Facebook is launching an ActivityPub service to compete with Mastodon. There’s a lot of talk in the webs about locking them out. Big words but not much power behind them. They will, in an instant, have millions more users than the existing ActivityPub services will, so the question really is — who’s being locked out? It was never a good bet that the architecture of ActivityPub would somehow be able to resist Silicon Valley-scale social networks. That doesn’t mean there are no answers, just that bluster isn’t one of them. You have to think. #
7 Dec
Scripting News Annex 
Posted by Kevin C. Tofel on December 7, 2006 at 11:08 am
Dave, CinemaNow offers HDNet content but I believe it’s in WMV-HD format, so no Mac. I thought at one time they partnered with a video hosting company for streaming Space Shuttle coverage, but I can’t recall. Best bet: hit up Collette Carey at HDNet directly = ccarey@hd.net. She’s always been very helpful to me in the past with HDNet questions.
Posted by Wes Felter on December 7, 2006 at 12:29 pm
You can buy some HDNet shows on disc, but playing them might not be easy.
http://hdnet-store.stores.yahoo.net/hddvds.html
http://hdnet-store.stores.yahoo.net/wimehddv.html
Posted by Fred Zelders on December 7, 2006 at 1:37 pm
My opinion: UniveRSS is a horrible thing!
I wonder how Microsoft could ever have been thinking this is something users were waiting for. (http://www.fredscapes.nl/2006/12/01.html#a1275)
It’s a pitty they put effort in creating this ‘thing’ instead of putting more effort in solving the problem they created by deciding that Microsoft (IE7/Windows Vista) only will download podcasts or other enclosures larger than 15MBenclosures from servers that support partial downloads with the HTTP Range header. (http://www.fredscapes.nl/2006/12/07.html#a1282)
Posted by Hieronymous Coward on December 7, 2006 at 4:19 pm
UniveRSS isn’t from Microsoft (it’s from a Swiss developer, but linked to from microsoft.com as a developer sample), and it isn’t for Vista only. It runs on Windows XP just fine if .NET 3 is installed on it.
Posted by Ahsan on December 7, 2006 at 8:06 pm
With http 1.0, if a download was interrupted, you would have to start all over again; you couldn’t just resume downloading from the point of interruption.
Http RANGE, which is part of the W3 Spec for HTTP 1.1 (and which was linked to by from the post), enables partial downloads so that only the parts that haven’t been downloaded will be requested from the server.
This has been around for at least six years.
You wrote: “It surely seems Microsoft could handle the buffering without requiring a special feature from the server. How much thought did they put into this?”
Well, clearly MS didn’t put much thought into it, because they used a feature that’s been available for years, rather than trying to reinvent partial downloads.
Posted by Scott Fletcher on December 7, 2006 at 9:07 pm
Ditto on the utter eye-porn of UniveRSS, though you haven’t really suffered until you try it with a laptop track pad. You WANT to be able to gesture through the space, but you are forced to manipulate the clunky 3D space by dragging a wet sponge through the myriad of meaningless levels of obstacles just to get to a… wait for it… a text list of news items. Lesson re-learned: “It’s about meaningful density, stupid.”
Posted by Fred Zelders on December 8, 2006 at 12:03 am
@Hieronymous: Where can we find this additional information?
The article on http://www.microsoft.com/emea/msdn/thepanel/featured/universs.aspx does not say anything about this and there are no links to sources giving this additional information.
So it seems to me as if Microsoft is claiming this development itself.
Posted by Scott Trotter on December 8, 2006 at 12:21 am
Regarding the 15MB enclosure limit in Microsoft’s RSS Platform, this post begs the question: have they actually encountered this situation “in the wild,” or is it just an FYI. If it’s true as they say that most servers support Range operations then this is no big deal. Far from being a “special feature,” the Range header is an integral part of HTTP 1.1–although the spec implies that support for byte range operations is optional.
The other unanswered question raised by Microsoft’s post concerns the plausibility of the Denial-of-Service attack that they are defending against. In my opinion, regardless of whether they are reacting to an actual DOS attack vector, or proactively preventing a potential attack, it seems like a reasonable tradeoff to require that large enclosures be served from range-capable servers.
And one more thing…If such a DOS exploit does in fact exist, or is at least plausible, it seems reasonable to ask other aggregator vendors how well their products would resist such an attack.
Regards,
–scott–
Posted by Hieronymous Coward on December 8, 2006 at 12:31 am
> Microsoft is claiming this development itself.
Well, according to http://www.futurecom.ch/de/references/projects-de.asp they did it for Microsoft EMEA. As of now, they’re hosting the ClickOnce file too.
The source is available from download.microsoft.com, though — under something called the MS Permissive License.
http://www.microsoft.com/emea/msdn/thepanel/default.aspx is quite clear these are just dev samples: “Each featured project provides the applications as well as the source code and explanation in order to help you leverage the new technologies and build better user experiences.” RSS readers are Microsoft’s favorite dev sample app these days :-\, one ships with Visual C# Express IIRC.
Posted by Michael Maclean on December 8, 2006 at 1:32 am
AFAIK pretty much every modern HTTP server (since about 1997 or so) supports HTTP Range requests. It’s used for resuming downloads, mainly. It’s the same feature that Firefox uses if you pause and then resume a download. Also, GetRight and things like that rely on it.
Posted by Ross on December 8, 2006 at 2:35 am
Microsoft say *should*, and this is probably for your own good. If the download gets aborted with 50bytes to go – do you want them to *have* to download all 15+Mb because the server doesn’t support range requests?
Range just allows them to request the remainder of the file… http://www.vbip.com/winsock/winsock_http_08_01.asp
Posted by Eran Sandler on December 8, 2006 at 3:03 am
As far as I know, the 15Mb limit (without the RANGE http support) comes not from the new RSS framework, but from a service in Windows called BITS (Background Intelligent Transfer Service – or something like that) which is in charge of downloading things (like updates from update.microsoft.com) without hogging your bandwidth.
If your computer is idle and you are not working on it, it will download faster and if you are working it will throttle itself.
Niall Kennedy also wrote about it here: http://www.niallkennedy.com/blog/archives/2006/12/ie7-windows-rss-enclosures.html
I suspect it didn’t cross their minds when they worked on it because they wanted to reuse BITS since it is relatively a good idea for doing so. The limitation was probably discovered later on when someone that has an HTTP server that does not support RANGE had a large enclosure.