mjashby wrote: ↑07 Dec 2019 11:31
...
If people don't make use of such facilities [provided backup facilities], even when they are provided, then I can't see a strong argument for individual commercial application developers spending development time on building and maintaining individual backup systems for anything other than 'zipping' copies of the essential data produced by their applications?
...
I can see this argument - I provoked this original sub-thread in the light of my perception that we get a lot of queries about people either recovering from a systems crash or transferring their FH environment onto a new computer. In these cases the root of the problem is that people have not made use of FH backup facilities (and as you point out, in some cases, not used the OS inbuilt facilities).
I suspect with increased broadband speeds and "capacity" more and more people will be getting "everything backed up to the cloud", which
possibly means in the case of the crash they can restore their entire system.
(My italics)
In the corporate world, with controlled "clean" configurations, a post crash full restore is probably easiest and quickest (which may be important). In the consumer world, where speed is not necessarily quite as important, I wonder whether post crash the best thing to do is to:
- rebuild a clean version of the OS,
- install clean versions of those applications (that the user still needs) to give a clean foundation and
- then restore user data.
The question then is whether "user data" (should) include user configuration data - or whether that negates the idea of creating a clean foundation!
Without constant back up to a NAS or the cloud, full systems backups can be time-consuming and laborious - and those sort of things don't get done! There is also an argument that your backup should be to an offline detachable media so that it cannot be attacked by malware, particularly ransomware (which if it can find your hard disk, it will also find and lock any attached backup storage!). For consumers, there is a lot to be said for a routine where say once a week you plug in a USB drive and within say half an hour you will have copied sufficient to enable you to do the sort of rebuild noted above.
I have spent considerable time on my environments (previously Windows 7, now Linux) ensuring that configuration files (and other details) are in paths that get backed up. That made changing OS a remarkably easy operation.
in terms of the time that system developers should spend, I am not sure how much it is a case of spending time developing backup routines as designing their systems such that configuration data is put (or copied) somewhere where it can be easily backed up (whilst being safe from numpties trying to directly edit it! Such things have to be, not just idiot proof but also, clever-proof).