[Mageia-marketing] Mageia product naming/scoping

Bradley D. Thornton Bradley at NorthTech.US
Tue Mar 15 21:27:27 CET 2011


-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160

Hi Romain,

I'm glad you brought this on up. I've been dreading it. For the most
part, much of this topic has already been decided, but FWIW, I will
comment anyway, as this is a great concern to me, and as far as I'm
concerned, the most important aspect of whether Mageia will ultimately
succeed as a major distro to be taken seriously.

Just to point out, I haven't had any use whatsoever for Mandriva, having
only installed it just a few times over the years, so I could be sure as
to why I recommended EVERYONE to NOT use it.

Conversely, I was a MAJOR proponent and advocate for Linux Mandrake, and
my company was responsible for somewhere in the neighborhood of 25,000
desktop installs at various Fortune 500's and others in the United
States, including Merril Lynch, ARCO, AIG, and a bunch of others.

Then came that damn IPO...

I think we all know what happened after that. But until then, as far as
I was concerned (Having completely abandoned the Redhat camp after RH
5.2), it was the ONLY RPM based distro I championed for the desktop.

We erradicated an awful lot of wYNdOZe in the enterprise with Mandrake.

As far as servers, however, I am not a proponent of RPM based distros -
nor will I be. I'm just not your man if you want someone to tout the
benefits of Mageia in the datacenter or back office. It ain't gonna
happen in any of my shops.

Yes, I'm quite prejudiced, yet this is not to say that I don't think
Mageia or Mandrake would work well as a server. It does evidence my
preferences for (IMNSHO) more suitable distros to serve that purpose.

And for that purpose, I'm kind of a
Slackware/debian/Sorcerer/Arch/Linboard kinda guy (Yes, I've been
installing and running Linboard Linux in the enterprise and datacenters
since I created it in 1992, and in 1993 it became available on the
shelves of many prominent stores via the retail distribution channels).

Occasionally, yet very rarely, I will actually settle on CentOS, but
that is mostly for datacenter installs for hosting companies, as has,
for the most part, only to do with the nice integration between
WHM/cPanel for their customers.

Other than that, I have virtually no use for an RPM based distro for
anything other than the desktop (with the possible exception of SuSE).

I'm hoping that Mageia will fill that gap that has been, IMO, missing
from the desktop distro offerings since Mandrake met its demise and
became Mandriva.

This is not to say that I think Mageia is not a good product for the
back office or datacenter - I've just been through RPM hell too many
times over the last 15 years and don't see it as a sustainable model -
and from my perspective, it has never proven to be suitable for that
kind of that environment as far as I'm concerned (SuSE included).

Remember, I come from a Sys III / BSD background. I was raised to
believe that a server should be capable of being maintained for ten or
more years, which is exactly what I have when I run Slackware, debian,
xBSD, or Solaris.

Therefore, my perspective might be quite different (and undoubtedly is)
from others here on the Mageia team.

But before I briefly address your subject below, let's look at the other
side of the coin before people get the wrong idea about what I expect
from Mageia and where I think (Hope) it is going to shine.

First, I'm hoping that Mageia will make for the wonderful, award winning
desktop distro that Mandrake was (as I've already indicated).

Second, Slackware and Debian, although my first choices usually for
enterprise servers, are really not paricularly well suited for desktop
deployment for the average joe (Well, debian can be). For example, I'm
writing this email on a pure Slackware64 -current laptop  - Something
that I definately would almost NEVER recommend to anyone but a seasoned
sysadmin.

Third, and I might get flamed for this, but Ubuntu just sux green
weenies. I'd rather give someone a Knoppix LiveCD than see them run that
bastard distro. Again, I admit I'm quite prejudiced when it comes to
choosing distros for a particular application.

Okay so what do we have now? Several reasons why I just don't see any
use for an RPM based distro (with rare exceptions for SuSE and CentOS)as
an enterprise server, and several reasons why I feel my favorite server
platforms are not suitable for massive desktop deployment in the enterprise.

One thing that disgusts me (Aside from RHEL), are the enterprise
offerings from Mandriva. You can download some "Enterprise" versions of
the distro, and not pay for support (WTF does support mean anyway?), but
you aren't, according to the website entitled to updates - that is just
INSANE!!!

And then, there other, in my opinion, utterly useless versions of the
distro that are lacking in some way, because of the inference that they
are somehow mission specific, crippled, or otherwise lacking in
capability (even though they're really not).

So, to address your topic of addressing the, "...listing and naming of
the different instances of our product,...", I offer the following, as a
roster of what *I* would like to see :)

1.) Of course, a full featured, Mageia LiveCD (I don't personally have
much use for these, but recognize and recommend that such an entity
exists for many reasons, too numerous to list here.

Things like LibreOffice, and other *end user* software should be
included in this packed image.

2.) For lack of a better term, a *utility* or *minimal* LiveCD. I know
that leaves a lot of wiggle room for people to figure out what should be
included in such a spin.

A plethora of *utilities* should be included, including things like
Wireshark (Ethereal), EtherApe, Filezilla, CLI and forensic tools, but
not going overboard with that because distros like Backtrack 3/4 already
have narrow focus on that front.

3.) A Mageia Minimal offering - Like debian's Netinst - a minimal spin
that you can use to quickly get an OS deployed and finish/customize over
the wire and/or via scripts - such a spin makes large scale enterprise
desktop delployments not only possible, but friendly to the VARs and
Integrators who will be rolling out the Mageia desktop  en masse in the
enterprise.

4.) A Mageia Full: this should be the shining gemstone of planet Mageia.
a complete set of userland applications, office and productivity suites,
games, whatever will make people go, "Man that Mageia is really kewl and
full featured right out of the box!".

Lot's of bells and whistles for the home user and corporate desktop
alike, something that will give that Ubuntu crap a run  for its money
from a team that is organized similarly to that of the debian project
(as we are), instead of some corporate sponsored, special interest
driven, distro - like Ubuntu, Redhat, and Mandriva are (Yeah,
unfortunately, SuSE falls into this category too, although there is
OpenSuSE too).

Finally, and no *handy* distro would be complete without it, a USB stick
version of Mageia, pretty much the same thing as Mageia Minimal, or
Mageia LiveCD, but something that IT persons and Internet Cafe patrons
alike, can carry on their keychains and boot from quickly, to surf or
access Cloud resources.

In my opinion, Whether any kind of Enterprise support packages are
offered or not for Mageia, there shouldn't be ANY DISTINCTION WHATSOEVER
between one of the above spins of Mageia and support subscription ready
versions - IOW: Specific *Enterprise* or otherwise specialty spins of
Mageia *Should NOT exist* - the distro should be able to stand upon its
own in the community and be equally appropriate for those who want to
self-administer with community support or paid subscription support.

I'll retell this one story briefly here again:

One day, a member of the development team in the medical industry
company where I worked came to me and asked if he could have resources
on a production server for one of our upcoming products that I had
previously provided him a custom staging server for (Watch out for Java
programmers, they want ALL of your memory LOL).

Since we were moving, wrt this particular suite of apps, from beta to
limited production for final elbow checking before general release and
shrink wrap, I told him I could give him space and resources for his
apps to run on one of our older, less used forward facing servers.

Well this particular server was an RHEL box, and the software didn't run
because of certain aspects of his suite, to make a long story short,
that needed some specific support in the Linux Kernel.

So I pushed him back three or four days and said I would compile him a
kernel on that box that would meet his needs and then no prob.... Until
one of the VPs came to me and asked if this was okay to do.

"But of course it is", was my general response. I mean, what could be
more natural than compiling a custom kernel?

But the VP urged me to contact RH and ask them if is was *OKAY* to
compile and install a custom kernel on a box with an RH Enterprise
support subscription.

I grudgingly obliged his suggestion, and was subsequently informed by
Redhat that we could NOT run this software (meaning, I couldn't install
my own kernel).

I had never heard of such a thing (um... Well, back in the mainframe
days, sure).

Basically, I was informed by the Friendly Redhat support person that in
order to run a custom kernel, that I had to provide Redhat with the
particular specifications and THEY would compile the kernel for me.

The argument has been made on several sysadmin lists that since you are
paying Redhat to hold your hand and do everything for you, that you
should just let them do it.

Okay, but then why do I need to be drawing a salary if Redhat can, and
will, do everything that I was contracted to do?

I thanked the nice gentleman at Redhat and went about my day. in the
late afternoon, when the second shift of my admins started to arrive, I
gathered everyone up in a conference room and told them that on Friday,
they would all need to bring sleeping bags, a bunch of sandwiches,
bottled drinks, and a couple changes of clothes - and be prepared to
live on premise for the entire weekend.

Come 06:30hrs Monday morning, there wasn't a single living Redhat
machine left in the entire enterprise, and the developer had his
production machine ready to launch his RC1.

The moral of the story is like this - If we create a distro that has a
*spin* which requires an enterprise support contract like Redhat or
Oracle (basically the same distro, but Oracle's enterprise support
package is cheaper), then we will alienate massive amounts of
administrators, and since Redhat already has this particular market
cornered for the idiot pencil pushers who for some stupid reason make
such support a requirement for the adoption of a vendors product, Mageia
will fail miserably - and I will abandon it as well.

On the other hand, if we simply provide enterprise support as a value
added option for some, or any of our distro's various spins, then I'm
all for it - even if such a model isn't in the works.

Preferably, I would like to see a Mageia certification program, and
links to vendors that provide enterprise support for Mageia
installations, but regardless, the Redhat model is a big loser for
Mageia if it is adopted.

Another way to derive income for the project would be to bundle third
party, *non-free* software suites with optional or mandatory
licensing/support packages. I won't personally use any software like
that, but I have nothing against it and the publishers of such software
will pay the Mageia project handsomely for such exclusivity, access to
their wares, and the advertising space on our websites.

Okay moving along now :)

As for the actual *Naming* for our products... Well that's really not
important to me, as long as the average joe is able to make a
distinction between what the various product spins of our distro above
are, and which one is right for them.

I'm really hoping that we end up with a product that I can confidently,
and proudly recommend on the enterprise desktop which I have no
hesitation to use, in the campaign to erradicate wINdOZe.

To summarize the most important points of what I expect to see emerge
from our product line:

1.) No particular spin of our distribution should be crippled (like some
Mandriva spins are), yielding an installation that is either full
featured from the start, or with regards to a minimal netinst spin of
our product, completely capable of seamlessly becoming the full featured
version with a couple (okay maybe six or seven) clicks of the mouse.

2.) If there is to be an enterprise support offering, it is something to
augment the user's installations of Mageia, irrespective of the
particular spin they chose to install, and without the requirement that
any *special* version of Mageia need be created for such purpose.

3.) We should have a LiveCD and USB stick version which fits on a single
CD, and on a USB flash drive of less than or equal to 1GByte. Those
versions should also have an exceedingly easy and DIRECT migration path
to install the *live* system onto the hard drive, yielding a complete
Mageia installation that conforms to item #1 above.

4.) We should have all of the above offerings in a strictly 32bit
version of Mageia, for backward compatibilty support for older, or
crippled hardware, while our flagship products should be 64bit multi-lib
enabled.

This last point might need some clarification. While I am, essentially a
Slackware stalwart, and have been for over 15 years, we slackers
typically believe that one should have a pure 32bit version of a distro,
and a pure 64bit version that is multi-libe ready, but not enabled. I do
not advocate such a thing for Mageia - I believe that all 64 versions of
our product also be 32bit enabled - we're talking about two different
beasts here, and if our 64bit versions won't run 32bit apps out of the
box - we are doomed.

I hope that helps :)

Bradley.

On 03/15/2011 07:58 AM, Romain d'Alverny wrote:
> Hey there,
> 
> here comes another serious part: listing and naming the different
> instances of our product, for finale release (and we ought to have
> everything square for beta2 or RC release if possible).
> 
> Here are notes summarizing what we currently have available for
> delivery, given our history. There are missing use cases, perhaps
> (being more specialized). Maybe missing media as well.
> 
> http://mageia.org/wiki/doku.php?id=marcom:product_naming
> 
> Goal of this topic is double:
>  * study/propose/find a name for the product as such (Mageia? Mageia
> something? something?)
>  * refine/study/propose/find use cases to confirm existing delivered
> products or suggest new ones (new ones may not necessarily be
> available for Mageia 1, but will likely be tested just after for the
> next run).
> 
> 
> Advice, proposals, mockups welcome.
> 
> Thanks!
> 
> Romain
> _______________________________________________
> Mageia-marketing mailing list
> Mageia-marketing at mageia.org
> https://www.mageia.org/mailman/listinfo/mageia-marketing

- -- 
Bradley D. Thornton
Manager Network Services
NorthTech Computer
TEL: +1.760.666.2703  (US)
TEL: +44.702.405.1909 (UK)
http://NorthTech.US

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Find this cert at x-hkp://pool.sks-keyservers.net

iQEcBAEBAwAGBQJNf8uvAAoJEE1wgkIhr9j3aAQH/3vNin1edDHzMAsrqVshtmAI
P2D3I0pkJt8NUUYI/cXfmH6V+JHT21ug53qVhGXwd6bW5W7UTxDOVuJZu47laqzc
LWiA99kH6fjgCE5/aZyWfFwhPBbEmY9DOOHNnaLdBxV9ehH2yz9x9v9kbrpewPTn
7xjUU5pQkcCddihP69jYgLhoyT5rc0kI2Fx0fD76bWeFoLzPmtOaiaR8VFwTinn2
ghXSrHeVDXrk21IDmqxTzO1Qcdsbs+AEXwYkYf/DbkY9+wIf47IReR/lPzFShNrN
zIXHF4mnww7oSgptywVTdSD+RJF9U2vjDcQasrl67ZFdTOEQBuKVFRA32C4KOOE=
=N5ED
-----END PGP SIGNATURE-----


More information about the Mageia-marketing mailing list