Hi,
>>"Rob" == Rob Browning <[EMAIL PROTECTED]> writes:
Rob> I thought about this, but unless you can actually hold up the dpkg
Rob> run, you still have ugly cases. Consider:
Rob> dpkg -i calc*.deb
Rob> dpkg --purge calc
I see. Well, there are indeed cases in which things may me
Jason Gunthorpe <[EMAIL PROTECTED]> writes:
> specificly the part where I said Joey has installed an override so
> that [EMAIL PROTECTED] always goes to the list.
Well, whatever, it still strikes me as very silly. You're putting a
random unrelated individual (Joey) in charge of where the mail fo
On 28 Jun 1998, James Troup wrote:
> Jason Gunthorpe <[EMAIL PROTECTED]> writes:
>
> > Is it OK to set the maintiner field on APT to be
> > '[EMAIL PROTECTED]'?
>
> Umm, colour me confused. Won't that cause a horrendous loop? IIRC
> [EMAIL PROTECTED] is generated from the Maintainer field; so
Manoj Srivastava <[EMAIL PROTECTED]> writes:
> Hmm. I think I would prefer 2.0.7-2.0.8pre1 or some thing to 2.0.7~1
Right. I'm going to do this for now.
> 2.0.8~pre1 is even better, I think.
And hope this is implemented soon.
--
Rob Browning <[EMAIL PROTECTED]> PGP=E80E0D04F521A0
Hi,
It just dawned on me that you may have meant that the ~ have
special meaning to the _humans_ and tell them that the numbers after
the ~ are meant to be the next version up.
Hmm. I think I would prefer 2.0.7-2.0.8pre1 or some thing to
2.0.7~1, just because the former is more
Manoj Srivastava <[EMAIL PROTECTED]> writes:
> Can emacsen common lock the dir (/var/lock/emacsen.lock) and
> not only look for the dpkg lock but also itself? (put the process id
> in there so that one may look for /proc/ to see if the lock is
> stale.
I thought about this, but unless y
>>"Martin" == Martin Oldfield <[EMAIL PROTECTED]> writes:
Martin> Rather than treating the ~ as introducing a special case which always
Martin> compares lower than something else, does it help to make numbers to
Martin> the right of the ~ add fractions to the integer part ? Since the
Martin> v
Martin Oldfield <[EMAIL PROTECTED]> writes:
> Rather than treating the ~ as introducing a special case which always
> compares lower than something else, does it help to make numbers to
> the right of the ~ add fractions to the integer part ? Since the
> version number is now a rational number you
Hi,
>>"Rob" == Rob Browning <[EMAIL PROTECTED]> writes:
Rob> I have a couple of concerns.
Rob> First of all, what do you do about multiple dpkg runs. The
Rob> problem here is that (without support from dpkg), dpkg finishing
Rob> is not the same thing as all the install/remove scripts being
Rather than treating the ~ as introducing a special case which always
compares lower than something else, does it help to make numbers to
the right of the ~ add fractions to the integer part ? Since the
version number is now a rational number you can have arbitrary numbers
of releases before hittin
Manoj Srivastava <[EMAIL PROTECTED]> writes:
So to summarize, you're advocating using Depends: to order installs
when absloutely necessary, but otherwise put the dependencies in
special files in the add-on packages, and then use a menu-like trick
to run the add-on package install/remove scripts (
Jason Gunthorpe <[EMAIL PROTECTED]> writes:
> Is it OK to set the maintiner field on APT to be
> '[EMAIL PROTECTED]'?
Umm, colour me confused. Won't that cause a horrendous loop? IIRC
[EMAIL PROTECTED] is generated from the Maintainer field; so
[EMAIL PROTECTED] would point at... [EMAIL PROTECT
I recall there was quite the discussion some time ago on this topic, I do
not recall the out come.
Is it OK to set the maintiner field on APT to be '[EMAIL PROTECTED]'
? (joey has installed an alias so this always goes to the list) The only
reason I am interested in this is so that bug reports ca
On 27 Jun 1998, Manoj Srivastava wrote:
> Hi,
> >>"Jason" == Jason Gunthorpe <[EMAIL PROTECTED]> writes:
>
> Jason> I was thinking simply adding a new rule that says if you have
> Jason> two groups and one group is a ~ character then the ~ character
> Jason> looses in all cases.
>
> Gr
On 27 Jun 1998, Manoj Srivastava wrote:
> Hi,
> >>"Lalo" == Lalo Martins <[EMAIL PROTECTED]> writes:
>
> >> Placement of pre-epoch is an irrelevant implementation detail.
>
> Lalo> Actually, no. If they're in the right side of the upstream
> Lalo> version, dpkg can keep the current left-to-ri
15 matches
Mail list logo
