost of Snowy #2 power: $2B for 2000 MW = $1,000 per KW.

Assuming a consumer would halve their air-conditioner consumption on demand, 5KW to 2.3KW = 2.7 KW saved. This would have to cost less than $2,700 to be cheaper than Snowy #2. Retrofitting DRED may well cost more than this. However, smart air-conditioners may require no extra hardware. The difficulty would be making it feasible for the consumer to enable the function, without requiring a technician to visit.

Raspberry Pi Computer,
photo by By Ayaita
(Own work) [CC BY 3.0],
via Wikimedia Commons
The computing power needed to provide the DRED function, could be provided by a computer equivalent to a $50 Raspberry PI. The computer needs no touch screen as it can instead use the consumer's smart-phone for the interface. The connection to the energy supplier can be via WiFi to the household hub. The system could be made fail-safe (and prevent the consumer cheating it) by having the unit programmed to operate on low power, until it receives an encrypted signal from the power company to switch to high power.

However, how many households would need to have DRED? The Snowy Hydro 2 scheme is proposed to produce 2000 MW. At 2.7 KW saved per household, that requires almost three quarters of a million households: 8% of the 9m in Australia. Also it is unlikely that these householders would be willing to give up half their air-conditioning for long periods. But such a system might be used during the annual peaks and when there is a problem with regular supply, rather than have fossil fuel stations on standby. So I propose the Snowy 2 Scheme include a 2000 MW virtual power station to be operational by the end of 2021 at a cost of $200M.

Saturday, March 18, 2017

Publishing a Kindle eBook with Amazon.com

My book Digital Teaching In Higher Education, is now available as a Kindle eBook from Amazon. But setting this up was not easy.
Having successfully published paperback, hardback, PDF and ePub eBook versions of my ICT Sustainability and Digital Teaching books, with Lulu.com I thought I was all done. Then I discovered a note from Lulu to say that Australian based books are not published for the Amazon Kindle. So went into Kindle Direct Publishing (KDP) and started the process.

Content Creation


The KDP process is similar to Lulu. You enter the details of the book (title, author ...). Then upload the content, design a cover, select a price and how you would like the book distributed.

One feature KDP has, which Lulu does not, is spell checking. I was surprised to find 28 spelling errors reported in my manuscript for one book. Some of these were not errors, but were proper names and acronyms, but seven were errors. It is a little embarrassing to misspell "unviersities" (universities) in a book about higher education. ;-) 

Creating the eBook with KDP required some changes to the word-processing document I had used for Lulu. I have the content of each section of the book stored in a separate file as HTML code (the same files are used, unchanged for the web site). These HTML code files are imported into a LibreOffice master document (.ODM). There is one master document with a table of contents which is exported as a PDF file for the hardback, paperback and PDF eBook editions. There is a second master document with the same chapters, but without  a table of contents, for the ePub edition (as ePub has a separate contents file). Lulu does not accept .ODM master documents, so I export it to a normal word-processing document (.ODT) and Lulu converts this to ePub, adding the table of contents.

Table of Contents


KDP does not accept LibreOffice files and when I tried one exported to .DOC, it did not generate a table of contents,page breaks for chapters, or insert images. Exporting to the newer .DOC format produced a better result with the page breaks included, and images, but still no table of contents. The ePub file generated by Lulu also formatted well using KDP, with page breaks and images, but also had no table of contents.

So I modified the .ODM master document to add a table of contents (ToC) and then saved as .DOCX). This ToC less detailed than the table of contents for the printed book. The printed ToC is nine pages long, which would be tedious to scroll through on screen. So the eBook version omits sub-chapters, reducing the ToC to one page. Also I omitted page numbers from the ToC, as Kindle eBooks don't have page numbers.

Next time I revise the ePub ebook, I will have to decide if I include this abbreviated ToC. The ePub ebooks do not normally have the ToC within the text of the document, it is supplied via a separate XML file (in this case that file is generated by Lulu). But I would prefer to have just one master document to generate both the ePub and Kindle eBooks. Also it might be worth using a similar abbreviated table of contents in the print edition (with page numbers) before the full ToC (it can be daunting for the casual reader to see nine pages of contents).

One quirk of the table of contents function in LibreOffice is that I found I had to "update" it each time I opened the ODM document. If I did not do this the links did not work.

Pricing and Distribution Options


KDP Select is an option where your book is made available for rent to subscribers, however this only available if you do not publish the ebook elsewhere. As my ebook is already available trough Lulu, I did not select this option.

KDP Pricing Support offers to help suggest pricing by analyzing the manuscript and comparing it to sales of similar books. However, I could not get this to respond (it is still in Beta.Kindle books are proved with either s 30% or 70% royalty. The details of are complex, but the 70% royalty seems to be for lower priced books (up to US$9.99), mine is US$6.90, so I chose that option.

Print Book Option

After I published my Kindle edition, KDP offered the option of a print edition. As I already have this through Lulu, I did not try the option. 

Author Page

Like Lulu, Amazon.com offer an author page, where I uploaded a photo and bio. Lulu automatically links the books you published, whereas Amazon alsos you to search for others, where you are the author, but not the publisher. Amazon will also display recent posts from your blogs on the page. One troubling aspect is that some third parties are selling some of my books on Amazon for many times the retail price. These are books I have not yet listed on Amazon directly. Hopefully when I have them up myself, this will stop.

Publishing is Slow and Complex

Both Lulu and KDP offer good on-line services for publishing a book. However, with either (or both), this is still a complex, time-consuming and error prone process. Both offer to have some of this work done for you (at a fee). By doing the process myself I appreciate that publishers do earn then fees. ;-)

Sunday, March 12, 2017

Change Canberra Light Rail Project to Bus "Metro"

The city of Brisbane has decided to abandon plans for a track based metro and instead upgrade its existing bus transit system. I suggest Canberra should similarly change the currently Light Rail Project to use buses. The roadworks already underway for light rail could continue, but without track or overhead power installed. The money saved could be spent on hybrid articulated buses to run on the new dedicated roadway. If light rail becomes viable in the future it can be installed. However, new technology will likely make light rail obsolete in the next ten years.

Thursday, March 09, 2017

Experiments to Optimize Web Advertisements

Google offered to run experiments to see how allowing more Adsense advertisement categories on my web pages could increase revenue. So I am running an experiment unblocking "Events, Shows & Cultural Attractions", to see if makes a difference. But I am not going to unblock sensitive categories, such as "Weight Loss Supplements".

Wednesday, March 08, 2017

Cyberwar: The Most Hideous Form of Warfare of the 21st Century

Greetings from the great hall of the Australian National University in Canberra, where Dr Mohammad Sabah Al-Salem Al-Sabah, former Deputy Prime Minister and Minister of Foreign Affairs of the State of Kuwait is speaking on "Regional Security in a Turbulent World: A GCC Perspective". Dr Al-Sabah referred to cyber-warfare as "The most hideous form of warfare of the 21st Century". Dr Al-Sabah pointed out how terrorist groups were now very well versed in the use of the Internet.

However, in my view, there is nothing in the nature of cyber-warfare which makes it worse than the use of conventional weapons and certainly not in the same category as chemical weapons. In teaching ethics to computer students at the ANU I used a Hypothetical on Cyberwar as an example. As with other weapons, the users of cyber-weapons should be judged by the effect intended.

Monday, March 06, 2017

Print and eBook Compatable PDF Creation using LibreOffice

After discovering that PDF eBooks are popular, I set about creating PDF version of my two currently published books, with minimum of effort. The book chapters are HTML files which are imported into a LibreOffice master document. There are actually two master documents: one for the print edition with a table of contents and one for the ePub edition without. The master document for the print edition is exported as PDF and uploaded to LuLu for printing. The master document for the ePub edition is exported as an ODF Text document (ODT) and uploaded to Lulu for conversion to ePub.

The obvious thing to do was to export the Master document used for ePub as PDF. The conversion worked and there was an index included in the PDF file. However, the reader of a PDF file normally expects there to be a table of contents in the document, emulating a printed work. So I then tried exporting the master document used for the print PDF.

The print PDF document looked fine on screen, but the table of contents was not hypertext linked (unnecessary for the printed edition). The reader may have not realized PDF provides a separate table of contents. So I created a third master document with hypertext linked table of contents. But this would add an addition step and I realized I could simply include hypertext links in the table of contents of the master document intended for printed books. The hypertext links appear dark blue for on screen, so they can be seen. I decided not to include the usual underlining of hypertext links, so when rendered for the printed book, the reader can't see what was a link.
I have now implemented this for:


 
 

There is still a problem with some pages being blank to allow correct pagination for print, which looks odd in an eBook. There are ways to have a PDF reader ignore blank pages, but I have yet to work out how.

 
 
 

Sunday, March 05, 2017

Most Successful Self-publishing Sales Channels

After self-publishing my new book "Digital Teaching In Higher Education" I decided to put out a second edition my my preview work "ICT Sustainability". I used Lulu.com for both books. With Lulu it is possible to produce electronic and paper editions of the one work. However, keeping track of changes between multiple formats is error prone and time consuming. Which editions are worthwhile in terms of sales? The answer is: paperback, hardback and PDF eBook. It is not clear if ePub is worth the effort.

For Digital Teaching I have paperback, hardback and Epub (electronic) editions. These all contain the same text. The hardback and paperback editions use the same PDF file for the interior, but have different cover files (as a hardback cover is larger than a paperback). The Epub uses a different interior file which does not have the table of contents (that is generated by the Epub system). I maintain one set of common files with the chapter contents in it, common to all versions, but then have to manually update each version on Lulu with the files. What I would like to do is upload the interior content and then have Lulu generate the paper and electronic editions, but it does not do that.

Previous I produced two electronic editions: Epub and PDF. The PDF edition is only sold by Lulu directly, not through electronic bookseller, who prefer the Epub. I can't just use the paper edition's PDF file for the eBook PDF, so this would require an additional manual step. But is it worth it in terms of sales? I had thought PDF was redundant: who would want PDF when Epub is so much better. But many people think of PDF and are not familiar with Epub.

LuLu provides some summary reports of sales, and also the ability to download details as a spreadsheet for analysis. For one edition, the analysis showed twice as many sales of PDF as Epub ebooks:

Format
Percent
Paperback
51%
Hardcover
25%
PDF
17%
ePub
8%

What was also surprising was that hardcover sales were half that of paperbacks. I was expecting hardcovers to sell one tenth the number of paperbacks, given the different in price. Paperbacks sold one at a time, hardcovers were multiple sales, which makes me suspect these were going to library suppliers.

So having PDF and Hardback editions is worthwhile. What is less worthwhile is the Epub edition. The print and PDF editions require similar skills to produce. The Epub requires a different set of skills and more than doubles the effort required for relatively little extra return. It is only because I have already invested the effort of understanding Epub I am continuing with it.

Also I looked at where sales come from:


Region
Percent
AU
33%
Other
19%
US
16%
UK
11%
NL
8%
CH
5%
FR
4%
EU
1%
IE
1%

Unfortunately Lulu has no record of where19% of the sales were. But for the remainder, Australia is not a surprise, nor is the US, UK, or China. What is surprising are the sales in the Netherlands and France.

One thing to keep in mind that each step in the distribution chain takes a cut of the revenue. Sales directly via Lulu earn much more than those which are then through another distributor. So while almost as many sales were through Ingram as Lulu, the revenue through Lulu was twice as much:


Channel
Sales Earnings
Lulu
44%
54%
Ingram (Print)
42%
25%
Amazon (Print)
3%
12%
iBookstore
11%
8%
Nook
1%
1%

Only some formats are accepted for third party distribution. I have found the "US Trade"  size (6 by 9 inches) a good option for textbooks. Selling outside Lulu also requires another step in the process: you have to purchase a proof copy and check it. If there are corrections to be made another proof is needed, delaying distribution by another week. What I do is delay that distribution step for a few weeks, only selling the book through Lulu, to make sure I have the book right.