On Wed, 23 May 2007 21:17:50 -0400 Douglas Allan Tutty <firstname.lastname@example.org> wrote: > If I gpg a tarball today with whatever algorithm is current, in 10 > years that algorithm may be long cracked. Will the gpg authors keep > support for it? Perhaps. Just yesterday I had a similar problem which pretty much combined all the troubles mentioned in this thread. It turnmed out I got lucky seven times in a row. Here goes the story: I tried to fire up my ancient parallel-port Iomega ZIP drive to rescue some data from old disks. I was happy to discover that I had been far-sighted enough to store the drive together with the disks in a plastic bag (1). Luckily, the current 2.6. kernels still have the ppa driver (2), and luckily modern desktop PCs still have parallel printer ports (3). The first problem was to get a reliable electrical connection through the dust-ridden ancient connectors. After unplugging and re-plugging the cable several times I got it to work (4). The next problem was to make the ZIP recognize the disks. I remember from back when I used it regularly the drive had started to play funny and make a lot of clicking noises, probably due to dirty heads. Obviously this hadn't gotten better during the past five years. It took me several minutes of kicking the drive and loading and unloading kernel drivers to get each disk to mount (5). Having finally mounted the disks, I discovered that they contained PGP-encrypted tarballs. When I tried to decrypt them, GPG told me that they were encrypted using IDEA which wasn't supported any more due to patent issues. Fortunately I was able to download an IDEA module from a Danish server that ran fine with GPG (6). One more problem I would have had is that I might have forgotten my old passphrase. Fortunately I've been using one and the same passphrase for over 10 years now, which in itself isn't so good, so I could finally decrypt all my data, re-encrypt it with GPG and commit it to my current backup scheme (7). --D.
Description: PGP signature