Web search privacy, Skein/Threefish,Cloud hosting (ep.io), VPS-servers, 2FA, Google Sites / JavaScript fail

Post date: Apr 22, 2012 4:19:02 PM

  • Checked out duckduckgo.com's and startpage.com's Privacy & ToS documentation in detail. DDG seems to now use AWS as global front-end CDN, which uses advanced caching to deliver search results faster. Cache hits are now very fast, but misses are still quite slow. Unfortunately I do quite many searches which are sure cache misses, by using 3-5 keywords and even with those using 1-2 excluding keys.
  • Read lot of documentation about Skein hash and Threefish cipher, downloaded Python PySkein and tried it. Yep, looks promising. I think I'll opt for Skein/Threefish when it's ready and proven. Oh well, only time will prove chiphers and hashing.
  • Watched some of my test projects go down, which were hosted with ep.io. - Ouch! Cloud hosting companies come and go. That's why choosing IaaS could be safest option, with proper backups. If you have built everything on proprietary PaaS you might be in deep trouble when hosting company simply quits. Situation is even worse, if you're small SaaS customer. There is absolutely nothing you can do, except find new SaaS service / provider and which that your data won't be lost.
  • Server tuning, now I have my test server running in Helsinki and configured to handle all my needs, I did drop all other VPS servers for a while. I also had test servers in Scranton,PA and London,UK for a while. I did some performance testing and cloud management console testing. Those did seem to work fine. I just don't need servers in those geographic areas right now.
  • Email security: This is a nice article. I really do recommend 2FA for everyone. There is only one big mixup, even if email access would be secure, it doesn't make email transport secure. So it's still important not to send anything confidential using email.
  • I just found out that Google Sites doesn't provide any links to older news / blog postings, unless pages are browsed using JavaScript enabled browser / application. I tried to wget my whole site, to back it up and started to wonder, why I got so few pages. After a while, I noticed that most of my blog entries were missing. Then I started to look what's causing that. Yes, reason was clear.
    • Unless JavaScript is enabled, older blog entries can't be accessed, without knowing exact url or adding offset parameter to blog-url.
    • I solved it by creating new blog-offset-links page. I really wonder if Google couldn't do any better? It seems that this search giant doesn't know anything about web-publishing?
    • It's quite unfortunate to see many sites, where navigation doesn't work properly or at all without JavaScript.
    • These navigation flaws could really hinder badly indexing of many many web sites by different search services.

I have a few longer topics in my mind, but it seems that today isn't the day to write about those. - I'll be back with those topics. Now I have to listen a few security now's.