* shameless plug * For longer than most such tools, we've been offering what our customers call the "easiest scraper of 'em all": https://feedity.com Try it out and see if it works for your needs.
33,377 karma · joined August 2, 2007
Blog: https://www.nilkanth.com
Social: https://x.com/voidmonk
* shameless plug * For longer than most such tools, we've been offering what our customers call the "easiest scraper of 'em all": https://feedity.com Try it out and see if it works for your needs.
In the book “Nudge: Improving Decisions About Health, Wealth, and Happiness”, Prof. Cass Sunstein and Richard Thaler talk about the science of choices and defaults:
The human brain is amazing, but it evolved for specific purposes, such as avoiding predators and finding food. Those purposes do not include choosing good credit card plans, reducing harmful pollution, avoiding fatty foods, and planning for a decade or so from now. Fortunately, a few nudges can help a lot.
http://www.australiangeographic.com.au/news/2013/01/aborigin...
"Thrive is the best way to receive key insights to running your business better."
* Seamlessness - it has an in-built processing pipeline (using queues, service bus etc.), whereas with Lambda you'll have to do a little more work to connect the components e.g. S3, SNS.
* Scalability (although both AWS and Azure offer auto-scaling) - Lambda has an execution timeout of 60 seconds for functions. If the functions are I/O bound then WebJobs is more suitable, else Azure Batch offers better CPU bound work support.
* Ease of deployment - 'drop and use', which I don't think Lambda offers out-of-the-box because there seems to be dependency on Node.js?! The WebJobs dashboard under Azure Portal is a great way to track invocations and logs.
* Cost - WebJobs comes free with App Services (Web Apps)
The one and only shortcoming of WebJobs seems be the lack of recurring tasks (but there's a workaround by adding the job back to the queue on its completion).
Further reading:
http://www.troyhunt.com/2015/01/azure-webjobs-are-awesome-an...
http://www.troyhunt.com/2015/04/orchestrating-massive-parall...
http://www.hanselman.com/blog/IntroducingWindowsAzureWebJobs...
An on-premise alternative can be Hangfire, Sidekiq or Quartz.
* ServiceStack - full-featured Services Framework https://github.com/ServiceStack/ServiceStack
* ServiceStack.OrmLite - Fast, Simple, Typed ORM for .NET https://github.com/ServiceStack/ServiceStack.OrmLite
* BCrypt.Net - A .Net port of jBCrypt implemented in C# http://bcrypt.codeplex.com/
* Stripe.net - .net api for Stripe https://github.com/jaymedavis/stripe.net
* TuesPechkin - .NET Wrapper for the wkhtmltopdf library (to generate PDFs) https://github.com/tuespetre/TuesPechkin
Also check this list for Oz VPNs: http://www.vpnanalysis.com/5-best-vpns-for-australia/
Investment makes sense (and you'll get a much better deal from investors) when you have recurring revenue AND you have to grow fast to the next stage (e.g. from a small team of 2 to a team of 5-10, from a few servers to many, from a local office to globally, from self-serve online sales to face-to-face B2B/Enterprise sales for which you need sales people etc.)
In short, evolve organically and delay outside investment until absolutely necessary for rapid growth.