Friday, September 28, 2007

Using SSL/HTTPS for Gmail and Google Apps

Today, I was thinking about Google Apps and wondering whether it used SSL to encrypt email. Now the Gmail POP/SMTP facilities use encryption for any data transferred between the mail client and Google’s mail servers, so I would have thought it natural that SSL was used to encrypt any email sent via the web interface as well. As it turns out though, that isn’t the case; firing up my trusty copy of Fiddler, I sent an email to myself using the Google Apps web interface and monitored the resulting HTTP exchange. This is the result (this is the POST request transcript - the relevant bits are in bold).

The same applies to Gmail itself. I know for most people this is a non-issue as email is generally insecure (as it gets transmitted unencrypted from mail server to mail server). I was thinking about local malicious users or sysadmins viewing the email I am sending or receiving. There are quite a few cases I can think of when I’d like to know local sysadmins can’t read my mail: a public internet cafe (in China for example, or another country free speech not a given), a corporate VPN, a non-encrypted (or WEP-encrypted) WiFi hotspot, etc… 

With Gmail, the solution is to access Gmail via https://mail.google.com rather than https://mail.google.com. In this case, a secure session is maintained throughout and all communication between Gmail and the browser is encrypted.

With Google Apps, the solution is to log in using a URL like this:


However, I couldn’t get https://mail.mydomain.com to forward to the above, which is a shame as it’s my main gateway to Google Apps (and others’ too I suspect - and therefore possibly a minor security issue for sysadmins administering domains using Google Apps…).

posted by Nick at 3:50 pm - filed in Uncategorized  

Saturday, September 22, 2007

Dvorak touchtyping - week 3

Well, better but still far from perfect. I broke the 50WPM barrier on the home row this week, at 98% accuracy too, but I also let up on the regular practise a little (I must have practised two or three times this week for an hour or so at a time, down from 6 the two first weeks at an hour and a half or so each). I’m still nowhere near as good at any of the other exercises though, and I still catch myself looking at the keyboard quite regularly.

Interestingly, I noticed myself being rather verbose in some of by emails this week, presumably as a subconscious reaction to the joy of finally being able to type at something approaching reasonable speeds. I’m also starting to get a feel for the positions of the keys other than the home row without having to look at them too, although I’m noticing that I frequently get the ‘e’ and the ‘o’ mixed up. Another interesting thing is that I seem to be in a different mindset when typing freely and doing the exercises; for some reason, some of the keys I remember well typing freely don’t seem to make it over to the exercises - I guess the different visuals (the terminal for dvorakng and Windows for daytime work) disassociate the two in my mind, which is a shame…

posted by Nick at 2:48 pm - filed in dvorak  

Sunday, September 16, 2007

Dvorak touchtyping - week 2

Well, there has been progress, but it has been slow. I’m on around 40-45WPM on the home row (which accounts for 70% of keys in English), but my error rate is still high at round 95-96%. The WPM drops dramatically once I leave the home row though, at around 30-35WPM for anything involving the top row (I haven’t even tried any exercises involving the bottom row). At least the frustration of typing letters at what seemed like 2WPM has abated, so I can stop resorting to using the mouse as much as possible and actually write emails without crying…

posted by Nick at 2:47 pm - filed in dvorak  

Saturday, September 8, 2007

Dvorak touchtyping - week 1

So I’m going to try to learn how to touch-type again; this will be my second attempt, the first being about three months ago. The difference is that this time I’m switching to Dvorak cold-turkey; the last time I had also tried to learn Dvorak, but I think that the constant swiching between Dvorak and Qwerty was what ended being discouraging, as progress was very slow.

It’s been just over a week now, and the first few days were among the most frustrating I remember having in a long time. The double-switch of Dvorak and touch-typing proved very difficult. On the first day, it took me almost 20 mins to write a 5-line email sheerly due to my typing speed; it got to the stage that I was not emailing people back because the frustration at not being able to type was so great. It’s not so bad now; I can type at a reasonable speed, I’m guessing around 10-20 WPM, which is still horrendously slow but at least workable. I had set myself an ambitious goal of 80 WPM in a month, which right now doesn’t seem achievable, especially given my still very high error rate - still worth trying though…

posted by Nick at 12:49 am - filed in dvorak