lmi
[Top][All Lists]
Advanced

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

[lmi] Munging "email addresses" in makefile patch attachments [Was: [PAT


From: Greg Chicares
Subject: [lmi] Munging "email addresses" in makefile patch attachments [Was: [PATCH] Integrate wxPdfDocument into lmi build system]
Date: Fri, 07 Aug 2015 02:55:31 +0000
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Icedove/31.3.0

On 2015-08-07 02:16, Vadim Zeitlin wrote:
> On Fri, 07 Aug 2015 02:05:59 +0000 Greg Chicares <address@hidden> wrote:
[...]
> GC> BTW, grabbing patches from the nongnu.org email archives is convenient
> GC> (especially if real email and lmi development are separated by a VM
> GC> barrier), but the archives perform interesting substitutions:
> GC> 
> GC> - cd $(cache_dir) && [ -e $@ ] || $(WGET) $(WGETFLAGS) 
> --output-document=$@ $(address@hidden)
> GC> + cd $(cache_dir) && [ -e $@ ] || $(WGET) $(WGETFLAGS) 
> --output-document=$@ $(address@hidden)
> GC> - cd $(cache_dir) && $(ECHO) "$(address@hidden) *$@" | $(MD5SUM) --check
> GC> + cd $(cache_dir) && $(ECHO) "$(address@hidden) *$@" | $(MD5SUM) --check
> 
>  Ah, this is very thoughtful indeed. As you might have probably guessed, my
> answer to all these problems with patches -- as well to anything else
> related to the source code management -- is git. If you used it (even just
> via git-svn, as I do), you could pull my changes from anywhere (here, your
> other machine, GitHub, ...) and apply them easily.

But I'm applying the patches in an msw VM, which I want to keep quarantined.
I could open a channel through the VM barrier and access my email directly,
by exposing host and client filesystems to each other, but then the barrier
would be much less secure. Well, I suppose you could say that a git-patch-only
hole would be uneconomical to exploit. But I think instead I'll just enable
cutting and pasting across the barrier: it's an unlikely exploit, especially
in the debian --> msw direction, and I can paste into a here-document without
thinking too hard.

>  Otherwise all I can propose is to just make the patch files available
> somewhere on the web, would it be helpful if I did it for the next patch
> series?

No, thanks, it's not worth your time. Grabbing patches with an msw browser
works smoothly almost all the time. We rarely have patches that change lines
containing at-signs. And this automated "coding rule" prevents damage:
    // Obscured email address.
    taboo(f, "address@hidden");




reply via email to

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