An interesting approach to storage.  but if you have such a cool idea backed by good tech…why call yourself “Nutanix?”  i chortle whenever i say the name out loud.  maybe that’s just me.


very enlightening, easy to comprehend, and sufficiently able to blow minds…Van Jacobson explains what the future of the internet could be.

/8’s and the United States military

here’s an interesting list of all the organizations in the world that own /8 ipv4 netblocks:


interesting to see the different companies on this list…the ones that were there “in the beginning” kinda thing.

also, the US DoD owns not 1, not 2, but 12 /8’s.  TWELVE.  That’s about 200 million publicly routable internet addresses.  they may not show up in the public global routing table…but they are around…and likely being used somewhere.

makes sense.

makes sense.

Windows Afterthought

Windows is the ultimate afterthought Operating System.

Anything useful in Windows, was added as an afterthought.

I tried writing a batch script and needed a “sleep” command.  It wasn’t until Windows 7 that “timeout.exe” was added as a native command to accomplish what I needed to do.  Before this, intelligent folks were actually shoving things like this in the middle of their .bat scripts if they wanted to pause for 5 minutes:

ping -n 301 >nul 2>&1

Apparently there is a “sleep.exe” or something, but it’s a separate download or some baloney.

the next release of Windows should be called Windows Afterthought™.

unlawful wiretapping

in the US, and probably other complicit nations, the NSA is unlawfully intercepting internet traffic and indexing it.   yep, all of it.

IXmaps is a site that attempts to graphically represent where your data goes on the internet (in North America), and at what junctions it is being stored and catalogued.

But wait, they couldn’t possibly have the means to do that…that’s gotta be exabytes of data to sift through…right?!  hey, no problem.  the private sector is hard at work at providing the means.



the future of the internet

Van Jacobson…made the internet possible back in the 70s and 80s with his tcp flow-control algorithms and header compression techniques. I first learned about him when working with dial-up modems, PPP and building RADIUS servers at some ISPs in the late 90s.

Example 10-3. RADIUS configuration for direct connect accounts

Anna    Password = UNIX-PW
    Service-Type = Framed,
    Framed-Protocol = PPP,
    Framed-IP-Address =,
    Framed-IP-Netmask =,
    Framed-Routing = None,
    Framed-MTU = 1500,
    Framed-Compression = Van-Jacobson-TCP-Header

Now he’s onto some next level stuff at PARC…and i love it.  clear 45min in your day to watch the first video of his talk.  This will be the future of the internet and networking architecture.  if this is embraced with the same openness of the original networking specifications, it will go far and profoundly change the internet experience forever.  i only hope corporate interests and greed don’t get in the way.


Other links:




this is my IE9 user-agent string:
"Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; Trident/5.0)" 
this is an IE7 user-agent string:
"Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 6.0; SLCC1; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR 3.0.30729)"
and yet, youtube (Google) manages to screw this up.  i wonder how they come to this conclusion.

this is my IE9 user-agent string:

"Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; Trident/5.0)" 

this is an IE7 user-agent string:

"Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 6.0; SLCC1; .NET CLR 2.0.50727; .NET CLR 3.5.30729; .NET CLR 3.0.30729)"

and yet, youtube (Google) manages to screw this up.  i wonder how they come to this conclusion.

Google Chrome is all hip with error messages.

Google Chrome is all hip with error messages.

subdomains ad infinitum

while watching this awesome youtube vid (because Def Leppard is totally awesome, shaddup), I came across this ridiculous subdomain in my squid logs - and yes I regularly “tail -f” my access.log, ok?


Every subdomain in that hostname must mean something very specific.  o-o?  v17?  I wonder if it’s dynamically generated.  probably, eh?  i wouldn’t be surprised if there’s some GSLB at play here.

So what is that?  A sub-sub-sub-sub-sub-subdomain?  I would love to see google’s zonefiles and overall DNS design and architecture, because this intrigues me.  Unfortunately, google won’t allow me to initiate a zone transfer.  funny, that.

DiG 9.7.3 <<>> @ns2.google.com axfr youtube.com
; (1 server found)
;; global options: +cmd
; Transfer failed.
robots.txt and the walled garden

facebook is the quintessential walled garden.  the internet thrived in the beginning because of it’s open nature.  then facebook came along and wanted your personal information.  statements like this first paragraph by Zuckerberg are strangely counterintuitive.  facebook didn’t make the world more open and connected.  the internet did that already.  if anything, facebook has done a great job at caging off a large chunk of the internet and asking for more personal information from us than anything before or since - and making a fortune with it.  facebook maintains one of the more comprehensive robots.txt that i’ve ever seen…presumably to keep out the spiders that would otherwise crawl facebook and make it, y’know, more open.

think in perl.

perl is a masochists scripting language.
sure, it can do wonderful things in the realm of text-parsing, but navigating the syntax of perl is akin to white water rafting. with no raft. in lava.

random perl error of the day:

Useless use of a variable in void context
lynx to the rescue

discovered today that using “lynx —dump” over “curl -o” or “wget -O” to save text from a website is preferable if you’re planning to do some parsing on the output. GNU sed doesn’t produce expected results with the curl or wget output. something to do with lynx writing linefeeds versus curl/wget writing carriage returns:

lynx.txt wget.txt differ: byte 61, line 1 is 12 ^J 15 ^M
Kobo browser user-agent

So, facebook.com thinks my Kobo Touch is an Android device, and renders accordingly. It even says “Using Android? get the FB app for Android!” (paraphrasing). It does this because the browser on the Kobo sends this as it’s user-agent string:


“Android” is listed ahead of the actual browser platform - AppleWebkit. User-agent strings are read left-to-right by web servers. Afaik, the Kobo Touch doesn’t run the Android OS, but it does run Linux.

The history of browser user-agents is gross:


Packet INternet Groper.

Packet INternet Groper.