- From: Karl Dubost <karld@opera.com>
- Date: Tue, 6 Nov 2012 08:59:42 -0500
- To: Sebastian Zartner <sebastianzartner@gmail.com>
- Cc: Boris Zbarsky <bzbarsky@mit.edu>, www-style list <www-style@w3.org>
Le 6 nov. 2012 à 08:22, Sebastian Zartner a écrit : > Instead people should be better prepared for what's coming next and support for the prefixed properties should be kept for one or two engine versions before they are completely dropped to give people time to adapt their code. The key of your paragraph lies in the last few words. "give people time to adapt their code" That's the principle. The sad story is that it is not always happening. It really depends on the social and economic infrastructure of the Web project. Here are cases where things derail (JS, CSS, server side user agent sniffing). * common Web agency projects with a fixed term contract. Once the project is delivered, any modifications to the project will cost money. The client is not paying for additional modifications. * Browsers with low market shares in the country *where* the developers are living/working are ignored to the point that sometimes the unprefixed fallback is not even considered. More than we should expect. :/ * Libraries being part of frameworks and not being updated. You can include user agent sniffing libraries here. * Libraries not maintained anymore but used on many Web sites. The site is online without maintenance. Valuable information but unusable for some browsers. The code gets updated sometimes in the following cases: * The Web site is an active startup, Web application, etc. * In-house management of the Web site (big companies) * very rare Web agencies with maintenance contract based on yearly contract. -- Karl Dubost - http://dev.opera.com/ Developer Relations, Opera Software
Received on Tuesday, 6 November 2012 14:00:27 UTC