As I hinted towards in last nights "From An Experts Point of View [via Oleg Tkachenko]" post I have been trying to put together the pieces of a project that is complimentary to Oleg's XLinq.NET project. The site is http://bLinqX.NET, and is based around del.icio.us Social Bookmarking, with a cronjob set to run a transformation process, polling the del.icio.us server at :15 and :45 past each hour for the tags "Linq", "XLinq", and "bLinqX" [used to tag content that is related to Linq and XLinq but not necessarily taggable as either.]
[NOTE: Another *HUGE* thanks to Uche for helping sort out the crontab issues I encountered this morning, and as such get this portion of the project working perfectly.]
At the moment you can access the result of each of these transformations at:
http://blinqx.net/linq.xml
http://blinqx.net/xlinq.xml
http://blinqx.net/blinqx.xml
So beyond lightening the load a bit on the del.icio.us server, what else is planned for this site?
There are a couple ideas I am playing around with [and I would certainly love to hear about any other ideas anyone might have] The most simple of these is to render the latest feeds as part of the front entry page. The feeds can also be used to extend the services offered by del.icio.us (e.g. add a ratings system, add extended commentary capabilities through a blog and/or wiki interface as well as a forum in which the content contained at each particular link can be expanded, discussed, sample code posted, rants, raves, etc... Using a combination of client and server side XSLT, using XMLHTTP as a workhorse to get and post xml between the client and server. etc..., etc..., etc...
The one thing I don't want to do is to go past this general premise of sharing bookmarked links and extending the capabilties of del.icio.us with features that is in no way recreating or competing against what del.icio.us itself offers the community. I also don't want to compete against what Oleg is putting together, which seems it will be focused more towards developing content that in turn would be bookmarked on del.icio.us and tagged with one of the above tags to then use this site to simply access the links directly, add commentary, rate the link, etc...
While I don't want to limit creativity, if it goes beyond this scope then my guess is that Oleg's XLinq.NET project would probably be the better place to focus these ideas, keeping this site lean and mean and able to act as a central place to access the latest and greatest Linq and XLinq related content, building tools and what have you that will enable such things as mentioned above without ever actually having to visit the site and instead either pulling down the XML feeds for client-side processing, interacting with the server via REST or SOAP to fine tune the data that is being sent, post comments regarding links using the Atom Publishing Protocol and accessing responses to such posts via the associated Atom feed, etc... etc... etc...
If you have an idea that fits into this space, GREAT! I'd love to hear about it. If its a great idea, but goes beyond this scope, Oleg already has a simple form for you to post your idea to him directly. Between these two sites, my guess is that if you have interest in Linq or XLinq, you will be well served by both.
Regarding the site data feeds
As mentioned these feeds are currently updated at :15 and :45 past each hour so setting your reader to poll these feeds twice an hour, 30 minutes apart, will coordinate nicely with this configuration. del.icio.us specifically requests[1] that you do not poll more than every 30 minutes. If it seems that the difference between the XML returned when polling twice is minimal, at best, I will probably reduce the feed updates to once an hour as to free up that bandwidth for someone else with greater bandwidth needs. If that happens, I will let you know so you can adjust your reader to poll the bLinqX.NET server once an hour if you have it set to two. If you happen to take notice that there are even less updates than would require hourly updates, please let me know as the more of a community effort we can make this the more we will all benefit from one another while keeping our bandwidth usage to the bare minimum required to have an effective tool at our disposal.
When I get some more time I will update the site with content, create a client-side transformation to render the data feeds within your browser when you visit the site directly, create Atom 1.0 feeds, add the blog and wiki engine, etc... With each major completion I will update this sites main feed to let you know.
In the mean time, enjoy!
---
[1] : See the bottom part of the about page, which states below the Feeds heading and just above the API heading [2] :
ยป Please do not poll any single RSS feed more often than every 30 minutes. RSS feeds are not updated more than twice an hour, and you will receive an error if you try to crawl more frequently.
[2] : With xml:id recently reaching a recommendation can we please start adding xml:id attributes, or even just id [technically speaking 'html:*/html:@id'] [@name is deprecated, use html:*/@html:id or xml:*/@xml:id and you'll be all the happier later on when this becomes the norm] attributes to our XHTML? Doing so will allow us to create client side tools which can create links that point EXACTLY to the content we are refering to using the typical hash + id that has been around since the beginning (well, at least #nameofanchortag has been around since the beginning] Thanks! :)