Information in specific types — including reviews — exposed using microformats, RDFa or microdata has already been used by Google for over 2 years, they call it "Rich Snippets", and it does improve the quality of experience for users, assuming that you equate an increase in clickthroughs to mean that the user percieves that page to be more useful compared to other SERPs results (and anecdotally I always go for results which include rich snippet information gleaned from pages with the required semantic enhancements).
This announcement is not the proposal of a new technique, but rather the extension of one which is already working and is a good thing for the web.
The HTML data-* attributes are intended for private data only; i.e., to store data to be used as configuration for a Javascript plugin but which does not hold semantic value and cannot be represented as actual content, whereas microdata (which is what theyre using and is also part of the HTML5 specification) is meant only for describing how the content of the page maps to some schema.
Now that Amazon S3 provides Website Endpoints and index documents [1][2], it is probably fairly trivial to set this up either using an S3 sync tool, or an S3 mount such as ExpanDrive, without resorting to CloudDrive or EC2.
You would need to set-up CNAMES for your subdomains though, and it will not work with the primary (non-www.) domain.
The most useful thing to come out of this post for me is the knowledge of dummyimages.com, and specifically the aliases for the IAB standard ad sizes such as http://dummyimage.com/leaderboard/E/C. Great work by Russell Heimlich on that site, and the source code is MIT too if you want to run it locally.
Thanks I'm glad you like the ad sizes. My goal was to make it pretty hard to get a 404 while sticking to a format that's natural and easy to use. Originally dummyimage.com only had size parameters then peopel clamored for color and custom text thanks to another Hacker News post.
Traffic was unexpected, I was expecting the site to travel within a small group of friends and colleagues who were in on the joke. I uploaded the site to a mediatemple (dv) server, on which Apache's MaxClients is set to a paltry 20. Whilst the server appears to be fine, Apache is unable to satisfy the requests, and I am unwilling to up the MaxClients because normally (mt) automatically disable the server.
It appears that most of the requests were being made to the ~8 images on the homepage, so I've moved these to an S3 bucket and the site is now responsive again. Hardly HIGH TECHNOLOGY, but any port works in a storm (and AWS is typically my 'any port').
I've also repointed nameservers to run via a https://www.cloudflare.com/ free account which should provide a reverse caching proxy for the homepage at the least, but will take some time to kick in.
The lesson to learn — one that I would expect most would not need to learn — is that launching a website (even in jest) in a 5 minute break in-between real work with real deadlines is not advisable.
And to second the parent post, I don't feel comfortable with this being posted to HN at all, and certainly not with it ranking #1/#2. I fear someone may have timed it perfectly to start Eternal February.
There is nothing wrong with your HIGH TECHNOLOGY. Yes, it is S3, and it is kind of boring because from the customer's perspective it just magically works, but you know what they say about things that are indistinguishable from magic... ;)
There is also nothing wrong with your flock of kittens on HN. Into each life there must inevitably come some kittens. We all know the first rule of HN: When kittens appear on the top ten, the Apocolypse is nigh. Unfortunately, the second rule of HN is that too many of us, when presented with a button clearly marked APOCALYPSE, cannot always resist the urge to press it just once just to see what happens next:
Are you the famfamfam guy from the Silk icon set? If so, let me say thank you, you did a huge service to the community by releasing those high-quality icons at a time where there were no decent free icon set available for web apps.
I just wanted to clear something up. A lot of our customers increase MaxClients settings, and we don't usually disable their servers for doing so. You can actually go into our community wiki to see how to do it. Other than that, good luck with S3!
I dunno man. I need placeholders, but there was something missing... But I love kittens even more, so the choice was obvious. This just made my site mid-design look exponentially cuter (meow).
Mark here, I threw placekitten.com together this weekend after a journey on a train in which I could not access placehold.it and few weeks of jokes about the concept with colleagues.
I was not planning on posting to HN, as I thought that a kitten-meme related site might not be the most appropriate (looks like the wisdom of the crowd has decided otherwise).
I can't claim that the server will stand up, so I would refrain from using the site for anything critical (not sure why you would need/want to), or that caching is entirely correct (I'm sending far-future expires, but I'm sure that I could improve given some time).
Glad that the internet likes kittens though. Confirmed a suspicion I had.
It also appears that I should have taken that extra couple of hours to host the site somewhere other than a small mediatemple server, you're all hitting HTTP client limits at the moment.
Wondering where you got all the pictures from? Is this a gray area of copyright? I'm asking because I've tried to source large amount of pictures in the past will little luck.
You could assume that the time allocations could be broken down as follows (billable by the hour), which makes it more believable that the total cost could be approx £585.
* 1 hour - Account handler discusses with request with client, adds to studio traffic scheduler
* 2 hours - Icon development by designer
* 3 hours - Developer wrestles with old website platform to try and add relevant <link> to every template
* 1 hour - Project Manager reviews and discusses with client
* 1 hour - Developer deploys to production environment
Making 8 hours at approx £75/h (£600) which all sounds perfectly believable for a London digital studio.
One of the issues with the low barrier to entry to web development is that a class of people exist who believe that the job is entirely a process of making a website (or a favicon) and uploading it via FTP, maybe even installing Wordpress. Complex client relationships, business/deployment processes and old web systems do not factor into their equation for cost. It is important that these people exist, otherwise the cost of being involved with the web would be astronomical for businesses at the lower end, but there _are_ good industry practices for quality work and jobs required at the higher end which they are putting in jeopardy with sensationalist headlines such as this one. Some clients have more available money, but require much more effort than throwing up a few plugins and shooting them an email once it is done. Nor is favicon design a case of taking one of my Silk icons are throwing it through a dynamicdrive favicon maker.
That said, there are still a few examples of UK website procurement which goes beyond what you might call value for money, such as Birmingham Council's website which cost £3m (http://www.birminghampost.net/news/politics-news/2009/08/04/...), although my concern with that website was not so much the cost but the quality of the work (from a quick scan I don't have the same concerns about the ICO website).
I agree, I (with rather bad timing) I spent part of yesterday scheduling over a month's worth of tweets into CoTweet, the ability to not have to manually enter a date and time repeatedly would have saved a large amount of time.
However, after watching the video I wished. I don't want to enter my information without at first being able to see what the limitations would be on the free plan. The text on the site doesn't say anything about non-free plans/pricing other than that they exist. Careful scrubbing along the video does suggest its 2 a day (which would be fine for a free plan), with a max of 5 in the buffer which is so low as to completely ruin any plans of swapping it in for CoTweet and makes me much less willing to enter my email address.
May I suggest placing the plans/pricing information on the site? There is a page for it, but it appears to be mostly a duplicate of the homepage without any extra information.
Hi Mark, a fellow brummie, I think we might even have met - I did a bit of work for Andy Higgs not too long ago.
Firstly, my apologies for the lack of pricing information. There was pricing information until about 20 minutes ago, at which point having watched in real-time the bounces off the pricing page I wanted to try something drastic - just a big button. It was all I could do in the short time I had whilst I had some traffic I could see the effect with, but it certainly improved the signup rate. Whether I am now attracting people who will never pay is another matter.
Just so you know, the pricing was:
Free = 2 tweets per day, 5 in your Buffer
Standard $5/m = 10 tweets per day, 50 in your Buffer
Max $20/m = unlimited
After some initial feedback, I am considering perhaps increasing the limits on the free plan so that people can get a real feel for the service. The paid versions also allow you to use your own Bit.ly details to track the clicks on your links.
My apologies again, but it was an experiment I had to try. I'd love to chat to you about how Buffer could be useful for you and answer any questions you have. Feel free to drop me an email at joel@bufferapp.com
These numbers sound about right to me. In the UK, with ~2 million engaged viewers, we see around a 1-2% conversion when a URL is mentioned and shown on the screen (off the top of my head 20,000 users in a 30 sec period would be a fair average), typically 3-6% conversion over the course of a 1 hour show.
The important thing is for the website to exist and be online within the 5 minutes of the URL being displayed.
If the presenter reads a URL and also explains some activity that can be done on the site, you can get anywhere from 2x to 40x the traffic that just showing or just saying the URL will create, it depends on the activity and also which presenter is reading out the URL.
Displaying the URL alongside a screenshot or a video of website will net you 4+ times the number of visitors that a simple piece of text displayed or read, but not as much as a presenter telling viewers to go to the website.
A broadcast on a delayed +1 channel gets you 10% of the above traffic.
Its simple, but it works well, especially to pick main theme colours, mrspeaker rightly points out that doing so in a tiny preview window rarely gives you a proper indication. I'm not entirely sure that it actually fits into a designer or css workflow, but it _is_ enjoyable. Plus it appears to be HTML5, so it has electrolytes.
A few thoughts (without necessarily providing any solutions!):
* While interesting that the logo changes colour, I find that it nearly always picks a gradient which is extremely uncomplimentary to the background. It would be less of a distraction if it was only white/black.
* When you click to lock the colour, it would be nice if it could leave a marker to show you where your choice was, so that you could make minor adjustments more easily (I guess the luminosity makes it a little tricky!).
* The links for the colours at the bottom of the app say rgb(239,16,146), which would be great for CSS to copy-paste the correct format, but when you click it actually enters '239 16 146'. The hex code is the only one which gives me a useful value straight into the clipboard.
* I missed the note about luminosity at first (I count myself as a typical user in that I do not read instructions), it makes sense and works quite well when you actually start scrolling, but is a little hidden.
This announcement is not the proposal of a new technique, but rather the extension of one which is already working and is a good thing for the web.