So the end of the sprint is slowly approaching and today it’s the last day of actual work. We had daily wrapups of what has been done and what needs still to be done and so we had yesterday (even with some demonstrations). There’s certainly some cool stuff been done and I will give some examples (not complete! I will compile a complete list with results with limi later today).
So first of all Locking comes to mind. Raphael Ritz and Jean-Francois Roche have been working on this.
It basically enables you to automatically set a lock when editing a page and releasing it again when saving it. If you are intending to edit a locked page you will get a warning via the portal status message.
For the Workflow engine Jan Murre and James Small showed us a permissions screen which shows all the states and every defined permission for every state and it’s settings. These can be changed there and thus makes it very easy to have an overview of which states have which permissions. Will probably save us a lot of clicking.
Danny has been working on the sharing page which we made extremely complicated during the last UI spring ;-) So now everything will get simple again (which is Danny’s TM anyway :-) but without losing functionality. Watch ploneTV episode 2 for seeing the mockup of how it will look. It will also use the Überselection Widget when it will be available.
Speaking of that, Doug Winter showed us his work on the Überselection Widget. The scope of that seems to be growing as he now implemented some sort of query language for doing queries of any data source.
The widget itself basically shall enable you to select a subset out from very large data sets like 1000s of users of huge catalog query results. This query language can be used to query users, the catalog or whatever you want and it will converted during execution into e.g. a catalog query.
The display of the (possibly still large result set) will be done by using Rico (openrico.org), a JS widget for dynamically loading new chunks of data as you scroll. He also showed this as a demo.
Alec Mitchel showed us his work on Viewlets. They basically seem to work in Plone now. So you can define them in the configure.zcml and then use them inside templates (as I did not get that completely I will not go any deeper into it before not interviewing him ;-)
Some other cool stuff was shown by Andreas Zeidler. He has been working on Link integrity and demoed us what he has been doing so far. So if you have a page and another points to it via an HTML link and you want to delete it you will get a warning. This is implemented by parsing the HTML and checking with the reference engine. This really looked cool as you can be sure then not to break any links.
Screencast will probably follow.
Martijn Pieters fixed the history bug and we are all very happy with it ;-) But Danny was thinking about enhancing the UI for this. So his idea was to show the actual changes inside the view. It was discussed how this can be implemented (e.g. with some AJAX based solution). I am very much looking
forward to it, sounded really cool.
Geoff and Ben were talking about the state of Cache-Fu. This is a tool for providing all the means for doing proper caching with Plone. It even spits out a Squid config and work is being done in doing some sort of a wizard to ask people about their system and doing the best for caching based on that.
Here a screencast will probably also be the way to go.
Then limi was presenting a proposal by Paul Everitt which sounded very cool. What Paul’s proposal enables users to do is to take a Plone site and put some other design over it easily. So designer can go and create a design with dreamweaver or whatever and this will based on some filter on the apache level the Plone output will automatically be converted to the new design. So it’s not actually a part of Plone but a separate thing which can even be used to skin external modules like mailing list managers and the like. It works by assigning IDs to certain parts (like „this is the navtree“, „this is the content area“) and these will get replaced by some Apache extension with the actual Plone part of it (which then can get further skinned with CSS). I hope I understood that right and I will try to get limi talk to me about it a bit more. But it sounds like a very cool idea and I hope Paul will get much support in making it happen!
The last thing to mention is probably AJAX. There have been lots of discussion how to implement
it in Plone. Quite quickly the decision of using Prototype has been made but some more discussion
was nevertheless happening. I didn’t actually get the problem I must admit and limi also said that he
had problems getting it but so far we reached sort of conclusion. I will try to get this sorted out with the involved guys today as I want to know what I have to do right now to make AJAX work in Plone.
There has been more stuff going on here but as my battery is going low I will find an end here and you might look out for the complete sprint report we will do later today!
(and excuse and typos as I was a bit in a hurry ;-)