Web Development

2nd
Feb , 2014

Our GIT client of choice: Git-Tower

Understanding GIT can be very confusing for newcomers to get their heads around, and Git Tower completely solves that problem for us.

At Volume, we take version control very seriously, and in 2013 we undertook an analysis on the various version control frameworks available such as GIT, Subversion and Mercurial. We were looking for a system to use collaboratively throughout the team, which included an easy learning curve for the less technically minded.

It didn’t take long to decide on GIT as the framework of choice due to its community support and branching/workflow methodology. The branching in particular provides us with a high level of flexibility and scalability when dealing with testing and deploying multiple site revisions for our clients.

Although we don’t use the distributed architecture of GIT to its potential, we leverage a very lean central GIT cloud service such as Beanstalk, which we push and pull our committed data to. This allows us to focus on refining our product development, knowing that GIT will take care of itself with little maintenance.

With that in mind, we tried a many GIT clients whom we won’t name in this post, but we absolutely love working with Tower [Mac. OSX only at this stage].
Its main advantage is the ease of use and clean interface. Rather than only the web programming team using Git, we were able to train all levels of our team how to view branches, track projects and deployments.

The interface is split into 2 sections: A sidebar of all the datasets available and a tabbed navigation/content block to view commit status, previous GIT logs, and browse the folder structure. The sidebar consists of Branches [local branches], Tags [Release versioning], Remotes [origin/remote branches], Stashes [test code you can’t bear to lose] and Submodules [Add-on repositories within a repository]. By hiding most of the secondary GIT commands, the user interface doesn’t initially overwhelm users due to its slick look and feel. It gives them familiarity by actioning most commonly used commands, such as push, pull, commit and merge in big buttons.

Our team relish GIT’s easy to manage branching functionality. It allows us to develop and test new features and plugins/modules simultaneously on different branches which can be deployed and tested independently of one another in a staging environment.
This flexibility means we can integrate features into our production environment in a non sequential manner.

 

Understanding GIT can be very confusing for newcomers to get their heads around, and GIT Tower completely solves that problem for us.

In under a couple of hours after initial installation and configuration, we had the whole team working collectively and committing project data.

GIT Tower is a paid application, but it delivers a great return on investment for those looking to streamline their version control strategy.

www.git-tower.com/

15th
Jan , 2014

Low cost android digital signage solutions reviewed

We’ve all seen them. You may not be consciously aware that you’ve seen them, but you have seen them nonetheless....

We’ve all seen them. You may not be consciously aware that you’ve seen them, but you have seen them nonetheless. They’re almost everywhere you look. I’m not talking about bogans or hipsters, I’m talking about digital signs.

One of our favourite clients, ARB, makers of fantastic gear for 4 wheel drive adventuring, currently have 46 retail stores around Australia. They want to advance their current digital signage setup – which comprises an LCD screen and and Apple TV at each store – showing a simple slideshow, to something a bit more feature rich and targeted. They didn’t want to over-capitalise on infrastructure as they had already purchased the screens not much more than 12 months ago, and they already had wireless network connectivity setup and running.

Now, if this was 5 years ago, maybe even a couple of years ago, their options would have been extremely limited. Back then there was such a limited number of signage options available that we had to custom build a system for another client using flash, xml and a number of external APIs just to display some simple information on a screen in their office foyer.  Hardware involved a plasma screen (which subsequently suffered from burn in quite badly), a fairly small form-factor pc and a 15 metre RGB cable. At the time we were quite proud of it, but by todays standards it was pretty embarrasing.

Now it’s different. With the emergence of thumb size computers, made possible by the pioneering efforts of SoC (system on a chip) manufacturers such as ARM, and the development of operating systems such as Android that support them, it is possible to set up a sophisticated, geo-distributed digital signage solution for very low cost.

This is where I wanted to focus my efforts. We had the screens, we had the network connectivity, we just needed low-cost, reliable player hardware and an intuitive, preferably cloud-based content management solution to manage the campaign creative. Finding the hardware turned out to be relatively easy. Finding the software was quite the opposite.

Over the last few weeks, I’ve researched nearly 50 different digital signage systems and/or content management software solutions. Very few stood up to their own hype.

In the end I honed my selection criteria to the following;

  • Must support non-proprietary Android player hardware
    Some of the companies out there are charging US$1000 for hardware that can literally be purchased on eBay for US$60 or less. I know this because i literally stepped through the specs one by one. Now obviously some of this is a player software licensing fee but I find that extremely hard to swallow. When we’re talking about 46 stores, almost US$50,000 for player hardware that is worth about US$2760 over the counter is just madness.
  • Player setup must be fool-proof
    No offence to the guys on the ground at the ARB stores, but we simply can’t afford to be walking through a complicated player setup 46 times over at different times of the day and with differing levels of technical know-how. The player hardware should be able to be plugged in, powered by the USB port on the TV, and the player software either pre-installed or easily installed via Google Play, and be able to be configured to run on device start-up.
  • Must have an intuitive, easy to use content management system.
    This one turned out to be the downfall of 99% of the systems out there. I could not believe the lack of quality that existed in most of the CMSs that I tried. Some had their own complicated methodologies that they didn’t explain via context sensitive help, some were built entirely on Adobe Flash runtime which while not a deal breaker does impact clients ability to manage content remotely given the lack of iOS support, and some simply didn’t work as promised.
  • Must support user encoded Video, YouTube and Images as an absolute minimum, with HTML5 and custom templates as a bonus.
    ARB didn’t need all the bells and whistles or need to support every single file format under the sun. In my experience, the systems that tried to be all things to everyone fell over in spectacular fashion because the system became so complicated that it was almost unusable.
  • Must have content scheduling and player grouping / targetting, with geo-targetting a bonus.
    Having 46 stores distributed all over the country, ARB needed to be able to push different content out to different regions or individual stores at any time hence player grouping, either manual or by user defined geographical regions was a must.
  • Must be well supported.
    Because there are so many providers out there, some having existed for a while and some being the new kids on the block, ongoing support was important. A bricks and mortar presence always makes me feel a little more confident.
  • Must be reasonably priced.
    This is the ‘how long is a piece of string’ element to the whole story. Obviously different solutions can hit different price points for many varying reasons, software complexity, proprietary hardware, staff numbers and SaaS running costs – just to name a few. I don’t want to have to pay, or have my clients pay, for functionality or services that they will simply never use.

Ok, so what didn’t make the grade?

Almost everything.

And what did?

Revel Digital and TargetR.  These two systems ticked most of our boxes.  Revel’s player software just works, you can buy their customised hardware which comes pre-loaded with the player and a 2 year warranty for not much more than you can buy the same raw hardware on eBay, and the cloud based CMS system is intuitive and powerful. Not only that but behind the initial simplicity of the interface, there is power to be harnessed. A full scripting engine, player audits and reports, the list goes on. The ability to manage multiple client accounts under one agency account is an added bonus.

TargetR’s player is also very robust and works very well.  There is also a lot of configuration behind the scenes if you need it, including rotation, video quality etc.  TargetR’s CMS system is quite a sideways shift from Revel’s, but once you get your head around the concept of parent and child channels and the targeting itself it’s quite good.  TargetR is more ‘techy’ than Revel so you will most likely need to read the documentation from top to bottom before diving in to add your players and content.

One of the biggest downfalls of TargetR’s system is that the zone layout is locked to the player installation, not the content.  In other words, if you want to have a sequence of content that includes a full screen video, followed by a period of time where the screen is split into two columns and running different sequences – for example, today’s event information in a column on the left and specials from the cafeteria on the right – you can’t do it.  I really should be able to change the zoning of a player in realtime during the sequence.  If most of your content is fullscreen then this isn’t an issue but it is something to be aware of.

Either of these systems are great for the money, but the decision might rest on how savvy your users are and whether you need content specific zoning.   We did find some caching issues on the Revel player during our testing which we didn’t find with the TargetR player and so this was what finally made our decision.

Final observations

There are a few systems out there that would work very well for a single store up to a SME, at very little cost. There is no point spending $1000 per player licence, or purchasing ridiculously priced custom hardware and bloated software with confusing licencing models if you just want to run a few screens and remotely manage content. OnSign.tv when it finally comes out of beta is a great little tool which is simple in principal and just works but it’s not a pinch on Revel yet.  Targetr is a great system but it’s CMS interface is not as easy to use.

During this R&D journey, at one point I got so frustrated with the offerings that I started working on initial specifications of our own system. I didn’t care that there was already hundreds of systems already in existence, there just seemed to be nothing out there that did what I needed it to do.

Revel is almost exactly what I would have built from a CMS point of view.  TargetR also has another nice little ‘sister’ system that they’re working on called DisplayStacks.  For small signage setups this is a fantastic option as you manage the signage direct on the device – and it’s free.

A not-entirely comprehensive list of the systems I looked at, and if possible, trialled:

11 Giraffes, Scala, 42media Oxygen, Stinova DMS, Dynamax digitalsignage.NET, SMIL-e, Rise Vision, CoolSign, signagelive, SignChannel, DisplayIt!, NEC Live, SignActive, OpenSign, iSignager (QNap), BrightSign, VizzGo, Oxygen, ComQi, SherVision, FireSign, OnSign.TV, Instillo, Capital Networks (Audience), AirTel, X20 Media, Adserve, DISE, TargetR, BroadSign, bitShuttle, ParkMedia.tv, Revel Digital, firmChannel, MyOpenSign.

 

There are others, but if i didn’t put them on the list it means they didn’t warrant the effort!