Why iXsystems Bought PC-BSD

weixin_38048698 2006-11-11 08:39:36

http://www.onlamp.com/pub/a/bsd/2006/10/23/ixsystems-pc-bsd.html

Why iXsystems Bought PC-BSD
by Dru Lavigne
10/23/2006


iXsystems is a leading provider of high-performance computing clusters, blade servers, rackmount servers, and storage solutions based on FreeBSD, NetBSD, OpenBSD, and Linux. iXsystems also recently announced its acquisition of the PC-BSD operating system.

I had the opportunity to interview Kris Moore, founder and lead developer of the PC-BSD project, and Matt Olander, CTO of iXsystems, about the acquisition.

Dru Lavigne: Tell us a bit about yourself and PC-BSD.

Kris Moore: I reside in Vancouver, WA. (not Canada), will be 26 years old in October, and am married with two boys. I enjoy building computer systems in my spare time, as well as playing video games, such as RPGs, RTS, or shooters on the Xbox 360 (my Xbox Live ID is Piett134).

PC-BSD is designed to be the first true "desktop" offering based on FreeBSD. With a graphical installer, and the custom "PBI" package management format, PC-BSD strives to be the most user-friendly and stable computing experience possible.

Matt Olander: I have been working at iXsystems as the chief technology officer for four years. Previously, I worked at BSDi as the IT director. When I'm not playing around on a FreeBSD box, I'm either reading or riding my recumbent bicycle, which will hopefully soon have an embedded FreeBSD system attached for a webcam, GPS, and wireless internet access.


Dru:PC-BSD isn't exactly a fork of FreeBSD as it provides a set of desktop extensions to the stable version of FreeBSD. However, it is a wholly integrated operating system with its own release engineering process. Can you describe your motivations for developing PC-BSD and why you chose to create a separate project?

Kris: One of my goals when creating PC-BSD was to not create another "fork", but instead keep the underlying OS 100 percent FreeBSD for compatibility and stability reasons. However, with this said, I felt that for usability reasons, it was important to just go beyond a add-on graphical installer for FreeBSD, and instead create a system which is complete out of the box, with tools, and utilities for managing a system without command-line interaction.

Also of importance was creating a new package management system, which doesn't work with a dependency model, and instead allows applications to be entirely self-contained. In our experience, one of the most frustrating things for casual computer users is trying to install an application which breaks your dependency chain.

Dru: PC-BSD has seen substantial growth in its user-base over the past year. Does the project track number of downloads or any statistics regarding users such as their age, location, and what they use their PC-BSD systems for?

Kris: The community response to PC-BSD so far has been overwhelming. With each release we've seen massive downloads from our website, with over 100,000 separate downloads of our 1.2 version to date. As for usage, we do not necessarily "ask" our users what they are doing with the system, but a check of our forums would seem to indicate that many users are running PC-BSD as their home or work desktop system.

Dru: One of the attractions for FreeBSD users is the ports and packages collections. Describe the motivations behind creating the PBI system and give a quick overview of its advantages for those who haven't heard of PBIs. Is it proving to be popular or are most users sticking with ports and packages?

Kris: The primary motivation behind PBI was enabling Windows or Mac users to easily migrate to PC-BSD without having to learn a new package management scheme. Often the whole concept of having to install dependencies for every application is confusing and daunting to casual computer users, who just expect things to work. (i.e., When I double-click, it just runs, I don't want to know what GTK is!)

That said, the FreeBSD ports and packages system is excellent for power users and those coming from the Linux world, and many of our users still make use of it. However, for the casual user, it is unrealistic to expect them to learn how to manage this, for fear of "breaking" their base system. This is why PBI was developed as an alternative, allowing users to install applications with ease, that live completely independent of their base operating system.

Dru: Matt, you've been involved in the FreeBSD community for quite some time and have witnessed first hand companies that provide direct support to the FreeBSD project as well as products and support to FreeBSD users. Do you find there is more or less company interest in FreeBSD then 10 years ago? Does interest tend to be within a certain industry?

Matt: Yes, there is definitely more interest in FreeBSD now than there was 10 years ago. I find more people recognize the term FreeBSD or BSD during casual conversations. I used to have to go into a detailed explanation of operating systems in general and that seems to happen far less now. We can thank the explosion of Linux and open source in general for that, I think.

As for certain industries showing an increase in interest, FreeBSD has always been dominant in the ISP and web hosting industries but we're really seeing a noticeable increase in interest in the OEM appliance industry. Especially as concern over the GPL license and the protection of a company's intellectual property grows, we field more and more queries about developing software on a hardware platform that supports the FreeBSD operating system.

Due to the nature of the BSD license, I think that it has been a double-edged sword for FreeBSD adoption and popularity. A corporation that runs their appliance or solution on a FreeBSD platform may not be so willing to share that information with the general public because it doesn't carry the hype that Linux does and if it doesn't benefit the company to share that information, why should they?

Dru: From both the PC-BSD and iXsystems perspectives, what were the reasons for considering a collaboration and what benefits do you perceive from the acquisition?

Kris: Right now from the PC-BSD perspective we are at a tipping point, where with the right resources and marketing we can begin to push into the business sector, to directly compete with Linux and Windows. iXsystems has a great reputation for their FreeBSD solutions, and understands the marketplace from a business perspective, so bringing PC-BSD on-board seemed like a perfect fit.

Matt: From a server manufacturing perspective, we want to help our customers adopt the platform that is right for them. Of course, we ship quite a few systems with some flavor of Linux pre-installed, but it tends to be because the customer is looking for a commercially supported operating system. If iXsystems can assist increasing the adoption of FreeBSD, we think that's a good thing for us as well as our customers. A higher adoption rate of FreeBSD encourages vendors to provide documentation so that the appropriate drivers can be created and supported. It also attracts developers to the project so that FreeBSD continues to progress and be the modern, stable operating system platform that it is currently.

Dru: Acquisitions of projects are often viewed negatively by those involved in open source and are often seen as a "sell out" or a slap against open source way of doing things. Why do you think this acquisition is a good thing and how will it benefit the BSD community?

Kris: I understand that among the open source community there sometimes is a anti-business philosophy, however I think that both open source and business can coexist peacefully. This merger will help address something that I think has hindered the BSD community in the past, which is having a solid business backing. One of the things that has sometimes prevented companies from adopting BSD is the confusion about where to obtain support. Who do you go to for help? By merging PC-BSD and iXsystems, we hope to provide businesses with a professional and supported alternative to Linux and other platforms.

Matt: Near the end of BSDi's corporate life, we were able to get most, if not all, of the relevant code from BSD/OS out into the public domain for use by the various BSD projects. I think it really depends on intention. Since the PC-BSD project is open source and will remain so, I can only see good things for the project as a whole from this acquisition. I also think the anti-business sentiment is far less prevalent in the FreeBSD camp than it is in some of the Linux distros that I've seen. People have to make a living. If this drives further BSD adoption and creates even more of a need for experienced BSD developers and system administrators than there is now, everyone in the FreeBSD community will benefit.

Dru: Are there plans to release a commercial version of PC-BSD or to provide commercial support for PC-BSD?

Kris: I know we definitely plan on commercial support for PC-BSD, to ensure that businesses and individuals have somewhere to turn when they need assistance. As for a specific commercial version, I'm unsure how that will play out in the long term.

Matt: Definitely commercial support for PC-BSD and FreeBSD in the near future. I really don't see any advantage in a closed commercial version of BSD at this time. Our core business is server and OEM appliance manufacturing. The end result of this deal will be that FreeBSD, the operating system that we test, install, and in many cases configure for many of our customers, becomes a more stable, reliable operating system with system management tools that perhaps it didn't have before.

Dru: Now that funds are available, are there any plans to implement features which were formerly just on your wish list?

Kris: I'm hoping that we may now begin to tackle some more of the improvements for business/server clients, as well as things such as a x64 port and better hardware support.

Matt: Eventually, iXsystems would like to sponsor some features in PC-BSD/FreeBSD that make it a more competitive modern operating system. Things like a shared filesystem, clustering tools, iSCSI support, etc. are all things that FreeBSD could benefit from having.

Dru: Give us a sneak peak into the new features that will be available in PC-BSD 1.4, due out in November.

Kris: Well, were actually still working on the 1.3 release of PC-BSD, which will sport a brand new system installer, with more powerful features, and the ability to "upgrade" your previous PC-BSD releases. Among the other planned features down the road for 1.4 will be a firewall editor, a system-services GUI, and more.

Dru: Anything else I forgot?

Matt: No, but I'd like to refer to an article written from an IBM developers perspective while evaluating FreeBSD. He said, and I'm paraphrasing here, that FreeBSD is the operating system that Linux was meant to be. I think, by that, he means an up to date Unix-like OS with advanced package management that is easy to install and maintain. FreeBSD is a modern operating system that is reliable, stable, mature, and performs incredibly well under high-traffic situations. Throw in the business-friendly BSD license and BSD makes a very sensible choice for end users.



Dru Lavigne is an instructor at Marketbridge Technologies in Ottawa and the maintainer of the Open Protocol Resource.
...全文
61 4 打赏 收藏 转发到动态 举报
写回复
用AI写文章
4 条回复
切换为时间正序
请发表友善的回复…
发表回复
内容概要:本文系统研究了基于交替方向乘子法(ADMM)的多主体综合能源系统分布式协同优化方法,旨在实现多个能源主体之间的高效、隐私保护的能量协同管理。通过建立各主体的局部优化模型,并引入一致性约束条件,利用ADMM算法将集中式优化问题分解为多个分布式子问题进行迭代求解,从而在保证各主体数据隐私的同时达成全局协同优化目标。研究涵盖了模型构建、算法实现、收敛性分析及数值仿真全过程,并提供了完整的Matlab代码实现,有效复现了相关学术成果,具有较强的可验证性与可拓展性。; 适合人群:具备一定数学建模能力与优化理论基础,熟悉Matlab编程,从事综合能源系统、分布式优化算法、电力系统调度等相关方向研究的研究生、科研人员及工程技术人员。; 使用场景及目标:① 掌握ADMM算法在多主体能源系统协同优化中的建模与实现方法;② 构建支持隐私保护的分布式能源调度系统;③ 利用开源代码开展算法改进、性能测试或应用于实际系统的二次开发与科研论文复现; 阅读建议:建议结合凸优化与分布式计算的基础知识,深入理解增广拉格朗日函数构造、对偶更新与收敛判据的设计原理,在代码调试过程中重点关注惩罚参数选取、变量分裂策略及通信机制的实现细节,以全面掌握算法的核心机制与工程应用要点。

478

社区成员

发帖
与我相关
我的任务
社区描述
其他技术讨论专区
其他 技术论坛(原bbs)
社区管理员
  • 其他技术讨论专区社区
加入社区
  • 近7日
  • 近30日
  • 至今
社区公告
暂无公告

试试用AI创作助手写篇文章吧