All of lore.kernel.org
 help / color / mirror / Atom feed
From: Gary Bisson <gary.bisson@boundarydevices.com>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH v2 2/2] package/mfgtools: drop package
Date: Tue, 14 Apr 2020 10:14:04 +0200	[thread overview]
Message-ID: <20200414081404.GA330240@p1g2> (raw)
In-Reply-To: <037e6ef02ea9e47f92daed0ac03bbe23231aafcb.camel@embedded.rocks>

Hi Jorg,

On Tue, Apr 14, 2020 at 09:47:48AM +0200, J?rg Krause wrote:
> Hi Gary,
> 
> On Tue, 2020-04-14 at 08:46 +0200, Gary Bisson wrote:
> > Hi Jorg,
> > 
> > On Mon, Apr 13, 2020 at 10:24:06PM +0200, J?rg Krause wrote:
> > > Hi Gary,
> > > 
> > > On Mon, 2020-02-10 at 17:26 +0100, Gary Bisson wrote:
> > > > Hi Jorge,
> > > > 
> > > > On Thu, Jan 09, 2020 at 08:10:20PM +0100, J?rg Krause wrote:
> > > > > As suggested in [1] the package mfgtools is dropped.
> > > > > 
> > > > > NXP did replaced the old mfgtools with the version number 0.2
> > > > > enterily with the uuu (Universal Update Utility) which is somehow
> > > > > named mfgtools 3.0 although the version scheme for the uuu tool is
> > > > > 1.xx.yyy.
> > > > > 
> > > > > As the old mfgtools scripts are not compatible with the new uuu
> > > > > tool and as imx-uuu goes hand-in-hand with imx-uuc, which we ship
> > > > > for the target, the mfgtools package is dropped.
> > > > 
> > > > I know I'm the one who pushed for that commit to happen but now want to
> > > > mitigate the claim above.
> > > 
> > > Thanks for your feedback and sorry for my late reply.
> > > 
> > > > True uuu depends on imx-uuc and mfgtools (1st of its name) too. BUT by
> > > > looking at imx-uuc, the two use cases are separated into 2 different
> > > > files:
> > > > - uu.c: daemon that uses utp protocol to communicate with host
> > > > - ufb.c: daemon that uses fastboot protocol to communicate with host
> > > 
> > > In fact, I am using mfgtools without imx-uuc, but with U-Boot.
> > > 
> > > > So I think it should be safe to have both host clients co-existing for
> > > > now. Maybe uu.c will be removed at some point in the future but I guess
> > > > NXP will leave it there for a while.
> > > 
> > > So, lets just update mfgtools to latest version (1.3.154), right?
> > 
> > No, mfgtools 1.x.y really is 'uuu', not mfgtools any more.
> 
> It is very confusing, how NXP deals with this. From my understanding,
> mfgtools is the top level name for the image deploying tools.

It is confusing yes. And NXP doesn't really deal with it, they just
said "here's a new tool, let's forget about the old one".

> In mfgtools 2 there was a seperate Windows and Linux build of the
> 'MfgToolLib', the linux command line utility was 'mfgtoolcli'. The
> linux branch ended with v0.02, the windows branch with 0.06 -> 2.8.0.

Yes the versioning is also confusing. Buildroot automated email keeps
reminding me that mfgtools isn't at its latest release whereas it is,
new 'versions' are just for a different tool really.

> Both branches are abandoned and replaced by libuuu and the command line utility uuu.
> This new library and tool are introduced of an evolving mfgtools 3.0.
> 
> So, indeed there are two different tools, the old and abandoned
> mfgtools 2 and the new mfgtools 3 (a.k.a uuu).
> 
> Note, that for the mfgtools 2 the tag v0.02 was only labeld as "Pre-
> release" on Github.

Yes, but Yann's thoughts were that since it was in Buildroot for a long
time, people might have scripts relying on mfgtools, and therefore might
be useful to keep it around [1].

Then at the time I made the point about imx-uuc which wasn't valid. Now
that the build is fixed, I agree that we should keep mfgtools as-is.

> > So since 'mfgtools' doesn't have any more build issues, I'd say we keep
> > it as-is for people that use it. So the idea is _not_ to drop it.
> > 
> > Then just have your uuu addition patch, so basically both tools
> > co-exist in Buildroot (as they serve different purposes).
> 
> If we really want to keep the mfgtools 2 package, we have to decide how
> to name the new package:
> 
> a) imx-uuu
> b) mfgtools3
> c) mfgtools_uuu
> d) something else
> 
> I'm voting for b) mfgtools3, as the github project name is still
> 'mfgtools' and the project is described as "uuu (Universal Update
> Utility), mfgtools 3.0".

I'd personnaly prefer a name with uuu in it as mfgtools is only
mentioned for legacy reasons from NXP. But then in every doc they only
say uuu [2]. So from a user perspective, I'd look for 'uuu' in
menuconfig.

I wish NXP would have switched to another repository...

Regards,
Gary

[1] http://lists.busybox.net/pipermail/buildroot/2019-June/252440.html
[2] https://github.com/NXPmicro/mfgtools/releases/download/uuu_1.3.154/UUU.pdf

      reply	other threads:[~2020-04-14  8:14 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-01-09 19:10 [Buildroot] [PATCH v2 1/2] package/imx-uuu: new host package Jörg Krause
2020-01-09 19:10 ` [Buildroot] [PATCH v2 2/2] package/mfgtools: drop package Jörg Krause
2020-02-10 16:26   ` Gary Bisson
2020-04-13 20:24     ` Jörg Krause
2020-04-14  6:46       ` Gary Bisson
2020-04-14  7:47         ` Jörg Krause
2020-04-14  8:14           ` Gary Bisson [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200414081404.GA330240@p1g2 \
    --to=gary.bisson@boundarydevices.com \
    --cc=buildroot@busybox.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.