rss

iBooks - Exciting Times

Nimble Works — Wed, 25 Jan 2012 10:46:59 +0000

Exciting times, we now live in that age of the interactive book. You can now enhance the experience of reading your book, add videos, questionnaires, tests, music and so much more. This has been possible if you built a custom application and added in all these features, but that was prohibitively expensive to do well for most publishers.

Enter iBooks, Apple announced on January 19th a new iBook format and tool suite to allow you to create fantastic, interactive books to publish and sell for the iPad.

I’m incredibly excited to announce that NimbleWorks is offering iBook publishing services now. With the ability to add 3D elements, web pages and presentations the possibilities for iBooks are enormous. No longer limited to stayed, boring textbooks the new format is ideal for so many more applications.

Product Catalogues - think demo videos and 3D interactive renders
Training Material - interactive questionnaires, presentations and videos
Portfolios - video, pictures, webpages.
Guide Books - Use the location services to offer information about what is around the reader.

NimbleWorks is incredibly well placed to offer iBooks authoring, with extensive experience in iPad and web development. If you want to know more about the possibilities or simply want to have a no obligation chat about iBooks feel free to get in touch.

January

Nimble Works — Fri, 30 Dec 2011 11:22:48 +0000

1st of 12

In January I’m going to complete the backend server side compnent for my personal advertising newtwork. I like many web people have quite a few domains that sit around doing nothing. I’ve been placing handpicked affiliate links on there just to see if I get any bites. I think I’ve earnt less than $5 from it so far. Still it’s not always about the money is it?

Currently the system is all javascript based and simple picks an advert and text from an array at random. The adverts are in the style of “The Deck” or Fusion ads, so one image, no animation, a link and 140 charachters of text.

The pplication I’m building will need to allow an administrator to upload an image, text and link. There will need to be start and end dates for the advert to run along with the option for a single adverto to be a road block. Road blocks are where one single advert is displayed for 24 hours.

The application will need to rotate adverts evenly and record statistics on how many times each advert is displayed. It will be worthwhile tracking which website is displaying the adverts.

These statistics will be displayed in a graph format to the administrator, or downloaded as a CSV if there is enough time to jam it in.

I’m going to build this in Ruby using the Sinatra framework, it’s light easy to use and very flexible. For the graphing I will use Raphael.js, I’ve used it for a similar application in the past and have been very impressed. I’d also like to get to know it a little more. It will either be hosted on Heroku or my production server.

This is all fairly safe technology that I’ve used before there is nothing new here. I could have built this in node.js, I do want to do a project with node. I’ll probably use MySQL as the database, MongoDB or one of the other sexy NoSQL databases might well be more appropriate. The reason I’m usng tried and tested technology though is to get the app completed as quickly and painlessly as possible. I want this first side project to be as quick and as painless as possible to trick me into thinking that this year of side projects will be easy.

Memories and attitudes are based on the first and last experience. First kiss, first day at school, first race, last race, last day at work. To associate 12 in 2012 with a positive result in my mind will make me more motivated to complete that harder months where I have to battle new technologies into submission.

12 Apps in 2012

Nimble Works — Fri, 23 Dec 2011 16:00:18 +0000

Last year I geared up to build 12 side projects in 12 months. I grossly over estimated my available time and my ever changing work situation. The apps I tried to build were complete applications that in retrospect would have taken a month of fulltime work to build. This year 2 things have changed, firstly 12 apps of 2012 sounds so much better, secondly I’m building 12 features or demos rather than applications.

They will most likely be small applications like tweetCMS. I’m curently interested in some less practical areas of computer science like Natural Language Processing, Neural Networks and Genetic Algorithms. I’d like to have a go at building my own RSS Feed reader that is a bit cleverer than most. Some months I might just build additional functionality onto this project and call it an app.

A while ago I built and advertising network in an afternoon, it would be good to go back and build a decent backend for that with statistics and some more advanced features.

The most important thing about this annual slog through side projects is keeping everything timeboxed to at most 20 hours work. In reality I don’t have more than 10 or 20 hours a month spare, so to deliver a project every month I need to plan effectively and be realistic.

So on the 1st of everymonth I will post a specification of the app, what I intend to achieve, technologies I’m using, expected outcomes. Twoards the end of he month I’ll publish the app and a quicj synopsis of trials, tribulations and possibly jubilation at actually shipping it.

Want to join in? I’ll also be using the #12in2012 hashtag on twitter if you want to follow along or block it out.

MODX Mobile Publisher

Nimble Works — Wed, 23 Nov 2011 13:18:39 +0000

Last week I was so frustrated with the number of steps that it took me to publish an article on my website I build a desktop publisher. Horribly site specific, it's a beauty. In a few dozen lines of code I can publish either an article or a link to my site. The brilliant Official Guide by Bob Ray and the great documentation by the MODX team made it a snip.

As a developer MODX is an A-Grade CMS, but in Revolution we had to take a few hits for the greater good. There is less control over the Manager interface as the ManagerManager plugin hasn't been implemented. As I would expect, with out feeling entitled obviously, is that the MODX developers are improving the manager interface. Sure enough, Revolution 2.2 comes with some seriously good Manager tweaks and performance upgrades.

The community has stepped upto the plate, most notably Mark Hamstra with a mobile version of the Manager Interface that is nearing finalisation. Exciting stuff.

I want what Pareto promised. I want to be able to do 80% of my work with 20% of the work. 80% of the time I log into the manager to simply post an article. I would much rather be able to do that from my iPad or Mac where the article is written.

Building an AdNetwork

Nimble Works — Wed, 09 Nov 2011 23:34:18 +0000

Another wekeend project from NimbleWorks was to build an Advertsing network. Turns out it's incredibly simple todo if you don't want any features.

Inpired by the best niche advertising networks out there at the moment the plan was to build a single image per page, with 140 charachters of text and a single link.

We need three things for this to work:

Inventory
A delivery mechanism
Space to display the adverts

Inventory was going to be the hard one, we quickly settled on affiliate links for products that we thought would be useful or that we use ourselves here at NimbleWorks. A quick scoot round the services that we signed up to and ther were half a dozen affiliate links.

The delivery mechanism of choice for all of these ad networks is javascript and is fairly easy to implement. All you need is to deliver randomly an image, a link and small ammount of text.

function theAd(){
	var adSpace = document.getElementById('adSpace');
	var randNum=Math.floor(Math.random()*(ads.length));
	var toInsert = '<a href="'+ads[randNum][1]+'"><img src="'+ads[randNum][0]+'"></a>
<span>'+ads[randNum][2]+'</span>';
	adSpace.innerHTML = toInsert;
}

What we have here is an array of adverts, we select at random one of them and inset the HTML into a div with the id of "adSpace" on the page.

All in all 20 minutes work to put together. What took the most time was finding and collating the affiliate links and updating a few website to now display adverts.

Obviously what you don't get with this is any kind of statistics. Idealy the javascript code would request the advert from a webservice that tracks impressions and sites requesting it. All in all the process of delivering adverts is very simple, the hard bit is finding the advertising space and selling to the right companies. In the end the more highly targeted the advert the higher the posibility there is that a viewer will click and buy.

Speeding Up Wordpress

Nimble Works — Tue, 01 Nov 2011 21:39:51 +0000

One of my hobby horses and something I like to rant about at any moment is page speed. For years there has been plenty of research supporting fast load times. Optimal target being 0.4 seconds, and a maximum of 4 seconds. Researchers at Amazon even found that for every 100milliseconds your site slows down, you lose 1% in revenue. Along came Google, who just happened to mention that page load times influence their search calculations but only by 1% or less.

I recently was chalenged to optimise a Wordpress site and have had to take a deep dive into all that is Wordpress optimisation. It's a fun world, and like an onion one that take many layers to uncover. This isn't a complete guide but a list of the immediate things that I did to get a site from 10 seconds down to 2. During this test I also ran the optimisations on a test site, which was consistently runing at between 0.4 and 0.7 seconds.

The site at the time was loading at between 6 and 10 seconds which is an incredibly poor performance. The first step was caching plugins. I prefer to use W3 Total Cache as it is easy to install and integrates well with MaxCDN, my content delivery network of choice. It is widely acknowledged that installing a cache plugin for WordPress is as essential as having a password for site administrators. The cache plugin brought the page load time down to 4 – 7 seconds which is more acceptable but still far from ideal.

Optimising WordPress serves two purposes; firstly it improves a visitor’s impression of your site. Secondly I don’t like angry phone calls and emails from clients at 1am. A spike in a website’s popularity could end up crashing the server. A lot of these optimisations will reduce server load which in turn will enable us to reduce our hosting costs and ensure resilience under heavy loads.

The following research took about 3 weeks, lots of late night reading and an example WordPress install to test on. We made the decision to move away from a shared hosting provider to a managed VPS, where theres is more control of the configuration, server loads and number of sites hosted. Being on a shared hosting package is okay until someone on that server gets busy and then everyone suffers. Previously on the clients website there were parts of the day where load time was distinctly worse. On the current VPS (virtual private server) we have guaranteed dedicated resources with bursts of extra resources available if needed.

Serious WordPress optimisations fall into 7 steps and are basically different levels of caching:

Set browser cache headers
MySQL caching
Installation optimisation (removing unwanted plugins)
URL Structure Optimisation
Use HTTP compression
Reverse Proxy (nginx)
PHP Caching

Caching Plugin.

The W3 Total Cache plugin does a lot of heavy lifting for us. Immediately we can set it to combine and minify (reduce extra text in documents) which reduces the number of documents that browsers need to download. The plugin can also set the browser cache headers for any documents. As a result if someone revisits the site they get a much faster load time.

MySQL query cache.

On the test server with a clean cPanel install there was no MySQL query cache setup which was a bit of an issue. After some fiddling around I arrived at a level that the server was happy with and didn’t cause the server to crash. MySQL optimisation is a whole industry in itself. It’s a complex task and I am by no means an expert. There are certainly still gains to be made here. For the small test server I set the ‘query-cache-size’ to 128M for a server with 1024Mb of RAM. Set a key_buffer for 20M and finally the table_cache to 128M. Simple options to set, but if they are set to high the server will crash under load, so be careful.

Remove unwanted plugins.

Uninstall any unwanted plugins and then uninstall any wanted plugins that are not absolutely necessary. Finally check through the head of your page, I found some plugins adding in additional copies of Javascript libraries. Not cool plugin, not cool. In any places like that remove the additional library calls, or seriously think about using another plugin that achieves similar results.

Change Your URL structure.

I found this very interesting article through the Forrst community about not using /%postname%/ in your URL structure. Using this tip it really helped me get my test site down from 0.8 seconds to 0.5 seconds a page. If you don’t fancy reading the article, it says to stop using words as the start of permalinks. Use /%post_id%/ instead. For exactly why and who says so (WordPress developers) read the article. Good news here is that Wordpress is clever enough to redirect old URLs to the new URL so you won’t lose any SEO value. Happy Days!

HTTP Compression

Turning on HTTP compression from the web server means that less data is flowing down the pipes. It’s a quick way to increase the load times of a page. If you really care about IE6 then be careful as you’ll need to turn off compression for most of it. I’m using cPanel and this has the option built in. Apart from that it’s a well-documented bug so I am sure a web search will find a tutorial.

That’s it for optimising WordPress. The next two steps don’t really affect page load times except if you’re running in heavier traffic scenarios. The next steps are therefore more about getting the servers ready for high page loads. I want my site to perform reasonably well on the off chance that we get up to 100 concurrent users. Under those conditions the more hardware you can throw at it the better, but really you should be still expecting 4 seconds tops.

Reverse Proxy

Everyone loves Apache, THE web server, well yes and no! Increasingly the world’s busiest sites are looking to smaller lighter more efficient web servers to perform under high loads. nginx does the trick; it’s well respected, well supported and super-efficient at serving up static files like images, pain html, css javascript etc. I still run apache to deal with the PHP and have noticed a significant number of concurrent connections that the test server could handle when running nginx.

PHP Cache

I’m using eAccelerator as it comes as a build option in cPanel but APC and XCache are two other options. eAccelerator et al improve the performance of PHP under load as objects are cached to disk or memory rather than being processed each time. As such they don’t have much effect on low traffic sites.

What Else

So I’ve not talked about image Sprites for CSS images that can improve page loads. I’ve also not talked about using tools like Memcache as we are not trying to cope with hundreds of thousands of hits a day. Under the right circumstances I’d look at trying to move some assets onto a CDN (content delivery network) but that has to be under the right circumstances as the additional DNS lookup can have a negative effect as I found out on the test site. All in all I’m fairly pleased with the current setup, the test server delivers at between 0.4 and 0.7 seconds and client site page loads somewhere in the 1.5. On the clients site they are however using a visitor chat plugin that, at times, adds another 2-3 seconds to the load time. As it is pulled in via Javascript and pops up when ready it doesn’t have a bad impact on usability.

BikeCoffeeBike | A Bit Of Fun

Nimble Works — Thu, 29 Sep 2011 11:40:09 +0100

BikeCoffeeBike.com is a new site quietly launched this week. It's a site for us to recommend all thigs bike and coffee to the wider world, we spend a lot of time geeking out on coffee and bike equipment. We've tried a lot of these things and broken a fair few 'pro' gadgets. Time to share that experience and knowledge with the wider world. Ther ewill be a daily update, 365 days a year of items that you *need* to have in your life.

At the moment we're quietly tweaking and adjusting things, like twitter integration, as the site has been built on a whim and late at night there are still a few things that need adjusting.

So check out Bike Coffee Bike

Write a comment

Required fields are marked with *.

No More NimbleTodo

Nimble Works — Sat, 23 Jul 2011 19:46:19 +0100

It must suck to be Appigo right now, absolutely suck. Apple announced at their worldwide developer conference that they were adding into iOS 5 (release in Autumn (or 'fall') 2011) a new ToDo list. Imagine those ToDo lists that you have to pay £5.99 for, like Appigos, now only imagaine it looking better, being easier to use, and of course FREE.

Now imagine if you've spent the past 4 years building up a business around one single idea, a ToDo list for the iPhone. You're going to take that badly, like Appigo did. I know they took it badly, they sent me a strongly worded email from a man with many letters after his name claiming that I've infringed on their trademark, TODO. Checking their site I noticed they've also redesigned the site and spend a lot of time talking about consulting services. They are scared.

I pulled the Nimble ToDo, I can't be bothered to play their stupid game. I can't make a better app than Apple.

I wish Appigo and their trademark all the best in the coming difficult times ahead for them.

Write a comment

Required fields are marked with *.

A New Website For TrackTime

Nimble Works — Mon, 17 Jan 2011 11:35:45 +0000

original by lancejohnson on flickr

After much work and umm-ing and ahh-ing I've built new sites for the main NimbleWorks products. This should make it easier to discover and find these applications.

TrackTime

TrackTime my time logging software for creative Mac professionals has moved to gettracktime.com its looking rather good over there and has a nice easy to remember URL.

Nimble ToDo

Nimble ToDo my iOS to do list software now lives at it originaly intended home - nimbletodo.com again another big visual step up.

Why the sites? How are they built?

Both of these sites use commercial templates that were purchased for a relitively small ammount of money. How ever I have them both running on different platforms. I wanted to try a little A/B testing with the two platforms, first nimbletodo.com is running on what seems now to be the industry standard Wordpress. gettracktime.com is running on my CMS of choice MODx, and my experience of developing the two is rather different.

Firstly I consider myself to be a dab hand at MODx using it everyday for the past 18 months or so in my PHP contracting work and as such I have a small repetoire of usefull snippets and code that I can call on to make MODx do some very clever things. On the TrackTime website the most notable bit of code that I use is on the fly thumbnail cache generation for all the screenshots. If I was to use word press I would have to generate thumbnails manualy for all the images. Next up is the ease of integrating Googles A/B testing into MODx - it takes 5 minutes to set up the system and be running, this is a BIG bonus if you want to do any sort of split testing / optimisation.

Caching MODx comes with some great caching out of the box it really does I have no worries about it being overloaded, the caching also has a benefit of making page load times fast. Both sites use lots of small graphical elements that slow page times up so there are improvements there.

Nimble ToDo took me one hour to set up over coffee on a saturday morning, the TrackTime site took 2 weeks of on and off work, but would have take 2 days of full time commitment. So TrackTime took a larget upfront investment, but it is a considrably more extensive site. I had to create 3 templates from the sample HTML that I had purchased, but there are still a few tweaks to make. Nimble ToDo I'm a little disappionted in to be honest, the Wordpress template that I purchased is shockingly bad. I'm going to have to go back and hack the code quite extensively to get any sort of SEO performace out of it.

And there we go, I have mentioned SEO, at the moment MODx gives very fine levels of control over the SEO performance of every single page, incuding internal redirects and 301 permanent redirects if I want. I get to control meta tags per page, sitemap priorities, rel_nofollow in any page, no index, etc etc. Wordpress comes with some good SEO plugins but I'm not sure that they give you such a fine level of control over ever page.

Ultimately SEO performance depends on the work that you put into it, making sure titles are set, identifing longtail keywords, meta tags correctly filled out. Currently neither site has all that completed, so the comparison is a moot point.

I'm going to be interested to see how well the two sites perfom, Wordpress as always was stupidly easy to setup and get running, hence its popularity, but I find it quite frustrating as every plugin adds unnecessary weight to the pages and getting a site 'just so' requires hacking horibble mish-mashes of php and html. MODx on the otherhand does require a steep learning curve, quite a bit of up front investment to get the site off the ground, but offers a much cleaner way to operate. PHP and HTML are seperated, templates are very very clean, and adding extras like A/B testing for conversions or optimisation is a breeze.

I prefer MODx as a system and fully expect gettracktime.com to out perform nimbletodo.com but its the home of a higher value asset so more time will probably be invested into it.

Contradictions, thoughts and comments are willingly accepted below

Write a comment

Required fields are marked with *.

Version 1.4.5

Nimble Works — Mon, 10 Jan 2011 15:23:15 +0000

I've just updated TrackTime to version 1.4.5 which is a shame, and a pleasure. Over the Christmas period I've been working on a new feature for TrackTime as well as some bug fixes, unfortunately the new feature intended for 1.4.5 is taking longer than anticipated to implement. So in 1.4.5 you get all the critical bug fixes.

The Future:

Right now I am working on the ability to retrospectively add time to a project from the time line. This is a fairly often requested feature, and one that I would like to have in there myself. As of the 10th of January the feature is about 60% complete. There are also a few minor bugs to squash before 1.4.6 is ready for general consumption.

In other news version 2 of TrackTime, the long awaited super improved dream version (in other words major update) of TrackTime is taking shape, currently work is taking place on two fronts, firstly a refreshed user interface - bringing TrackTime bang up to date in terms of look and feel. Secondly and most importantly the architectural changes to TrackTime to improve the overall performance. TrackTime 2.0 is set for release at the moment for the first half of 2011, if all goes to plan.

The App Store:

It seems appropriate to mention the app store at this point, currently TrackTime isn't available on the app store. The plan is to get the current version out of the door, and start working on a version that Apple will accept into the app store. I think that the app store is the best thing since sliced bread for indie Mac Developers giving us all access to a much larger market, this in turn results in more resources to produce better applications. Expect to see TrackTime on the app store shortly.

TrackTime 1.4.3

Nimble Works — Wed, 10 Nov 2010 14:55:53 +0000

TrackTime 1.4.3 received a minor update today in the form of updated help files, credits and two improvements to the application tracking. TrackTime 1.4.3 and upwards will no longer track the Login Window or the screensaver.

FireFox Update

I was hoping, really really hoping that the Firefox team would update the AppleScript data in 3.6 or even in Firefox 4 (currently in Beta) so that we could again have URL information from them. Alas todate the betas of Firefox 4 don't have improved support.

Solution Coming:

I'm working on what I consider a less than ideal solution at the moment, a Firefox plugin, that will re-enable full Firefox support. This update will ship by the end of November at the latest. <- its printed I have to keep to it now.

That Firefox problem.

Nimble Works — Mon, 09 Aug 2010 20:21:16 +0100

A Number of users have got in touch regarding Firefox not tracking websites. Firefox 3.6 broke support for AppleScript which TrackTime uses to get URL and Title from Firefox. This bug has been reported and as yet has no fix date within Firefox.

So its not cool just not to track the data from Firefox, Google Chrome doesn't have full AppleScript support and yet its working. As yet the methods that were added into TrackTime 1.4.2 for Chrome support don't work with Firefox. Thats probably a good thing as it can be unstable. The good news with chrome is that the version 6 developer preview there is better AppleScript support and plans for Chrome releases before the end of the year to have Safari like AppleScript support.

I digress, Firefox 3.6 what to do? Well at a minimum I can log page titles in the upcoming 1.4.3, that isn't a problem. TrackTime 1.4.3 is scheduled for release before the 21st of August, that gives me time to work out a temporary solution while we wait for Firefox to be fixed.

TrackTime Launched

Nimble Works — Tue, 27 Jul 2010 14:35:08 +0100

Nimble Works’ latest package, ‘TrackTime’, does exactly what is says on the tin. Offering speedy and accurate time-recording, it guarantees to save the user time and money.

‘TrackTime’, designed by Mamooba Software, is the finest retrospective productivity tool for the Mac, allowing users to see how productive they really are – as opposed to leaving it to inaccurate guess work or time-consuming counting.

It monitors application usage, iTunes history and web browsing, to allow the user to pass through multiple sites, from internet to program, without the task of manually recording links. Most importantly, however, TrackTime allows users to assign time spent to projects to help in billing clients or simply understanding how they use their day.

"Acquiring TrackTime adds another arrow to our quiver, as we strive to build the best tools for independent professional Mac users. Nimble Works is really excited about the future of TrackTime, 1.5 is scheduled for release later this month. We are also keen to hear user feedback for planning the feature set for version 2" says John Nye owner of NimbleWorks.

All current owners of TrackTime can expect unlimited support and 1.x updates that they received previously - so as far as current users are concerned, it's business as usual.

For more information on TrackTime, visit www.nimbleworks.co.uk/tracktime. For the press kit, visit www.nimbleworks.co.uk/pr where information on review licenses can be found.

June - Update

Nimble Works — Wed, 23 Jun 2010 14:57:45 +0100

WiFi MAC Address (itunes link) launched on the 21st of June Globally. WiFi is a true iOS app, designed to solve a very specific problem. What is my MAC address? it isn't very often you need to know the MAC address of your device. When you do invariably you can't remember where the MAC address is listed in the Settings. I couldn't. WiFi MAC address solves that problem.

WiFi MAC is released as a free app, it didn't take much time to code or design, it small simple and light. I'm releasing it free as a way to learn the intricacies of the dreaded app store. Which I'm pleased to let you know was a pleasant experience. So far with zero promotion there have been 156 downloads in 2 days, far exceeding my expectations.

In other news, the next iPhone product is coming along mighty fine, just in the final throws of beta testing. Next up on the schedule is preparing the dreaded but vital marketing information, but this has to be postponed there is a very very exciting development that will be announced shortly concerning our line up of Mac Desktop software.

Introduction

Nimble Works — Wed, 09 Jun 2010 16:15:52 +0100

NimbleWorks, a small independent Mac developer run by John Nye, at the moment we have a small iOS utility WiFi MAC awaiting approval in the app store. Now that iPhone OS 4 has reached the GM stage there is a lot of work to do on the next iPhone App that has been under development for an embarrassingly long time, it's a completely different and more beautiful beast than what was intended, so it hasn't been all wasted time.

Towards the end of 2010 there will be a desktop MAC application that is currently taking shape in notebooks and sketches. Of course there is the iPad, what a superb device I'm itching to build something for that, there are ideas that haven't reached even the sketch pad yet so who knows. Join the mailing list to keep up to date, or subscribe to the RSS feed when it is up and running later today.