Beyond the “LiveUSB�

[netzen]

Beyond the "LiveUSB"
Hi everybody,

this topic here is more or less a continuation of the following topic:
http://www.knoppix.net/forum/viewtopic.php?t=30397
taking a different direction than the original there.

It's basic an exchange of information around the idea of "LiveUSB" concept.

I'm sure the information that we can exchange on the subject can benefit not only the Knoppix users and develop but it can benefit anyone who has interest on the matter.

First I like to share the way I personally look at the concept of a "Live" system running out of an USBKey and its relation to a more traditional system. This diagram represents my current view of the "live" stuff.



The diagram shows (at right) a small GNU/Linux Project I'm starting around the concept of "subito operating system", which is the third block at the right side.

I like to invite everybody to participate on this "dialog" here about this subject.
Anything related to "liveCD" and "liveUSB" will be interesting.

My first take on the subject is that perhaps we should start to use the term "Native USB" instead of just "liveCD on a USB Key", which, in my opinion does not translate the full potential of the idea in a correct manner. The current 6 series of Knoppix, in my point of view is part of a "Native USB" rather than a (just) "LiveUSB".

Please, participate and share your ideas.

Thanks all,
Valter

[netzen]

Hi, my second take on the subject is around the GNU/Linux "marketing".

I believe that this following diagram illustrate well the options we have every time we want to talk about GNU/Linux with somebody that currently don’t use it and don’t have technical knowledge of PCs and etc.



hires here: http://farm4.static.flickr.com/3109/...fed629b2_o.jpg

The way I think is that if we focus on the "live" and "virtualization" modes as the introductory way to talk about GNU/Linux, we reduce and almost eliminate all initial barriers that a first-time user may have. In this sense we, introducing GNU/Linux at beginning using the "live" or "virtualized" modes, can help make it more popular because it avoids the technical needs that are part of a "dual boot" or "native HD" install.

It is not a bad idea to show to the new potential users that he/she has 4 alternative ways to "install" a new OS, and 2 of these are completely safe and very easy.

As for the virtualization mode, the VirtualBox is currently, in my point of view, the best alternative to use for such purposes.

Valter

[netzen]

Hi, this is my third take on the subject here. This time the arguments are around the whole FOSS (free/open source) and possible areas of opportunities (from now). And only part of this argument is related to "USBOSes" (usb O.S.).

This is just a small but I believe, very important point on effort and where part of if should be going. The reasoning will be very simple and I believe can make a good point.

Here is the point: Today 1% of the (PC) users use Linux, and the other 99% don’t.



Then, who is this "1%"?

The answer to this question, in my view, is quite simple. GNU/Linux is a greater success amongst IT professionals, technical oriented people and hobbyists alike. Because of its flexible tools and structure, GNU/Linux became a weapon for this group of users to achieve higher and powerful usage for information technology.

Much more can be done for this "1%" of users, and it will, as GNU/Linux walks towards maturity in several areas it becomes much more powerful in the hands of people in this "1%" group.

So, here is the main point of my "third take":
There are opportunities "there" on the "99%" group that today may benefit from GNU/Linux but for several reasons are not achieving a happy relationship with it.

I believe that the present and future of FOSS relies on putting part of the effort to "achieve a relationship" with the today’s non-users, changing a little from the current (and past) approach which is almost 100% focused on the "technical group" of users.

It is not the best example I can put here but take a look at the following picture.



We can definitely argue that this is "Linux" for the "99%".

Of course, the correct is that this is "Free and Open Soft for the 99%". And iPhone uses more than the BSD foundation, but the argument is that these examples of devices here can be successfully used by the "99%" group without demanding higher "technical" knowledge.

One particular important notion here is that, the "server operating system" and the "desktop operating system" are not the only ones where GNU/Linux can aim for, and these examples of devices given here show another area of importance because it can reach the other "99%" of users that can have huge benefits from FOSS, but currently are, for several reasons, not enjoying this powerful tools and resources.

In this following diagram I’m trying this argument, using the word "local" and "power" to the server and desktop installation scenarios.



hires: http://farm4.static.flickr.com/3062/...bcb881dd_o.jpg

One consequences of my argument here is that the "Native Live USB" is as important as the "Desktop" and "Server" categories are, and with further attention and development can be of great value for the "other 99%" of "us and them".

Valter

[Capricorny]

I think this is interesting, and targeting the 99% really IS something to go for. But I think it will, at least to end users, look more like Windows/Linux than one might think. And practically all the heavy application software will be based on standard versions.

I also think it is worthwhile trying to do such things on the basis of Knoppix, at least in the first place. And, starting from collaborations of technically minded users, we will have a resource base. I don't think I'm that untypical myself: In a pinch, I could do most systems maintenance tasks myself, but I'd rather prefer to avoid it, if it's possible. But I would be happy to participate in projects to develop a ready-made system "for almost all".

[netzen]

Hi all.
In this four remark about NativeUSB (LiveUSB) I like to emphasize the fact that a NativeUSB (and beyond) represents a departure from the traditional "desktop era" in the information technology.

[Mainframes]
We can think as (big) computers as the first era in IT, with focus on the hardware and system software specific for that hardware.

[Desk PC era]
We can think as the Apple II creating a new era in IT, the era where people can have a computer in their desk.

[Mobile Render Devices and Cloud Processing (Computing)]
The appearance of the "web2.0" and the creation of the iPod and the iTunes Music Store can be used as the "mark" when the so-called "cloud computing" appears. Then, the creation of the iPhone upon the iPod Platform consolidates the idea with the introduction of the "render device" as counterpart of the "cloud computer".

Render Devices like the Kindle, Google Nexus One, iPhone does not need to have massive power processing necessarily, they can achieve great results by combining mobility with the "cloud" that they "talk to". In fact, the combination of the "cloud massive power and resource" with the mobility and convenience of the "render device", they can achieve a breakthrough experience in results when compared to the "desktop PC".

This "breakthrough" in experience doesn't come from the power of the CPU or the sophistication of the (traditional) O.S., rather, the innovation achieve unprecedented results by combining relative simple and low power render devices in one side with network of relative simple and cheaper networked "PCs" at the other side. Perhaps none of the sides are using sophisticated Operating Systems to achieve its functionality (although it may be used).

What we have here is a relative higher form (sophisticate) combining simple components, creating something bigger (higher) than previous possible.

So, instead of:
Exploring powerful CPUs and sophisticated Operating Systems.

We can explore the idea of:
Explore the power of combining LESS powerful CPUs and LESS sophisticated Operating Systems.

As this following picture suggest, I believe that we can use Google, Amazon and Apple (what they are doing) to get inspiration and insights into this "cloud/render device" paradigm, and try to capture ideas to put into practice.



Hires: http://farm5.static.flickr.com/4014/...f1ba3d42_b.jpg

So, to conclude this four point I like to put the idea that Apple, Google and Amazon are already exploring the idea of different conception for what and OS mean, and by creating SMALLER and packing LESS functions into the OS, it still possible to achieve higher level of results by exploring the combination of small pieces connected together by the net.

Valter

[netzen]

I also think it is worthwhile trying to do such things on the basis of Knoppix, at least in the first place. And, starting from collaborations of technically minded users, we will have a resource base.

Hi Capricorny.

One thing that I am motivated to try is a remaster of 6.2 with the "bare minimum" software infrastructure to run the VirtualBox.
Something like "USB Knoppix VBox Player", so people can "use it" as the basis for build all sorts of FOSS (free/open source) USB systems.

I succeed in "opening" and "closing" a remaster following the instruction on this site few weeks ago and I will be looking more at it later.
I'm excited with this because it combines TWO things I believe is "hot" stuff:

a) VirtualBox virtualization
b) NativeUSB (LiveUSB)

These 2, combined into one "bare minimum Knoppix VBox Player" can potentially create a tool for further exploration of the GNU/Linux in new ways...

[Definition for "NativeUSB"]
Perhaps the starting point to explore the current topic is to have a more precise (or at least one) definition for what a NativeUSB system is.
As the first diagram (on the thread) shows, I have the idea but I don't have a good definition for it.
Because I'm very interested on the topic, I keep taking notes and conducting little experiments, so, when I arrive at some useful definition for the "NativeUSB" and its scope, I will made it public and putting here...

Valter

[netzen]

Just want to put a complement to the last remark (the four) taking some notes of some features of the "cloud" and some of the "render side", so that it may help people to think about these two categories of "OSes" and what kind of roles they may play.



Hires: http://farm5.static.flickr.com/4035/...65f30fa9_b.jpg

Perhaps, if we can list a whole range of (desired or intended) features for these two categories we can end up with better ideas about what kind of practical implementations we should be targeting.

For me, there are few basic things that are very clear (and for me, source of inspiration):

a) The "desk top" paradigm reaches a peak and limit to deliver more value.
b) Cloud structures can achieve transcendent higher "computing stuff" results.
c) 24h connected and 24h battery organic mobility opens up a "new universe" in software.
d) For GNU/Linux, there are opportunities on the cloud side and on the render side.
e) The "cloud" uses the (traditional) notion of "server" but goes far beyond it.
f) The "render device" relies on the notion of user interface but goes beyond it.
g) The combination of "cloud" and "render device" IS A GREAT WAY to explore innovation.

Valter

[woodsmoke]

The above discussion has been quite nice and I don't really have anything that would meaningfully add to the discussion.

However, I'd like to relate something "tangential" that happened maybe five years ago.

A fellow down in Texas(USA) spent some money advertising on radio and in the newspaper that he would give a "live cd" to anyone who requested one.

Don't know the numbers because he didn't report them anywhere that I could find, but supposedly the response was....small....

So the question that was asked by some was...

We in "Linuxland" know what a "live cd" is but probably NObody in Windblowsland knows what a live cd is.

Now.........admittedly, the term "live cd" is a cool term, that is a fack jack as Bill Murrey said in the film...

But.... maybe there might be a rethink of "terminology" and a focused effort to put adverts into "free places" such as Craig's list, local newspaper forums, yada, yada, that would use a "quick" term that is more understandable by non Linux people while still retaining the "coolness" factor.

That might help, at least somewhat, get people in the door as it were so that other factors in this thread could then help funnel them further in.

I don't intend this to be a diversion of the main thrust of the thread, which is a good discussion, it is just a comment.



woodsmoke

[netzen]

We in "Linuxland" know what a "live cd" is but probably NObody in Windblowsland knows what a live cd is.
Now.........admittedly, the term "live cd" is a cool term, that is a fack jack as Bill Murrey said in the film...
But.... maybe there might be a rethink of "terminology" and a focused effort to put adverts into "free places" such as Craig's list, local newspaper forums, yada, yada, that would use a "quick" term that is more understandable by non Linux people while still retaining the "coolness" factor.
woodsmoke

Hi.
I think you are right and this seems to be THE KEY point for GNU/Linux to appeal to the "99%" group, Windblowseans, as you put...

No matter great the technical argument might be, for the "99%" group what we should be doing is to put a great dosage of effort on the process of communication and use a "language" effective for "them" instead of a more technical arguments that is being used inside the LinuxLanders groups.

In other words, perhaps THE key question is to depart from a "technical advantage approach" to a "mundane, why not use it too"...

I think that you raised the central question.
At least in my point of view, this is where the greatest challenge is and where the effort cannot be ignored...

Valter

[Capricorny]

My take on this is a very practical one in the first place: What is the minimum necessary to get things working?
I think as a "Live USB" current Knoppix CD installed on a stick is about as close as we get right now.
It has a minimalistic approach (e.g. LXDE), still you can do most of what you want right out of the box (e.g. Java installed).
I would like to see arguments for stripping it still further down. I don't see what big advantages we get.

In ordinary user space, premises are changing quite rapidly, some points:

• * Live CDs never were a great idea for actually running programs, too slow, too static.
  * The practical minimum for a software base nowadays is about 2GB installed, and growing. So current CD can be considered minimal.
  * USB3 will, in practice, solve the remaining problems with USB access speed. SSD USB3 disks will be fast enough.
  * 16 or 32 GB sticks will be enough for programs and personal user data for quite some time.
  * "Installs" in the Knoppix form of copying images to disk and loop-mounting them, is extremely useful and will only work better and better.
  * Remastering may actually become more usual, to cater for customizing and upgrading. User interests are diverse, and diverging, "install everything" is not a viable option.
  * Using services in an intelligent way, for instance a LAMP stack, may really help ordinary users.
  * We should NOT distinguish between "ordinary" and "expert" users until we really have to. "Experts" are getting increasingsly fed up with system maintenance and installs.
  * Configuration tweaks on standard software versions are essential, special versions must be avoided as far as possible.
  * With vmware workstation running under Knoppix on a memory stick, running Windows is within Linux USB-land, even without Wine, so no need for further fuss about that.
  * There may be a convergence going on, with different Debian-based Linuxes on mobile devices, like the Nokia N900. Utilizing the potentials in this may be essential.
  * Getting Wine to run important user software like Photoshop and video editing programs (especially when Linux versions are not easily available) is gaining in importance.

I think Klaus has moved in exactly the right direction with Knoppix, so we can safely build on that as a standard basis.
From there, I think that a minimum-maintenance "expert" version could be the way to go. Think, for instance, of enabling an interested user to re-create Quantian with minimum work.
Look at the reasons why most remastering versions are not updated.

I would like to add that making ISO images will still be a good idea for quite some time - they just have to be considered and used as starting point for tailoring.