Congratulations, I hope everyone got everything in on time OK etc etc. It’s all over – Woop!
Looking forward to the Grad Show to properly get to see everybody’s work, cheers to the people who tested Talkboards in Uni over the night before the deadline. Right now it’s behind the impserver firewall, so you can only play with it properly on campus – you’ll be able to connect alright outside of Uni but it just won’t work properly (and not actually tell you why 
)
When it’s been marked I’ll get back on it and use RTMPT (tunneling) to get it working from everywhere.
- May 8, 2007 – 6:56 pm
- Posted in adaa, fmp
Spent the past day and a bit putting together the web interface for Talkboards.
It’s purposely very blank, with the intention that the attention is only really ever on the Flash elements. More than that, the Flash itself begins completely blank and it’s the user group’s task of building and colouring their own interface, so the minimal web design I hope conveys this further more, that it’s only when the user enters their own environment do they feel they’ve truly ‘arrived’.
The problem is there may actually be quite a bit more interaction with the HTML-only/non-Flash parts of the site than I first expected. There’s obviously unavoidable processes – registering accounts, logging in, browsing for boards to then enter etc – but it seems that by being so blank and essentially text-only, it may not be so much of an engaging experience as could be possible, or I would like, before initiating the Flash environments.
Anyway, have a look:
Comments appreciated.
I’ve also got a session-based CSS selector which I’m not sure whether to implement in the final version or not. Basically the blue-coloured parts could be changed to other major colours used in the Flash elements, again, so show that the web-end really isn’t what the project is about and if anything, what it is about, is individuality and customisation – i.e. the ultimate conveyance of the Flash parts – and with these CSS choices it’s demonstrating it’s possible in the HTML too, that at least this bit is trying it’s best.
In other news, a while ago I submitted my project for the Adobe Design Achievement Awards, today I got an e-mail through saying it’s made it into the semi-final stages of the competition. There’s 62 entries left, so I’m not too sure if it’s ‘made it through’ to the semi-finals or these are just all the entries that have been submitted in total.
Hopefully it’ll be well received.
Congratulations to those who completed the 24hr Sponsored Work-a-thon, I hear a good amount of money was raised by those taking part.
Doing my part instead from home, I’ve made a fair amount of progress in various areas of the site. I’ve got a lot more PHP queried data being passed to Flash, those sections are now able to determine the environment initialisation based on who’s logging into what boards – essentially with regard to access levels, profile information, the admin’s specs and set up etc.
I’ve also established a linear log of chat history in the Flash interface, quite simple to create but I think necessary. Half of the brief I outlined intended the project to converge features of a number of ‘traditional’ communication platforms. This chat log not only offers a more straightforward means for returning visitors to catch up on activity during their time offline – similar to message boards or Facebook’s news-feed, for example – but it also introduces new users with a little familiarity, the non-linearity of the project could easily be confusing at first.
The ever-huge principal task of migrating all static Flash to shared objects via Flashcom continues but the majority is almost complete. Most of the shared objects are now persistently and remotely stored, organised and automatically sectioned by room instances. I’ve can now successfully query this info on re-entry to regenerate the rooms to their previous/continued states.
I’ve also combined some SSAS with multiple netConnections to handle the user’s data transfer – their connections in whole, rooms and connected user lists, webcams and now a centrally defined and independent server-time these all run by.
Plenty more to go, got to sort out image handling next. both inside and outside of Flash. For user profiles on the HTML sections and calculating some kind of pixel/size/quantity allocation for those uploading into communities.
So the first production deadline is looming, I’ve just put together the mother of all to-do lists, it seems my production schedule (from the minor project) is pretty much accurate, but it’s cutting it close.
I now have a better – working – webcam to test with, some prototyping with that has been successful so far. Unfortunately I’m not able to attend the 24hr Sponsored Work-a-thon tomorrow, but at least with the new cam I can be there visually if not physically.
Plus, I recently received an e-mail regarding my project’s proposal for a Rhizome commission and I’m pretty happy to say it’s made it into the final round of voting. Apparently it was very well received by the community and it’s now one of thirty-five projects subject to ranking and voting from Rhizome members and a special jury to work out the rest.
If anyone here is a member and susceptible to bribery and/or flattery I’d very much appreciate your vote 
- April 18, 2007 – 12:40 am
- Posted in fmp
Following up a previous post, I’ve now managed to fully render the virtual networks connecting users and boards in all directions.
Last Friday I went to Matt Northam with some issues nesting repeat regions to achieve this, basically I had an SQL query returning the ‘boards’ of which user was part of, based on session variables, then needing to use those results in a second query matching other users who are also part of these user groups – the first post explains this a bit better. Anyway we couldn’t fix it, the first repeat region would execute but the ‘inner’ would go unrecognised.
Later on though, I tried again instead with completely creating the second recordset (the inner query) actually within the outer repeat region, meaning the SQL would be executed each time the loop is initialised, guaranteeing it being refreshed – and it worked. It was then an intricate case of appending and aggregating arrays of the recognised users into larger container arrays – essentially because one user could relate to another user through multiple connections – then looping through these to display them properly.
With some tweaking this was replicated for the Immediate Network based on finding users connecting to groups that you have created, then formatting everything so usernames and board names are links to relevant pages.
A lot of conditional statements and some hand-coded PHP and SQL were needed but the end result is quite useful.
Also on Friday I spoke to Matt about some problems passing PHP variables to Flash. With absolutely no reason that either of us could find, Flash was accepting some variables – static, sessions and POST variables – and completely ignoring those defined by the GET method, which was needed. In the end we used the <param> and <embed> tags to pass variables directly, rather than using the Actionscript of before to read an echo output from the PHP. GET variables worked fine this way.
This now means I have a complete communication between the SQL through PHP to Flash with a pretty clean retrieval from the database when I need it. The way in which Flash Media Server assigns instances to .swf visits means I can also handle the large majority of user authentication through the PHP/HTML that I’ve been developing so far, perhaps only to use further Actionscript authentication as additional verification.