PDC demo Dilemma


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial"> href="http://www.simplegeek.com/">ChrisAn and I ( href="http://www.gotdotnet.com/team/dbox/default.aspx">Don was already
snoozing) spent some time Friday on the bus back from the conference
discussing demos for the PDC and we reached a dilemma. Maybe you can help us
out.  It should be of no surprise to
you folks that we are in very active development of Longhorn right now. style="mso-spacerun: yes"> A month+ ago we forked off a build for
the PDC and have been stabilizing it (taking only very limited fixes) to get it
ready for the PDC.  In parallel the
dev team has been slamming out new features and, in some cases substantially
cleaning up the programming models. style="mso-spacerun: yes"> So for demos during the PDC should
we: />


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial"> 


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial">Plan A: Use only the bits we had out
to attendees and speak to the fact that the code isn’t the final model. style="mso-spacerun: yes"> This lets you repeat exactly the same
demo on your laptop at the event. style="mso-spacerun: yes"> 


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial"> 


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial"> style="mso-spacerun: yes">  — or –


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial"> 


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial">Plan B: Use the latest and greatest
internal build that shows off the programming model as we expect it to be when
we ship.  Provide demos\samples in
the SDK that demonstrate the same features but will work on the PDC bits. style="mso-spacerun: yes">  This shows you the right long-term
programming model so you can see clearly how we are evolving the
platform.


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial"> 


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial"> 


style="FONT-SIZE: 10pt; FONT-FAMILY: Arial">Love to have your votes, thoughts
and comments.

Comments (52)

  1. Roy Osherove says:

    Clearly – the most accurate information wins out. If I were attending, I’d want to see the latest up to date info on the programming model. At home I can play around with a nice demo. Sure, It will bit somewhat of a disappointement, but it’s bettr to see the "real" bits rather than to see the "old" bits and knowing that the real ones maybe really different in lots of ways. It would revoke the reason of coming in the first place – getting latest up to date information.

  2. Corrado Cavalli says:

    I’ prefer you to show us latest bits and gave us samples that will work on PDC bits.

  3. Jeff Key says:

    Show us the latest. We’ll have plenty of time before RTM to play w/the PDC samples.

  4. Carlo Cardella says:

    Plan B, of course! : )

  5. Marco Russo says:

    Plan B: we have not to ship products on Longhorn very early πŸ™‚

  6. Same here: Plan B. It is good to know where Longhorn is heading to, even if we’ll only have a limited version to play with afterwards.

  7. Christophe Lauer says:

    IMHO, Plan B is definitely the best option. It gives developers a long-term view whereas in some cases, the PDC build may miss some expected features and would lead developers to create workarounds or build similar features on their own.

  8. Corey Gouker says:

    Definitely Plan B…

  9. Joe Duffy says:

    Absolutely most definitely plan B. A nice, polished, Marketing-approved version might be nice to impress the PHB’s, but we geek-types want to see a few blue screens. πŸ˜‰

  10. Plan B sounds best. I’d like to see the cool new stuff with a few comments during demos on how it’s improved from the PDC fork.

  11. Dewayne Christensen says:

    Plan B from Outer Space

  12. Plan B, but be explicit in your demo as to which branch you’re coding to.

  13. At the risk of being a "me, too"-er, I’d opt for Plan B. No sense in providing information that you know is inaccurate.

  14. chris says:

    I guess the question thats begging to be answered is, How different is the branch? does the newer code simply contain more api than the older code, or are we talking about interface changes?

    if there have been actual design changes since the branch, and those design changes are going to stick, then by all means, show demo’s based on the agreed upon design; but if its just a matter of things being a bit more polished, i would rather be able to run a demo, un modified, on my own hardware.

  15. Scott Swigart says:

    As far as accuracy is concerned, I would guess that it’s statistically irrelevant which you choose, as I’m guessing that the delta between month-old-bits to RTM vs. today’s-bits to RTM, are within a rounding error of each other.

    That said, it sounds like you can show cooler stuff with later bits, and PDC is all about showing cool stuff, so I vote for B.

  16. Atul says:

    B — enough said

  17. Jay Welshofer says:

    Plan B!

  18. Kenny says:

    Plan A.
    why? I felt sorry for Plan A — no one voted.

  19. Keith Hill says:

    Plan B. Give us the latest and most accurate (at that point in time) info.

  20. Joel Semeniuk says:

    Plan B of course..

  21. Jimmy layton says:

    Plan B.

    Cause why see a slideshow of what you can run yourself.

    Show everyone what’s coming up!!!

  22. dannyR says:

    Huh? You forked around a month ago and now you think you have the model you’re going to ship? And Longhorn is still how far away from shipping?

    What’s plan C?

  23. Luka Debeljak says:

    Definitely plan B!

  24. Ian Hanschen says:

    Maybe cover the newer design for the demo, then quickly cover the changes between that and the PDC build, so the developers know what to do if they want to monkey with the samples.

  25. Michael Zill says:

    I the PDC bits are 4 week before the show starts so outdated that it does not make even sense to build the demos on it – so whats the value of giving these bits out to the attendees?

  26. Greg Hurlman says:

    It seems I’m a bit late for the party! I had a case made in my thoughts before I got here… but now, it’s been covered 4 times already.

    So, um, B. :o)

  27. DonXML says:

    I’ll be the rebel and vote for plan A. The main reason, the FUD that will occur after protyping something that people can’t play with. There’s a lot of misinformation going around already with what is or isn’t in Longhorn, and just wait until you protype something publicly that isn’t in the released bits. The FUD will be 1000 times worse, especially with the blogs. At least if we had the bits we could kill some of the FUD.

    DonXML

  28. Chris Jackson says:

    Plan B, but you should definitely point out what portions are exclusive to the new bits.

  29. MarkSA says:

    My two cents is for B.

    Those who will throw FUD arround will do it regardless of what you do.

  30. Kollen Glynn says:

    Plan B. I want to see as far in the future as possible because that’s ultimately what we’ll be building against.

  31. Olivier Hault says:

    Plan B, but provide two Longhorn versions on CDs, a stabilized with no SDK nor samples (for testing Longhorn) and a latest internal build and SDK with samples and demos (for alpha dev)

  32. Dmitriy Zaslavskiy says:

    Plan B

  33. Anonymous says:

    Plan B

  34. Matthew Douglass says:

    Well my preference would be to get both versions of Longhorn because they both serve important purposes.

    But what I need as a developer is something I can look at and start playing with as if it were real — as such, as nice as running the demos at home would be, give us a version with the APIs the way they will be (or at least close). I will can’t code without that and I’ve already seen the demos anyway.

  35. vsub says:

    Want both C Aguilera (plan A) and B Spears (plan B) !!

  36. I vote for Plan B. Show us the latest bits.

  37. I vote for Plan B. Show us the latest bits.

  38. Adrian Paleacu says:

    NO SIR! plan A.

  39. Brian says:

    can anyone remember in memory database?

  40. RobertK says:

    I attend PDC (as opposed to TechEd) because I want to measure the directional progress being undertaken at MS not because I want to see a polished RTM! So show me the bits and pieces and show me code (large font please) explain to me the thoughts behind it all ! I view the get togethers and BOFs at PDC as an opportunity for the MS-zealots (like me) to give some feedback (if you want it or not)!!!!!!

  41. Thomas Lee says:

    Plan B – and a promise to ship us newer bits within 90 days of PDC.

  42. pompi diplan says:

    Time permitting, I would combine both: A to encourage the audience about our commitment to stable, robust code and B to entice their curiosity. But an A-exorcism of fears (or prelude) cannot hurt before the B-experience.

  43. Da5id says:

    B. Architecture decisions attendees make based on what we see @ PDC impacts product design decisions. Correctness wins over stability, we know we are getting early bits, and we can give feedback on how we will utilize the final form so it can be evolved into goodness for all.

  44. Lamont Harrington says:

    How about a Plan "C"? Altough I’m an advocate of promoting the "vision" of a long term programming model/architecture, I do empathize with those who’d like to see workable, fairly stable, and most recent source code. Being that there has been a great deal of confusion it seems over when some of the technologies to be presented at this year’s PDC will RTM, perhaps delaying the release of the bits until the latest builds can be produced and distributed to attendees would be a better approach. There are those customers out there who would like to plan new product/service offerings around the "vision" and shouldn’t be ignored. An alternative is to provide access to Virtual PC images of the demo environments and make them available through some delivery mechanism.

  45. Wasp says:

    I write from Italy, and I won’t attend at PDC, but… plan B!!!
    :o)