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

Re: [ttylinux:390] ttylinux end-of-life



Nothing FUBAR about it, though I understand the rationale.   I've used ttylinux effectively on several projects, and I've always thought it hit a sweet spot in the embedded linux space between the behemoths like Yocto and the uber-embedded tools.    Good luck with the new project.


On Mon, Jun 10, 2013 at 7:09 PM, djerome <[email protected]> wrote:
I've decided to end ttylinux, at least for me, for a few reasons:

=> The original reasons for ttylinux have been long gone. It became an
exercise in minimalism that
     can't be maintained with current revisions Linux and Glibc as
they're almost too big for older 486
     computers with limited RAM. I also bloated it past where it
should be. ttylinux-i486 is the only
     variant that is true to what ttylinux should be, and I can't
revise it with newer Glibc or Linux kernel
     or it will be too big.

=> I so bloated the build system...

=> Other than the i486, the ttylinux variants are too big to be called
ttylinux and they are too
     small to be useful.

=> No "market". Serious embedded folks use things like BuildRoot or
PTXdist. The virtual and cloud
     folks seem to be moving on to other things, like Cirros for
OpenNebula. Small Linux users have
     Tiny Core Linux, Puppy Linux, etc.

=> Interest seems to be waning. By the end of 2010 downloads were
about 70 /month with a server output
     of a few GB. By the end of 2011 that was several hundred d/l and
a few 10s of GB per month. By the
     end of 2012 it was over 1000 d/l and consistently over 50 GB per
month. For the past few years I've
     got several email per month about something not working or needed
help.
     Now, for the past 3 months, I've had no email, and d/l have been
down with just a few GB being
     served per month.

I think ttylinux is starting the end of its life. It doesn't fit a
niche and its use seems to be declining.

For anyone wondering: ttylinux has never been a commercial venture for
me. I've never earn money
from it and I spend a little each month for the web site hosting. I
wouldn't have it any other way, this
is not a complaint at all.

I'm not sure where to take ttylinux next. There's always bugs to fix,
but other than that I think it
has run its course, at least for me. I've fubar-ed it enough. The
things I want to do for a Linux build
system do not fit with the thing that is ttylinux.

***** Read This *****
The web site http://ttylinux.net and http://ttylinux.org will be gone
by July 10. ***Pull down what you
want from the web site now.*** There really isn't any time for posting
another release, even though I
am updating parts of the build system (in the GitHub repo).
If someone wants the ttylinux domains, then I'll transfer them, if
that is possible; otherwise wait
till after July 10 to grab them.

I'll won't delete the GitHub repo. After I am done with my current
updates in the x_2013a branch I
will merge them into the main branch, hopefully be this weekend. After
that I'll make easy fixes
that are sent to me.

If someone really wants to keep ttylinux going, then please do. Wait
till I'm done with my current
updates, then you can clone the git repo and I'll delete mine, and
then you can create a new one,
unless there is some easy way to transfer a GitHub repo. ***** But
read more first.

********** Maybe a New System to Replace ttylinux
Starting in July, I plan to focus on making an embedded-style [eglibc/
uClibc/(maybe glibc)
and busybox] Linux cross-building  system, free from "smallest
possible Linux" constraints,
for my BeagleBone and i486 PC.
My new cross Linux build system will:
1. be in GitHub
2. use crosstool-ng, I will submit patches to that project for
anything I need, and hope they take it
2. use ./configure, make, make install to put itself into /usr/local
4. let me start with an empty directory, initialize a cross Linux
build, "make getcfg",
    maybe "make (menu?)config", then "make dist" or "make packages"
"make fsys" etc.
5. be developed in Debian 6 s as to work with older tools
6. make proper use of fakeroot to build the packages
7. maybe do better at getting package files installed into sysroot
using unionfs, for building target packages
8. have no web site maintained by me, no target system ISO releases
maintained by me

Well, this is what I am thinking. Either this or spend the summer
swimming, grilling, and drinking
beer.

Cheers.

--
You received this message because you are subscribed to the Google Groups "ttylinux" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.