livecheck behaviour

[John Patrick]

I had similar issues with livecheck and the maven ports.

It works out the set of potential matches you have defined in the livecheck
rules. If the install Port version or the latest Portfile contains a
version that is not within that set then it does exhibit odd behavior if
technically it's a newer version.

Personally I feel the current Port version / latest Portfile should be
prepopulated into the potential matches before livecheck does it thing,
this way your example would work. As I could see other issues if say a Port
was distributed via mirrors and livecheck is using a mirror which is
hours/days/months behind. As if installing and a mirror is behind it will
produce a 404 and so it will failover to another mirror. Maybe livecheck
should also failover to another mirror if the current version isn't in the
potential matches...

John



On 19 November 2013 12:50, Peter Danecek <Peter.Danecek at bo.ingv.it> wrote:

>
> Hi all,
> I am just experimenting a bit with livecheck, and wonder if the behaviour
> below is really the desired behaviour.
>
> --- snip ---
> DEBUG: Port (livecheck) version is 0.8.5
> DEBUG: Fetching https://pypi.python.org/packages/source/o/obspy/
> DEBUG: The regex is ">obspy-(\d+\.\d+\.\d+)\.zip<"
> DEBUG: The regex matched ">obspy-0.8.0.zip<", extracted "0.8.0"
> DEBUG: The regex matched ">obspy-0.8.3.zip<", extracted "0.8.3"
> DEBUG: The regex matched ">obspy-0.8.4.zip<", extracted "0.8.4"
> py-obspy seems to have been updated (port version: 0.8.5, new version:
> 0.8.4)
> --- snip ---
>
> I would argue 0.8.5 > 0.8.4 so ideally this situation would be recognised.
> Is there some way to influence the evaluation, e.g. to result "seems to be
> up to date" in such a case.
>
>
> NOTE: In this case this actually is not really needed (it was only a
> test). But I am looking at some other situation. Where `updates` got into a
> different directory. So in that case a somewhat different behaviour would
> be helpful.
>
> It would also be interesting to check for the presence of some regex, in
> case it is not to judge that the port is up-to-date. Would this be possible
> somehow?
>
> ~petr
>
>
>
> _______________________________________________
> macports-dev mailing list
> macports-dev at lists.macosforge.org
> https://lists.macosforge.org/mailman/listinfo/macports-dev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.macosforge.org/pipermail/macports-dev/attachments/20131119/5978cdbf/attachment.html>