lmi
[Top][All Lists]
Advanced

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

Re: [lmi] Building wx from trunk


From: Vadim Zeitlin
Subject: Re: [lmi] Building wx from trunk
Date: Fri, 24 Oct 2014 19:34:53 +0200

On Fri, 24 Oct 2014 16:19:35 +0000 Greg Chicares <address@hidden> wrote:

GC> On 2014-10-20 20:01Z, Vadim Zeitlin wrote:
GC> > On Mon, 20 Oct 2014 15:48:06 +0000 Greg Chicares <address@hidden> wrote:
GC> > 
GC> > GC> (1) Clone the wx repository--only need to do this once.
GC> > GC> 
GC> > GC> cd /opt/lmi
GC> > GC> git clone https://github.com/wxWidgets/wxWidgets.git
GC> > 
GC> >  Just a note: you've previously mentioned that access to some domains may
GC> > be restricted from some developers machines, I hope you can add github.com
GC> > to the white list of allowed domains.
GC> 
GC> "Sein Ermatten ist das des Gladiators nach dem Kampf,
GC> seine Arbeit war das Weißtünchen eines Winkels in einer Beamtenstube."
GC> 
GC> Like us, the author also worked at an insurance company.

 So you're implying that he was not at all as original as I naïvely
thought? Or that you might write a book of aphorisms soon? I'd like to read
the latter...


GC> >  Now for the promised answer: the canonical way to identify a commit it 
git
GC> > is indeed by using its SHA1. To find the SHA1 you want to use, you can
GC> > use "git log" (to see just the latest commit, "git log -1")
GC> 
GC> Great. That's easy enough.

 Most people still prefer symbolic names, which my other suggestions would
allow you to use. But if 40 digit hex numbers don't bother you, this will
definitely work.

GC> I'm surprised that a truncated SHA1 is apparently accepted. I see
GC> it on the console when I 'pull', e.g.:
GC>   Updating e897361..9650c5b
GC> The part on the right, "9650c5b", seems to be the first seven hex
GC> digits of the new SHA1:

 Yes, git abbreviates SHA1s to the shortest unambiguous (in the scope of
this repository) prefix, but not less than 7 or 6 (depending on the
command...) characters long.

GC> With only seven hex digits, the chance of a collision is 16^-7, and I
GC> anticipated that a tool like git would take a more cautious approach.

 It's just smarter than you think (a feeling almost as common with git as
the feeling that it's the stupidest program in the world :-). It will use
more characters as soon as the prefix becomes ambiguous, but this just
happens relatively rarely.

 For example, I've only recently, after ~5 years of daily git use,
including with some really huge repositories, have seen my first 9 digit
commit.

 Regards,
VZ

reply via email to

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