Domain: aillon.org
Stories and comments across the archive that link to aillon.org.
Comments · 5
-
Re:wireless support on linux is horrible
Ndiswrapper is unfortunately a way to work around what is to all intents and purposes broken hardware. You were cheated when you bought that thing. It's always going to be an administration pain in the ass until the manufacturer co-operates with the developers by providing specs. I know you probably don't want to hear that having spent your money (and time) on whatever that card is, but my advice would be to sell it and get a fully supported one. (I actually had to do exactly the same thing).
W.r.t. NetworkManager in FedoraCore4 the developer (Christopher Aillon) says himself that the current incarnation sucks and he's released a new, better version that you might be interested in trying out. It's working like a champ for me, and I had resorted to doing all my wireless configuration on the commandline in FC4.
Hope you get your situation sorted out. I feel your pain, but if you spend another $40 on a card that's supported, use OpenVPN on the Linksys router (flash the router with OpenWRT, it's simple and gives you real encryption as opposed to the lame-ass WPA which is crackable), then you'll be in clover.
-
Re:Where is the source code?
It's in CVS, but there's a warning that one should get the source package instead—the source package that is not available. Apparently, others noticed this as well.
-
Re:Can we start linking the original articles?
But if they'd linked to Chris Ailon's blog, people might have noticed this update:
Update 20050522 19:25:24 -0500: the slashdot article is misleading.That links to a blog post where he rebuts the slashdot article that supposedly represents his views. ^_^
-
Bashing?
Doesn't look like Mr. aillon was bashing them based on his response
-
Re:Why not re-examine http?
>Pipelining already works, most clients just don't use it.
Actually, there are lots of compatibility issues do to how some servers handle it. See here
My proposal is to create a new HTTP, that in order to claim support... you must support it.
> Nearly all web servers already support gzip. HTTP 1.1 supports arbitrary compression protocols.
> Any server/client can add 7zip support right now, just put it in the "accept" header.
Only a handful implement it. It's very under used. CPU is cheaper than bandwidth at this point. 7-zip tends to be faster, and better compression.
This is why http 1.1 sucks right now... because nobody takes advantage of the ability to get performance, because they are afraid to break things.
We need a protocol that is designed for performance. Transfering plain data is archaic and unnecessary....
we can also just go back to plain-email... aka pen and paper. But is that efficient?
XML isn't the problem... it's the pipeline it uses. HTTP needs to branch a 2.0, with strict standards to adhere to. It needs to be geared towards performance TODAY, not for compatibility with 1995 webservers.