Skip to content

Category: Uncategorised

Pi-Hole

Pi Hole

I had my Raspberry Pi lying around doing nothing. I’ve had a version 1 since they were released and I’ve had a version 2 for a while as well (in fact, I’ve been through two version 2’s). I’ve had this one in the draw for a while now doing nothing – it’s not really seen the light of day since I purchased my Synology DS116 as I had been using it as a home server and accessing my documents via SFTP outside the house.

However, after I stumbled across the Pi Hole project, I decided that it was time to dig it out again. Whilst the NAS is excellent (DS Photo allows me to backup mine and my girlfriends photos from our phones, DS Video allows me to stream videos to my Chromecast, DS Audio streams all my music to my iPad and it can act as a VPN when I’m out and about, though this is less of a benefit now I have a years subscription to Private Internet Access), it doesn’t allow me to block adverts which is what Pi Hole does. Quite successfully.

Setup was easy and I had the Pi Hole up and running within 30 minutes or less. Setup was as simple as downloading a fresh copy of Raspbian, installing that, remembering to enable SSH as the new versions of Raspbian have SSH turned off by default

For headless setup, SSH can be enabled by placing a file named ‘ssh’, without any extension, onto the boot partition of the SD card.

Once this was done, I updated the Pi and then downloaded the Pi Hole software using the command

curl -sSL https://install.pi-hole.net | bash

This installed everything I needed and set it up. All that I then had to do was to set my router to give the Pi a static IP address and change the DNS settings on my devices to point to the Pi.

Once I’d done that, I’ve been able to enjoy ad blocking across all my devices.

The Pi presents everything in a web interface, so you can keep an eye on the drive at any point of time.

Untitled

This shows the DNS requests that are being made and you can even drill down into what sites are trying to be accessed.

Untitled

This shows the top domain name requested by devices and the top advertisers being blocked. You can also see what devices are requesting the most domains by IP address.

The downside I have is that the DNS records on my router cannot be changed, so I’ve manually had to change the details on each web connection. This isn’t the end of the world and actually can prove handy where I might want to allow an advertiser to track me (such as if I’m using a cashback site). However, it does mean that devices like my work laptop, ones that I cannot change the DNS settings on (in this case, due to admin lockdown), will not benefit from the blocking whilst on my network.

However, overall, the Pi Hole was easy to setup and manages to block a lot of adverts – browsing on the iPad, it’s surprising how much cleaner various websites are. There have been a single instances where the website has appeared to be broken due to the blockage and this was solved by white-listing the site. Whilst this should stop the need for running an adblocker in my browser, I’ve still got uBlock installed.

It’s certainly worth a try.

New Domain Name Provider

This month, one of my domains was up for renewal. Currently I own four domains and they’ve all been with 123-Reg for many years. However, the service I’ve had from them and the communication in response to a query was pretty poor – I realised that I’d had to contact them the year before regarding a renewal as well! So all in all, I decided to take my business elsewhere.

I’ve moved all the domains to Mythic Beasts, a UK based company that I’m hoping will provide a better service! I’m also considering moving my email and hosting there for a one stop shop, considering that Fastmail have now removed file hosting (though, I still have it as a grandfathered account) though in fairness, staying with Fastmail might be a good option, as it offers some better features that I might get with another host, such as an app and contact sync.

Gigabyte Brix BXBT 2807

I was trying to install Windows 10 on the recently purchased Gigabyte BRIX-BXBT-2807 that I got the other day and I was struggling.

It would boot the Windows 10 USB up without any issues and would then allow me to go off and install it. Windows copies the files to the disk and then reboots as any normal install.

On boot, it would then however sit there at the boot screen and wouldn’t even boot into Windows to continue the install. Uh-oh.

After playing around, it turned out that the way to solve it seemed to be to boot into the BIOS and switch the Operating System mode to Windows 8.x (rather than Windows 7).

Simple.

But not really well documented anywhere that I could see!

Trigo Garmin Mount

Last week I found out about a Garmin mount from Taiwan called the Trigo. This mounts to the stem of the Brompton where the handlebars meet.

This overcomes one of my issues with my current setup for the Brompton in that my GPS starts rotating around the handlebars occasionally and it has also caused some surface marks to the handlebar. Putting the Garmin onto the handlebars also means that I have to rotate the Garmin mount so I can attach the Garmin anyhow as the top of the stem gets in the way.

I was slightly surprised to find the Trigo had arrived today – I thought I was going to have to wait a bit longer for this to arrive.

The package contained the following (well, not the Garmin!)

Trigo in it's packaging

Trigo Kit

The Garmin mount had to be purchased separately.

Fitting

Fitting the Trigo was really simple. The handlebar and stem screw needs to be unscrewed and then insert the mount into the gap and then screw it up again.

Mount on the bike

This gives the mount mount right in the middle of the bike.

Trigo Mounted

Putting the Trigo on means just screwing the mount point onto the mounting block. Simple and easy with the attached.

However, this is where I came across the issue that it would effect the fold.

Effecting the fold

The mount catches on the hub. It turns out that it just needed the screw on the Trigo mount to be lessened a bit, letting me move the mount.

Garmin mounted

Folded with Garmin attached

I’ll give this a go for a while and see how I get on. The mount seems to be perhaps a bit loose to allow me to move the mount forwards and backwards.

The mount itself is ideally mounted in the middle of the handle bars and is ideally located for viewing. I’ll give it a go for the next few weeks cycling to work and see how I get on. But it does look like a good mounting point. Time will tell.

Zim Journalling

Zim and Journalling

I’ve been using Day One for probably close to five years. In that time, I’ve built up quite a database of journal entries.

However, with the release of version 2 of Day One, they have moved to running their own sync service – which I can understand and I can see the benefits of, as it allows them to add in additional features to the software that using iCloud or Dropbox sync wouldn’t let them.

But having had a couple of outages in the sync service over the last few weeks, it’s something thats made me think again before I continue using it.

By chance, I discovered Zim and decided that it could potentially suit replacing Day One.

Zim

Zim is a desktop wiki software. You enter data into Zim under different pages and you can link them all together using links, like Wikipedia.

It allows files to be attached to each wiki page and allows easy creation of a personal wiki. Pages can be prevented from editing if you just want to browse and the whole wiki can be exported as a website (the Zim website itself has been written in Zim). Think of it as Evernote but you control all the data as without a sync solution, all data remains on your computer.

Zim has a journal add in installed by default. This add in allows you to add journal pages to a wiki (where you might want to record experiments or whatever on a particular day). This happens to make it a perfect journaling software, as it allows me as the user to create a workbook that is purely journal entries if I so wanted – arranging the pages into a date and time format. It supports a number of journalling options – you can have a page for each day (which is how I have it set up) or you could have it setup as a single page for each week, adding each day on to the bottom of the file. It’s a bit more flexible than Day One in that regard.

Replacing Day One

In its basic form, Zim is perfectly capable of replacing the desktop version of Day One out of the box (and it’ll run on Windows, Mac and Linux 1). Entries are stored as plain text files and images can be either be attached (and stored in a relative path) or can be stored anywhere else on your computer and linked to.

Zim allows for images to be attached and displayed within the entry list and unlike Day One, you can have as many pictures attached as you want 2. In fact, you aren’t just limited to pictures as you can also attach anything else that you might want with a specific date. The only difference I’ve found to Day One is that you have to take care to ensure that the file is attached, rather than linked. It’s a small difference but if you link a file, it links to the original file on your hard drive, whereas attaching it will allow it to be stored in the Zim folders and linked as a relative link. Attaching a file is my preferred option, as it allows me to work on different operating systems and not worry about moving the original file.

Rather than use the Markdown language I’m used to using, Zim uses DokuWiki style markup. This is a bit different from using Markdown, specifically with regarding to the headings. However, using the desktop app, I’ve found that I’ve not had to write in markup language at all, as the default shortcuts let me edit the files without knowing the language and this seems to work well. This means I only have to know the syntax when working on a mobile device (see later) and then I have to manually add the syntax, as I haven’t found a text editor that works with DokuWiki markup.

I’ve managed to get Zim working on my Mac by using PyGTK and running it direct from the source code. That seems to work well enough for it to run, though there can be some issues when changing the window size (I’ve found it easiest to increase the window size from the top right corner, rather than the lower right which seems the more natural way of changing a windows size).

Using Zim keeps the data within text files and means that I retain control of the information. It also means that I can sync my journal using whatever system I want to, be that Dropbox or BitTorrent Sync or even accessing via SSH file system – I retain full control and I’m not relying on the Day One sync service which hasn’t seemed to be the best since I started using it.

Mobile Editing

New Entries

As all the data is stored in text files, it means that the files are small and fast to sync, but importantly, can be edited in the iPad or iPhone using a text editor. It means that I can use both my mobile devices to write diary entries as I could on the computer. In fact, this is fairly easy to set up – just point Editorial at the journal folder in Dropbox and you can type away 3.

However, when Zim creates each text file for a page in the wiki, it adds metadata to the start of the file. For a journal entry, this means the date and the file creation time. Editorial by default does not add this information and it would be a pain for me to write this in at the start of every journal entry. Editorial comes to the rescue however, as it allows the creation of template text files.

Text file template creation in Editorial, allowing variables to be used

By allowing date (and time) variables to be inserted into the text file, I can create the Zim metadata header automatically when creating the file, even having the file named correctly for Zim when I’m back at a computer.

An issue with this method is when the date changes between months. Zim in my setup, creates a folder for the Year > Month and then places the text files within that (one per day).

Folder Layout of Zim Journal

My current template does not take into account the month and therefore will create a text file in the folder with todays date. It means that at the start of a new month, I manually have to create a folder for the month and then create my text file. I’ll look into changing this in the future but at the minute, it works ok.

Image Attachments

Attaching images is properly the biggest thing I need to do to a journal article. I’d attached quite a few images to my Day One journals, so I needed a way of doing the same to Zim. I managed this by using a mixture of Editorial and Workflow.

Editorial has a powerful automation tool built in, but it’s not the most straight forward tool to use. I’d class myself as techie, but I’m not a Python coder.

That’s where Workflow comes in – it lets me create a workflow visually. This lets me upload an image to Dropbox and link it back into the Zim post in Editorial.

Attaches an image to the current Editorial post

You can find the Workflow here.

The downside to this is that I have to be online to upload the image and that the workflow requires that I’m adding the file on the day of month that I want the image to appear on. Zim will automatically create a folder for attachments in the same name of the file you’re editing (so if you’re editing the Home page and add an attachment, it’ll add a folder called Home and then link the file as a relative link to that). This does mean that if I’m offline, I’ll have to wait to add the image and I’m likely to find it easier to then use the desktop version of the app to add images.

Conclusion

So not using Day One means that I use some of the functions that they’ve recently built in, such as IFTTT integration, but these can be overcome to a degree by using IFTTT Dropbox integration, allowing me to create text files within Dropbox using IFTTT and I can use Hazel to automate the process a bit more if needed.

However, in general, I’m currently happy with the journalling process and not using Day One. It’s not as pretty, but I’ve got a solid syncing process, rather than the occasional flakey option of the Day One sync. It’s not as pretty and not as well put together, but I find it still lets me journal and that’s the main thing, regardless of what platform I’m on. The Windows editing and the ability to back up my own posts is something that I’m keen to keep.


  1. Compared to a the Day One app which only works on the Mac – considering I use all three operating systems, Zim lets me enter data regardless of which operating system I’m using at the time.
    [return]
  2. Day One Version 1 (or Classic as it is now known) only allows for a single image per entry, Day One 2 allows for up to 10 pictures per entry.
    [return]
  3. Whilst you can only use a single folder for Editorial, my text editing needs for everything else on the iPad has switched to Ulysses. Ulysses is about to add Dropbox file syncing support for the iPad version, so I’ll switch it to that when it’s released.
    [return]

Goodbye Dropbox!

I’ve done something this week that I don’t think I’ve done for almost five years – I’ve removed Dropbox from my Mac.

*gasp*

I know, how could I?

Well, I think it’s slowly been a long time coming. Since I bought my Synology1, I’ve been using that more and more for various items. Up until recently however, it hadn’t come close enough to replacing Dropbox fully. However, I think the time now has come. Here’s some of the reasons why I can move away.

Onedrive

I guess the big step change has been the addition of Onedrive. My new job has an Office 365 subscription and this includes Onedrive for Business. It’s installed by default on my work laptop and we are encouraged to save all of our personal (work) files to that so that if we ever lose (or break) the laptop, we just get a new one, connect to Onedrive and carry on where we left off. Slowly, our server will be replaced with shared folders in Onedrive as well I believe.

This has played a large part in getting rid of Dropbox. I used to use Dropbox to sync my Zotero library between my MacBook and my work laptop – now that I can download and install the Onedrive app for Mac and access my business account (following these instructions which involves some Terminal jiggery pokery), I’m able to sync my these files between the two devices with no issues whatsoever.

I have the client installed on my work laptop, my home Windows laptop and my MacBook, so everything is within reach, including all my text notes. For a while, I had issues accessing the Onedrive folder on iOS for some reason – I managed to connect to it whilst outside of work, so I can upload and download documents as I need. It’s not quite the same as the auto sync feature built into many of the notes applications on iOS but I’ll to live without that unfortunately.

Day One

I’ve been a Day One user for many years now (based on my journal, five years!) I now journal everyday – sometimes multiple times a day, but a max of Day One and notebook (the notebook normally whilst at work). Up until recently, everything I wrote in Day One was synced to Dropbox – I could use iCloud and I could have used Day One Sync when they released it, but only Dropbox filled me with confidence. With Dropbox, I could also backup my own files using whatever backup tool I wanted to use as the files were stored in a folder in Dropbox on my computer (this is in addition to the Dropbox “backup” ability to restore deleted files from within 30 days).

However, Day One version 2 was released at the start of February. This brought in a range of features such as the ability to save more than one picture, the ability to have multiple journals and other features. However, the trade off was that you would have to move to the Day One sync service over that of Dropbox or iCloud. However, it’s a “price” I’m willing to pay for the new features (as both the iPhone and Mac apps backup my journals daily, which will be included in my Time Machine backups on the Mac).

So that removed that app from needing Dropbox.

1Password

Since I’ve been using 1Password, it’s been able to sync with Dropbox, iCloud or via wi-fi directly with a Mac. I’d always used Dropbox, as I already had it installed and I was syncing my passwords to my work computer (1Password for Windows). However, I can’t install 1Password at work now, so I don’t need 1Password files to sync with anything other than Apple devices and therefore I’ve moved my vault into iCloud instead.

I’ve an Android phone and I use Windows at work, so how do I access my passwords? At the minute, I’m using Lastpass on the PC as this is installed via a Chrome plugin and doesn’t require me to run anything on the computer that I can’t install. I’m not the biggest fan of Lastpass on Chrome though – it’s not a nicely tied in as it is on Firefox. This isn’t to much of an issue as I’m slowly leaving Chrome behind – it’s bookmark and syncing is great but there’s been a number of issues I’ve had on different websites with it. It’s also rare that I need to access passwords on my phone and if I do, I usually have my iPad or work iPhone nearby that has 1Password installed and working on.

If iCloud didn’t work out, there is an option for wi-fi sync with a computer and I can also use another cloud service to sync between desktop computers (and perhaps Android) using folder sync. But as I can’t install the 1Password app on my work machine, I’ll stick with Lastpass (or, I could go down the route of Keepass Portable).

You Need A Budget

This ones a tricky one – all my budgeting and money management is done in YNAB. Whilst they’ve announced version 5 of the software and this is a subscription based model with a web service for sync, I’m not keen on the idea as yet. It seems that the version 5 apps for iOS and Android don’t support offline access and version 5 doesn’t yet currently have all the same features as the existing version 4. So I’m sticking with version 4 – which uses Dropbox as it’s syncing backend.

I’ve tried some alternatives – I’ve gone back and tested budgeting in Moneywiz! (but it’s not as intuitive, straight forward or simple as YNAB’s budgeting) and a number of other alternatives, none of which seemed to meet my criteria, one of which is having different budgets for myself and a shared budget between me and my girlfriend. I was stuck with using Dropbox sync.

On the mobile apps, Dropbox is accessed via the API so it doesn’t need the Dropbox app installed. However, on the computer, this is required – you need to download the files into your Dropbox folder to access. I looked stuck at keeping Dropbox, purely for YNAB. I did consider setting up Cloud Sync on my Synology and having it download all the YNAB files and then I could access the files over a network share. In the end, I removed Dropbox and YNAB from my Mac completely – I’ve a work around 2.

YNAB Report - Only on Windows and OSX

The desktop application does more than the standard Android and iPhone apps – for a start it lets you create budgets and lets you move money around and assign it to each item you want to budget for. It also provides reports and balance reconciliation. This doesn’t happen in the iPhone or Android app. However, the iPad app does allow you to move money into each budget. So in general day to day use, I can use the iPad to do the management that I need to,

Here’s where I cheat a bit – I’m keeping Dropbox installed on my Windows (home) laptop. I have a Lenovo Thinkpad at home that runs Windows 10. My girlfriend uses it for working from home and I’ve got it as an extra laptop if needed. And in this case, it’ll still have Dropbox running on it and this will let me use it for performing tasks on YNAB that I can no longer do on the Mac, such as if I need the reports or to reconcile the accounts. I’ll see how I get on in this fashion – it may end up that I start using Expandrive to access the files directly on Dropbox, or I may use the Cloud Sync function on my Synology if I feel I need to, but I think this will be perfectly fine.

BitTorrent Sync

BitTorrent Sync was released a while ago and I’ve been using it on and off for periods during that time. However, late last year, I ended up buying a Pro licence. With the release of BTSync 2.3, the purchase of the Pro licence really pays off as I now have selective sync on Mac, Windows and Linux devices but I also have the ability to create an encrypted offsite folder. I’ve been using my Synology NAS as an always on node in my sync network. BTSync on the NAS has one issue – it runs as the admin user, but this hasn’t stopped me editing and changing files as a user – I can upload via SFTP without any issues.

Sync has been working fine in syncing my files from my MacBook to my NAS. It falls down where firewalls block access (i.e. work). I need to see if this is because access to the BitTorrent relay servers is being blocked or the traffic protocol – I imagine as it’s Bittorrent traffic, it’s not allowed. However, I can try setting the MacBook up with a specific server (as the NAS will have one thanks to Synology’s DDNS service) and I’ll have to open a port on the router. Outside of that, it works perfectly fine.

I’ve just set it up to backup the photos from my devices via the apps. That took some getting used to. It’s a special implementation of the sync protocol – BitTorrent try and explain it here. I’ve found so far that it’ll send the images to the node and then if it’s deleted on either of the devices, it wont be removed from the other node. So currently, I have my iPad, Work iPhone and personal Android phone to send my images to the NAS and MacBook Pro. Images I delete on the phone will only be deleted on the phone and not on the other devices. I can also set up Hazel to move and rename images from the folders into my pictures folder on my MacBook, where as, on the NAS I’ll just keep it as an image backup. This seems to be working nicely, and deleting something doesn’t delete from all the nodes. On Android, I can use this to backup any folder, not just the camera roll, so I could backup camera exports or whatever I wanted to keep a copy of – I’ve an app that automatically backs up APK files of the apps I install and everytime they update. The latest update also allows folders on the SD card to be included (useful in this case, where my camera app saves my photos to the microSD of the phone).

Free File Sync

It supports SFTP uploads so I’m able to sync my work laptop (I keep all my files on an encrypted USB on my work laptop so I’m able to access them from work regardless). SFTP access is allowed at work, so my data is able to sync with the NAS. Why have Dropbox when I can use my own server with 3TB of space for an initial outlay of two years worth of Dropbox data? The Synology DDNS service is pretty robust and Virgin have rarely had hiccups with my connection and I’ve got a decent upload speed – it’s not as fast as Dropbox but once I’ve done the initial sync, it’s not an issue.

Conclusion

A potentially misleading title, as I’m not getting rid of Dropbox completely, I’m just really relegating it to a cloud based disk for my YNAB data.

Synology makes it easy to run a NAS. On occasion, I’ve felt that a standard PC would perhaps be a better option for a server – perhaps in the future I will replace it with a Mac Mini or similar that’ll act as a file server as I’ll be able to do what I want with it then (such as install all my backup software directly on that, rather than on the laptop). However, that’s for later consideration (the Mac Mini draws 6W at idle!), as the Synology is currently able to meet all my needs. Posts such as Coding Horror’s ones about low powered x86 computers convince me that this might be the future way to go (though a Raspberry Pi 2 is relatively OK for home use). I’d be able to run almost anything I want (Windows included if I was that way inclined) – however, it would be a bit more intensive to maintain. Sometimes running a server from the command line does get tiring and I wish I could use a GUI.

But in the meantime, I think I can pretty much wave goodbye to Dropbox for the time being.


  1. A Synology DS115j, which was the lowest in the range – next time, I might get a better model, though this has been perfectly adequate so far.
    [return]
  2. If you can call it that…
    [return]

Plain Text Blogging

For the past two years, I’ve been writing my blog in Markdown and plain text – in fact, almost everything I write at home is in plain text format – using nValt, Byword and Marked2 I’m able to write everything I need (blog posts, emails and magazine articles) 1 2. With these three apps, I’ve been happily able to work on my MacBook, Windows laptop, work laptop, iPad, Amazon Fire tablet, Android phone and iPhone 3 – plain text documents synced to the cloud via Dropbox and accessible anywhere from pretty much any device I happen to be using.

This has worked nicely – I’ve no real complaints about the system (other than nValt is Mac only, so I have to resort to workarounds to search my notes in whatever text editor I’m using). The only thing that I would like is to edit my blog in Markdown. Which, I kind of already do as I use MarsEdit to post my blog posts to WordPress and that allows me to write in Markdown and then post to WordPress where it is converted into HTML. Whilst this works, I also want to be able to keep the data I write myself – putting it into WordPress is actually a bit of a pain. Moving hosts means I have to download databases and files and back them up regularly – considering I backup my local files easily, trying to remember to backup the website or even rely on the fairly problematic WordPress automatic backup plugins is a bit of a pain.

On this basis, I’d been looking at moving my site to a static site, using only a flat file system and no database. I’d have to be able to write in Markdown and ideally, I’d be able to self host the images and the HTML files. It would mean that I could host my site easily on whatever device or host I wanted. For example, I’d be able to use my Fastmail account to host the files as they have static site hosting (for low ish traffic websites) included as part of the annual fee. Otherwise, Github allow you to host a website with them and many other locations have static hosting. I could even use my Synology NAS to host them as well if I wanted – whilst the Synology will also take PHP and MySQL compatible files, I went for one of the lower models and therefore, I’d probably be taxing the CPU a lot if it wasn’t just a static site and I don’t want it interfering with my own use of it whilst at home!

So the quest for a static blogging site started – MarsEdit was something that I liked the idea of, writing in a program but this then pushes posts to websites. It wouldn’t have been to much of a problem, but MarsEdit doesn’t keep a copy of all the posts, only the most recent. Whilst I’m not the most prolific writer, I still wanted to be able to have a local copy that I could post elsewhere or move if I wanted to at a later date.

Jekyll / Octopress

So I looked at moving to Jekyll and Octopress and these seemed to be ideal – initially on first glance. They both generated static sites and would have allowed me to continue writing in Markdown. But they require installing Ruby (whilst it is installed on my Mac, the version isn’t new enough to run Octopress) and this was far more technical than I was wanting to do. I followed the guidance to try and install it, but it proved problematic and I decided that it was to much hassle. I wanted to be able to concentrate on writing and publishing, rather than messing around with code. I am a geek, but I don’t code day to day and therefore this seemed to be perhaps a bit of a step to far for me. To much time maintaining and trying to get the blog online and in a decent format before I could actually write anything!

Ghost

Ghost appealed to my nature of publish and write rather than concentrate on looks so I signed up to the website to give it a go. Whilst I was able to use it fairly well, it seemed that all the interaction was within the web editor and most of my writing is as far away from the web editor as I can get! This obviously wouldn’t stop me writing my articles within my text editor of choice and then copying and pasting into Ghost. However, I couldn’t quite get on with it – not entirely sure why.

And then, as it’s a premium software, there’s a monthly payment. In fairness, this goes towards the development and hosting costs, so it’s not like it’s unwarranted, but for the amount of blogging I actually do, I thought that perhaps it was a little steep. Of course, I could host it myself. But it’s node.js based again and it seems a complete PITA to actually setup, not knowing anything about node. It’s cheaper to get a Digital Ocean “droplet” and use their one click install Ghost to get up and running, rather than use the Ghost lowest package (but then, there are other perks that the Ghost platform does that Digital Ocean don’t do, like auto updates etc). Either way, I don’t think I can quite justify paying the Ghost fees and I don’t want to have to run my own server and maintain the system – whilst I could do it, it’s not something that I want to spend ages updating and trying to sort out. And there is the issue of backing it all up again!

Voodoopad

Voodoopad isn’t really a static blog publishing software, it just happens that as a by product of its design, it can act as a static blog publishing engine. It’s primary role is a wiki type software for the Mac that allows you create your own wiki space – linking different notes together, inserting pictures etc. It can export to a large number of formats, a webpage being one of them. This standard web export is good, but will export the document as a wiki type site – that is, no order to the chaos you’ve introduced into the document by adding articles! However, it does have a static site publishing feature – in fairness, I think this should be renamed into static blog generating as this is what it does – it creates an index page with your last x (selectable by the user) number of notes presented as articles and then generates an archive page with all the previous notes linked to as individual pages and the archive is sorted via date created. It also creates an RSS XML file for RSS readers to connect to.

In this fashion, it acts as a static blog generator with a number of limitations:

  • The date of the post is the creation date of the note within the software
  • The date of the note is not editable

With those two issues in mind, it makes it difficult to insert previous blog posts (so it would be ideal for new posts going forwards, but a bit of an issue otherwise).

To get around the issue above, I managed to set my computer time and date to the specific date that I wanted and then created the note for the post. This worked but was a bit long winded. The next issue I ran into – not all of my posts were published when I wanted to. You have to select which posts to publish and I’d made sure that all the ones I wanted were selected. However, it would only export five or so articles which made it problematic for publishing!

In the end, I tried to resolve the issues but I just couldn’t find a way of getting it sorted and therefore I decided to move on (with some regret, as it looked exactly what I was after).

Tumblr

Tumblr is synonymous with the sharing of GIF’s and short articles, however, I’ve used it in the past for writing and publishing. My personal blog and then my cycling blog were both published on Tumblr before I moved off to WordPress. There were a number of issues I ran into using Tumblr as my blog software – namely the uploading and displaying of images for a long text post. I had been using MarsEdit to write my post and then upload all my photos/pictures to Flickr as there is a good integration between MarsEdit and Flickr. However, I was finding that occasionally Flickr would throw a hissy fit and not display the images how I wanted or would error out completely.

MarsEdit seemed to be ideal – I thought it was keeping a copy of all my posts but it was only keeping the most recent ones. Whilst all my posts were started in nValt, it would have been nice to have a copy in MarsEdit that I could then move to a new blog if I wanted to. In regards to backup, I ended up linking Tumblr to Evernote with IFTTT and using that to automatically send my Tumblr posts to Evernote!

Overall, Tumblr was probably the best blogging software that I’d used. Simple and easy with little to no hassle. It was even easy adding my own DNS records to have it point to my own Tumblr blog, though I could only seem to have a single domain point to a single blog, regardless of how many Tumblr blogs I could edit/write. In fairness, as a free service, this seems reasonable enough to limit it.

Hugo

Hugo could have appeared further up the list as I tried this after Jekyll and Octopress to be honest – but I gave up. After some more playing around with it though, I think I finally cracked it and have been able to get it working quite nicely.

It initially appealed to me – I just have to download the pre-compiled binary from Github and away you go. Well, initially. I was able to get a site up and running in seconds (as promised) but then I wasn’t able to get it uploaded to a website. I wasn’t even (at the time), able to generate a website – I could only use the Hugo server to display the pages! Whilst they state that this is perfectly fine for various sites, I wanted to export my items as HTML files and then I can use whatever web service I want.

Well, second time round and some editing of files and reading of data, I’ve managed to sort Hugo out to the point where I can produce a site and produce some output (as that’s what you’re reading now!)

It took some figuring out, but I think I managed to get there in the end. Finding out where and how to add images was an issue – even now, I’m having to use absolute paths to reference them within documents to get them showing up nicely in both the Hugo local server and the web server (it might be ok if I used the Hugo webserver to display my blog, but I’m using a hosting account, not a VPS where I have access to the processes on the server). And even with some figuring out, I’ve still got some issues to iron out – for example, my current theme is opening all links within new pages when viewed on my webserver, but this isn’t the case when viewed locally on the Hugo server. I’m not currently sure if this is due to my webserver setup or the theme itself.

If I want to change the theme in the future, I just build the blog using the new theme and all the content is create afresh using the new layout! Pretty simple!

As it gives each page its own folder (by default), I was able to replicate the WordPress content layout for the backend so I should be able to link to all the old pages from WordPress without any issues. this layout happens to be quite a nice layout (each post is in a folder for the year, then month) so I’m pretty happy with that. There are some edits I have to do to get previous images displaying correctly though but in general it works.

I’m just hoping that Google is now able to hit the webpages and start scraping it as a couple of my pages come fairly high up in Google search for specific items. The current theme allows me to use Google Analytics, but I’m not currently using it. It seems to be working nicely in both mobile and desktop browsers so that’s good to know.

Overall, it seems to be working pretty well, but I just have to change how I write some posts now – with having to give an absolute path for my images, I have to make sure I get the URL correct, as I’ve had a number of images fail to show where I’ve missed typed the URL! I’m sure Keyboard Maestro or Alfred can help with that!

Conclusion

So in the end, I’ve switched over to using Hugo – it suits my workflow quite nicely, though it does mean I do have some duplicate content on my Mac – namely images and posts. The post content I’m not to concerned with as the files are only a few Kb and I don’t post sufficiently for that to build up to unreasonable levels really. However, images could be an issue, but I’ve dropped them all down to 640×480 for display on the website (partly to cut down on the bandwidth and partly because this seems to be a reasonable size!) and I’ve run them through ImageOptim which drops the size of them (in terms of my PNG’s, quite considerably in terms of percentage though it does take some time to do so as ZOPFLI and PNGOUT aren’t the quickest tools to run – though they can be de-selected). If this begun to cause issues, I could always install Expandrive and have Hugo write directly to the host server via a virtual FTP drive – that would also save me uploading the information to the server after I’ve built the website at least!

With a few small niggles to resolve (adding a search bar to the website and wondering why it opens each page in its own tab), Hugo seems to be working quite nicely.


  1. In fact, the only thing I’ve been unable to write in Markdown is a letter. Perhaps with some conversion into LaTeX, it would work, but then I could just use LaTeX to write the letter itself – I’ve a few already written that I could use as a template.
    [return]
  2. Actually, I haven’t converted my invoices from Pages format into Markdown either, but these can remain as they are for the time being I think.
    [return]
  3. Not all at once! Over the course of two years, I’ve owned a lot of tech…
    [return]

Garmin Vivoactive Continued

So another article on the Garmin Vivoactive!

As you know, I’ve done a few other articles on the Garmin Vivoactive (here and here). However, I can add a bit more to the continued “series” now! I’ve tested the device on Android (a Wileyfox Swift, a British mobile phone brand) and can now compare this to iOS, which I had been running it on.

The Wileyfox is running Cyanogen OS, so results might be slightly different on other Android phones with different brands.

There are some differences between how the device is seen on Android and on iOS.

Settings

Settings is where the first difference is spotted.

iOS and Android Settings

The iOS menu doesn’t have an option for default music player and it doesn’t have the option for smart notifications.

The default music player setting is as you might expect – you can change what app the controls on the watch will adjust – as Android lets you change the default music player, this allows you to control Spotify or any other audio program.

The smart notifications section is a bit different.

Smart notifications

On Android, you can change the settings of the watch to only allow notifications from specific apps, whereas on iOS, it sends everything and the only way to prevent certain apps is to prevent certain apps from showing notifications. This gives far more control over the content that you want to see on your watch and perhaps be disturbed for. There are a number of default options built in but depending on the Android ROM you’re using, you might have to add in some apps you might consider as important – on my Cyanogen phone, it didn’t have the Cyanogen default email client or text message app included by default so I had to add these. However, be aware, you can suddenly swamp yourself with messages as well. I turned on Google Music so I could set it to download music and notify me when done or similar, yet it would inform me whenever Google Music changed track which was a bit to much, vibrating every couple of minutes whilst I was driving home from work!

Syncing

The Vivoactive syncs to the phone via bluetooth (or low energy bluetooth to be exact). This works on both iOS and Android.

However, the implementation of bluetooth between Android and iOS is different. On both, you have to pair the phone via the Garmin Connect app, which makes sense as you’re associating the app with bluetooth device. On iOS, this is then seen via the phone bluetooth devices list – iOS forces everything to be seen via the main bluetooth menu and can be shared between apps. It makes adding my Wahoo heart rate sensor fairly easy.

Yet, on Android, it seems that bluetooth devices can be linked to a specific app and don’t necessarily show up on the Android main bluetooth devices list. In the image below, the Vivoactive is paired to each device but in the Android one, it doesn’t show up as being attached or paired, whilst iOS is displaying it within the settings.

Bluetooth Devices

Perhaps this isn’t an issue in general, with the exception that one of the perks of using a smart watch with Android is that you can disable the lock screen if connected to specific bluetooth device. For example, my car is connected to my phone (when I’m in it) and therefore I can set the lock screen to be disabled whilst I’m in the car. Likewise, my smart watch, as a bluetooth device, could also be setup to disable my lock screen – a perk that’s worthwhile as the Android device I’m using doesn’t have the equivalent of the Touch ID sensor allowing access with my fingerprint.

That is the theory anyhow – as the device doesn’t show up in the standard Android menu, it cant be set to unlock the phone! It means that you’d have to use an Android Wear device (or maybe a Pebble?) – either way, the Vivoactive isn’t able to act as a trusted device to unlock the phone. This can be set for locations if you want, but I’m testing a third party app (Delayed Lock) that does it based on wifi connections, as well as other items, such as bluetooth devices.

One of the issues I had with the Vivoactive was that you could pull down to refresh on the iOS app but this wouldn’t actually force a sync – in fact, there was no way to force a manual sync on iOS. This meant that I would be limited to the standard iOS refresh rate, which seemed to be as and when it feels like it. On Android, pull to refresh works as it should and even opening the app starts a sync with the watch.

GPS

One comment I had was that at times, the GPS could be slow to acquire – once it’s connected, it seems to work perfectly fine (compare last weeks walk recorded on the Vivoactive and todays’ walk, using the Wileyfox Strava app). Zooming in you can see that the GPS tracks on the Vivoactive are a lot more accurate than the phone. Both were in similar positions (the watch on my wrist and the phone in my pocket).

It doesn’t seem that communicating with the phone helps the watch acquire GPS signal any quicker. It’s a shame as perhaps the Garmin app could help determine the initial location and then the watch could get a more accurate fix from the GPS satellites. After all, the phone is able to use mobile mast triangulation and wifi databases to help get a location before the GPS kicks in. I’d like to see Garmin bring the speed of the initial start down if possible. It is quicker if you enable GLONASS as well, but I think this drains the battery fairly rapidly – I’d look into testing this in the future.

Overall

It seems that either Android allows for better compatibility with a smart watch (which wouldn’t surprise me) or the Android app is better coded and doesn’t suffer some of the same issues as the iOS app. Either way, it seems that unfortunately, the Android phone I’ve got is just better suited to being used with a smart phone – I’ll just have to determine if I can leave my iPhone behind and go back to Android again!

In the week I’ve been testing the phone and watch, Android has been consistently better with the watch than iOS which is disappointing. It’s hard to pin down exactly where the issue is as well – iOS limitations or poor coding on Garmin’s behalf. Considering the Android app works perfectly, I’m almost tempted to say this is an iOS issue. Which is a shame, as I’ve begun to find a smart watch is a useful tool.

LBC: London to Brighton Again

This weekend I headed down to Brighton with some of the London Brompton Club.

I’d offered this ride to some of my work colleagues as we did this ride last year and it was a great ride, thanks to David’s excellent navigating skills, and a couple decided to come along this weekend. I also offered it to the LBC as well, though I was aware that this would clash with the Dunwhich Dynamo that was being ridden that evening. In the end, there were three from the Brompton Club that turned up – Anne and David and Tom, who I hadn’t met before but it was nice to meet. From work, my mate Dan ended up coming along – slightly nervous as he hadn’t done a long distance ride for quite some time. I did feel that perhaps I was the odd one out – I was the only one riding a Brompton, everyone else was on full size bikes!

We met up at Richmond Costa where the weather was pretty good – sunny and not quite as warm as it had been at the start of the week. However, it was reasonably windy with the wind coming from the South that would make life difficult for the ride down. Yet, we left in good spirits.

Quick stop for food and drink

Loki having a rest

We carried on and got to Headley where we stopped at the village hall and stores for a quick ice cream and drink.

Headley Village Hall

Once we got past Dorking, we ran in to trouble. Dan had been struggling with a bit with some of the hills prior to this point and unfortunately we left him behind slightly as we climbed the hill. As we got to the top and started down the otherside, I got a text that he’d punctured. I rode back to meet him as he’d spoken to me the day before and said he’d left his pump in Scotland. I managed to fix the puncture and inflate the wheel – it seems my purchase of a Topeak Mini Morph at the end of last year was a good one. But after it was fixed, Dan decided to leave us and head back to Dorking station.

At that point, I rushed to catch up with the others, who had carried on to Newdigate where we would be eating lunch. It turns out I wasn’t that far behind the others and I got to the pub about ten minutes after them.

After lunch, we carried on. The road continued towards the big finale at the end of the road, Devils Dyke. The hill appeared on the horizon and slowly grew in size…

I been perhaps struggling a bit up until now – the final stretch before Devils Dyke was a killer last year and was a drag this time around as well – perhaps the heat had been getting to me both times at this point. However, with Devils Dyke coming up rapidly, I found my second wind – I didn’t want to be beaten by it. Last year, I had to put a foot down about halfway up where you have to turn right off the main road and up towards the golf course. This time, there were no cars coming and so I was able to make the turn and carry on up the hill. And finally reached the summit with no foot down.

And with that, we headed down into Brighton. The ride down as ever was pretty quick and before we knew it, we were at the seafront. We didn’t wait around and we headed directly to the station, as all of us (with the exception of me) had booked tickets. I made the mistake of trusting to my annual Gold card for travel discount and ended up paying over double what both Anne and David had paid for both tickets! Perhaps I should consider pre-purchasing next time!

Managing to secure a bike spot on the train, we headed back to Clapham where we parted ways. I considered cycling to Wimbledon but I’d had enough for the day, so I got the train back to Wimbledon. And with that, another London to Brighton had been completed.

Phone GPS

On the London to Whitstable ride, I made use of my Samsung S5. This let me make use of the apps on my phone, rather than my old Garmin unit for showing directions.

I sold my Garmin recently – this was partly down to the fact that I could use the Samsung for navigating and partly because I never seem to have the Garmin with me, but my phone goes everywhere. The Samsung is waterproof (and I’ve tested it – though, I’m not sure if it’s water resistant or water proof – there is a big difference!). However, there were a number of issues of using a phone as a navigation device.

Phone Navigation

The first issue I had to resolve was how to attach the Samsung to the Brompton. Obviously the Garmin came with the required kit to attach it to the bike as that’s where it was supposed to go. But the Samsung required some third party equipment to attach it. I spent some time looking around for a device and I came to the conclusion that either the Quadlock or what I went for in the end, the Tigra case for the Samsung would do what I wanted. There was little difference between the two but I went for the Tigra on the basis that I could see that I could charge the phone whilst it was mounted (I only assumed you could do it with the Quadlock) and that the mounting system looked more stable. The Quadlock seemed to be the same as Garmin – using elastic bands to secure the mount to the handlebars, whilst the Tigra used a tightened strap. It should be noted that whilst the Tigra can be fitted and removed from different bikes, it is harder to do so than on the Quadlock, so if you’ve different bikes you want to be using the phone on, you might want to investigate the Quadlock in a bit more detail.

Mounting to the Brompton was easily done and doesn’t really affect the fold on an S type. There’s some degree of touching but this hasn’t caused an issue yet.

Phone mounted to the Brompton

Phone mounted and displayed

Tigra mount on folded Brompton

Tigra with folded Brompton

With the phone mounted, the other two issues to address were navigation and battery life. Using the screen of the phone and the GPS will cause the battery to deplete pretty rapidly. On the London-Whitstable ride, the first two hours were on battery alone and the battery dropped by 50 percent – clearly not enough to get to the end of a 11 hour day (as the Whitstable trip was). So I’d purchased a battery pack on Amazon to plug into my phone. This would live in my Mini O bag that tends to go everywhere (at least in longer rides) with me and I’d run the cable to the handlebars. That was easily solved. Since writing this, Amazon have recalled the battery pack I’ve linked to above as there was a risk that is was damaged. Using damaged Li-Ion batteries is not recommended!

Battery pack in the mini O bag

The navigation issue could prove to be an issue – Google Maps offers turn by turn navigation but doesn’t have a full set of cycling routes built into it so it makes it harder to navigate. I looked at Strava and Ride With GPS as the navigation apps – Strava because I was already using it for day to day tracking and Ride with GPS because I used it for programming my routes, so it would seem the best option for riding the routes as well.

Strava

The Strava website features a route builder for creating and riding of cycling routes, but it does note that this is only in Beta at the minute, so I guess a number of teething issues are to be expected. However, it performs reasonably well for an item that is in Beta and the limitations are probably due to the mapping software they use (Google Maps), rather than the Strava website itself.

Creating a route in Strava is a bit of a pain – no importing from elsewhere so you can’t copy anyone else’s route and import it to follow without planning it yourself. However, you can (if you have a premium subscription, add other peoples rides on Strava as a route). I tried recreating Mark’s Whitstable route on Strava and was left plotting the route manually. However, issues arose where there were one way streets with counterflow bike paths and with cycle paths themselves (such as an area where we crossed the Medway on a cycle/footpath). These weren’t on the Strava/Google Maps route planner and the software would try to do a detour to get to where we wanted to go, clearly not ideal.

I tried it on a different route to see how the directions would work – how to get to cadets, something I know well as it is through Welwyn itself. Strava was a bit disappointing in that regard. The map was purely a “breadcrumb” trail – no directions or warnings, just a route shown on the map and you take responsibility for keeping yourself following the track. Rudimentary I guess but it does work. This is what happens on the Garmin when you put in a GPX file to follow, rather than a TCX file (in my experience anyhow). This means you’ve got to keep an eye on the track often, in comparison to if it was turn by turn with warnings. It also means that you have to keep the phone display on constantly to view the route.

Overall then, it’s not ideal, though it’s big plus is that it’s free. It would be nice to see some turn by turn directions and import but at the minute, it provides a basic route finding functionality. I know the routes for the RAFBF Challenge Ride are on Strava and Strava are mapping partners. However, I might be looking at other alternatives – though if it’s all on roads, I might be ok.

Strava (Left) compared to Ride With GPS (Right) whilst riding

Strava compared to Ride With GPS in the route selection

Ride With GPS

Ride With GPS offers the ability to plan your cycling routes online and then download them to your phone or GPS device to follow. It also offers cue sheets so you can print and have a copy as a paper copy in case your batteries die on your navigation device.

When planning your route, you can use a mixture of Google Maps. Open Street Maps and Ride With GPS’s own maps. This means that you can probably get exactly the route that you want or need as the Open Street Maps and RwGPS maps are actually very good with cycle paths and footpaths, something that Google seems to be missing in places.

In addition to the route planning, RwGPS let you follow directions to get to where you need to go. This route finding is only open to those with a premium RwGPS membership, which is available for a small fee monthly or yearly. This gives some benefit over and above using a Garmin as the maps are updated automatically for you and you can use one website/app to track rides and to direct you. However, I can understand some peoples reluctance to pay for a service when with the Garmin, you can pay for the device and then don’t have to pay anything for additional maps (if you use the freely available maps – Garmin and Ordnance Survey maps require ongoing payments for updates).

The RwGPS app took some setting up and getting used to. To start, it can override the Android devices functionality of turning the screen off. However, this seemed to need some playing about to sort out, as it could be set to be always on, only when the device is navigating or following a route and it could be set so that it only lit up when the directions appeared on screen. Likewise, you can adjust whether the track follows the directions (so your direction of travel on the screen is always up). When riding, I find that occasionally I didn’t hear the warnings for the turns so I ended up missing them, however, I adjusted this for the Mosquito Practice Ride that I did and I ended up hearing the warnings then – the warnings are displayed on the screen.

The warnings and turn by turn directions are ok, but they can (especially in the case of warnings of off course), take up the whole screen and make life difficult to correct your course without pressing the dismiss button.

A feature of RwGPS that attracted me to it was that it can download the routes you ride and store them offline – this includes the maps. This means that you can go out cycling and be happy that if you don’t have a mobile data signal, you can still find your way around. However, this seemed a bit awkward until I figured things out – on the Whitstable ride, I left data on and I managed to lose signal along the way – the maps disappeared but the route remained as a red line on a grey background. Not the most useful tool then, but would allow me to follow the route. I find when doing a later ride, that the trick is to switch off data. This then forces the app to use the download map tiles and I didn’t have any issues then at all.

For me, RwGPS is a much better tool for navigating with. The route finding was better online and the maps are much better for following, though it comes at a cost. However, Strava does display information better for just wanting to know your riding stats, such as time, distance and speed as it has a single page for this, whereas RwGPS displays the data along with a map.

Conclusion

Using a phone as a navigating tool on one hand beats using a Garmin – the screen sizes of the smart phones I have (iPhone 6 Plus and Samsung S5) are significantly bigger than the Garmin and so allow a much greater ability to see where I’m going. However, they’re let down by the water “proofness” of the phones themselves and the battery life. Without a battery pack, using a phone to navigate might see you run out of juice before the end of the ride. I found that using the battery pack above, I was able to charge my phone whilst I was riding and it would end up slowly charging the phone as well (so it was supplying more power than the phone was using whilst cycling).

Whilst you might have a smart phone, a mapping Garmin is likely to cost you £200-£300. At that rate, you can use RwGPS paid for version for about 6-7 years (£30 a year for RwGPS) before you’ll equal the cost of the Garmin. You also have the ability to upload and download data to the phone on the fly if you need to or to upload your Strava rides!

Accuracy seems reasonable to me – it allows me to navigate successfully and that’s the big question. I have noticed that Strava uploads that peak speeds seem perhaps a bit far out and that occasionally the tracks in Strava seem a bit less accurate using my phone than perhaps if I’d used a Garmin. I don’t think that’ll put me off to much though.

I think for now I’ll continue with the phone – after all, I always have my phone with me, but I rarely carried the GPS.