Yesterday I was fooling around with python and trying to serve
some XML to a Cisco 7960 I have. I spent a lot of time getting
my web server set up to serve a dynamic XML page which I could pop
my phone book information into and when I finally got it done it
wouldn’t work. Searching online I found out that the latest
firmwares from Cisco, 9.*, has a couple of bugs in it dealing with
XML. Unfortunately I don’t remember where I found this information
but I can give out the sparks notes on what I saw.
When I was served the page from a web server, if it responded
HTTP/1.0 OK, no good, the phone would only accept if it was
HTTP/1.1 . I was getting some cryptic 404/BTXML errors but I could see
that the files were fine by loading it on to the production
server as a static file. Sure enough the phone was able to parse
it, so I actually moved the dynamic parts to the production
server to see if it would work as well.
After moving it, the phone can read it fine but I am getting some
errors that look like encoding problems. I am going to have to
look into this some more later, but here is what I have tried so
far as well as some notes about my setup which I will release
after I get all the problems worked out.
Set the encoding in the \<xml …\> tag that prefixes the generated
XML for the phone book.
I am using minidom.Document from python to write the XML files
and I have noticed that the ends of the strings are mac like, but
I don’t think this is a problem since apache served the downloaded
file to the phone just fine.
This has potential to be very cool, but until it gets worked out,
When you first turn on your computer the BIOS kicks in and searches for instructions on your master boot record or MBR. From there it is able to locate the boot loader which in turn figures out which kernel to run and then loads it into memory and that is the point you start actually running your flavor of UNIX. At this stage your computer is just running a kernel and has no interactivity or running services.
I have been playing around with http://wolframalpha.com lately and it has a lot of power behind it. I mostly use it for doing math stuff when I don’t have a calculator in front of me and it can even do some stuff I cant do on my ti-89 like Laplace transforms. When it returns a result it often time gives a comparison of other things that have the same unit. As an example, I asked it about 15 gigs and it immediately figured out what I was talking about and told me how big that was in relation to a blueray disk,60%, as well as how long it would download at different speeds. All this information got me thinking about what kind of questions it can answer and what it cant. The questions it cant answer fall into 3 main categories and while I could never list all the things it can answer I can easily list what kind of questions I want it to answer.
“average size of an iphone app in the apple app store” - while this query would take a lot of work to acquire the requisite data I think it could be answered one day by a computer but it is mostly a issue with data mining apples app store which they probably wouldn’t like. This is just a data issue and it could be worked out in time since right now it can answer similar questions like “average income of a person”
“meaning of life” - this is an obvious one that cant be answered, it falls into the category of being incalculable. When a question is asked that there is no data to the next best thing that can be done is a interpretation which is exactly what Wolframalpha does. In this case its knows the answer to the ultimate question of life the universe and everything but not the question itself so it displays that. 42.
“what is the best way to …” - Wolfram was designed to answer questions with facts and not opinions so this type of question should never really be handled
Wolframalpha is very cool right now and I hope that they can expand the focus of their domain as it is mostly geared to math type questions at the moment. This is understandable since their expertise has been math software and since that is where knowledge is the most quantitative. While this article was mostly to cover what it cant do, you should check out the examples page on the Wolframalpha site to see what it excels at and hopefully by the time you read this article it will be out of date.
My filesyserver at home is a opensolaris box using raidz under zfs and right now it is having some problems. Recently I had to replace a drive that did not fail but it was making some clicking noises. After repplacing it the first resilver left me with an error about a snapshot that i didn’t need being damaged. okay i can live with that, so i just need to delete the snapshot and bring the pools redundancy back up. well, it reslivered everything, but it didn’t seem like the pool was running completely. i cleared the error and it started to resilver again. same thing happened again but now i am getting a warning for something else <0x5c>:<0x64d62> and now the resilver is going past 100% and copying more then the amount of data i have. some posts online suggest disabling snapshots because they are keeping it from completing. didnt work so i reboot the computer and when it comes back online it starts resilvering automatically. again it goes to 100% and keeps working but this time it takes two files with it, the one before and <0x2b3>:<0x64d62>. looking to another post i see to upgrade opensolaris and try again, i do that, but now i am starting to get worried about playing fast and loose with my data. i still hvent fixed the problem but i want to have some assurance i am not going to make it worse.
Just sent to PROD a copy of my personal site. PROD is a 20$ vps currently hosting two websites and it is running open source software which is important. thanks to jamescarnley[http://jamescarnley.com] for doing some beta testing.
subscribe via RSS