> On 03.07.2015 08:28, Ady via Syslinux wrote:
>
> > Maybe we should rather wait?
> ...
>
The link I posted in my email was very much relevant.
> > FWIW, I am still against releasing 6.04-pre1 yet.
> >
> > Regards,
> > Ady.
> >
>
> http://www.syslinux.org/archives/2014-October/022689.html
>
> <quote>
> I would like to do a short cycle for 6.04, because 64-bit support for
> ext4 has come up as a very high priority. Therefore I do not want to
> stuff too many things into the 6.04 release but hold them for 6.05.
> </quote>
>
>
> Wishful thinking gets you nowhere.
> :)
I'm sorry but I don't understand what you mean.
Since 2014Oct, no actual effort has been made to solve / improve that 64-bit
support for ext4 in Syslinux, and the "very high" priority was changed
later to
"maybe some day someone with the necessary skills and knowledge would care
enough
and do something about it".
The development pace was slowing down even before 2014Oct, and stalled (almost)
completely for several months after the release of 6.03. The only reason v.6.03
was
released was because Debian requested it, and at the time there was no realistic
expectations of solving additional bugs in some "soonish" period of
time. There was
no ETA (of any kind, time-based or otherwise), and that's still the
situation
today.
The "stable" release (v.6.03) came with even slower development than
before it,
comatose in fact. The same already happened before, and I don't like the
idea of
slower development (yet again). In theory, a release should not affect the
development pace; in practice, it does, negatively.
After 8 months, we finally have public access to official development code /
repository again, which was not publicly available in an official manner since
2014Oct, just a few days after the release of v.6.03.
There are still relevant regressions, and discussions about improvements are
still
ongoing; slowly, but they exist. I know for a fact that some initial attempts
for
additional patches are sitting somewhere, waiting for the relevant developer(s)
to
share them in this Syslinux Mailing List (please emphasize _fact_, no wishful
thinking about that, but I do wish they would actually take the necessary
steps).
I don't want (future) users claiming that they are using "6.04"
and they have a
problem, when there is no factual evidence that some important bugs are really
solved. There are some improvements, and some assumptions, and some partial
success
cases; that's not *solving* bugs - it might hardly be called a workaround in
some
cases.
Additionally, I have requested (more than once, even partly in bugzilla) for a
more
accurate / better identification of versions. Until now, users would see
"6.04" and
they would assume "it is all the same, there cannot be much
difference". Similar
expressions are seen since 2014Oct: "how much it could affect the resulting
behavior if I use upstream official 6.03 pre-built binaries, or some package, or
some kind of re-build". The results are different in many cases, and the
current
versioning accuracy / details are not enough help for troubleshooting.
I think it is also appropriate for me to express now and here my appreciation to
everyone who is (and has been) involved in The Syslinux Project.
Going back to the real matter in question, all I am suggesting is to wait (as
oppose to investing time in potentially blind searches) for Adam to investigate
further, considering that his later tests are succeeding, without reverting any
prior commit (and under the same Fedora Rawhide environment).
Even then, I am still opposed to releasing 6.04-pre1 yet.
Regards,
Ady.