New hire orientation

From FedoraProject

(Difference between revisions)
Jump to: navigation, search
m (From Max's talk)
 
(One intermediate revision by one user not shown)
Line 1: Line 1:
== Max's notes ==
+
[[MaxSpevack|Max]] is in charge of the ''Open Source, Fedora, RHEL, and how it all fits together'' section of New Hire Orientation in North America.  The talk is meant to be about 45 minutes + questions, and is delivered without any slides.  It is also full of questions and audience interaction.
  
* Max's notes (internal Red Hat only, sorry!) are [http://commarch.usersys.redhat.com/~mspevack/orientation-2009.txt here].
+
The notes below represent:
 +
* The material that needs to be covered, though the order and the examples sometimes vary.
 +
* The material that is general enough to be documented outside of the Red Hat firewall.  
  
== Mel's notes ==
 
  
=== From Max's talk ===
+
<pre>
 +
RHT -- 3k
 +
Fedora -- 10k
 +
GOOG -- 20k
 +
MSFT -- 93k
 +
ORCL -- 100k
 +
IBM -- 400k
 +
Sourceforge -- 2M
  
* The way Red Hat's software is made is the open source way.
+
* The genius of open source is its model.
** Fedora as supply chain to RHEL
+
* The software commons is Red Hat's supply chain.
** Licensing
+
** Why is the open source way different?
+
* Communities are not the borg! There are many of them, with many people, with many views. There is no uni-mind.
+
* Red Hat's mission: to be the trusted open source leader.
+
* On the importance of "giving back" to our upstream communities - it's our responsibility to do so, and also how we can steer and influence people who have free will over what they want to do, since we can't order volunteers to work on things; the idea itself needs to have merit and also manpower behind it.
+
* RHEL is a 7-year maintenance load.
+
** Relationships with people in the FOSS community interested in that maintenance and able to help with it makes all the difference.
+
** Engineering metric of success: how much can you ''make happen,'' not how much can you ''do yourself''? Communicate with the community!
+
* FAQ: How do you keep quality control if anyone can contribute?
+
* Note: RHEL/Fedora is a stability/innovation split strategy.
+
* FAQ: How do new packages get into Fedora?
+
* The open source way lets us preview what customers want. Actually, they make what they want, and then pay us to sell it to them.
+
** SELinux
+
** AMQP
+
* FAQ: How many contributors to Fedora are there?
+
* Open question: what's the JBoss equivalent for this?
+
* FAQ: What's the motivation for contributors to contribute? It helps them do their job better. For instance, mmcgrath before he worked for Red Hat.
+
  
=== Activity ideas ===
+
=====================================================================
  
''These are not meant to be comprehensible to other people, they're just to remind myself of things I'd like to try. [[User:Mchua|Mel Chua]] 20:26, 31 August 2010 (UTC)''
+
VALUE PROPOSITION
 +
 
 +
"What the customer gets for what the customer pays."
 +
 
 +
* RHEL & Fedora --> why wasn't the old Red Hat Linux sustainable?
 +
    * Boxed set every 6 months = FAIL
 +
    * 18 month support cycle = no time for ISVs = FAIL
 +
    * Moving too fast for enterprises = FAIL
 +
 
 +
* Subscription to an ecosystem, and a community.
 +
    * Our best customers are in partnerships with us.
 +
    * We ourselves are a member (albeit large) of that community.
 +
        * NO RH --> Gimp, Mono, text editors, etc.
 +
        * MINIMAL RH --> Firefox, OpenOffice
 +
        * MAJOR RH --> Spacewalk, kernel, glibc, SELinux, RPM, Yum
 +
        * TOTAL RH --> Deltacloud, and a challenge of moving away.
 +
 
 +
* Commitment to community proportional to tech roadmap success.
 +
    * Open roadmap lowers risk of failure.
 +
    * For Red Hat, Fedora lowers R&D risk.
 +
        * Run several bets in parallel and choose the best.
 +
            * libvirt -> Xen, KVM, etc.
 +
            * Modularity and option value!
 +
    * Lower risk of technology irrelevance through Fedora.   
 +
 
 +
* OSS community --> many projects w/ distros as magnets.
 +
* Fedora --> dev, integration, packaging, QA, distribution.
 +
* RHEL --> certification, compliance, support, etc.
 +
 
 +
* Engaging the consumer earlier is more efficient and sustainable.
 +
    * Feedback arrives more quickly.
 +
    * Advanced Development -> Design -> Prototype -> Feedback
 +
 
 +
* Using Fedora allows you to see the future of RHEL.
 +
* Participating in Fedora allows you to create the future of RHEL.
 +
* Technical election (best of today and best for seven years).
 +
 
 +
* Fedora is where Red Hat generates potential value.
 +
* RHEL is where Red Hat distills that potential into products.
 +
 
 +
=====================================================================
 +
 
 +
RIVER OF CODE
 +
 
 +
* Explain upstream vs. downstream.
 +
    * Upstream leads to efficiency --> AMQP & SELinux
 +
* Compare RHEL's position to that of RHEL rebuilds.
 +
* Value to the customer comes from strong upstream associations.
 +
 
 +
=====================================================================
 +
 
 +
FREEDOM = CHOICE = POWER
 +
 
 +
* Internet routes around censorship.
 +
 
 +
* Interoperability of data.
 +
    * Patchwork
 +
        * Many 1:1 relationships.
 +
        * Promotes vendor lock-in.
 +
    * Open Standards
 +
        * Many:1 relationship
 +
    * Internet is the best example of OSS and open standards
 +
 
 +
* Compare to FSF 4 freedoms
 +
    * Royalty-free
 +
    * Immune to vendor capture
 +
    * Freely available specs & collaborative public review
 +
 
 +
* Internet speeds it up, but it's throughout history.
 +
    * Scribes vs. printing press.
 +
    * Assembly lines and tools become standard, industrial revolution.
 +
    * Railroad tracks becoming uniform, enables trade and connections.
 +
 
 +
* Open standards leads to commoditization.
 +
* Cost of switching is zero.
 +
 
 +
=====================================================================
 +
 
 +
LEGAL
 +
 
 +
* Public domain -> BSD -> GPL -> Trade secret
 +
    * GPL provides an ever-expanding commons.
 +
 
 +
COPYRIGHTS                          PATENTS
 +
* Actual implementations            * Potential implementations
 +
* Clearly documented                * Poorly documented
 +
* Clear protection instantly        * Poorly analyzed and granted
 +
 
 +
=====================================================================
 +
 
 +
CONCLUSION
 +
 
 +
* Red Hat model provides:
 +
    * Mindshare hedge
 +
    * Leverage and efficiency
 +
    * Finding talent -> no age on the internet!
 +
    * Consumer integration early on, including customers & partners.
 +
 
 +
* OUR CUSTOMERS TRUST US BECAUSE WE ARE THE GOOD GUYS.
 +
</pre>
 +
 
 +
== Activity ideas ==
 +
 
 +
''These are not meant to be comprehensible to other people, they're just to remind myself of things I'd like to try if/when I lead orientation. [[User:Mchua|Mel Chua]] 20:26, 31 August 2010 (UTC)''
  
 
* Give all but one table a puzzle and a lead.
 
* Give all but one table a puzzle and a lead.
Line 50: Line 144:
 
* bug report / RFE
 
* bug report / RFE
 
* community of practice
 
* community of practice
 +
 +
[[Category:Community Architecture]]

Latest revision as of 03:29, 22 February 2011

Max is in charge of the Open Source, Fedora, RHEL, and how it all fits together section of New Hire Orientation in North America. The talk is meant to be about 45 minutes + questions, and is delivered without any slides. It is also full of questions and audience interaction.

The notes below represent:

  • The material that needs to be covered, though the order and the examples sometimes vary.
  • The material that is general enough to be documented outside of the Red Hat firewall.


RHT -- 3k
Fedora -- 10k
GOOG -- 20k
MSFT -- 93k
ORCL -- 100k
IBM -- 400k
Sourceforge -- 2M

* The genius of open source is its model.
* The software commons is Red Hat's supply chain.

=====================================================================

VALUE PROPOSITION

"What the customer gets for what the customer pays."

* RHEL & Fedora --> why wasn't the old Red Hat Linux sustainable?
    * Boxed set every 6 months = FAIL
    * 18 month support cycle = no time for ISVs = FAIL
    * Moving too fast for enterprises = FAIL

* Subscription to an ecosystem, and a community.
    * Our best customers are in partnerships with us.
    * We ourselves are a member (albeit large) of that community.
        * NO RH --> Gimp, Mono, text editors, etc.
        * MINIMAL RH --> Firefox, OpenOffice
        * MAJOR RH --> Spacewalk, kernel, glibc, SELinux, RPM, Yum
        * TOTAL RH --> Deltacloud, and a challenge of moving away.

* Commitment to community proportional to tech roadmap success.
    * Open roadmap lowers risk of failure.
    * For Red Hat, Fedora lowers R&D risk.
        * Run several bets in parallel and choose the best.
            * libvirt -> Xen, KVM, etc.
            * Modularity and option value!
    * Lower risk of technology irrelevance through Fedora.    

* OSS community --> many projects w/ distros as magnets.
* Fedora --> dev, integration, packaging, QA, distribution.
* RHEL --> certification, compliance, support, etc.

* Engaging the consumer earlier is more efficient and sustainable.
    * Feedback arrives more quickly.
    * Advanced Development -> Design -> Prototype -> Feedback

* Using Fedora allows you to see the future of RHEL.
* Participating in Fedora allows you to create the future of RHEL.
* Technical election (best of today and best for seven years).

* Fedora is where Red Hat generates potential value.
* RHEL is where Red Hat distills that potential into products.

=====================================================================

RIVER OF CODE

* Explain upstream vs. downstream.
    * Upstream leads to efficiency --> AMQP & SELinux
* Compare RHEL's position to that of RHEL rebuilds.
* Value to the customer comes from strong upstream associations.

=====================================================================

FREEDOM = CHOICE = POWER

* Internet routes around censorship.

* Interoperability of data.
    * Patchwork
        * Many 1:1 relationships.
        * Promotes vendor lock-in.
    * Open Standards
        * Many:1 relationship
    * Internet is the best example of OSS and open standards

* Compare to FSF 4 freedoms
    * Royalty-free
    * Immune to vendor capture
    * Freely available specs & collaborative public review

* Internet speeds it up, but it's throughout history.
    * Scribes vs. printing press.
    * Assembly lines and tools become standard, industrial revolution.
    * Railroad tracks becoming uniform, enables trade and connections.

* Open standards leads to commoditization.
* Cost of switching is zero.

=====================================================================

LEGAL

* Public domain -> BSD -> GPL -> Trade secret
    * GPL provides an ever-expanding commons.

COPYRIGHTS                          PATENTS
* Actual implementations            * Potential implementations
* Clearly documented                * Poorly documented
* Clear protection instantly        * Poorly analyzed and granted

=====================================================================

CONCLUSION

* Red Hat model provides:
    * Mindshare hedge
    * Leverage and efficiency
    * Finding talent -> no age on the internet!
    * Consumer integration early on, including customers & partners.

* OUR CUSTOMERS TRUST US BECAUSE WE ARE THE GOOD GUYS.

[edit] Activity ideas

These are not meant to be comprehensible to other people, they're just to remind myself of things I'd like to try if/when I lead orientation. Mel Chua 20:26, 31 August 2010 (UTC)

  • Give all but one table a puzzle and a lead.
  • Tell the no-puzzle table to clear their table. They are Red Hat. Roles:
    • Project leader - ship on time
    • QA - no blank spots in the quilt
    • Engineer - orange border
    • Engineer - green corners
  • Other teams come in with parts and reqs
    • orange border
    • green corners
    • some sort of dinosaur in the middle (mid-round RFE)
  • QA goes and "patches" and ships

Terminology:

  • requirements
  • patches
  • "upstream" (pile)
  • mid-round contributions / code dumps
  • bug report / RFE
  • community of practice