Calendar was broken 2

Posted by dgtized Thu, 21 Jul 2005 08:28:00 GMT

I still need to upgrade to the super fancy version of the calender that was posted as a patch on typo.leetsoft.com, but it was loosely based on the calender system running on my version of typo. Unfortuneatly somewhere in my modification I had forgotten to include the critical piece of code that read: 

[""] * (self.class.civil(year,month,1).wday)

Which needed to be prepended to the array of dates in the month to ensure the calender started on the right day of the week for each month. Perhaps sometime next week I will dissassemble the patch that followed mine since it doesn’t seem to work as a patch set anymore. It would be nice to have a calender with AJAXy goodness and all that nonsense.

If anyone is interested I based all of my calendar code from a snipet that appeared on comp.lang.ruby:

calendar.rb
require 'date'
require 'enumerator'require ‘date’

require ‘enumerator’


class Date

  def days_in_month

    self.class.civil(year, month, -1).day

  end

  alias dim days_in_month

def weekdays
(1..dim).to_a
end
def cal_print
banner = ABBR_DAYNAMES.map{|d| d[0..1]}.join(‘ ’)
puts "#{MONTHNAMES[month]} #{year}".center(banner.size)
puts banner
wd = [‘ ’] * wday + weekdays
wd.each_slice(7) do |slice|
puts slice.map{|d| d.to_s.rjust(2)}.join(‘ ’)
end
end

end


month = (ARGV[0] || Date.today.mon).to_i

year  = (ARGV[1] || Date.today.year).to_i


Date.civil(year, month).cal_print

Perhaps the above comment about the missing line makes more sense in that context. Also hopefully someone determines a way to use syntax to auto-hilight code in typo.

On the weirdness of wget, dvd iso's and md5sums

Posted by dgtized Wed, 20 Jul 2005 21:50:00 GMT

So continuing with my issues in getting fedora core 3 installed on this server at work I decided it would be wise to try downloading the i386 dvd image to assist with the install. First I made the mistake of trying to let firefox download the image, apparetly firefox doesn’t like files over 2gb, it started counting negative sizes and negative rates, I guess there is an overflow in there somewhere.

So I decided to switch over and use a tried and true favorite, wget. So I downloaded it, burned it, and the burn failed. Startled I decided to check the md5sum’s on the iso image to ensure it was an uncorrupted image. Sure enough the image was corrupted, the md5sum didn’t match.

So I tried downloading the image again. Now mind you initially it took about half an hour to download each image at about 1.5megabytes a second from University of Southern Florida. Luckily I noticed that wuarchive had a local copy which allowed me to download it at about 8 – 9 megabytes a second across the local connection. I will now present the fruits of my labors:


[649 comstocl@jeff ~/Desktop]$ md5sum FC3*
cd776e133f39ac721246063ef030d7be  FC3-i368-curl.iso
7416173f3ba8e0e2e97101c85e796330  FC3-i386-DVD.iso
ae1f8b88e9d5905fb153d2f179f4cb5c  FC3-i386-DVD.iso.1
582533c1c28a7f4f474cc3d0fc5b8f28  FC3-i386-DVD.iso.2
4d01e89bf0a44e8a545e616429eafba5  FC3-i386-DVD.iso.3

[654 comstocl@jeff ~/Desktop]$ ls -l FC3*
-rw-r--r--  1 comstocl comstocl 2.3G Jul 20 15:44 FC3-i368-curl.iso
-rw-r--r--  1 comstocl comstocl 2.3G Jul 19 19:03 FC3-i386-DVD.iso
-rw-r--r--  1 comstocl comstocl 2.3G Jul 19 18:36 FC3-i386-DVD.iso.1
-rw-r--r--  1 comstocl comstocl 2.3G Jul 19 19:25 FC3-i386-DVD.iso.2
-rw-r--r--  1 comstocl comstocl 2.3G Jul 20 16:30 FC3-i386-DVD.iso.3

I tried to use curl once but clearly that didn’t help much. The intended md5sum for all of these is ca49964739f84848ca78fc03662272fb. All but the curl one was downloaded with wget, and as you can see each has a different md5sum. I am running on a 64bit machine, so I consider it possible that md5sum or wget is somehow broken as a result. But it is rather odd nonetheless. I’ll take a look through the bug reports at some point, maybe something will pop out. Clearly I am just not meant to have this iso image. I have downloaded and verified iso’s of FC3 before, though only of the x86_64 iso, so I’m not quite certain what is different now.


/">Valid CSS