[Top][All Lists]

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

More stability needed in our Rust packages, for IceCat 60

From: Mark H Weaver
Subject: More stability needed in our Rust packages, for IceCat 60
Date: Fri, 21 Sep 2018 23:35:17 -0400
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/26.1 (gnu/linux)

Hi Danny,

address@hidden (Danny Milosavljevic) writes:

> dannym pushed a commit to branch master
> in repository guix.
> commit ec944a1de97c8a523d6e1089b8c4430424297f93
> Author: Danny Milosavljevic <address@hidden>
> Date:   Fri Sep 21 15:42:56 2018 +0200
>     gnu: rust: Remove unused "ar" setting.
>     * gnu/packages/rust.scm (rust-1.20)[arguments]<#:phases>[configure]: 
> Remove
>     unused "ar" setting.

Thanks to your hard work on our Rust packages, I was able to update our
IceCat package to a pre-release of 60.2.0.  I recently pushed it to
master and Hydra is currently busy bootstrapping the chain of Rust
compilers in order to eventually build IceCat.

Unfortunately, I see that you've pushed some more small patches that
have already rendered those builds obsolete.  The words below are not
complaints -- far from it -- but are merely meant to illustrate a
practical problem we now face.

It took my fastest laptop about 36 hours to build the chain of 5 rust
compilers required to get to the latest Rust release, and to build
IceCat.  By the time I had gotten IceCat built and tested, you had
pushed a few more improvements to our Rust packages, rendering my
testing obsolete.  I decided to push it anyway, since the IceCat update
is a critical security update.

Since then, I started my fastest laptop building the new Rust compilers,
and also generated an evaluation on Hydra to build the new IceCat.  Both
of these build efforts are less than half through the Rust bootstrap,
and now I see that I'll need to cancel these jobs and start again.

Of course, I'm extremely grateful for all of the work you are doing on
this, but it has also become apparent to me that if our IceCat package
is to be effectively usable, we'll need to drastically reduce the
frequency with which our Rust compilers are modified on the master

How would you feel about doing further Rust work on a separate branch,
and periodically (maybe 2-4 times per month) merging the collected
improvements into master?


reply via email to

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