vdongen wrote: >>As I've pointed out elsewhere, patching old versions without changing >>the version number is so stupid it leaves me boggling. But I guess in >>future I'll write into advisories: "warning - your vendor may be such >>a >>moron that you can't tell whether you are vulnerable or not by the >>version number, so I advise building from source or switching to a >>vendor with a clue". > > I have to disagree on this, the way debian patches the current versions > of the stable distribution is a good thing in my opinion. > Instead of upgrading the software, they backport the fixes in the > current version. > This prevents getting new problems with compatibility and such when > inplementing new versions. > New versions of a certain package mosty require updates of other > packages and/or rewriting config files. which is something that > requires lots of testing before applying on a production machine. > Which is time you mostly don't have when a problem is found.
Please pay attention. I am not complaining about the practice of backporting fixes, which I wholly support. What I am complaining about is doing it in such a way that both the user and (particularly) the original author of the software cannot tell that it has been done. Cheers, Ben. -- http://www.apache-ssl.org/ben.html http://www.thebunker.net/ "There is no limit to what a man can do or how far he can go if he doesn't mind who gets the credit." - Robert Woodruff _______________________________________________ Full-Disclosure - We believe in it. Charter: http://lists.netsys.com/full-disclosure-charter.html
