Robert A. Uhl

Backup tip

Way back in 1994, Rob Horn gave an excellent tip: randomly restore one file from backup every week. If you have data which you care about (and you do, whether you know it or not), it needs to be backed up. And if it’s being backed up, you need to test that backup system regularly to ensure that you’ve not forgotten something or been bitten somehow. Restoring one random file per week is a good practise to follow. Read more →

Backup tip

Way back in 1994, Rob Horn gave an excellent tip: randomly restore one file from backup every week. If you have data which you care about (and you do, whether you know it or not), it needs to be backed up. And if it’s being backed up, you need to test that backup system regularly to ensure that you’ve not forgotten something or been bitten somehow. Restoring one random file per week is a good practise to follow. Read more →

Predicting the future

Back in 1989 Alan Kay (the creator of SmallTalk) gave a talk on predicting the future to the 20th anniversary meeting of the Stanford Computer Forum. It has many insights, but my favourite is that we predict evolutionary changes, not revolutionary ones. Read more →

Predicting the future

Back in 1989 Alan Kay (the creator of SmallTalk) gave a talk on predicting the future to the 20th anniversary meeting of the Stanford Computer Forum. It has many insights, but my favourite is that we predict evolutionary changes, not revolutionary ones. Read more →

Unix makes computer science easy

I just found an article which details how Unix took a lot of difficult computer science stuff and made it easy for an end-user. Even in 1979, Unix had a lot of great capabilities. There are two problems: first, Unix hasn’t progressed very much since then; second, noöne else has even gotten this far (or at least, hasn’t gotten this far and succeeded). By now we should all be using Lisp machines; instead we’re all using technology that’s older than I am. Read more →

Unix makes computer science easy

I just found an article which details how Unix took a lot of difficult computer science stuff and made it easy for an end-user. Even in 1979, Unix had a lot of great capabilities. There are two problems: first, Unix hasn’t progressed very much since then; second, no-one else has even gotten this far (or at least, hasn’t gotten this far and succeeded). By now we should all be using Lisp machines; instead we’re all using technology that’s older than I am. Read more →

Effective emacs

Jacob Gabrielson wrote a nice blog post giving some tips for effective use of emacs. I’ve managed to reduce my emacs start-up time from six seconds to one, which ain’t too shabby. Read more →

Effective emacs

Jacob Gabrielson wrote a nice blog post giving some tips for effective use of emacs. I’ve managed to reduce my emacs start-up time from six seconds to one, which ain’t too shabby. Read more →

Off-the-Record Messaging

Typical cryptographic software provides assurances that a message was sent by a specific sender; this is generally what one wants. But what if one prefers plausible deniability? Off-the-record messaging provides encryption and authentication for the duration of a conversation, but after the conversation is completed anyone can forge messages within that conversation, thus one can deny that one ever said anything. This could be useful for whistleblowers, those living in oppressive regimes and so forth. Read more →

Off-the-Record Messaging

Typical cryptographic software provides assurances that a message was sent by a specific sender; this is generally what one wants. But what if one prefers plausible deniability? Off-the-record messaging provides encryption and authentication for the duration of a conversation, but after the conversation is completed anyone can forge messages within that conversation, thus one can deny that one ever said anything. This could be useful for whistleblowers, those living in oppressive regimes and so forth. Read more →