Re: cvs: temporary file handling fixes
Derek Robert Price <[email protected]>
| Newsgroups | gmane.comp.version-control.cvs.bugs |
|---|---|
| Organization | Ximbiot <http://ximbiot.com> |
| Message-ID | <[email protected]> |
Solar Designer wrote: >However, looking at 1.12.1, I notice that the only two scripts which >will now use mktemp (if enabled at configure time) are cvsbug and >rcs2log, and the uses by cvsbug are buggy in that the file name in >$TEMP will be re-used multiple times. Yes, Red Hat has this bug in >their patch too. > >I don't understand why you consider our fixing the other scripts in >contrib/ and the documentation misguided. > I forget why, I'll see if I can find time to review them again soon. >>The fixes that >>might be usable are going to need at least ChangeLog entries to >>accompany them, >> >> > >Obviously, but: > >- it doesn't make sense to write full ChangeLog entries before we know >the fixes are even getting in (and I don't expect you to include them >without any changes at all); > Well, yes it does when I can't figure out the purpose of your changes. A more complete abstract would help immensly in this case as well, but if I can't decipher the reason for any part of a patch when reviewing it, I find ChangeLog entries can be useful. >- CVS is just one of over 120 packages in Owl and we're primarily >concerned with making our distribution better; we also like to share >our changes with upstream maintainers, but we can't afford to spend >much extra time on the integration of our changes upstream. > If I don't understand the reason for your changes I am hardly going to incorporate them. If you plan on continuing to maintain a distribution of CVS, I expect it would be useful to you to have those changes incorporated upstream. >>some may need more documentation or tests in sanity.sh, >>and all will need to have their purposes explained more fully to be >>accepted. Please see the HACKING file in the top level of the CVS >>source distribution for more on how to submit patches. Please note in >>particular that they should be sent to the <[email protected]> mailing >>list and not directly to me. >> >> > >This all is fine with me (although I won't necessarily have the time >to submit any of this officially), but it doesn't make a valid >procedure for reporting security problems and proposing fixes to them. >In particular, I was looking for a (security) bug reporting address >that wouldn't automatically reach a public mailing list, -- but it >seems you find unsafe temporary file handling to be a minor enough >issue to be discussed in public. This is OK with me, but I thought >that some vendor-sec members could prefer to handle it differently. > > Again, sorry for bouncing a possibly sensitive email to bug-cvs so quickly, but unless clearly and believably labeled as sensitive, it is practically a reflex action for me to bounce emails about CVS from senders I don't recognize to [email protected] when they contain patches. I get a lot of email and have little enough free time as it is. Derek -- *8^) Email: [email protected] Get CVS support at <http://ximbiot.com>! -- There are no absolutes.