--
You received this message because you are subscribed to the Google Groups "FubuMVC Development Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fubumvc-deve...@googlegroups.com.
To post to this group, send email to fubumv...@googlegroups.com.
Visit this group at http://groups.google.com/group/fubumvc-devel?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
However I ran into issues with IE
Why the need for two accept types? Accept: "application/xlsx,application/json; charset=utf-8"
This is a little awkward, but here's how it's currently working and the reasons for the two accept entries.Short answer: Basically what Jeremy said - you're asking for application/xlsx, but in lieu of that you'll take application/json too. The xlsx export media writer happens to always return json. You can actually leave the application/json off and Chrome and FF work as expected. However I ran into issues with IE not treating the response as json without the accept header which can lead to more awkward code on the client.Long answer: By design, the export Endpoint gets triggered when application/xlsx is requested. However it always returns json with a url that can be used to download the generated xlsx file. This is so we can generate a temp file on the server and use the built-in DownloadFileModel that fubu provides to download the file. By design it's a two-step process: 1) request the file generation 2) download the generated file. One of the things we wanted to have it do is just tie into an existing Endpoint Action. That's great - however things get awkward when that action also has a view associated with it, or multiple other media writers, such as the traditional conneg xml/json writers. For everything to play nice in all browsers, the client needed to request the two accept types, and the Media Writer - ExcelMediaWriter<T> - needed to provide those two as Mimetypes as well.Hope that helps!Joe
use Html conventions heavily to generate the two pieces of edit in place with swappable strategies for actually applying the changes to the backing persistence.
I like the edit in place mechanics in slick grid a lot better than whatever awful JS library we used for Blue's edit in place.