bug-ddrescue
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bug-ddrescue] Suggestions to rename "logfile"


From: Antonio Diaz Diaz
Subject: Re: [Bug-ddrescue] Suggestions to rename "logfile"
Date: Wed, 17 Jun 2015 18:27:43 +0200
User-agent: Mozilla/5.0 (X11; U; Linux i586; en-US; rv:1.8.1.4) Gecko/20070601 SeaMonkey/1.1.2

Thanks to those who answered, and also to those who thought about it but perhaps couldn't find a good answer. :-)

I think the best suggestion is 'blockfile', proposed by Martin Bittermann. All the others either don't make sense for some use of the file, or are too long.

After all, the logfile is mostly described in the manual as a "list of data blocks".

I'll comment the other suggestions below:


Martin Bittermann wrote:
I'd rather keep the current name ;-)
Of course there is the pitfall for new users who think "I don't need a logfile, won't read it anyway..." but the documentation makes it pretty clear it is a bad idea to run ddrescue without logfile.

Another reason against keeping the current name is that ddrescue can produce two kinds of "real" logfiles (see --log-rates and --log-reads). These files really record events, while the current logfile just records the effects of those events (if any) and records nothing if used as a 'domain logfile'.


Robert Trevellyan wrote:
> mapfile
> rescuemap

The problem with these is that 'map' is only mentioned in the ddrescue documentation in relation with the '--test-mode', while 'block' is mentioned in about all places related to the logfile.


Adrien Cordonnier wrote:
If you change the logfile name, I suggest additional modifications at
the same time:

Well, if I change the logfile name then there will be no logfile, and users won't have the excuse that they don't need a logfile. ;-)


Felix Ehlermann wrote:
If the user is writing to an image file ddrescue should just automatically
create "<outfile>.resumestate" next to the outfile unless --no-resume-state
or a different location for the resume-state is specified.

In addition to creating unrequested files, which I think ddrescue should never do, this could create more problems than it solves. For example, how could ddrescue differentiate an existing "<outfile>.resumestate" created in a previous run from one stale file that the user forgot to delete?


This way unexperienced users will have a logfile if they don't read through
the manual.

Lets see how many inexperienced users dare not to use a 'blockfile'. ;-)


Ketil Froyn wrote:
In any case I think it's unnecessary to call a file something with "file".

I find it self-documenting. For example, which of the following two commands find you easier to understand?

  ddrescue [options] in out [sectormap]
  ddrescue [options] infile outfile [blockfile]


That's all. Further comments are welcome.


Best regards,
Antonio.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]