I just recently started working on a small pet-project to keep sharpening my skills on the whole stack in my free time. One of the things I decided to go with is angular-material. It's an official Angular implementation of Google material design and it uses flexbox layout instead of the grid layout we're used to seeing in frameworks such as Bootstrap or Foundation. We talked a bit over it and since we will most probably use angular on most of our future projects, we decided to keep an eye on the whole thing so we could start using it when it feels production-ready.
Big data has been a hot topic for some time in our Industry, and Dovetail has of course been exposed to the challenges of creating, consuming and reporting on large data sets. InfluxDB is our new friend on this journey.
Without going into the details of the project - Kit, John & Trevor have been working with a new client recording large amounts of information from sensors, then creating valuable, responsive reports on the fly for end users.
We created the initial prototype using SQL Azure, but quickly found this would not perform sufficiently as a long term solution. The team did some research into time-series databases and settled on InfluxDB as a great next step.
InfluxDB is developed specifically to handle time series data and automatically summarizes data at specified intervals, resulting in much faster reports from large datasets.
Since we're all now humming the song now, we might as listen to it - https://www.youtube.com/watch?v=GDpmVUEjagg (Great tune).
It took me a bit of time to upgrade certain part of the site (mostly the blog), but it was worth it as we're finally running our own website on the latest version of umbraco.
I posted about the release of version 7 almost a year ago. The main upgrade is to the backend UI which has been given an overhaul. There is new functionality we still need to take advantage here, but I'm liking what I've seen so far.
I've started collecting links that go past my Twitter feed, Facebook stream, Slack channels and many other news sources (read: distractions). These are my notable saves from this week. P.S. this won't be weekly, I'm not a machine!
Designing for Website accessibility. A nice checklist of, mostly obvious, but often overlooked details on designing for visual impairments and other disabilities. The internet is is like any other public space. It should be open and accessible to all who wish to use it.
IBM's new smart chip. We still haven't figured out why our brain is so powerful. Okay, we (not me, scientists) have a fair idea why it's powerful. I find it interesting that the faster chips are the ones trying to mimic a neural network.
The feature guy. You don't always make software better by adding features. Sometimes, taking away features or polishing existing functionality is a better use of time.
Balancing bike stations. A discussion cropping up here a lot is the difficulty in keeping the Dublin Bikes stations with bikes and also with spaces. Turns out it's a lot more complicated than we thought, and it's not quite been solved yet.
Today, I wanted to create a clone of a SQL Server Azure database. I was looking at various ways of doing this, including exporting and re-importing the database, but thankfully, there is a much easier way.
I ran the following SQL command against the master database on the SQL Azure server:
CREATE DATABASE New_database
AS COPY OF Old_Database
For more details, check out the official documentation and this Idera blog post.
How to Simplify Input with Steppers. I've always been a fan of Luke Wroblewski. He simplifies UI and input problems, and demonstrates why they shouldn't have been complicated in the first place.
Our mobile apps and touch screen systems are pretty good as they are, but they will be given the LukeW treatment in their next iteration. As we all know, there is always room for improvement.
See the entire series of videos, they're short and well worth your time, regardless of your industry.
In last week's blog post I bemoaned the fact that search engines had conspired to give Dovetail a lot of useless traffic; namely people looking for sample testimonials. That post piqued my interest in what people wanted with our "sample testimonials".
"Surely they're just satisfied customers looking for inspiration to write their own genuine testimonials", I thought. I was wrong.
I started by taking snippets from our testimonials, replacing the word "Dovetail" with the wildcard character (*) and performing a search. Here is a sample. Our testimonials were being stolen.
Some people used our testimonials as starting points for their own unique variation. They'd copy a couple of sentences from our testimonial and then they'd take it from there:
Others would be a bit lazier and just replace "Dovetail" with their own company name:
Others were lazier still. They didn't even manage to remove "Dovetail":
My personal favourite is a copy of our testimonial from Inland Fisheries Ireland, a very specific testimonial:
As you can imagine, I began to doubt all testimonials as I went through this process. I have been assured, however, that all of Dovetail's testimonials are real.
Last night myself, Martin, Garrett and some friends from the Dublin software development scene had the privilege of witnessing the launch of Teamwork PMs latest iteration - the birth of Teamwork.com.
A little Googling will tell you just how much they spent on that domain name, and there is discussion about how crazy it seems. But I think the important message which can be taken away from this is that the guys have huge pride and confidence in what they are doing. They are here for the long haul and to continue to solve real problems, really well. The polish of the new product and the quality of their marketing website certainly backs up that message.
Having fun with the height difference between Martin and Peter McCanney. Photo taken from Mossy's eye level.
Left: Garrett Heaver and Martin Wallace at the Teamwork.com launch in the gravity bar. Right: The gravity bar as seen from the Guinness Enterprise Centre, Dublin 8.
In fact it's here, in RC mode. We haven't installed it yet, but we are very excited about some of the upgrades that come with newest version of Umbraco. Umbraco is our go to Content Management System and allows us to integrate with our custom software seamlessly for our clients.
We love umbraco. We've used it as our CMS since version 3. In fact, one of our clients still uses version 3 and it's still working perfectly for their needs. But over the years, the backend management site has become quite cumbersome and was definitely in need of an overhaul. The workflow for the backend starts to creak when a site/blog starts growing in size. The new version promises to solve a lot of those problems by using more of a wordpress style for handling large news archives or blogs.
Management of media and images is also drastically improved and we know that these features will be welcomed by a lot of our clients.
Our own marketing website (this one, the one you're reading right now) will likely be our testing ground to get used to the new UI and features of Umbraco 7. Once we're happy we can deliver it as part of great software solutions to our clients with it we will begin the rollout/upgrade.
Recently I've been doing some software consultancy work on a payments system, to do with SEPA integration, and ISO 20022.
As part of that I've come across Steve Muench's Muenchian Method of Grouping in XSLT. It's been around for 13 years, but this is the first time I've needed it. I won't describe it here - I couldn't improve on Jenni Tennison's description here - I just want to state well done Steve. Given the constraints of XSLT 1.0 it is a very neat, elegant and efficient way to solve that particular issue.
Thanks very much Steve!