Sign in or 
Share this
Web Unconventions
This talk will be happening on the 2nd of June
But do we really know how to develop web applications or are we all in a cul-de-sac of web conventions. Is it really necessary to fire up a browser to test your application? Surely AJAX and Javascript is no more or less testable than anything else. Isn't it time we forced our clients to move to IE or up sticks to a new browser altogether?
MVC? Yesterday's news, now we have MVP or VROOM. Why write any Javascript when you can write a Javascript generator in your preferred development language. Or perhaps we can write the entire application in Javascript? Can we use Javascript to generate our CSS?
On the data front we now have document databases, graph databases, XML databases. Why do we try and store objects to the database when the web isn't made up of objects?
Why are CMS systems so complicated and so common? Surely we can just generate a website in the same way as we generate Controller code in Rails and its clones. Why the intermediary of the code in the first place?
To try and tackle some of our web conventions and propose some new "unconventions" we will be assembling a crack team of web architects and practioners. They will be proposing a new best practice or the repeal of an outdated one and then we'll throw open the questions to audience. Grill the panel, ask them what a "Transaction modelled in the Domain" means, after all they are meant to be experts.
Felix Leipold (Felix's Blog)
Dan North (Dan's Blog)
Alistair Jones (Alistair's HTML in Java project Hypirinha)
Ian Robinson (Ian's Blog)
Carlos Villeia (Carlos' Blog, Carlos on GitHub)
Post-Event Links
Ian talked about Hypermedia and using it in practical circumstances of designing a message bus. You can get more information on REST and and Ian's ideas from his website.Synopsis
Welcome to the new century where we all know how web applications are developed. You have your MVC web framework with your ORM on top of the relational database that is used as a datastore. You test it with Selenium and you swear at Javascript and IE.But do we really know how to develop web applications or are we all in a cul-de-sac of web conventions. Is it really necessary to fire up a browser to test your application? Surely AJAX and Javascript is no more or less testable than anything else. Isn't it time we forced our clients to move to IE or up sticks to a new browser altogether?
MVC? Yesterday's news, now we have MVP or VROOM. Why write any Javascript when you can write a Javascript generator in your preferred development language. Or perhaps we can write the entire application in Javascript? Can we use Javascript to generate our CSS?
On the data front we now have document databases, graph databases, XML databases. Why do we try and store objects to the database when the web isn't made up of objects?
Why are CMS systems so complicated and so common? Surely we can just generate a website in the same way as we generate Controller code in Rails and its clones. Why the intermediary of the code in the first place?
To try and tackle some of our web conventions and propose some new "unconventions" we will be assembling a crack team of web architects and practioners. They will be proposing a new best practice or the repeal of an outdated one and then we'll throw open the questions to audience. Grill the panel, ask them what a "Transaction modelled in the Domain" means, after all they are meant to be experts.
Panellists
The panel will give a short talk on a web topic close to their heart. After that the panel will be open to any questions the audience have on web-related architecture and development.Felix Leipold (Felix's Blog)
Dan North (Dan's Blog)
Alistair Jones (Alistair's HTML in Java project Hypirinha)
Ian Robinson (Ian's Blog)
Carlos Villeia (Carlos' Blog, Carlos on GitHub)
rrees |
Latest page update: made by rrees
, Jun 5 2009, 1:41 PM EDT
(about this update
About This Update
Edited by rrees
33 words added view changes - complete history) |
|
More Info: links to this page
|
| Started By | Thread Subject | Replies | Last Post | ||
|---|---|---|---|---|---|
| rrees | CouchDb Question | 0 | Jun 6 2009, 9:17 AM EDT by rrees | ||
|
Thread started: Jun 6 2009, 9:17 AM EDT
Watch
During the evening I answered a question saying that you can query historic revisions of documents in Couch. This answer was the right one, the map-reduce is done over the latest revisions of documents.
A more detailed explanation of dealing with historic revisions will be appearing in the CouchDb wiki but for now the brief answer is that if the data is relevant you should store it under a different key in the document (for example Surname and Maiden Name), taking advantage of the schema-less nature of the data or you should separate the historic data into another document and then link via the id of the document representing the previous information. This would then allow you to work back to the main document. For example for addresses you could store each person's address in a different document and then have an array field in the main document that holds the ids of previous addresses in historic order. You can go to the CouchDb IRC channel to discuss the issue in more detail. |
||||
Showing 1 of 1 threads for this page
|
webunconventions_felix.ppt (Powerpoint Presentation - 1,122k)
posted by rrees Jun 5 2009, 1:44 PM EDT
Felix Leipold's Slides talking about DRY in webdevelopment
|
