summaryrefslogtreecommitdiff
path: root/2009/foss-silicon_manufacturer-elce2008
diff options
context:
space:
mode:
authorHarald Welte <laforge@gnumonks.org>2015-10-25 21:00:20 +0100
committerHarald Welte <laforge@gnumonks.org>2015-10-25 21:00:20 +0100
commitfca59bea770346cf1c1f9b0e00cb48a61b44a8f3 (patch)
treea2011270df48d3501892ac1a56015c8be57e8a7d /2009/foss-silicon_manufacturer-elce2008
import of old now defunct presentation slides svn repo
Diffstat (limited to '2009/foss-silicon_manufacturer-elce2008')
-rw-r--r--2009/foss-silicon_manufacturer-elce2008/abstract.txt24
-rw-r--r--2009/foss-silicon_manufacturer-elce2008/bio.txt25
-rw-r--r--2009/foss-silicon_manufacturer-elce2008/default.mgp21
-rw-r--r--2009/foss-silicon_manufacturer-elce2008/embedded-openexpo.mgp517
-rw-r--r--2009/foss-silicon_manufacturer-elce2008/linux_netfilter_singapore_entertainment.jpgbin0 -> 640673 bytes
5 files changed, 587 insertions, 0 deletions
diff --git a/2009/foss-silicon_manufacturer-elce2008/abstract.txt b/2009/foss-silicon_manufacturer-elce2008/abstract.txt
new file mode 100644
index 0000000..ae094c3
--- /dev/null
+++ b/2009/foss-silicon_manufacturer-elce2008/abstract.txt
@@ -0,0 +1,24 @@
+How chip makers should (not) support Free Software
+
+Silicon manufacturers, or rather design houses play a key aspect in how well
+their products are supported in Free Software oparating systems such as Linux.
+In the early Linux days - more than a decade ago - it was normal to have
+completely public technical reference manuals for the silicon, enabling Linux
+community developers to write drivers for the chips.
+After chip design houses start to realize there is an economically significant
+Linux market, they try to use their existing workflow, processes and
+development model for proprietary operating systems and try to apply this to
+Linux. The result are in many cases binary-only drivers for certain Kernel
+versions and/or distributions or unmaintained, non-portable, coding style
+incompliant open source drivers for outdated kernel versions. Those kind of
+drivers are bound to create dissatisfaction within the Free Software developer
+community, among the Free Software users. Furthermore, they also result in
+inefficient use of R&D resources both inside and outside the chip vendor.
+
+Many silicon design houses still don't understand the Free Software and
+particularly Linux development model at all. This results in suboptimal
+support of their hardware products. In the end, customers are likely to buy
+from a different vendor.
+
+So what can chip design houses do to ensure excellent support of their products
+in the Free Software world?
diff --git a/2009/foss-silicon_manufacturer-elce2008/bio.txt b/2009/foss-silicon_manufacturer-elce2008/bio.txt
new file mode 100644
index 0000000..960ef74
--- /dev/null
+++ b/2009/foss-silicon_manufacturer-elce2008/bio.txt
@@ -0,0 +1,25 @@
+Harald Welte is a freelancer, consultant, enthusiast, freedom fighter and
+hcaker who is working with Free Software (and particularly the Linux kernel)
+since 1995. His first major code contribution to the kernel was within the
+netfilter/iptables packet filter.
+
+He has started a number of other Free Software and Free Hardware projects,
+mainly related to RFID such as librfid, OpenMRTD, OpenBeacon, OpenPCD, OpenPICC.
+
+During 2006 and 2007 Harald became the co-founder of OpenMoko, where he served
+as Lead System Architect for the worlds first 100% Open Free Software based
+mobile phone.
+
+Aside from his technical contributions, Harald has been pioneering the legal
+enforcement of the GNU GPL license as part of his gpl-violations.org project.
+More than 150 inappropriate use of GPL licensed code by commercial companiess
+have been resolved as part of this effort, both in court and out of court.
+
+He has received the 2007 "FSF Award for the Advancement of Free Software" and
+the "2008 Google/O'Reilly Open Source award: Defender of Rights".
+
+Harald is currently working as "Open Source Liaison" for the taiwanese CPU,
+chipset and peripheral design house VIA, helping them to understand how to
+attain the best possible Free Software support for their components.
+
+He continues to operate his consulting business hmw-consulting.
diff --git a/2009/foss-silicon_manufacturer-elce2008/default.mgp b/2009/foss-silicon_manufacturer-elce2008/default.mgp
new file mode 100644
index 0000000..a0fcfc2
--- /dev/null
+++ b/2009/foss-silicon_manufacturer-elce2008/default.mgp
@@ -0,0 +1,21 @@
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%%
+%% This default.mgp is "Xft2" oriented.
+%deffont "standard" xfont "serif"
+%deffont "thick" xfont "sans-serif"
+%deffont "typewriter" xfont "monospace"
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%%
+%% Default settings per each line numbers.
+%%
+%default 1 area 90 90, leftfill, size 2, fore "white", back "black", font "thick"
+%default 2 size 7, vgap 10, prefix " "
+%default 3 size 2, bar "gray70", vgap 10
+%default 4 size 5, fore "white", vgap 30, prefix " ", font "standard"
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%%
+%% Default settings that are applied to TAB-indented lines.
+%%
+%tab 1 size 5, vgap 40, prefix " ", icon box "green" 50
+%tab 2 size 4, vgap 40, prefix " ", icon arc "yellow" 50
+%tab 3 size 3, vgap 40, prefix " ", icon delta3 "white" 40
diff --git a/2009/foss-silicon_manufacturer-elce2008/embedded-openexpo.mgp b/2009/foss-silicon_manufacturer-elce2008/embedded-openexpo.mgp
new file mode 100644
index 0000000..700b740
--- /dev/null
+++ b/2009/foss-silicon_manufacturer-elce2008/embedded-openexpo.mgp
@@ -0,0 +1,517 @@
+%include "default.mgp"
+%default 1 bgrad
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+%nodefault
+%back "blue"
+
+%center
+%size 7
+
+How to do Embedded Linux [not] right
+
+Infineon / February 2009
+
+%center
+%size 4
+by
+
+Harald Welte <laforge@gnumonks.org>
+
+netfilter.org / openmoko.org / openpcd.org
+gpl-violations.org / openezx.org / gnufiish.org
+berlin.ccc.de / openBSC.gnumonks.org
+deDECTed.org / hmw-consulting.de / viatech.com
+
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Introduction
+
+Who is speaking to you?
+ an independent Free Software developer, consultant and trainer
+ 15 years experience using/deploying and developing for Linux on server and workstation
+ 12 years professional experience doing Linux system + kernel level development
+ strong focus on network security and embedded
+ expert in Free and Open Source Software (FOSS) copyright and licensing
+ digital board-level hardware design, esp. embedded systems
+ active developer and contributor to many FOSS projects
+ thus, a techie, who will therefore not have fancy animated slides ;)
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Introduction
+
+Why am I qualified?
+ The 'Linux community' POV
+ Former kernel subsystem maintainer (netfilter/iptables)
+ Initiator of OpenEZX project
+ Author of various drivers for embedded hardware
+ The 'embedded Linux done the rigth way' POV
+ Lead System Architect (SW+HW) at Openmoko, Inc.
+ Co-creator of Open Hardware + Software for RFID
+ OpenPCD, OpenPICC, librfid, libmrtd
+ The 'chip manufacturer' POV
+ Open Source Liaison at VIA Technologies, Inc.
+ The 'customer of consumer-grade embedded Linux' POV
+ Done reverse-engineering on hundreds of devices for gpl-violations.org
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Linux is everywhere!
+
+%image "linux_netfilter_singapore_entertainment.jpg"
+
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Linux is everywhere
+
+
+Linux is everywhere!
+ Linux mobile phones (Motorola, Openmoko)
+ In-flight entertainment systems
+ Embedded networking gear
+ DSLAMs
+ rack monitoring
+ Public payphones
+ ATM's / PoS / vending machines
+ Now even Fitness gear (Daum Ergometer)
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Strengths of FOSS [0/4]
+
+
+What are the true strengths of FOSS
+ Innovative and creative development
+ Security due to code review / bugreport / patches
+ Long-term maintainable code
+ Stable and reliable systems
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Strengths of FOSS [1/4]
+
+
+Innovative and creative development
+ easy-to-read existing codebase
+ standard (FOSS) development tools
+ thus, easy to modify and add features
+ community will build around great new features/apps
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Strengths of FOSS [2/4]
+
+
+Security due to code review / bugreport / patches
+ all the code is out there
+ many people are familiar with existing architecture
+ code quality requirements usually very high
+ community process allows quick and fast integration of bugfix
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Strengths of FOSS [3/4]
+
+
+Long-term maintainable code, because
+ there's a lot of attention on good software architecture
+ many developers are familiar with the shared/common API's
+ code quality requirements usually very high
+ all code in mainline gets maintained and updated
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Strengths of FOSS [4/4]
+
+
+Stable and reliable systems, because
+ code quality requirements usually very high
+ kernel releases are quite frequent
+ all mainline code is automatically ported to new releases
+
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Reality Check
+
+
+ So we should have the perfect world
+ tons of embedded Linux products
+ all of them maitainable, secure, stable
+ encouraging lots of creative work on top of their codebase
+
+ What is the reality
+ tons of embedded Linux products
+ none of the strengths of FOSS present in 99% of them
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Differences to PC Linux
+
+
+Differences to Linux on a PC
+ In the PC world, I can
+ download the latest kernel from kernel.org
+ compile + install it on almost every current+old platform
+ have an almost 100% chance that it will boot and support all the major peripherals
+ only some more obscure hardware might not have drivers yet
+ update at any time to the latest 2.6.x.y maintenance release
+ update at any time to the next 2.6.(x+1) release
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Differences to PC Linux
+
+
+Differences to Linux on a PC
+ In the PC world, I can
+ take about any major Linux distribution, based on my own preference
+ install and run that very distribution on about any hardware
+ distribution kernels are very close to mainline these days
+ benefit of regular security updates by distributions
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Differences to PC Linux
+
+
+Differences to Linux on a PC
+ In the Embedded world
+ every CPU/SoC maker runs their own kernel tree
+ often one kenrel tree per product, based on different mainline versions
+ ages-old base revisions
+ a neverending security nightmare
+ no benefit from recent new features in mainline
+ non-standard subsystems (e.g. different USB device or SDIO stack)
+ proprietary drivers cause lock-in to old kernels
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Differences to PC Linux
+
+
+Differences to Linux on a PC
+ In the Embedded world
+ I often do not have a choice of which distro to run
+ There might actually be no distribution
+ No regular security updates
+ Often no package management for deploying those updates
+ If there are distributions, they either need to use the kernel from the BSP (which is ages old) or creare yet another custom off-mainline branch/port
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Differences to PC Linux
+
+
+
+
+%center
+THIS SUCKS!
+
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+What does the vendor get
+
+
+So what do the embedded vendors get?
+ unstable software
+ security nightmares
+ unmaintainable code
+ no innovation
+ no user-contributed bug fixes
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+What does the vendor get
+
+
+unstable software
+ because the code really sucks in many cases
+ because they patch around problems rather than solving them
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+What does the vendor get
+
+
+security nightmares
+ because they use stoneage forks of the kernel
+ because they never contributed their code mainline
+ because those forks can never be merged back with mainline again
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+What does the vendor get
+
+
+unmaintainable code
+ because they have one fork of the code per device (product)
+ because their code quality sucks
+
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+What does the vendor get
+
+
+no innovation
+ because they try to hide their code (gpl-violations.org)
+ because their R&D environment is non-standard
+ weird cross-toolchains that nobody has seen before
+ weird filesystems with custom patches that nobody knows
+ because they add proprietary components to lock developers from adding features
+ e.g. the entire web-based UI for embedded networking gear
+ binary-only kernel modules that force people to use old kernels with no interesting new features
+ because it is, overall, way too hard to develop on/for their platform
+ because they don't disclose serial console and/or JTAG access
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Reality Check
+
+
+So why do they still do it?
+ there can only be one conclusion:
+ they never understood the real potential of FOSS!
+ all they do is try to compete with what proprietary competitors do
+ they never think about creating platforms, every product is distinct/separate
+ they have no interest in improving their products
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+What does the community get
+
+
+So what does the community get?
+ products that run some crappy fork of Linux somewhere under the hood
+ but which we cannot really touch/modify without a lot of effort
+ we face opposition from the product maker if we want to help him to improve
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+TODO (devica maker)
+
+
+What should the device vendor do?
+ stop thinking in terms of selling black boxes
+ defining products that take advantage of the true strength of FOSS
+ compete against the proprietary competition on a level that they can't match
+ give up the idea of defining all aspects of an appliance
+ rather think of building an extensible platforms and let community innovate
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+A note to chip vendors
+
+
+There are two types of customers
+ The Linux-aware customer
+ understands FOSS much better than you do
+ will share the criticism of this talk
+ will likely go to a competitor who understands Linux better
+ The Linux-unaware customer
+ who just uses Linux to save per-unilt OS royalties
+ who doesn't really care about most issues presented here
+ who will create inferior products
+ Linux-awareness is increasing, not decreasing
+ now is already late, but if you don't have proper FOSS support on your agenda now, you will likely loose the "openness competition"
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+TODO (chip maker)
+
+
+What should the chip vendors do?
+ engange in "sustainable development"
+ develop against latest mainline
+ make your development trees public (use git!), don't just release stable snapshots as BSP to your tier-one customers
+ actively interact with the community
+ learn how to play by the rules (coding style, use common interfaces, no proprietary drivers)
+ don't just do big code drops every now and then
+ since most of your customers' skill is very low
+ they cannot tell you what they actually want/need on a technical level
+ you have to understand Linux better than them, and better than the competition
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+TODO (chip maker)
+
+
+What should the chip vendors do?
+ Externally
+ Don't mistake FOSS as just a technology. It is a R&D philosophy!
+ Provide public reference manuals with no NDA
+ If you have to resort to NDA, make sure they are FOSS friendly
+ Ensure you don't license IP cores that conflict with FOSS
+ Internally
+ Draft a proper in-house FOSS strategy with clear goals
+ Don't expect your product mangers or engineers to know everything about FOSS without proper training
+ Hire people with strong community background into your R&D and managemnt to facilitate the know-how transfer
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Lessons Learnt (chip maker)
+
+
+ Chip Product Managers need to learn
+ There's more Linux use with their chips than they ever learn
+ Their customers are not just the tier-one customers
+ If you want to support Linux, do it the mainline way. If you support only N number of distributions, your 'N' will be growing and you'll be wasting R&D resources to support each one of them
+ That there is no single 'contact window' / entity for Linux
+ The big change is the FOSS development model, not the Linux API's
+ Linux-aware customers care not only about performance+price, but also about the quality of the Linux port code
+
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Lessons Learnt (chip maker)
+
+
+ BSP R&D Managers need to learn
+ Linux is not just a set of API's for their developers to learn
+ Linux is FOSS. The FOSS R&D model is different
+ Their engineers need to be encouraged to communicate
+ and thus, need real internet (git, mailing lists, ..)
+ have to get clear indication what is confidential and what not
+ FOSS is about cooperation. You don't need to reinvent the wheel
+ don't include your own 802.11 stack in your wifi driver
+ If you do not merge your code mainline, you end up in a maintenance nightmare
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Lessons Learnt (chip maker)
+
+
+ BSP Software Engineers need to learn
+ Linux is not just a set of API's
+ How and where to ask the right kind of questions
+ How and where to look for the latest code
+ Code is written to be read by other people, not just to execute
+
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+TODO (community)
+
+What should the community do
+ Provide non-partisan documentation
+ on FOSS advantages, proper FOSS development
+ hardware companies are intrested to learn, but don't know who to ask. If they ask a commercial distributor, then they get in bed with them, which is not the same as working with the mainline development community
+ Something like a mentoring program
+ take software/driver R&D by their hand and walk them through the mainline merge
+ Don't scare them away
+ They have to be taught about the communication / review culture
+ Your valid criticism to patches has to be explained
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Problems (chipmaker)
+
+ Patent licensing schemes
+ e.g. MPEG patent license doesn't at all work in a FOSS model
+ Technology licensing schemes
+ e.g. Sony Memorystick. Impossible for a chip maker to provide FOSS driver
+ Everyone in the industry has those very same problems
+ Chip makers should cooperate to present their case together to the respective licensors
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Problems (chipmaker)
+
+ Patent trolls
+ openly accessible documentation invites patent trolls
+ Why?
+ patent trolls rarely read+understand FOSS code
+ thus, open documentation increases the risk to be hit
+ However
+ this is a by-product of how the patent system currently works
+ if there is a patent violation in the product, then it exists whether or not there is documentation
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Outlook
+
+Outlook
+ We will see even more embedded Linux, e.g. Mobile phones
+ We will see even more restricted devices (Tivo-ization, DRM for code)
+ which go on _very_ thin ice even with GPLv2
+ which almost completely remove all freedoms of FOSS
+ We have some faint dim light at the end of a very far tunnel
+ e.g. projects like Openmoko, who truly see openness as a feature
+ e.g. chip makers who slowly open up a bit more
+ on the PC side, Intel defiitely is setting the best overall example
+ on the Embedded side, I see some - but way too small - imovement in major players (TI, Marvell, Samsung, Infineon, ...)
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+The Embedded SoC market
+
+In the Embedded SoC market
+ your typical customer is a chinese company with very limited technical skill and resources
+ almost all products are equal, market extremely competitive
+ thus, licensing proprietary OS or software stack is not feasible
+ thus, SoC makers get pushed to provide not only base OS/drivers, but full 'turnkey solution' reference design
+ which also means
+ your actual SoC customer does not understand Linux, neither technically nor the Open Source model (and its benefits)
+ if you are the first to provide a solution that fully embraces Open Source, your products will be more stable and feature-rich than the competition
+
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Questions
+
+ Questions to the audience:
+ Do you really enter a legal grey area and make life hard for yourself and your customers/users by doing binary-only kernel modules?
+ Do you want to ignore the security benefits from open source?
+ Can you really afford to not tap into the creative power and workforce of the FOSS community?
+ Do you really want to use Linux just as a cheap replacement for other OS, or do you want to embrace its true strength?
+
+ Competing with openness against proprietary competitors is competing at a level they can never match!
+%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
+%page
+How to do Embedded Linux [not] right
+Thanks
+
+ Thanks for your attention.
+
+ Some Time left for Q&A
+
+ Feel free to mail me at laforge@gnumonks.org
diff --git a/2009/foss-silicon_manufacturer-elce2008/linux_netfilter_singapore_entertainment.jpg b/2009/foss-silicon_manufacturer-elce2008/linux_netfilter_singapore_entertainment.jpg
new file mode 100644
index 0000000..91b839f
--- /dev/null
+++ b/2009/foss-silicon_manufacturer-elce2008/linux_netfilter_singapore_entertainment.jpg
Binary files differ
personal git repositories of Harald Welte. Your mileage may vary