Issues on Umit while identifying paths in a portable manner

["Adriano Monteiro" <py.adriano () gmail com>]

Hi Folks,

After sometime trying to hold the whole world by myself, trying to
develop a "one size fits all" approach to the localization of the umit
files, I figured out that it shouldn't be as such. I got many reports
everywhere from users stating that they were having problems with the
configuration file location, and also having problems whit old
configuration files. Those more experienced with umit, found out that
simply removing ~/.umit should solve most of these problems.
Unfortunatelly, a regular user won't figure that out so easily.

Yesterday, while talking to Fyodor about this issues, I realized that
I should avoid the "one code base fits all os" approach, and hardcode
some paths, leaving the rest of the work to package maintainers and
installation scripts, instead of trying to find the paths. But, I
would like to know from you your thoughts about this approach, and how
you guys have once managed to have it fixed.

Another problem that exists on Umit, and probably exists on most open
source projects, is that some configuration files may get old and out
of date. Installing new versions of Umit would require the update or
replacement of the old files. The last approach is the better one for
developers: there is no such thing as "rm -rf ~/.umit" and recreate it
again using the new configuration templates. But,  using this refined
technique, users will lost their target list, options added to
options.xml, profiles, umit preferences, and a lot more. The other
approach (updating the config files) is better for users, but can turn
into a nightmare to maintain, mainly after several updates.

Fyodor suggested me that I should replace the untouched files, and
keep the modified ones, in order to avoid users from loosing their
data. But also, this approach won't work if a new version of Umit
requires a differently structured options.xml, as we won't be able to
replace it if user has modified it.

My first thought was about upgrading the files, but that really seens
to be a pain. What do you think about these issues?

BTW, I'll be working on them this week, as we need to have these
problems fixed asap to get Umit to be sent integrated to Nmap on the
next big release.

Hope to hear from you soon,


Take care

-- 
Adriano Monteiro Marques

http://adriano-marques.blogspot.com
http://umit.sourceforge.net
py.adriano () gmail com

"Don't stay in bed, unless you can make money in bed." - George Burns

_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://SecLists.Org