Thursday, October 15, 2009

Aweekstweets October 11-18 2009

Aweekstweets October 11-18 2009

  • Oh, that's right, a Twitter phish hack screwed us royally at mid-day. Almost forgot. Cool how quickly a social network responds, moves on.about 9 hours ago from TweetDeck
    • JK2—With Twitter down (for me), I circulated (via Facebook) a (too late for me) link to an article in the trade press warning Twitter users not to reset their passwords. We (me and some other analysts) used e-mail to back-and-forth a status on the hack, and what we could do. I was (physically) sitting in an analyst summit room (in a plush hotel) at Oracle OpenWorld (in downtown San Francisco) with a bunch of other impacted Twitter-using analysts, and we pool notes on the whole thing. And we kept paying attention and asking probing (non-Twitter-related) questions to (physically in front of us) Oracle. We survived, and socialized through the whole thing. We even joked about it all.
  • RT @rwang0 @destinationCRM: "All interfaces are BI-driven. No need to buy a BI add-on #oow09" JK--Whither stand-alone BI for ORCL app shops?about 11 hours ago from TweetDeck
    • JK2—Embedded BI features in forthcoming Oracle Fusion App suite, per Larry Ellison preso at Oracle OpenWorld. Stand-alone BI will still be a necessity for many Oracle app shops. The BI that’s been embedded in Siebel etc for years is rarely any user’s preferred tool, and I suspect it will be the same for Oracle Fusion apps users. Besides, Oracle Fusion apps won’t necessarily be the whole app story for many users, given that commitments to many Oracle pre-Fusion apps will endure, as will usage of various SAP and other non-Oracle apps. And Oracle certainly won’t slack off trying to sell users on its own stand-alone Oracle BI Enterprise Edition suite.
  • RT @williammcknight: "[ORCL]..$10M challenge..ibm..for app..1/2 as fast as on exadata2"#oow09:JK--Bucks for DW bang: ORCL bets rival fizzlesabout 11 hours ago from TweetDeck
    • JK2—Gutsy, bold, classic Ellison stunt. Would be even gutsier if Oracle made it a continuing program. In other words, why not give any customer (not just one lucky customer) some specific discount on Sun Oracle Database Machine if they can prove that some deployed instance of the DW appliance of any rival (not just IBM) can perform at least half as well (on handling both mixed-query and data-ingest workloads) as an equivalently configured (as CPUs, cores, blades, memory, disk, I/O bandwidth) Oracle system. But Ellison didn’t break it down to these sorts of meaningful details. Besides, he knows he’d pay through the nose for that. If he knew what I know’s coming along, he’d watch his wallet closely.
  • #OOW09: Ellison going laser-sharp on Netezza as key rival. Netezza much smaller, of course. But Netezza has DW momentum that Oracle covets.about 13 hours ago from TweetDeck
    • JK2—Ellison and crew have done a good job of building and sustaining their own momentum around the Oracle Database Machine/Exadata product family. I suspect that the reason they announced Exadata v2 a month before this year’s OpenWorld (rather than at the show) was to keep the industry focus on Oracle’s progress in this market, and to shift attention from their frustrating, protracted, regulatory-stymied attempt to acquire Sun. In fact, last month’s Exadata v2 announcement was a pre-announcement: Oracle’s not yet shipping any units of the new machine, and they didn’t give any clear GA dates at the show either, though all indications are it will ship in the coming month. Ellison was hardselling near-futures, not an actual product that customers can take possession of now. BTW, Netezza is actually shipping TwinFin units, and has been since the product was announced a couple of months ago.
  • Ellison focusing on Oracle as provider of complete systems, in 1960s model of IBM's TJ Watson. IBM? Yep, but Larry not channeling Palmisano.about 13 hours ago from TweetDeck
    • JK2—Funny. Ellison’s pooh-poohing Lou Gerstner’s transformation of IBM into a professional services powerhouse, and Palmisano’s efforts to shift the company toward more of a verticalized, solution-focused vendor, and away from being the Big Blue that was primarily known for hawking Big Boxes. In this analyst’s humble opinion, the new IBM’s services and solutions focus is the key to sustained profitability in a market where the big-box integrated systems of yore (which Exadata machines sorta resemble) will be undercut by lean, low-cost commoditized competitors galore. Larry talks a good talk though.
  • At #OOW09, as at #OOW08, Oracle's biggest advantage is that CEO is laser-focused on this strategy tech, and doing relentless hardsell.about 13 hours ago from TweetDeck
    • JK2—Don’t underestimate the power of hardselling. Though I found the late TV-shouter Billy Mays hyper-annoying, he was beloved by his clients, who said he had an unparalled ability to move units en masse with his motormouth.
  • No doubt, Larry, that Exadata v2 is fast, scalable, HA and that flash cache is disruptive. But let's see yr "faster than IBM" proof. #OOW09about 13 hours ago from TweetDeck
    • JK2—He offered plenty of numbers and assertions. I think that “proof” is too open-shut a conversation-ending term when discussing performance of something as complex as a DW appliance. By the way, in the track session on Oracle Enhanced Hybrid Columnar Compression (a core feature of Exadata v2), the Oracle person on stage said (rapidly, in passing) that this new storage layer, in speeding queries, must of necessity put a performance hit on loads, slowing them down. I forget the exact technical reason (I’m reviewing my notes and their slides), but there you go. Robbing Peter to pay Paul. Larry didn’t mention this little, er, inconvenient truth in his preso. And it’s nowhere to be found in any of the marketing collateral.
  • Ellison and customers attest to 10-50x Exadata v1 accel vs. previous Oracle DWs; Jobs @ Apple said 60x. Exadata v2 v. v1: how much faster?about 13 hours ago from TweetDeck
    • JK2—Clearly, Oracle is most comfortable presenting the advantages of its latest DW platform (Exadata v2 this year, v1 last year) vis-à-vis previous generation of its own products. It even has Larry’s friends vouching for the same. Though clearly Ellison went to town with benchmarking v2 against rival offerings as well, these comparisons are highly, er, problematic, as evidenced by Oracle’s recent run-in with TPCC, under which Oracle had to withdraw published IBM comparisons, which of course put IBM in a sorry light but weren’t audited by an independent third-party benchmarking firm.
  • Oracle's rich saturated red color scheme is pure energy. The Sun logo is calmer blue. Exadata machine a steely light gray. Larry is coffee.about 13 hours ago from TweetDeck
    • JK2—Highly distinctive blend of primary and monochrome hues in corporate branding, extending to CEO attire and hot now hardware product. Bravo!
  • Here's the downlow from #OOW09, Oracle has circled back to DBMS/DW focus, but apps, middleware, virtualization, and platforms drive the showabout 13 hours ago from TweetDeck
    • JK2—Ellison’s detailed discussion of Oracle Fusion Apps was just as important (actually, potentially way more important) as Exadata in future revenue contribution. But far more boring, of necessity. SAP’s conferences revolve around this sort of stuff, which is of keen interest to various vertical-market customers, as it digs deep into the heart of their operational requirements. But few of these industry-specific or function-deep discussions will play in the Wall Street Journal.
  • Oracle did good analyst summit today at #OOW09. Had been prebriefed on most details. But they clearly have value-prop, customers, momentum.about 13 hours ago from TweetDeck
    • JK2—They had a half-dozen Exadata customers doing in-person testimonials. Showed logos of 20 Exadata customers. Hinted at much longer list of customers in production or on the verge (for v1) and strong interest in v2. My independent assessment, taking CTO- and DBA-level inquiries at one of the leading IT-analyst firms, is that Oracle isn’t exaggerating any of this. Exadata’s a hit. It’s the out-of-the-box success that HP wishes Neoview (on the market for 3+ years, in with which 1-year-old Oracle Exadata has approximately equal number of production customers) were.
  • Re this phish attack, I was a Twit, I admit. In my defense, @jpmorgenthal is to blame for leading me down the click-strewn path to perditionabout 13 hours ago from TweetDeck
    • JK2—Way too much coffee that morning. I was breathlessly clicking every stupid link in sight. Should have shut my f***ing laptop while listening to Oracle.
  • Oracle Exadata customers generally report in the vicinity of 10x acceleration vs. prior Oracle DW. #OOW09about 15 hours ago from TweetDeck
    • JK2—This is borne out by the many Forrester customers who are Oracle shops and have evaluated, and in some cases deployed, Exadata (v1 only, so far).
  • George Lumpkin says Oracle's done 100s of Exadata demos, wkshps, POCs, benchmarks around world. Plenty of eval units out there. #OOW09about 16 hours ago from TweetDeck
    • JK2—For sure. Every enterprise DW POC these days seems to include Exadata (v1, so far), and usually in a short list of no more than 3 vendors. Netezza even more often. Greenplum well-represented too.
  • George Lumpkin says Oracle has v1 customer wins, mostly in FY Q3 & Q4. Not giving out specific numbers. Showing slide with 20 logos. #OOW09about 16 hours ago from TweetDeck
    • JK2—Yes. It’s a fairly long sales cycle for Exadata, due to a) the technical eval complexity, b) the price tag, c) the weak economy, and d) soft IT budgets.
  • George Lumpkin discussing the many fundamental improvements in Exadata v2 over v1. #OOW09about 16 hours ago from TweetDeck
    • JK2—BTW, George is head of DW product management at Oracle. As I’ve said elsewhere, Exadata v2 has so many fundamental performance-enhancing improvements—including hybrid columnar compression, comprehensive in-database analytics and OLTP, and SSD/flash cache—that it’s even more significant than v1 for Oracle’s longterm differentiation in this arena.
  • Parallelize a query, or paralyze it? Query optimization with mixed workload mgt and dynamic resource governance makes all the difference.about 16 hours ago from TweetDeck
    • JK2—Parallelize queries across Exadata storage cells in conjunction with parallelizing them across extraordinarily fast flash cache—unbeatable combination. Expect competitors to introduce very similar fundamental improvements in the next 1-2 years.
  • Oracle 11g w/ Exadata as "foundation and infrastructure for applications," per ORCL's George Lumpkin, DW analyst summit. #OOW09.about 16 hours ago from TweetDeck
    • JK2—Intelligent parallel persistence (storage/cache) layer with comprehensive in-database analytic and transactional pushdown optimization: key to transforming DW into the new analytic app server platform. See my forthcoming Forrester report on this fundamental game-changing industry development. Other DW vendors have similar roadmap to Oracle. Some of them will announce some of these roadmaps, and new platforms, very soon. Stay tuned.
  • Proof is in the pudding, or proof is in the put-on? Slippery world of DW vendor performance claims, and what if anything is "proof." #OOW09about 16 hours ago from TweetDeck
    • JK2—Stop using that word in this context. Only mathematicians, logicians, and lawyers speak of “proof” in the classic “summary judgment” sense of the term. Everybody else, including computer scientists and engineers, speak of demonstrations. Show me. If someone can show me faster, even better.
  • Per Loaiza, Oracle "Beehive" (email, calendar etc.) on Exadata w/OLTP. Internally, ORCL biggest DBMS. Is Exadata becoming app server? #OOW09about 17 hours ago from TweetDeck
    • JK2—Very interesting “eat your own dogfood” story from Oracle. See note above about DW becoming analytic app server. As you add OLTP-ish apps to it, DW becomes key component of “transalytic” app server (I just made that one up).
  • RT @bwerther "GP's Polymorphic Data Storage: @CurtMonash http://bit.ly/1ymHMg http://bit.ly/W1zIL: JK--Is lo-footprint DW "ectomorphic"?about 17 hours ago from TweetDeck
    • JK2—GP = Greenplum. Polymorphic = similar in key respects to Oracle’s enhanced hybrid columnar compression scheme. Leading-edge DW vendors are gong this hybrid route. Pure-columnar and pure-row-based DW/DBMS vendors: take note. Combine storage/persistence approaches in distributed, role-optimized DW fabric: you’ll get better performance with more flexibility.
  • RT @CurtMonash "No risk in getting paid maintenance by customer ..otherwise get $0 rev from." JK--Yep: clear overlap with open-source model.about 17 hours ago from TweetDeck
    • JK2—If a DW vendor demonstrates the promised order-of-magnitude superior performance in the customer’s actual production environment, or in a realistic testbed environment, they should be able to charge a premium upfront on the hardware/software combo. If they don’t demonstrate up to promise, the vendor should sell at deep discount to list, or, extreme case, give to customer “free,” but with minimum customer commitment to premium maintenance contract. In other words, the open-source model syncs up with value-based pricing of DW appliances based on actual realized performance.
  • RT @Intelligentform @rschmelzer "analyst biz relic of..when info was limited in...availability" JK--No: analyst value is actionable contextabout 17 hours ago from TweetDeck
    • JK2—Nobody has ever turned to analysts for pure information. IT pros have always had plenty of ready sources for info. They’ve always paid a premium for expert people who can put 2 and 2 together, bring in integral calculus and advanced stats, and then use it to assess the tensile strength of whatever software/hardware/network superstructure you want them to help you lash together. That’s actionable context. The customers already knows basic arithmetic, and enough of the higher-order stuff to determine whether the hired analyst/consultant is a true wizard or a wan wannabe.
  • RT @CurtMonash "ParAccel challenge...If they fail challenge, "free" isn't valuable." JK--Agree, but DW vendor bears fair share of risk.about 17 hours ago from TweetDeck
    • JK2—DW vendors would think twice about blithely claiming “10x” query performance enhancement if they had to cut their price one 10 percent for every “x” they couldn’t deliver in the field. No extra “x”s from our new DW appliance? Then no charge! “Yikes!” (say the vendor’s shareholders). “Yippee! (for now—see what the competitor offers),” say customers.
  • RT @JeromePineau "ORCL...what will do with MySQL...Larry has no clue" JK--Disagree: Larry has clues crammed in vestpocket, keeps poker faceabout 17 hours ago from TweetDeck
    • JK2—Has anybody who’s ever observed Mr. Ellison, even casually from a distance, mistaken him for an innocent bystander?
  • Juan Loiaza on Exadata v2, at DBMS/DW analyst summit, #OOW09. Game-changing: full in-db analytic pushdown + flash/cache. Will go all-flash.about 18 hours ago from TweetDeck
    • JK2—Seriously important: the future all-virtualized, all-cloud, all-memory, all-latency, all-info, all-app DW platform. Stay tuned (to Oracle and many of its visionary competitors). Not a deep future prognostication. A 2009-2010 “coming soon” to a DW near you” story.
  • JK thoughts re "bang4budget" DW appl pricing: $/TB-usable sets "budget" baseline + premium for POC-customer-site vdr-demo 10x+ accel "bang"about 18 hours ago from TweetDeck
    • JK2—Let’s say the DW industry “budget baseline” settles for the time being at $20K/TB per the core DW appliance stack delivering some baseline query, load, and concurrency performance. If, dear DW appliance vendor, you can demonstrate , in the customer’s environment with their data/loads, an order-of-magnitude superior performance over that baseline, you can justify a higher per-TB price. If you can’t, see note above: extreme case of “same ol’ same ol’” performance means you give your DW appliance to the customer for “free,” but with customer lock-in to a multi-year maintenance and support contract. That’s clear DW “bang for budget.”
  • Oracle can't say anything on what will do with MySQL. This analyst's hunch: midmarket lo-cap lo-cost DW appl on Sun hw: MSFT-killer #OOW09about 18 hours ago from TweetDeck
    • JK2—Oracle is really and truly mum on this issue. Good for them. Their team was disciplined and circumspect to a person re this issue, at this year’s OpenWorld.
  • RT @nyuhanna: "I'll be presenting at #OOW09 today at Noon on how to develop a security plan..." JK--DBMS security. DW pros tend to overlook.about 18 hours ago from TweetDeck
    • JK2—Frankly, I’m surprised nobody addresses DW security under the broad heading of “data governance.” I was having this same discussion with Roxana from Oracle.
  • RT @raesmaa: Stress vs productivity during a work day [infographic] http://bit.ly/eICZt: JK--Productivity calms me, inertia unnerves.about 18 hours ago from TweetDeck
    • JK2—The post-productivity inertia calms me even more.
  • For Oracle 11g R2 RAC deploys, 8-node are customer "sweet spot," says Mark Townsend, w/ some on 16-node. Larger clusters uncommon. #OOW09about 19 hours ago from TweetDeck
    • JK2—Mark’s also high up in Oracle’s DW product organization (sorry I can’t recall anybody’s actual job title at this moment). BTW, as Oracle customers seek to scale out to petabytes in cloud environments, they’ll almost certainly expand to 100s and 1000s of parallel nodes in shared-nothing MPP grids. Hey Oracle: consider that a suggestion from this analyst for the inevitable “Exadata v3” (no, dear blog readers, Oracle has not announced, discussed, or even hinted at a v3—yet—awfully premature).
  • Oracle has no clear strategy for Exadata-based analytic appliances. Have componentry (DW, BI, LDMs, etc.) but no solution packaging. #OOW09about 19 hours ago from TweetDeck
    • JK2—Key weakness of Oracle’s Exadata platform products in a commoditizing DW appliance market.
  • Oracle has no clear strategy for Exadata DW in cloud, hosted by them or partners. Ellison's cloud positioning feels noncommittal. #OOW09about 19 hours ago from TweetDeck
    • JK2—Another weakness in Oracle’s Exadata product roadmap. But they’re not the only DW vendor keeping their distance from cloud. It’s more of a 2010-2011 topic. Customers aren’t screaming for it. But they are asking me about cloud DWs—and the most appropriate apps/roles for them—through Forrester inquiries. I’m sure Oracle will begin to address when they, and the market, are ready.
  • Listening to Oracle's Mark Townsend speak on DB 11g R2 & Exadata v2 at DBMS/DW analyst summit, #OOW09. Most interesting: flash/SSD storyabout 19 hours ago from TweetDeck
    • JK2—The days of the all-solid-state-drive DW appliance are approaching. No, they’ll be pricier and less capacious than all-disk DW appliances for at least the next 3-5 years, but they’ll be wicked fast right out of the starting block!.
  • DW value-based pricing based on upfront proven customer-site query accel would change market: put POC-aggressive vendors at advantage.about 19 hours ago from TweetDeck
    • JK2—See discussion above of “bang for budget” pricing. Speaking of POC-aggressive DW appliance vendors: I currently put Oracle, Netezza, and Greenplum in that category.
  • "Exadata v2 Raises Questions re DW Mkt" (http://bit.ly/3OmIig): JK--#Forrester seeing significant customer Exadata eval, growing deployment.about 19 hours ago from TweetDeck
    • JK2—Every working day I see this. You should see my inquiry schedule (but, no, I won’t/can’t show you)..
  • By using already-dated "Cash4Clunkers" term, ParAccel just feeding on FUD. Should change program name but keep cool trade-in allowance.about 24 hours ago from TweetDeck
    • JK2—Oh, and while ParaAccel’s at it, they should get their appliances into POCs everywhere, so more prospects can kick the tires. Last time I checked, they haven’t been anywhere near as aggressive as a former start-up, Netezza, that’s come a long way in a short time by making customer-site competitive-benchmarking POCs the focus on their sales dog-and-ponies.
  • "Univa Puts Oracle E-Business Suite In Cloud" (http://bit.ly/27RmID): JK--When's Oracle going to put Exadata in cloud?6:53 AM Oct 14th from TweetDeck
    • JK2—As you may or may not be able to tell by now, I’m reasonably confident that Oracle will put Exadata “in the cloud” in the coming 1-2 years, and will begin to put that cloud Exadata into its cloud-based transactional apps inn roughly the same timeframe.
  • "Oracle Execs Detail App Integ Strategy" (http://bit.ly/flka8): JK--If integrated ORCL suites "send you complete car," do cloud apps=zipcar?6:47 AM Oct 14th from TweetDeck
    • JK2---I suspect that as Moore’s Law kicks in on DW size, more of these multi-100-TB and petabyte-plus DWs will live only in public clouds. The data marts you provision in these environments for limited projects (e.g, clickstream, tweetstream, tickstream, eventstream mining for marketing campaigns, customer churn, and the like) and then promptly deprovision at project end? You’ll essentially be “renting” them: zippy zipcars that can sprout titanium wings and supersonic engines for enterprising low earth orbit and surveying Planet Petabyte.
  • Pay bucks for POC-demonstrated DW "bang" not silly. It's value pricing. @laurenk1802432: @JeromePineau @jameskobielus @merv @CurtMonash6:19 AM Oct 14th from TweetDeck
    • JK2—See “bang for budget” discussion above. Value-pricing for demonstrated DW performance. Not “trust us” pricing based on bold public marketing swagger by Oracle or any other DW vendor.
  • RT @JeromePineau @CurtMonash: "ParAccel's faster or free gimmick" JK--Not bad idea: if DW POC can't make good on 10x claim, give it to 'em.12:01 AM Oct 14th from TweetDeck
    • JK2—Often times, startups with less to lose and chutzpah to burn innovate in surprising ways: such as new pricing approaches that make great sense for the market as a whole. Sometimes, these same startups don’t fully realize what they hath wrought.
  • Problem w/ ORCL being so big is that individual analysts never get to grill CEO 1:1. Same w/ MSFT, SAP, IBM. IMHO, would behoove 'em. #OOW096:29 PM Oct 13th from TweetDeck
    • JK2—I’m an analyst who has enough chutzpah to at least request being allowed to enter the presence of His or Her Excellency.
  • Open discussion: is Exadata simply Oracle proprietary arch a la old Netezza? Of course, with TwinFin on IBM, Netezza's slightly beyond that5:15 PM Oct 13th from TweetDeck
    • JK2—Oracle says they don’t want to be in the commodity hardware business, but how can they avoid it after acquiring Sun? They can avoid it by taking Sun’s hardware entirely proprietary within Oracle-branded Exadata-based solution DW, analytics, transactions, and solution appliances—and by entirely cutting off all of Sun’s many OEM and ISV partners, including several other DW appliance vendors. I doubt that Oracle wishes to strangle that particular golden-egg goose. Besides, Oracle will need the high-volume unit shipments of a commodity-hardware in-house vendor to keep scale economies of procurement, production, and distribution low enough to keep costs of the Exadata hardware low enough to keep it at the $20K/TB threshold—and inexorable cheaper—to stay competitive.
  • What I like most about Oracle EHCC is ability to optimize both fast table scans and fast single-row lookups in single storage layer. #OOW094:35 PM Oct 13th from TweetDeck
    • JK2—In other words, optimized for analytics and transactions, OLAP and OLTP, DW and LOB.
  • Great discussion at #OOW09 on their enhanced hybrid columnar compression (EHCC) in Exadata v2. Huge differentiator!4:26 PM Oct 13th from TweetDeck
    • JK2—Their EHCC compression efficiencies are sensational!
  • RT @CurtMonash "M/R paradigm commonly violated...good thing.. it illustrates paradigm's strength" JK--Yep. Inline analytics the big paradigm11:55 AM Oct 13th from TweetDeck
    • JK2—See my forthcoming Forrester report: “In-Database Analytics: Heart of the Predictive Enterprise.”
  • Vendors talk about "building relationships w/ analysts." Remember that "relationships" are wildcards. We won't necessarily agree with you.9:20 AM Oct 13th from TweetDeck
    • JK2—But we will spend quality time with you!
  • Vendors talk about "building relationships w/ analysts." Remember that analysts are basically market researchers. First rule: loop us in!9:15 AM Oct 13th from TweetDeck
    • JK2—E-mail is fine. Nothing fancy necessary.
  • Vendors talk about "building relationships w/ analysts." Sounds way heavier than it is in reality. Basis for any good relationship: say hi!9:13 AM Oct 13th from TweetDeck
    • JK2—And if we forget your name, even momentarily, please don’t hate us. We’re working hard on connecting names and faces.
  • Catch my presentation on in-database analytics at next week's Teradata Partners conference, National Harbor MD, on the mighty Potomac!2:00 AM Oct 13th from TweetDeck
    • JK2—And meet Egidia.
  • Whatever became of Oracle Optimized Warehouses in their DW product portfolio? Superseded by the Exadata product in strategic importance.1:35 AM Oct 13th from TweetDeck
    • JK2—They never mentioned the Oracle Optimized Warehouses, as products or as partnerships, at OpenWorld. Not even once.
  • RT @marksmithvr: "Oracle CEP....not integrated into Oracle BI discussion yet" JK--Missing from both Oracle BI & Oracle DW (Exadata) strategy12:39 AM Oct 13th from TweetDeck
    • JK2—I bet they’ll ramp up the CEP story, deepening the acquired BEA technologies, in 2010. Why? IBM, Microsoft, and SAP focusing on it.
  • Modern computing began in WWII calculating ballistic trajectories. Now we do same on grander scale: predictive analytics: hit moving targets12:09 AM Oct 13th from TweetDeck
    • JK2—I think I aweekstweeted about this thought before. I forget. But, oh well, here it is again (or the first time, as far as I know).
  • Social network analysis: more & more we'll use predictive models to surface emergent patterns of future civilization: archaeology in reverse12:06 AM Oct 13th from TweetDeck
    • JK2—Important eval criteria for differentiating in today’s predictive analytics and data mining market.
  • @merv noted that #OOW09 put DBMS/DW in (my word) side (i.e., west) Moscone, and app/middleware in (mine again) main Moscones. Significance?11:59 PM Oct 12th from TweetDeck
    • JK2—Don’t read too much into it. Moscone West was jampacked too.
  • HP SAP partnership on DW. Neoview needs vertical solution packaging. They recognize that. Will work toward that goal. Details TBD.5:27 PM Oct 12th from TweetDeck
    • JK2—Same recommendation applies to Oracle. Same TBD applies.
  • HP's Livermore up to speak. Curious what she'll say, or not say, about the Exadata v2 machine, and whither the HP Oracle partnership. #OOW0911:35 AM Oct 12th from TweetDeck
    • JK2—She said nothing. What could she say?
  • RT @merv: " just be list of different apps and industries, or is there some thematic unity? no setup at beginning to tie it together #oow09"11:32 AM Oct 12th from TweetDeck
    • JK2—Oracle OpenWorld 2009 got off to a disappointing start, but built up to a reasonably satisfying climax with Ellison’s Exadata v2 hardsell and “$10 million challenge.” Gov. Schwarzenegger had a good, substantive speech as well. But the chief take-away—Oracle rocks and rolls over the competition with Exadata v2—was way too swagger-intensive to function as a unifying theme, considering that it had little direct relevance to the apps and middleware products that occupied more than two-third of Moscone Center this time around. On the whole, there was something lacking from the OpenWorld experience this year. Oracle was gracious and informative, though, at least with the analysts in attendance.
  • I'd like to see some discussion of Oracle Data Miner as an enabler for social CRM: demand planning, etc. #OOW09.11:31 AM Oct 12th from TweetDeck
    • JK2—One big disappointment for me, considering that I’m Forrester’s lead analyst in predictive analytics and data mining, and also considering how important this is for the future evolution of in-database analytics on Exadata platforms.
  • On the DW/DBMS side, Oracle announced DB 11g R2 and Exadata v2 a few weeks ago. This year's #OOW09 is mostly just educating on all that.11:18 AM Oct 12th from TweetDeck
    • JK2—Tactically, I think Oracle erred by not waiting till OpenWorld to announce Exadata v2. For starters, they’re still not shipping it yet, which can only breed frustsration and disappointment. Also, it could have been their banner announcement to add the necessary dramatic oomph to this, their big annual show. (no, I’m not convinced that the “$10 million challenge did the trick; see above). Furthermore, Oracle gave DW competitors (most significantly Teradata and IBM, whose respective conferences I’ll be attending in the next 2 weeks) a lead time to digest the Exadata v2 announcement and tailor their counter-messages.
  • Disappointed there are no splashy announcements from mainstage at #oow09. Ellison spoke last night. Judge for yourself: http://bit.ly/YwVhD11:16 AM Oct 12th from TweetDeck
    • JK2—Glad I was able to have a semblance of a Sunday at home (morning) and a relatively leisurely flight to SF, with a good night’s sleep. Reading about it the following day was more than sufficient.