SELinux appears to be an overengineered, overflexible, poorly deployed, poorly documented, jargon wrapped pieces of software with the dubious distinction of having taken the obscurity of Windows' security model to Linux, with a vengeance. In fact, its design reminds me of the good ol' Motif apps (maybe this IS a Unix philsophy argh), which popped up in all their blue black and square white glory and could become OH-SO-BEAUTIFUL, you just had to write an app-default file for them: remember the easy syntax, the breezy deployment (just copy it over in 15 different locations, with slight changes in capitalization and name hyphenation), and joyful surprise of the oh-so-totally-unpredictable effects that could be achieved. Users that turned to Windows and its GUI where called stupid lazy lusers then. The type of supporting arguments that are being used runs close to the ones of yore, just replace lusers with admin.
Another (forgotten) lessons from the past would be sendmail.cf which hardworking admins were supposed to write from scratch ($1 ). Now sendmail is much ridiculed, even after a sane configuration system makes it possible to never look at sendmail.cf. xauth and its worthy is another fond memory of something that could neither be understood or made to work (xhost +) until it was made fairly transparent.
ssh and X11 tunnelling, which used to require tens of arcane cli commands and a small chart of the involved IP adresses... Same tune, lazy-admins-i-d-fire-them-if-they-worked-for-me.
Hint: when the vast majority of the perspective users of something DO NOT WANT to even hear about it, then it is not them that should be faulted.
Greetings.
Had this debate popped up two years ago (one year ago, even) I'd have been more than supportive to the idea of keeping strong crypto etc...
After DMCA, cryptografic content protection, and other depressing news of this sort, I am not so sure that the advantages of keeping crypto in the market really outweigh the disadvantages.
It looks like all the sound and the fury is about a technology way too cumbersome for common Joes to use - with all the gain going to the corporate content behhemot du jour. So, why bother?
Consider that all the copy protection plans that will wrestle from us (the public) the control of our hardware, software and of electronically published media in general, hinge critically over the wide availability of strong encryption software and skills.
Slashdot Mirror
SELinux appears to be an overengineered, overflexible, poorly deployed, poorly documented, jargon wrapped pieces of software with the dubious distinction of having taken the obscurity of Windows' security model to Linux, with a vengeance.
In fact, its design reminds me of the good ol' Motif apps (maybe this IS a Unix philsophy argh), which popped up in all their blue black and square white glory and could become OH-SO-BEAUTIFUL, you just had to write an app-default file for them: remember the easy syntax, the breezy deployment (just copy it over in 15 different locations, with slight changes in capitalization and name hyphenation), and joyful surprise of the oh-so-totally-unpredictable effects that could be achieved. Users that turned to Windows and its GUI where called stupid lazy lusers then. The type of supporting arguments that are being used runs close to the ones of yore, just replace lusers with admin.
Another (forgotten) lessons from the past would be sendmail.cf which hardworking admins were supposed to write from scratch ($1 ). Now sendmail is much ridiculed, even after a sane configuration system makes it possible to never look at sendmail.cf. xauth and its worthy is another fond memory of something that could neither be understood or made to work (xhost +) until it was made fairly transparent.
ssh and X11 tunnelling, which used to require tens of arcane cli commands and a small chart of the involved IP adresses... Same tune, lazy-admins-i-d-fire-them-if-they-worked-for-me.
Hint: when the vast majority of the perspective users of something DO NOT WANT to even hear about it, then it is not them that should be faulted.
Cheers,
alf
Had this debate popped up two years ago (one year ago, even) I'd have been more than supportive to the idea of keeping strong crypto etc...
After DMCA, cryptografic content protection, and other depressing news of this sort, I am not so sure that the advantages of keeping crypto in the market really outweigh the disadvantages.
It looks like all the sound and the fury is about a technology way too cumbersome for common Joes to use - with all the gain going to the corporate content behhemot du jour. So, why bother?
Consider that all the copy protection plans that will wrestle from us (the public) the control of our hardware, software and of electronically published media in general, hinge critically over the wide availability of strong encryption software and skills.
Cheers, alf
- Planck's black body radiative formula (come up with quantum mechanics in a week);
- Galileo dropping weights from the tower of Pisa;
- Gnu emacs.
Cheers, alf