Power of Opensource!!!

26 views
Skip to first unread message

sudhendra seshachala

unread,
Aug 13, 2008, 6:55:14 PM8/13/08
to cloud-c...@googlegroups.com
One question is kind of nagging me quite a bit. May be it is just me. I always think, Web 2.0 (some might not like it, but let keep it this way for the sake of arguments)
was driven by lot of innovation in open source arena. There are tons of companies which are relying on opensource community including billion dollar organizations like Oracle, Sun, HP, IBM. No body can deny the fact that they are adopting and nurturing and even selling open source through co branding (using opensource libraries, DLLs)  to customers. But, some how there are lot of folks, I am talking to who are kind of ignorant and nervous of adopting something  based on open source. There are consulting firms who provide services on these products. There is a whole lot of community driving the open source development. To me, Perl, PHP, Python, and the whole stack of Apache is an opensource product is'nt it (Thanks to IBM alpha group).
 
Has any one faced similar apprehensions about open source from customers/friends/boss? If yes, how do you take them head on?
Is no one using anything in the cloud platform?
Is it only me in a place where there are tons of companies with oil revenues who just want to spend money like never before?
Is IT still nervous about Opensource?
 
Can any one shed light on this topic?
 
Thanks and Regards
Sudhi
 
 

Ray Nugent

unread,
Aug 13, 2008, 11:39:07 PM8/13/08
to cloud-c...@googlegroups.com
Sudhi, can you send me the name of your CFO???!!! :-)


Ray

Rich Wellner

unread,
Aug 13, 2008, 11:56:16 PM8/13/08
to cloud-c...@googlegroups.com
I've worked in open source for over ten years at this point developing
grid solutions, writing books and papers and consulting to a variety of
organizations in just about any vertical market you can think of. My
take is that IT has done a complete 180 from ten years ago. Then there
were only a few organizations willing to 'risk their enterprise' on open
source. Now it is only the last of the laggards who have that view. In
fact, I haven't run into a single open source averse company in the past
two years. Not that they don't exist, I'm sure they do. But they are
becoming increasingly rare.

rw2

Chris Marino

unread,
Aug 14, 2008, 12:04:02 AM8/14/08
to cloud-c...@googlegroups.com
Sudhi, you cover a lot of territory here (and probably better suited to a different group), but let me take a stab at the first one.
 
The apprehension you indicate was everywhere at first, but over the past 2-3 (5??) years knowledge and awareness has replaced fear and ignorance.  In large part by the efforts of the Redhat, mySQL, and other early open source companies that faced these issues early. The Open Source Initiave (http://www.opensource.org/) did some important work as well.
 
Also there is a conference that specializes in this exact area and you can tell by looking at how the program has evolved over the past few years how far things have progressed (at least in the US).  Overcoming apprehension and fear was the hot topic a few years ago.
 
 
 
 
CM

Moshref

unread,
Aug 14, 2008, 1:22:24 AM8/14/08
to cloud-c...@googlegroups.com

We should really thank IBM from beginning, supporting first Linux, and then porting many of its mainframe OS, as well as Application to Linux, and open source. IBM with its install base has put solid stamp of approval to open source!

They are the best reference to use, for anybody that has apprehensions about open source!  

 

Rgds,

Moshref

Ajay ohri

unread,
Aug 14, 2008, 6:11:46 AM8/14/08
to cloud-c...@googlegroups.com
Opensource is an artificial apprehension floated by vested stakeholders , who find revenues threatened by the whole concept of people coding for reasons they dont understand (like community not money).

I can share an example from an outsourcing company In India. To certify their facilities were very secure , the IT department adopted a policy of no freeware. This meant that when a friend of mine requested Google Analytics to be installed on a computer, he was turned down on the grounds that Google softwares are free , while Microsoft Desktop search is what they will install.

A lot of places , the IT policy etcs. are determined by administrators who are either technically not competent to understand the flux and flexibility of open source softwares or by risk management/legal who opppose open source on the grounds that "who will we sue if the opensource software goes bad"

Regards,
Ajay
http://decisionstats.com

Reuven Cohen

unread,
Aug 14, 2008, 8:58:10 AM8/14/08
to cloud-c...@googlegroups.com

We're working on several cloud projects for a number of fortune 500 firms. None of those companies had any issue with our open source platform.


Reuven
www.enomalism.com

Sasikumar

unread,
Aug 14, 2008, 9:04:54 AM8/14/08
to cloud-c...@googlegroups.com
We at CDAC Mumbai has set up an open source software resource centre
(http://ossrc.org.in) a few years back, to help nurture the OSS
eco-system in India. We did some applications using open source, and
explored a number of solutions which are very good for various
purposes. There are a lot of Indian companies using open source (often
meant to include only Linux machines!).

When you refer to OSS, there are 2-3 different dimensions: the use of
Linux/Apache/etc as the base for your IT infrastructure, using OSS
methodology for development of applications, and using OSS at the user
level (openoffice, moodle, globus, etc). The popularity varies widely
across these dimensions.

- Sasi

--
M Sasikumar, KBCS/ETU/OSS Divisions, CDAC Mumbai (formerly NCST) -
Navi Mumbai campus

Rich Wellner

unread,
Aug 14, 2008, 10:48:54 AM8/14/08
to cloud-c...@googlegroups.com
Moshref wrote:

We should really thank IBM from beginning, supporting first Linux, and then porting many of its mainframe OS, as well as Application to Linux, and open source. IBM with its install base has put solid stamp of approval to open source!

Directly relevant to the folks on this list, they've also put millions of dollars into open source grid/cloud research and software related to security, resource management, data storage, metadata management and virtualization.

rw2 (who's own research was at one time supported by big brother)


Ben Kevan

unread,
Aug 14, 2008, 8:22:03 PM8/14/08
to cloud-c...@googlegroups.com

I think the current biggest negitivity against Open Source (I am a Sr. Linux
Administrator and just about everything I run is open source) in the business
is that it's a security hole. Why? They say because the source is open,
someone can find an exploit, exploit it and not report the exploit. This
typically isn't the case as many eyes see the code, and exploits are fixed
all the time.

It "could" happen, but hey it happens just as often (if not more) with closed
source.

Ben


Jan Klincewicz

unread,
Aug 15, 2008, 8:04:49 AM8/15/08
to cloud-c...@googlegroups.com
Look at Xen and Cirix.  Xen was originally conceived as a hypervisor stack for a world-wide grid.  It is turning out to be a pretty viable commercial product (XenSever, Virtual Iron etc.) wrapped around an Open Source core.  The fact that large commercial enterprises have adopted it has only made the Open 'source comminity stronger (withness the phenomenal trajectory of improvements compared to an all-commercial proprietary approach vs. VMware.

The end result is a more robust, less-expensive, better-supported platform for all users and the commotidization of the core technology causing even Microsoft to value a hypervisor at less than twenty bucks.
--
Cheers,
Jan

Steve Chambers

unread,
Aug 15, 2008, 10:06:10 AM8/15/08
to cloud-c...@googlegroups.com

Let’s stop with the Open Source chest-thumping – it’s not what this group is about, surely?

 

Ignoring the bogus claims about large enterprise adoption of Xen, I’m confused about the claim of Xen’s “phenomenal trajectory of improvements compared to VMware?”

 

Have you recently compared the feature set of Xen to VMware?  Let me help you:

 

 

 

  • Microsoft - only just brought out Hyper-V V1, whereas our V1 product was out in early 2000.

 

I’m all for distributed and open development, and we do that with our academic and community source programs, but can we please back on course to Cloud?

 

Thanks!

 

Steve Chambers  |  Senior Architect  | VMware Technical & Enterprise Marketing  |  scha...@vmware.com  |  M: +44 7870 160976


From: cloud-c...@googlegroups.com [mailto:cloud-c...@googlegroups.com] On Behalf Of Jan Klincewicz
Sent: 15 August 2008 13:05
To: cloud-c...@googlegroups.com
Subject: Re: Power of Opensource!!!

 

Look at Xen and Cirix.  Xen was originally conceived as a hypervisor stack for a world-wide grid.  It is turning out to be a pretty viable commercial product (XenSever, Virtual Iron etc.) wrapped around an Open Source core.  The fact that large commercial enterprises have adopted it has only made the Open 'source comminity stronger (withness the phenomenal trajectory of improvements compared to an all-commercial proprietary approach vs. VMware.

David Medinets

unread,
Aug 15, 2008, 10:24:59 AM8/15/08
to cloud-c...@googlegroups.com
> VMware - 10 years of development, with full-time engineers now numbering in
> the thousands: http://vmware.com/products/vi/features.html

Did you really mean to say that VMware has thousands of software
developers writing code? Are you talking about 2,000 or 4,000? The
URL that you reference make no mention of the size of the development
team.

Reuven Cohen

unread,
Aug 15, 2008, 10:53:08 AM8/15/08
to cloud-c...@googlegroups.com

As a long time open source advocate I'm kind of bias. That said, we're
seeing a tonne of interest in our upcoming support for VMware's ESXi
in our open source Enomalism platform (AGPL). Most corporate users
have no interest in moving away from their VMware deployments. They
are looking at ways to reduce their costs. I think the general
consensus is if it ain't broke why fix it. That said, the pricing for
the higher end, feature bloated, centralized VMware offerings are not
exactly price conscience or particularly cloud oriented.

Let's for a minute assume someone is going to use VMware to build out
there public cloud (similar to EC2) using VMware Infrastructure
Foundation for 2 processors + Gold (12x5) 1 Year Support at $1,540.

Typically we're seeing an 8 core machine as the basis for our
deployments, so $6,160 per machine per year.

Hours in 1 year = 8 765 / $6,160 = .70/h an hour for vmware VI as a baseline.

At 70 cents per hour, it would be impossible to compete with Amazon's
baseline of 10 cents. We're also seeing the entry point for most
clouds in the area of 1000 servers. We're talking about a multi
million dollar investment in software. Most cloud deployments we're
seeing today are Xen based, not because of quality or ease of use, but
cost.

reuven

Reuven Cohen

unread,
Aug 15, 2008, 11:22:20 AM8/15/08
to cloud-c...@googlegroups.com

A few have pointed out that I forgot to include the division of the
amount of virtual servers which does reduce the cost.

70cents is per physical server, so 10 vm's per machine would be be 7
cents per vm, which is better.

ruv

Daniel Ciruli

unread,
Aug 15, 2008, 12:26:11 PM8/15/08
to cloud-c...@googlegroups.com

To bring this slightly more on-topic: I think that, in reality, cloud computing may have a side effect of making people care very little about whether their cloud vendor is using open-source or not.

 

One of the major ideas behind cloud computing is that the user no longer has to worry about what the infrastructure is running, right? If I write my app in Python and upload it to AppEngine, does it matter to me whether Google is using an open source operating system or their own, private version of Linux? Does the *infrastructure* under EC2 matter to a user as much as the uptime and the experience?

 

I’m certainly not sounding a death knell for OSS, nor for commercial. I guess I’m agreeing (a bit) with Steve – this just seems orthogonal to the discussion of cloud computing.

 

Dan

Khazret Sapenov

unread,
Aug 15, 2008, 12:45:25 PM8/15/08
to cloud-c...@googlegroups.com
On Fri, Aug 15, 2008 at 12:26 PM, Daniel Ciruli <danc...@hotmail.com> wrote:

To bring this slightly more on-topic: I think that, in reality, cloud computing may have a side effect of making people care very little about whether their cloud vendor is using open-source or not.

 

One of the major ideas behind cloud computing is that the user no longer has to worry about what the infrastructure is running, right? If I write my app in Python and upload it to AppEngine, does it matter to me whether Google is using an open source operating system or their own, private version of Linux? Does the *infrastructure* under EC2 matter to a user as much as the uptime and the experience?

 

I'm certainly not sounding a death knell for OSS, nor for commercial. I guess I'm agreeing (a bit) with Steve – this just seems orthogonal to the discussion of cloud computing.

 

Dan

 
I concur with Daniel's point, whatever is underneath is no matter to end user of cloud.
 
If open-source monkeys bring less valuable/attractive (or more expensive in terms of TCO) end product to cloud user, he would go with propietary banderlogs, led by Mowgli :) 

Jan Klincewicz

unread,
Aug 15, 2008, 3:33:16 PM8/15/08
to cloud-c...@googlegroups.com
What you say is very true for the ultimate CONSUMER of the cloud, the end-user. 

For the folks BUILDING the infrastructure for the cloud, scalability, and the economics thereof is hugely important.  I would think it matters a great deal to Amazon, Google and their shareholders on what the costs associated with their capital as well as operating costs are with regards to revenue (real or potential.)

Jan

On Fri, Aug 15, 2008 at 12:26 PM, Daniel Ciruli <danc...@hotmail.com> wrote:



--
Cheers,
Jan

Steve Chambers

unread,
Aug 15, 2008, 3:57:25 PM8/15/08
to cloud-c...@googlegroups.com
Just don't equate open source with free or best value for money :-)
Steve Chambers | Senior Architect | scha...@vmware.com | +44 (0) 7870 160976

----- Original Message -----
From: cloud-c...@googlegroups.com <cloud-c...@googlegroups.com>
To: cloud-c...@googlegroups.com <cloud-c...@googlegroups.com>
Sent: Fri Aug 15 12:33:16 2008
Subject: Re: The weakness of Open Source

What you say is very true for the ultimate CONSUMER of the cloud, the end-user.

For the folks BUILDING the infrastructure for the cloud, scalability, and the economics thereof is hugely important. I would think it matters a great deal to Amazon, Google and their shareholders on what the costs associated with their capital as well as operating costs are with regards to revenue (real or potential.)

Jan


On Fri, Aug 15, 2008 at 12:26 PM, Daniel Ciruli <danc...@hotmail.com> wrote:


To bring this slightly more on-topic: I think that, in reality, cloud computing may have a side effect of making people care very little about whether their cloud vendor is using open-source or not.



One of the major ideas behind cloud computing is that the user no longer has to worry about what the infrastructure is running, right? If I write my app in Python and upload it to AppEngine, does it matter to me whether Google is using an open source operating system or their own, private version of Linux? Does the *infrastructure* under EC2 matter to a user as much as the uptime and the experience?



I'm certainly not sounding a death knell for OSS, nor for commercial. I guess I'm agreeing (a bit) with Steve – this just seems orthogonal to the discussion of cloud computing.



Dan





From: cloud-c...@googlegroups.com [mailto:cloud-c...@googlegroups.com] On Behalf Of Steve Chambers
Sent: Friday, August 15, 2008 7:06 AM

To: cloud-c...@googlegroups.com

Subject: The weakness of Open Source



Let's stop with the Open Source chest-thumping – it's not what this group is about, surely?



Ignoring the bogus claims about large enterprise adoption of Xen, I'm confused about the claim of Xen's "phenomenal trajectory of improvements compared to VMware?"



Have you recently compared the feature set of Xen to VMware? Let me help you:



* VMware - 10 years of development, with full-time engineers now numbering in the thousands: http://vmware.com/products/vi/features.html



* Xen - ? years of development, with a dwindling Open Source following (thanks to KVM) http://citrix.com/English/ps2/products/feature.asp?contentID=683160



* Microsoft - only just brought out Hyper-V V1, whereas our V1 product was out in early 2000.



I'm all for distributed and open development, and we do that with our academic and community source programs, but can we please back on course to Cloud?



Thanks!



Steve Chambers | Senior Architect | VMware Technical & Enterprise Marketing | scha...@vmware.com <mailto:scha...@vmware.com> | M: +44 7870 160976






--
Cheers,
Jan




Anshu Sharma

unread,
Aug 15, 2008, 4:12:13 PM8/15/08
to cloud-c...@googlegroups.com

Also, remember that list prices are meaningless when it comes to large deals. You transition from per processor pricing to large scale ELA's at that level. Do you really think AT&T's of the world today with 1000s of servers pay billions to likes of oracle, microsoft. they pay 10s of millions, even 100s but its nowhere near list prices would dictate.

anshu
--
________________________________________________
Visit My Blog at http://www.anshublog.com

Jan Klincewicz

unread,
Aug 15, 2008, 5:04:25 PM8/15/08
to cloud-c...@googlegroups.com
Unless you're Amazon or Google ....

I realize I am new to this group, but I get the strong impression that most posters are looking at cloud computing from the consumer side rather than the producer side.

Having had equal-length careers on both sides of IT (Data Center management and vendor Systems Engineer) I have seen a number of "paradigm shifts", from Time Sharing, to PCs to LAN and WANs to Client Server to Web-based to Web 2.0, and looking through the eyes of an MBA turned techie, have seen a number of people get rich from each shift, and a much larger number get screwed.

Ultimately, Cloud-Computing appears similar to Outsourcing in that it provides a means for organization to shift the burden of providing IT infrastructure from Internal resources to External under the guise of eliminating an entire department which does not reflect the "core competencies" of the organization.

That is all well and good, and it would seem, on the surface, absurd for pharmaceutical companies and law firms to be in the business of providing storage and IP switching, when their money could be spent on Chemists,  Epidemiologists and Attorneys.

From a wider angle, however, "generic" infrastructure may not provide the compliance necessary to deal effectivelt with HIPPA, Sarbanes-Oxley etc. that is required in an increasingly complex and regulated environment.

If IT does, in fact, become commodized to a "cloud" the same as water or electricity, we will likely see a dramatic decrease in the number of graduates seeking careers in Computer Science.

Today, American companies like Microsoft, HP, IBM, RedHat and VMware create huge amounts of value, and provide high-paying jobs for not only engineers, but marketers, field sales, graphic artists, channel reps and resellers.

If organizations are willing to accept "generic" IT to make a few extra bucks  for shareowners any given quarter, those opportunities will dry up and dissapear rapidly.

It is highly plausible for companies like Google and Amazon to buy bare-bones motherboards, RAM and disk from Taiwan, and hire a few hundred REAL SMART admins whom know Linux, Xen, Python and Apache and run their data centers with zero support dollars going to vendors.

In this case, "value for your money" means a lot more to an unemployed admin with a dozen certs in his/her pocket when their jobs are not just outsourced, but ELIMINATED because what they studied and gained experience in for years has been absorbed by a "cloud" deemed "good enough" by some board of directors comprised of old men who use MACs at home because they still cannot figure out Windows.

Personally, I have made a decent living off of proprietary hardware ands software for most of my adult life.  I think there are many things achievable within IT organziations which owe a debt to the vendors whose innovations are driven by the profits they make as a result of their intellectual property.

At the same time, I am not so naive as to dismiss pure "Open Source" software as "weak" because it is not created by an army of thousands of developers.

To my knowledge, some of the best and most influential software successes (Unix / C / CP/M, BASIC, PASCAL, Visicalc etc.) were the work of a single individual or a pair of smart guys working in a garage, or coding freehand on yellow legal pads in assembler.)

The Internet has created a multiplier effect for this "lone genius" model, and the very concept of Open Source will naturally assure that software developed for the "greater good" will evolve more efficiently than code developed by a committee of computer-illiterate Marketing and Finance types looking no farther than next quarter's earnings to guide the development of their products.

I don't mean to rain on the cloud, but after a few decades in this business it is hard not to see history repeat itself.

Jan



--
Cheers,
Jan

Laurent Therond

unread,
Aug 15, 2008, 5:33:19 PM8/15/08
to cloud-c...@googlegroups.com
I think it goes back to the enablers vs. users divide.

Enablers and users have different contingencies to deal with.

Enablers want recurring revenues.
Users want services that bring more value than they cost.

It's why I believe users should not care about software or how those
valuable services are delivered.
In that frame of mind, Cloud Computing could be the future of computing.
But, as always, it won't be the solution to every problem, even though
countless enablers will try to market it to all.

David Medinets

unread,
Aug 15, 2008, 5:39:36 PM8/15/08
to cloud-c...@googlegroups.com
On Fri, Aug 15, 2008 at 5:04 PM, Jan Klincewicz
<jan.kli...@gmail.com> wrote:
> I realize I am new to this group, but I get the strong impression that most
> posters are looking at cloud computing from the consumer side rather than
> the producer side.

While you wrote a long eloquent message, I have no idea what your
point was. Are you complaining that companies will 'vanish' technical
jobs and they should not do so out of love for their fellow citizens?
You don't like outsourcing, even to 'american' companies? Think fifty
or a hundred years from now, will we still need millions of
programmers throughout the world? Of course not. Everything will be in
the cloud except for some small percentage of really paranoid
organizations like the Templar Knights or the Freemasons.

I especially liked the idea of old men sitting at home using their
macs because Windows is too complex, I have never heard of that
scenario. What is your source for this observation?

Reuven Cohen

unread,
Aug 15, 2008, 5:51:07 PM8/15/08
to cloud-c...@googlegroups.com
On Fri, Aug 15, 2008 at 3:57 PM, Steve Chambers <scha...@vmware.com> wrote:
> Just don't equate open source with free or best value for money :-)
> Steve Chambers | Senior Architect | scha...@vmware.com | +44 (0) 7870 160976
>

Steve, I agree with you on this. Open source is a license and is just
part of the overall picture.

ruv

Jan Klincewicz

unread,
Aug 15, 2008, 5:55:28 PM8/15/08
to cloud-c...@googlegroups.com
Understood Laurent, but how does that differ than what has been going on for years ??   You bet on the the wrong horse, you lose a lot.  I saw a lot of yin-yangs buy into Micro-Channel and OS/2 because "nobody ever lost their job choosing IBM <g>.

What happens if you select a Cloud and they evaporate ??  I agree, things like web-servers and e-mail are relatively simple to move out to clouds (assuming you have control of backups). If the  lifeblood of my organization were aqt stake, I'd have second thoughts ....

In a sense, clouds abstract the hardware and even many of the software layers, but the data is the real stickler....

Jan
--
Cheers,
Jan

Ray Nugent

unread,
Aug 15, 2008, 10:44:35 PM8/15/08
to cloud-c...@googlegroups.com
You have to isolate your API layer, make it data driven. Then switching Cloud providers should be significantly less painful.

----- Original Message ----
From: Jan Klincewicz <jan.kli...@gmail.com>
To: cloud-c...@googlegroups.com
Sent: Friday, August 15, 2008 2:55:28 PM
Subject: Re: The weakness of Open Source

Sam Johnston

unread,
Aug 16, 2008, 3:55:25 AM8/16/08
to cloud-c...@googlegroups.com
On Sat, Aug 16, 2008 at 4:44 AM, Ray Nugent <rnu...@yahoo.com> wrote:
You have to isolate your API layer, make it data driven. Then switching Cloud providers should be significantly less painful.

Exactly - I've got some (hopefully good) ideas about this that I've been working on but more on that later.

In terms of Open Source, I believe the vast majority of cloud computing solutions today are run on open source software stacks, or at least open source operating systems (eg Linux). On the client side a large and increasing number of users are using open source browsers (eg Firefox).

I don't see anything that will change this on the server side - at these scales you just need to be able to get in and tweak the stack from the network driver through the database itself and you've got a snowflake's chance in hell of doing that with most proprietary systems.

I also believe this trend will continue, particularly on the client side where the cost of the hardware will continue to drop to the point where software licensing cannot be carried on top of it - even the first generation of CherryPal, Eee PC and XO machines, at around $200, can't support a $100 OS license (plus apps etc).

In terms of the actual licensing, now that software is merely 'performed' rather than 'conveyed' the 'triggers' for most licenses requiring release of modifications don't fire so licenses like the Affero GPL are appearing to address this 'loophole'.

Finally, vendors who have found themselves in the unfortunate position of being in competition with open source products ought to be very careful about calling into question the value that it provides - we've heard it all before and are using it extensively anyway.

Sam

Steve Chambers

unread,
Aug 16, 2008, 4:55:49 AM8/16/08
to cloud-c...@googlegroups.com
We've been through this Providor vs. Consumer conundrum for years at vmware - we are disrupting the delivery model from the old, broken 'one app one physical server' model and moving to a shared model where your Consumer does not know which physical server their application runs on, how much memory they _really_ have access to, nor which disk nor which switch they are using.

Some of our direct customers (Providors) migrate their customers (Consumers) from physical to virtual without telling them: in the words of one senior manager at a global bank 'they don't ask how we generate their electricity so why should they ask about the server type other than its capacity, performance and availability?'

We hide features like high availabily, QoS and dynamic workload balancing ( is that the feature bloat we were accused of earlier in this thread? ) and many other 'implementation' details, and provide an open and well supported api for anyone to code to (we have a healthy ecosystem of community and commercial extensions) and we are actively involved in open standards like DMTF, OVF, VMmark.

So while vmware might not be a perfect cloud today, it exhibits several Cumulus characteristics already and is already changing Providor and Consumer attitudes to the delivery and consumption of IT services today.

There has been a lot of resistence to change but the friction is reducing and I hope that in some small way we have been doing our bit to facilitate a move towards the cloud.

Cheers
Steve
Steve Chambers | Senior Architect | scha...@vmware.com | +44 (0) 7870 160976

Sent: Fri Aug 15 14:33:19 2008
Subject: Re: The weakness of Open Source


Cameron

unread,
Aug 16, 2008, 12:29:53 PM8/16/08
to Cloud Computing
Jan -

> Look at Xen and Cirix

This is a very interesting one, considering that it has been rumored
that Citrix bought the people behind the open source Xen *at the
behest of Microsoft* in order to cause confusion in the virtualization
space until the Microsoft Windows virtualization solution had the few
extra years of time necessary for it to get out into the market ..

(I am obviously not speaking for any particularly company in the
space, since I've heard it rumored that my own employer sells a Xen-
based product and I'd hate to consider the multitude of possibilities
in terms of potential conspiracies that one could come up with given
that little detail ..)

Peace,

Cameron Purdy | Oracle
http://www.oracle.com/technology/products/coherence/index.html

Jim Starkey

unread,
Aug 16, 2008, 1:12:31 PM8/16/08
to cloud-c...@googlegroups.com
Steve Chambers wrote:
> Just don't equate open source with free or best value for money :-)
> Steve Chambers | Senior Architect | scha...@vmware.com | +44 (0) 7870 160976
>
>

Huh? Am I missing something? You can't equate open source with free,
but all open source software is free if you're willing to abide by the
license.

That said, cloud computing has the potential to do an end run around
most open source licenses, which only require publication of source or
changes when software is distributed. A cloud provider distributes no
software and is therefore immune to both GPL and MPL derivative
licenses. Google, for example, pays only a pittance for their lavish
use of open source software in their infrastructure. This is perfectly
legal, but not exactly what most open source developers had in mind.

What effect will this have on the open source community in the long
run? Hard to say.

Jan Klincewicz

unread,
Aug 16, 2008, 1:51:09 PM8/16/08
to cloud-c...@googlegroups.com
Under ordinary circumstances I am willing to accept just about ANY
conspiracy theory, but I don't think MSFT found it necessary to "buy
time.". They seem to show little interest in really supporting Linux,
and basically, IMO, need to devlop their own solution toi keep the
market from heading towards OS-less "appliances". I believe I heard
Ellison talk about that 10 years ago.

In any event. I think it is clear that the hyperviisor stack will soon
be pereived as having near ZERO value, and the Management tools and
utilities that add REAL High Availability (not some lame best effort
restart) will be where the money is (what will be left to scrap over).

In the end, it is also the Support Infrastructure (which if cloud
computing really takes of may be handled by the cloud providers) which
will be needed.

In any event, I do not see continuing double-digit profit growth for
Proprietary hypervisor vendors now that a monopoly has been broken. I
don't think Wall Street sees that either.

--
Sent from Gmail for mobile | mobile.google.com

Cheers,
Jan

Alex Kurganov

unread,
Aug 16, 2008, 12:32:04 PM8/16/08
to cloud-c...@googlegroups.com
The comparison to electricity providers is incomplete at best. Even
your sump-pump requires a backup battery -----that costs more then the
sump pump itself ----if you have some $10K worth of finished basement
value. A company business is worth significantly more. So, if one wants
to implement a cloud computing model, then a seamless mechanism must be
developed to switch-over to a local backup system. And that (having a
full backup system) may negate the whole cost advantage of cloud
computing except for very specific cases. BTW, the new energy prices
may enable local electricity generation model using solar and other
alternative energy sources. So, it is not clear at all that utility
model is economically superior even for electricity generation when oil
is at $100+ per barrel. I don't think we have proved that the economics
are there for a true, low risk (for the consumer) utility computing
model yet. And electricity grid analogies don't really help given
recent massive failures (of both the electricity and the computing
clouds), the increasing costs of remotely generated electricity. The
ability to replace the physical server model with the virtual one is of
tremendous value, but it alone does not make the computing cloud model
feasible at this stage.

Regards
Alex

Thomas Lockney

unread,
Aug 16, 2008, 2:56:46 PM8/16/08
to cloud-c...@googlegroups.com
On Sat, Aug 16, 2008 at 10:12 AM, Jim Starkey <jsta...@nimbusdb.com> wrote:

Steve Chambers wrote:
> Just don't equate open source with free or best value for money :-)
> Steve Chambers | Senior Architect | scha...@vmware.com | +44 (0) 7870 160976
>
>

Huh?  Am I missing something?  You can't equate open source with free,
but all open source software is free if you're willing to abide by the
license.

That's hardly true. Every system you operate has a cost -- the cost of operations, maintenance, etc. Not too mention the associated costs of integration if you end up needing to add on additional services. Further, if one piece of software is not free but is more efficient at resource usage than a free piece of software, you're looking at a very different picture over the long run. These are factors you MUST consider when running an IT operation. So, no, open source does not mean free.

And, FYI, I am a very serious proponent of open source solutions -- but I also know well that there are always costs involved.

~thomas

Massimo

unread,
Aug 16, 2008, 3:05:40 PM8/16/08
to cloud-c...@googlegroups.com
I agree Thomas,

this is why In MySQL we talk always about TCO (total cost of ownership):
is not only the saving in term of licenses or subscriptions (depends on
the business model) but you have to take care about the hardware, the
headcounts involved, hw is easy to find experts on the market, and so on...

Thomas Lockney ha scritto:


> On Sat, Aug 16, 2008 at 10:12 AM, Jim Starkey <jsta...@nimbusdb.com
> <mailto:jsta...@nimbusdb.com>> wrote:
>
>
> Steve Chambers wrote:
> > Just don't equate open source with free or best value for money :-)
> > Steve Chambers | Senior Architect | scha...@vmware.com

> <mailto:scha...@vmware.com> | +44 (0) 7870 160976

Jan Klincewicz

unread,
Aug 16, 2008, 4:11:48 PM8/16/08
to cloud-c...@googlegroups.com
Absolutely agree. The reason that a Red Hat or SuSe exist is that no
responsible organization would run their IT un unsupported platforms
UNLESS they are willing to bear the burden of hiring manu smart,
experienced and EXPENSIVE people. Amazon and Google seem to be
comfortable with that model.

That does not mean that using Open Source is INHERENTLY better or
worse than proprietary versions. The market decides whether for
logical or emotional reasons which products succeed.

It is difficult to fathom that Oracle pays no attention to MYSQL or
that Microsoft can ignore the growth of Linux.

Commercial prodcts may have more features where Open Soirce versions
may boast better performance. To disparage either is ludicrous.

--

Jim Peters

unread,
Aug 16, 2008, 4:35:46 PM8/16/08
to cloud-c...@googlegroups.com
It's not readily apparent from their website, but Oracle owns InnoDB, which is a very popular transactional engine for MySQL ...
--
Jim Peters
+415-608-0851

Jim Starkey

unread,
Aug 16, 2008, 4:50:52 PM8/16/08
to cloud-c...@googlegroups.com
Jan Klincewicz wrote:
> Absolutely agree. The reason that a Red Hat or SuSe exist is that no
> responsible organization would run their IT un unsupported platforms
> UNLESS they are willing to bear the burden of hiring manu smart,
> experienced and EXPENSIVE people. Amazon and Google seem to be
> comfortable with that model.
>
Yes, enterprise customers are more comfortable with formal support. But
they have a choice of hiring Red Hat, SuSE, MySQL, independent
consultants, or develop their own support capabilities. They may feel
most comfortable with the original developers, but that is their
choice. If you want a bug fixed in Oracle, you pay Oracle.

> That does not mean that using Open Source is INHERENTLY better or
> worse than proprietary versions. The market decides whether for
> logical or emotional reasons which products succeed.
>
Open source is almost never the technology leader. Note, for example,
the popularity of the slogan "Innovation Happens Elsewhere". Often,
stability trumps cutting edge. Other times, it doesn't.

You can expect open source database systems to tolerate clouds, but not
exploit them. Later, when other folks have lead the way, they will
slowly catch up.


> It is difficult to fathom that Oracle pays no attention to MYSQL or
> that Microsoft can ignore the growth of Linux.
>

Oracle pay a vast amount of attention to MySQL, enough attention to buy
MySQL's only transactional storage engine out from underneath them.
I've never given a talk at a MySQL user conference without an Oracle
vice president in the first or second row (nice fellow, actually).

And don't believe for a nanosecond that Ray Ozzie doesn't spend a good
deal of his waking time worrying about Linux and open source. Their old
strategy was FUD. Now it seems to border on peaceful co-existence.
>

Laurent Therond

unread,
Aug 16, 2008, 5:33:49 PM8/16/08
to cloud-c...@googlegroups.com
If I were to take a rather cynical view of the software industry, I
would say it is not in the best interest of software vendors to deliver
software that may solve a problem or a class of problems for years.
Simply because software vendors are totally bound to realize recurring
revenues, which are necessary to cover their operational expenses and
growth.

For consumers of software, it is foolish to focus on the intricacies of
software.
If you do, you end up in a rat race and a forever expanding IT budget,
where you find yourself advised to "Keep up with the Joneses". Hence,
you go along and stay hot and bothered with such details as RPC, CORBA,
SOM, COM, DCOM, JavaBeans, COM+, Jini, Web services, etc.
Each and every time, Gartner, IDC, Forester tell you (for a fee!) that
your enterprise must integrate these technologies or be at risk of a
severe lack of competitiveness.

I believe this is why Cloud Computing and SaaS are so appealing to
consumers of software. These approaches prioritize high level
requirements and SLAs vs. technical intricacies.

From a pricing standpoint, when Microsoft adopted their various
subscription models, the message was clear: Software in itself has no
interest ; it's the business value that software can deliver that matters.

Finally, I see vmware as a provider of tools that could boost the
development of Cloud Computing and SaaS. I don't think any business
person should focus how you do the things you do, but you are well
placed to be part of the solution they seek.

Jan Klincewicz

unread,
Aug 16, 2008, 6:10:58 PM8/16/08
to cloud-c...@googlegroups.com
Very true statements, and I agree that it behooves most organizations
to abstract as much technology as possible to shield themselves both
from the costs of change, as well as the complexity of IT "plumbing"
as it exists.

But the fact remains that the burden of dealing with change and
complexity is merely shifted to the "Cloud Provider" which is not
really that far removed from Outsourcing.

Let's face it, IT is a complex business, with a lot of moving parts.
If consolidation takes place and a smaller number of large IT shops
(cloud providers) evolve this will naturally result in a smaller
number of vendors offering a reduced number of technologies.

SaaS will undoubtedly be a boon to many organixations who should not
be in the IT business, but I fear from a larger perspective that
innovation will suffer without the competition which has existed in
the past century.

--

sa...@samj.net

unread,
Aug 17, 2008, 5:39:47 AM8/17/08
to cloud-c...@googlegroups.com
It's outsourcing with economies of scale. Outsourcing is about having
your arbitrary needs serviced by a third party. Though cloud is in
many ways the ultimate in outsourcing, it's only possible because all
the users are consuming exactly the same (commoditised) thing.

Also, there's nothing to say a cloud provider has to be massive.
TrustSaaS (and the other monitoring as a service providers) are not
(currently) run across thousands of servers but still play a critical
role in the cloud computing ecosystem.

Sam

Jan Klincewicz

unread,
Aug 17, 2008, 11:24:47 AM8/17/08
to cloud-c...@googlegroups.com
      I'm sure a large number of users can run their businesses on commodity platforms for common apps, like e-Mail or Sales tracking.  Perhaps cloud Providers will emerge to handle specialized verticals.  Unless they do, I don't see how the highly-regulated environments of Health Care, Pharma, Financial Markets etc. will be served.

      I've had many customers who were attempting to subscribe to ITIL best practices, and their Change Control policies alone prohibited taking advantage of a lot of  benefits that today's "dynamic" data centers can provide.

      It would seem to me that the economic benefits of cloud computing would tend to come from economies of scale, where the compute resources are purchased in volume.  If end-users need to manage multiple cloud providers, they are not gaining as much as they would from a one-stop shop.

Cheers,
Jan
--
Cheers,
Jan

sa...@samj.net

unread,
Aug 17, 2008, 7:13:56 PM8/17/08
to cloud-c...@googlegroups.com
On 8/17/08, Jan Klincewicz <jan.kli...@gmail.com> wrote:
> I'm sure a large number of users can run their businesses on commodity
> platforms for common apps, like e-Mail or Sales tracking. Perhaps cloud
> Providers will emerge to handle specialized verticals. Unless they do, I
> don't see how the highly-regulated environments of Health Care, Pharma,
> Financial Markets etc. will be served.

Compliance is expensive and providers like postini have already taken
care of this for you by attaining SAS 70 type II, webtrust and other
audits-this alone is a significant value add.

> I've had many customers who were attempting to subscribe to ITIL best
> practices, and their Change Control policies alone prohibited taking
> advantage of a lot of benefits that today's "dynamic" data centers can
> provide.

...and they surely (will) have competitors who won't be hamstrung.

> It would seem to me that the economic benefits of cloud computing
> would tend to come from economies of scale, where the compute resources are
> purchased in volume. If end-users need to manage multiple cloud providers,
> they are not gaining as much as they would from a one-stop shop.

Do you know which power station wiggled the electrons you just used to
send these 1's and 0's? No, and you don't need to. Something like 80%
of mine were nuclear but I'm certainly not about to build a nuclear
plant to benefit from the technology, and nor do I need to.

Sam on iPhone

Jan Klincewicz

unread,
Aug 18, 2008, 7:49:46 AM8/18/08
to cloud-c...@googlegroups.com
Are you implying that the analogy between providers of electricity (or water, etc.) and providers of IT services is a valid one ?  I've always found that comparison rather specious.

I may not care who GENERATED the electricity, but I am going to care about how I get single phase, vs. three phase, and how many KVA I get get to "n" number of racks, etc.

Cheers,
Jan
--
Cheers,
Jan

sa...@samj.net

unread,
Aug 18, 2008, 10:56:14 AM8/18/08
to cloud-c...@googlegroups.com
On 8/18/08, Jan Klincewicz <jan.kli...@gmail.com> wrote:
> Are you implying that the analogy between providers of electricity (or
> water, etc.) and providers of IT services is a valid one ? I've always
> found that comparison rather specious.

It has it's place and it's an efficient way of explaining to users
what all the fuss is about.

> I may not care who GENERATED the electricity, but I am going to care about
> how I get single phase, vs. three phase, and how many KVA I get get to "n"
> number of racks, etc.

Indeed even with a well defined commodity (the electron) there are
many variables. We can manage this complexity by standardizing on
platform neutral technologies like java rather than processor native
binaries and by breaking problems up to increase granularity, but
we're making progress in this direction with initiatives like OVF.

Jan Klincewicz

unread,
Aug 18, 2008, 2:02:57 PM8/18/08
to cloud-c...@googlegroups.com
Yes .. I hope to see Kensho come to fruition sooner than later.  Again, there seem to be two distinct camps in this group.... End-user oriented, and Provider-oriented.  Each has a radically different different take on cloud computing (like two blind men at different ends of an elephant trying to describe it...)   To that mix, I'd have to throw in the vendors, who would be pitching to cloud providers in the future as opposed to end-user customers.

This is pretty radical stuff , so I guess we'll have to see how things pan out.  I've seen the gamut of customers in my near three decades of IT, from early-adopters to adamant change-resisters.  Given that climate, this won't happen overnight....
--
Cheers,
Jan

Jim Peters

unread,
Aug 18, 2008, 11:56:08 AM8/18/08
to cloud-c...@googlegroups.com
electrons != sensitive data
--
Jim Peters
+415-608-0851

Vinayak Hegde

unread,
Aug 18, 2008, 7:06:53 PM8/18/08
to cloud-c...@googlegroups.com
On Thu, Aug 14, 2008 at 6:28 PM, Reuven Cohen <r...@enomaly.com> wrote:
> We're working on several cloud projects for a number of fortune 500 firms.
> None of those companies had any issue with our open source platform.

I have written a blog post summarizing the relationship between
different OSS licenses and the Affero GPL, which is of particular
importance to Cloud comuting and SaaS vendors. A lot of cloud
computing infrastructure uses Linux, the LAMP stack, Xen in addition
to many many tools for manging the network and writing and deploying
software.

http://thoughts.vinayakhegde.com/2008/08/18/cloud-computing-and-open-source/

-- Vinayak

Chris Marino

unread,
Aug 19, 2008, 8:31:08 AM8/19/08
to cloud-c...@googlegroups.com

Vinayak, I think you have a very optimistic view on the the efficacy of the the AGPL.  IHMO, it solves nothing.  I wrote about it here and here.  The gist of the issue is that it's just too easy to get around the provisions of this license.  From my original post:

This ambiguity leads directly to the trivial ways in which the license can be avoided: Take the AGPL’d code, rip out the ‘remote user interaction’ interface components and replace them a ‘local, machine interacting’ interface (i.e. a different API) and I’m done. If you don’t think this is possible, then we disagree on what a user interaction is and a machine interaction is. Which is exactly my point. A slightly more liberal reading of this license would conclude that a simple (local) proxy isolated the ‘remote user’ from the Work, which undermines the license in it’s entirely.

‘ASP loophole’ is too complex a problem to be addressed by the addition of a single paragraph in a document that had different licensing objectives. Or for that matter, by any license that covers only a portion of a complex, distributed service offering.

Given the  widespread interest in anything 'cloud' and the meager adoption of the AGPL, I think people are (not) voting with their feet.


__________________________________
Chris Marino
SnapLogic, Inc.

Really Simple Integration
www.snaplogic.com
650-655-7200

Reuven Cohen

unread,
Aug 19, 2008, 9:20:32 AM8/19/08
to cloud-c...@googlegroups.com
On Tue, Aug 19, 2008 at 8:31 AM, Chris Marino <ch...@snaplogic.com> wrote:
> Vinayak, I think you have a very optimistic view on the the efficacy of the
> the AGPL. IHMO, it solves nothing. I wrote about it here and here.
> The gist of the issue is that it's just too easy to get around the
> provisions of this license. From my original post:
>
> This ambiguity leads directly to the trivial ways in which the license can
> be avoided: Take the AGPL'd code, rip out the 'remote user interaction'
> interface components and replace them a 'local, machine interacting'
> interface (i.e. a different API) and I'm done. If you don't think this is
> possible, then we disagree on what a user interaction is and a machine
> interaction is. Which is exactly my point. A slightly more liberal reading
> of this license would conclude that a simple (local) proxy isolated the
> 'remote user' from the Work, which undermines the license in it's entirely.
>
> 'ASP loophole' is too complex a problem to be addressed by the addition of a
> single paragraph in a document that had different licensing objectives. Or
> for that matter, by any license that covers only a portion of a complex,
> distributed service offering.
>
> Given the widespread interest in anything 'cloud' and the meager adoption
> of the AGPL, I think people are (not) voting with their feet.
>

We've licensed our Enomalism software using AGPL. I would disagree
with your assessment of the AGPL. Any license can be "interpreted" in
a manor that suits your particular vantage point. The law in general
can be interpreted, that's why we have a legal system to help set
these kind of precedents. We've chosen the AGPL for Enomalism because
it generally dictates how we intend on our software to distributed and
utilized. Simply, "any users who interact with the licensed software
over a network needs to receive the source for any revisions made to
our source code". We also assume a good portion of the people who use
our Enomalism Elastic Computing platform will completely disregard the
licenses terms. We hope that a certain percentage will abide by the
license, even if its a small percentage.

Here are some pros and cons we outlined when looking at the AGPL.

* Based on GPLv3, but has an additional term to allow users who
interact with the licensed software over a network to receive the
source for that program

"…if you modify the Program, your modified version must
prominently offer all users interacting with it remotely through a
computer network an opportunity to receive the Corresponding Source of
your version by providing access to the Corresponding Source from a
network server at no charge…"

* Pros
o Provides more "protection" to commercial open source
software companies by applying the viral property to SaaS (hosted) use
o OSI approved
o Builds on the wide use and acceptance of the GPL (V2.0)
o AGPL is suitable for International use (GPL V2.0 was U.S. specific)
o Patents are explicitly addressed

* Cons:
o This is a very new license, not yet broadly understood or accepted
o Software under this license may not be combined with
software under several other OSI-approved licenses
o Copyleft clauses may inhibit adoption by (Open Source)
customers due to compliance risk
+ The SaaS vendors and Hosting Providers will tend to
avoid software under this license, if they think they might ever need
to modify it.

AGPL Background

* Note about "protection": the AGPL only requires a hosted
application or system provider to provide access to their source if
they have modified the software.
* Note that the AGPL is less "combinable" than the GPLv3. Here's
what GNU has to say about this:
o "Please note that the GNU AGPL is not compatible with
GPLv2. It is also technically not compatible with GPLv3 in a strict
sense: you cannot take code released under the GNU AGPL and use it
under the terms of GPLv3, or vice versa. However, you are allowed to
combine separate modules or source files released under both of those
licenses in a single project, which will provide many programmers with
all the permission they need to make the programs they want. See
section 13 of both licenses for details."
* The best chart I've seen on license compatibility with the GPL
is found at http://en.wikipedia.org/wiki/List_of_FSF_approved_software_licenses

GNU provides a good summary specific to GPLv3 at
http://www.fsf.org/licensing/licenses/quick-guide-gplv3.html

And the more specific details may be found at
http://www.fsf.org/licensing/licenses/gpl-faq.html#WhatDoesCompatMean
and http://www.gnu.org/philosophy/license-list.html

Reuven
www.enomalism.com

Vinayak Hegde

unread,
Aug 19, 2008, 10:27:26 AM8/19/08
to cloud-c...@googlegroups.com
On Tue, Aug 19, 2008 at 6:01 PM, Chris Marino <ch...@snaplogic.com> wrote:
> Vinayak, I think you have a very optimistic view on the the efficacy of the
> the AGPL. IHMO, it solves nothing. I wrote about it here and here.
> The gist of the issue is that it's just too easy to get around the
> provisions of this license. From my original post:
>
> This ambiguity leads directly to the trivial ways in which the license can
> be avoided: Take the AGPL'd code, rip out the 'remote user interaction'
> interface components and replace them a 'local, machine interacting'
> interface (i.e. a different API) and I'm done. If you don't think this is
> possible, then we disagree on what a user interaction is and a machine
> interaction is. Which is exactly my point. A slightly more liberal reading
> of this license would conclude that a simple (local) proxy isolated the
> 'remote user' from the Work, which undermines the license in it's entirely.
>
> 'ASP loophole' is too complex a problem to be addressed by the addition of a
> single paragraph in a document that had different licensing objectives. Or
> for that matter, by any license that covers only a portion of a complex,
> distributed service offering.

I think some of your objections have been answered in the comments of
your own blog post. AGPL is not perfect, but that does not mean it is
not enforceable. My focus on the intent of AGPL and not on dissecting
the AGPL itself. There are grey areas but do you want to bet your
business on the grey areas. People said the same thing about GPLv2
(that is not enforceable in a court of law). They were wrong. GPLv2
has been upheld in courts and damages had to be paid / changes ahd to
be done to get around it. There is a site tracks GPL violations.
(http://gpl-violations.org/).

> Given the widespread interest in anything 'cloud' and the meager adoption
> of the AGPL, I think people are (not) voting with their feet.

There is widespread interest in the cloud. But step back a second and
compare the deployment of desktop software to that of cloud computing
- the share of cloud computing is really miniscule. Cloud computing is
in it's infancy now. The Mozilla Public License or the Apache License
have less than 2% of the software licensed under them. But lot of
world-class software is licensed under them. Maybe AGPL will fail - no
one will adopt it. But it's too early to pass a verdict.

This is getting way too off-topic for this list.

-- Vinayak

Chris Marino

unread,
Aug 19, 2008, 12:18:43 PM8/19/08
to cloud-c...@googlegroups.com

Veering off topic into open source land again, but the actual number of decided cases involving open source licenses is extremely small. Sure, there have been violations, but most have been settled without the courts weighing in so to say that 'they were wrong' about the enforceability of the GPL in 'a court of law' is simply not supported by the facts.  Mark Radcliffe at DLP Piper blogs about this all the time. On the recent Artistic License case he commented:

"One of the frustrations of lawyers serving the open source industry is that they have few cases which interpret open source licenses. As Eben Moglen has pointed out, such cases are few because licensees need the license to be in effect to avoid copyright infringement."

The one case I can think of involved a case in Germany, which would not help very much here in US.   Ruven's analysis on why they chose the AGPL makes sense for them, but note that there were built-in expectations of violations as well.
 
This is just one of my pet peeves. I cringe whenever I hear someone says that the ASP loophole is closed by the AGPL, .
 
 
Now, I'm going to STFU....
CM

>-----Original Message-----
>From: cloud-c...@googlegroups.com
>[mailto:cloud-c...@googlegroups.com] On Behalf Of Vinayak Hegde
Reply all
Reply to author
Forward
0 new messages