BTW, the thing you quoted must be terrible! For the security. > --- src/skipstone.in > if [ -f ./skipstone-bin ]; then > exec ./skipstone-bin $@ > else Welcome, virus! (or just a thoughtless error) $ cat > skipstone-bin echo Hi i am a virus $ chmod a+x skipstone-bin $ skipstone Hi i am a virus $ rpm -qf $(which skipstone) skipstone-1.0.0-alt3 $ I think "exec ./skipstone-bin" shouldn't be there. And IMO updates for the stable branches are desirable.
I considered removing that but frankly didn't really bother. It's obviously crafted for running skipstone from build tree. If one has arbitrary code in locations which might be cwd when running stuff like gecko-based browsers, one usually has two _other_ problems racing for insecurity championship: - buggy gecko (plus NPAPI plugins like flash) - the path by which that code appeared on the system in the first place OK, I'll fix it -- but if you're really interested, I'd gladly hand the package so you could (co)maintain it :-)
Fixed in 1.0.0-alt4. I won't do updates though, it's not on my "actively supported" package list (there was one published in devel@ shortly before Server 4.0.0 release) and there are really lots of more important things deserving attention in probably each one of 200 packages I currently (co)maintain...
Thanks for the fix! (In reply to comment #1) > - the path by which that code appeared on the system in the first place Perhaps, by lack of attention: someone sent you a .tar, and it was there. > OK, I'll fix it -- but if you're really interested, I'd gladly hand the package so you could (co)maintain it :-) No, I'm not quite interested, I just reported problems I happened to notice, because I thought it would be better to report any problem I notice than to keep silent.
(In reply to comment #3) > > - the path by which that code appeared on the system in the first place > Perhaps, by lack of attention: someone sent you a .tar, and it was there. Then the problem is between the chair and the keyboard as it happens ;-) > I just reported problems I happened to notice, because I thought it > would be better to report any problem I notice than to keep silent. Уважаю.
[JT] (In reply to comment #4) > (In reply to comment #3) > > > - the path by which that code appeared on the system in the first place > > Perhaps, by lack of attention: someone sent you a .tar, and it was there. > Then the problem is between the chair and the keyboard as it happens ;-) Not completely true: a normal UNIX user shouldn't expect that executables from the working directory are executed. This feature was perhaps appropriate for skipstone developers (who understood the internals), but completely inappropriate for any other user, who just doesn't think about this possibility when working.
