Re: Why cookutils and no 4.0 yet

[ Thread Index | Date Index | More lists.tuxfamily.org/slitaz Archives ]




The main reason for tazwok was the depends check. With nearly 3000 packages its better for tazwok to know what is needed if its not in build depends. This allowed receipts to be more in line with archlinux PKGBUILDs. The makedepends is only used when needed. In slitaz BUILD_DEPENDS is used for almost every package cause it doesn't detect.

Just my 2 cents on this.



This is a great feature. I guess we will now have a great build app.
 
On Mon, May 9, 2011 at 8:11 PM, Christophe Lincoln <pankso@xxxxxxxxxx> wrote:

Hi all,

He is a small resume of the situation and why I'm working on cookutils
and not just make 4.0

   * From end of last year wok was fozen but update never stops
   * When I was in holiday in december/janury some people decided to
     use an experimental tools as main build tools, it broke a lot of
     things and delayed to release cycle.
   * Tank was crashing a lot and then FSF said they cant' handle the
     load so we had to find and put in place a new server (Thank again
     to Thomas, the box work like a charm) and this delayed again the
     release cycle.
   * When I realize that and realize the huge and incredibale work
     done on Tazwok in the same time, I thought "ok we cant revert that
     we must give this tazwok-exerimental it chance"
   * It's what we done and we used new Tazwok on Tank, but 3 month
     later new tazwok was still not running as a build bot and is
     difficult to maintain on Tank for me and other tank maintainers..
   * Worse, with last cooking we got a lot of kernel panic, this is
     very bad. With help on the forum we found that thes kernel panics
     was caused by a change in kernel processor type.
   * To fix this kernel issue I changed linux config, but then API
     change and a lot of packages must be rebuild (that normal and I
     was ready to rebuild them) so I wanted to rebuild toolchain on
     tank with tazwok and it fails 3 times. This was to much and the
     first reason to start new cookutils.
   * I all this time I got private mail and read here on the list
     complain about new tazwok to difficult to use and improve if you
     not a guru: second reason to code cookutils.
   * Finaly, we want a build bot working on tank before 4.0, because
     we want a build bot for stable and cooking, this let us send
     upgrade to stable easily and got packages builded automaticaly.

The goal is not the release for release but to release a stable
version, so we will release when ready. I think it's not a problem
since SliTaz have 5 years now and we NEEDED huge change is build system
because they was wrote when slitaz had 300-400 packages but handling
nearly 3000 package is an other story.

Now we need help to test cookutils and in a few times (say 10 days) we
will need packages testing. I also think Godane would apreciate help on
Tazwok since Gokhaley a less time to debug it now and also would
apreciate help on test slitaz source DVD.

Hope it's help to clarify the situation,
- Christophe

---
SliTaz GNU/Linux Mailing list - http://www.slitaz.org/





Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/