Extracted from a thread began on @OP+1jgxdnf0h
207 replies (most recent on top)
The tech world is full of people who don’t behave according to any main stream definition of the word “humble”, yet are always learning because they know the bar of excellence is always being raised. The higher their accomplishments and commensurate influence/power often the more this perceived lack of humility manifests, especially to those who don’t possess or understand the math/engineering mindset.
Unless you are a credentialed and licensed psychologist, preferably with a PhD and a specialty in NPD who has formally evaluated OS, then please spare us the armchair diagnosis of “narcissist”. Of course, even if this were the case, HIPPA and all sorts of professional codes would prevent you from disclosing such diagnosis.
Many of us who worked with OS directly over a significant period witnessed him behave with high degrees of situational/self-awareness and empathy in many situations. I certainly did. Most of us cannot imagine the amount of stress he was under as an executive and engineering leader at SAS.
"JG-centered design" set the tone for the entire future of the company.
A single CAS connection process could only run a single action at a time, but a client could certainly spin off multiple connections and therefore run multiple CAS actions concurrently. AFAIK this remains the case to date.
@14n+1jk94kx0x
I was never familiar with the CAS internals, and forgive me if I am mis-remembering, but I seem to recall that a user could only have one outstanding action running at a time. That seemed like a limitation, but again, that was me looking at it from the outside.
@14q+1jk94kx0x I agree, OS was placed in a difficult position.
And, even with seven years gone, I remember the IMDB. What was your take on it?
Spot on. JG started a lot of this by tinkering with prototypes he thought would advance SAS and then handing them off to build products on. The in-memory database (IMDB) is a perfect example.
My history in SAS R&D predates MVA and extends well into the Viya era. The summarized cultural play-by-play given by @wg+1jk94kx0x is very accurate. Scapegoating OS exclusively for Viya failure is pointless because there was an entrenched executive and political machine surrounding his efforts. The problem was/is systemic and existed prior to OS beginning his tenure at SAS.
CAS initially ran on local hardware — SMP and MPP grids, not even VMs. Design work was put into making CAS “cloud friendly”, but a cloud-first design was not pursued. It’s important to remember this was 2012-2015 when although public cloud computing was emerging yet not fully ubiquitous AND standard deployment infrastructure like Docker and definitely Kubernetes were nascent and/or evolving. In that timeframe, SAS as a whole, including the CAS team, were already behind on these technologies. At some point JG+OS decided to fall back to initially targeting CAS for primarily on-premise deployments.
CAS never was a good fit for cloud paradigms like AWS Lambda.
"@120+1jk94kx0x
Curious, specific aspects of CAS design led you to conclude this?"
I'll quote from an earlier poster on this thread
"@hb+1jk94kx0x wrote: 'Viya might be interesting cloud wise if a customer can actually run more than one user ;)'"
Other cases would require me posting code segments that would pretty much out who I am :)
Sounds like a typical description of the SAS product development process:
- Build a vanity project
- Search for customer use cases.
- Convince the customer that you know best.
The world no longer accepts this strategy, if they ever accepted it at all. This is why SAS is in trouble. Viya is just a scapegoat for a larger cultural issue.
Former CAS founding member here. In my experience, a key to working directly with Oliver and influencing design/implementation details went something like this:
- Go along with his idea, which he almost always coded up in no more than a day or two. * Pay close attention to how the idea integrated with CAS overall.
- Demonstrate your improvement with running code that either integrated with his code/idea or replaced it. The former was always the safer option, but the later was possible if it was a really good idea.
Oliver did not micromanage and rarely demanded status more than once or twice a week max. Hence, there was typically enough time to iterate over improvements, providing team members were keeping pace with the overall arc of CAS development.
“CAS …, but it was, IMHO, a flawed architectural design from the start. The threading and parallel processing design was somewhat crippled from the start.”
Curious, specific aspects of CAS design led you to conclude this?
I don't know if I believe that. When the CAS team was formed it included the same people who worked on LASR in the ACL and others who "had scores of person/years dealing with multithreaded data and compute multi-user servers". Oliver also invited some people who flat out turned him down, some because they didn't want to return to the "startup grindset". I'm also sure Oliver didn't invite some people because he didn't want someone telling him what couldn't be done, or how it should be done, which was (and is) almost a hobby with some in R&D.
"If you were not on the CAS team, you were not important and were not consulted, no matter what your background or experience."
Silo example.
whenever a new server was envisioned the people with experience were never consulted
This seemed generally true. If you were not on the CAS team, you were not important and were not consulted, no matter what your background or experience.
30+ years, principal, worked in several areas in R&D, took my VRBP in the first round. JG always paid me well, the cafes kept me fed, the RFC was da bo-b, and getting the VRBP was the cherry on the top. Worked 35-40 most weeks, more when I needed to. Don't recall ever holding up the schedule.
For various reasons I decided not to apply for the CAS team, but my limited interactions with OS were OK. There were some technical problems with CAS related to multi-tasking. Of the things mentioned in this thread, I worked on projects integrating LASR, SPDS, CAS, HADOOP and many other products with V9.
CAS seemed like a cool idea. but without any practical benefit to customers when it was first rolled out, and in my recollection and limited understanding, no organized plan to move forward to provide customer value.
No doubt many of the developers I respected ended up on the CAS team, but it was, IMHO, a flawed architectural design from the start. The threading and parallel processing design was somewhat crippled from the start.
Our customers wanted V9 value. They could have focused on V9 integration and added the new stuff once the paying customers were on board. Just my opinion, obviously.
No idea what the current state of CAS is, so I hope they got their feet under them after 2018, though it doesn't sound like it has reached the potential that was hoped
The sad thing is that there were people at SAS that had scores of person/years dealing with multithreaded data and compute multi-user servers but whenever a new server was envisioned the people with experience were never consulted.
@x1+1jk94kx0x Some of us read this thread to learn what happened with Viya and why.
This matters because Viya appears to be a critical turning point. If it halted the revenue decline, JG might have retained ownership.
But with a declining asset, the logical decision is to sell. Once Oliver left, JG began talks with Broadcom.
So I hope that the OP will continue the discussion. When it descends into Oliver-bashing, though, we agree that there are better things to do.
@x1+1jk94kx0x
Whyareyouhere?Godothosethingsyousaid.Blessyourheart.I'msurewe'llneverhearfromyouagain.
None of this matters. Want to make a difference? Go visit a loved one suffering from cancer. Try to patch up estrangement with a family member. Offer an apology if you wronged someone. Donate blood. Donate time to a food bank. Volunteer at a pet shelter.
I challenge folks to post more ideas.
You sound like someone who learned all of this second-hand, and didn't know Oliver or ever work with him.
Oliver didn't invent Viya. Armistead did, in discussion with Dr. Goodnight. And Oliver didn't decide to make SNAP (oops I meant LASR) a thing either, that was a 60 / 40, maybe 70 / 30, Goodnight / Oliver conversation about the next generation of SAS. The very fact that CAS exists as a replacement for LASR is an admission that SAS was listening to customer feedback.
You have to remember that no one here was privy to the conversations Oliver had with Goodnight about the division of labor, which changed over time, and which probably went along the lines of "You design and build the product, and leave the business to me" in the beginning to more "Take care of this business" when he was made COO in addition to CTO. Goodnight even assigned him a minder, and an experienced personal assistant to make sure his public communications weren't stupidly alienating because he utterly lacked soft skills.
It's become popular to bash Oliver for the same reason everyone trash talks their former boss, or blames everything on the co-worker who just left. But he was just a guy who was given much by his boss, and probably returned much in return, then left when it all became too much. He made a lot of people angry, bitter, or resentful, and they were afraid of him because of his proximity to Goodnight. But he wasn't "just another nerd who thought he knew best and had no business background what so ever [sic]". When I first arrived at SAS they told me that he was a "prima donna". But a few years later those same people were telling me: "Oliver is the future of SAS".
… the 35-hour people in SAS R&D …
I don’t know how many developers in R&D worked a “35-hour” week but I’ve got to call B.S. because I was in R&D development and nobody I knew fits that description.
"Can you all cut to the chase and tell me why OS and JG parted ways?”
I’ll pass on this rumor because I find it believable.
I was told that OS requested to hire 200 more developers to build out Viya, and that he left because JG refused.
It’s believable that OS realized he was not moving fast enough for his market. He may well have concluded that the 35-hour people in SAS R&D could not produce the quality or quantity of work that he needed.
The timing of this request would have been after the first buyout. So it’s quite believable that JG would have turned down any such request. At that time, he was already trying to shrink his company, not grow it.
Jim compared to John was always easier to talk to when casually passing by on campus. He was more welcoming. John would return a pleasantry but it seemed forced. Always wondered if he had Asperger's syndrome. No fault of his own if so.
Ten or so years ago I would have never dreamed SAS would be in the current pickle situation.
“ Q: Where were the enlightened leaders and world-class technologists who could have set SAS on the correct path?”
Apparently they were all ja--ing around on thelayoff.com
For you anachronistic “competence argument” to hold water it needs to include JG and most of SAS for the past 20+ years, because that’s when the evolution of innovation-to-market impedance mismatch, creeping staff mediocrity, bureaucratic bloat, and ineffective leadership began. In truth, further back than that, but almost no one could have foreseen it in the late 90’s and JG was only focused on 2-year forward tech time horizons back then.
OS did the best he could with who he had to work with and the constraints JG required him to operate within. How come no one with better ideas and a commensurate work ethic prevailed instead?
Q: Where were the enlightened leaders and world-class technologists who could have set SAS on the correct path?
A: They were working for the cloud hypers-scalers, hot startups, FinTech, etc. for a LOT more money.
"Why haven’t SAS biz + tech leaders managed to execute on new ideas that address the declining revenue?"
I suspect you are much younger than JG or JS hence do not understand the brain machinations of an elder person. Hence your question.
In poker terms, JG and JS have the desire to fold and it is stronger than their desire to continue playing their hand.
To put it another way, the safe bet is to keep reducing costs. That makes profit look better on a GAAP compliant balance sheet. Better profit means better chance at catching a buyer. An older person wants to get some things done before they expire.
To put it an even different way, the wise investment for elders is a risk averse investment. JG and JS are not young men and have no young heirs interested in taking over the SAS operations.
JG and JS are doing the best they feel is possible with the hand they have been dealt, either by circumstances or by their own making.
In a nutshell a senior's brain works differently than the brain of a younger person. Life does that.
If I was still there, I would personally thank Jim for the environment and the glory years. Neutral on John, never was interested on working on anything remotely JMP related.
SAS was the best job I ever had, and the jobs before SAS were very decent.
Good luck to all!!!!!
@vv+1jk94kx0x More dime store psychology lessons from anonymous people on thelayoff.com
Humpty Dumpty indeed.
“ In this regard the notion of "build it and they will come" certainly does NOT apply.”
Viya may not be a rousing success but build it and the will come is how virtually every major piece of software has come about, business users rarely know what the he-l they want. It typically comes from some “nerd” looking at how inefficient and narrow focused people are at doing their jobs,
It's a false dilemma. You can take the "narcissist" part as an insult, but it wasn't intended to be one. It's used as a clinical term to mean an insult to one's ego and vanity. Here it is in more simple terms:
Humpty Dumpty sat on a wall
Humpty Dumpty had a great fall
All the King's horses
and All the King's men
Couldn't put Humpty Dumpty back together again.
Humpty Dumpty sat at a great height; on top of the world, or the top of a corporation. They develop a big head from it. Some outside force knocks them from the great height, they fall to earth, and crack up; they realize they are fallible, and flesh and blood like the rest of us. That's the narcissistic mortification part. This realization utterly crushes them -- they implode like the Titan. They have hit "rock bottom", and won't be returning to sit atop the great wall. They now have to navigate the world like everyone else.
Humility is painful for some.
"Can you all cut to the chase and tell me why OS and JG parted ways?"
Have no idea why he departed. But safe to assume that if he truly loved working in the tech sector and had the skills for then he would still be in the tech sector making a good go of it.
I do know this for sure. Not much crying on Cary campus when he left SAS. My job role at SAS was so far removed from his that I can not say if that previous sentence is fair or not.
"Scapegoating everything on OS + Viya is getting so tired. A competent crew surely could have cleaned the mess up by now!"
It will take more than 4.5 years to unwind a 10 year mess.
even the best could not solve the innovator's dilemma
Can you all cut to the chase and tell me why OS and JG parted ways? Why why why : )
A competent crew surely could have cleaned the mess up by now!
A competent crew surely would not have gotten the company into this situation in the first place.
CAS + Viya Compute Server + Micro Services = Framework for building a range of software products.
OS departed SAS 4.2 years ago. You’ve had a friendly populist CTO every since, JG is sill at the helm and Building C culture survived. Gen Z has arrived in force and DEI is still alive. We hear that R&D has plenty of able minds. The S2 partnership is still intact, etc.
Hence, a BIG unanswered question remains:
Why haven’t SAS biz + tech leaders managed to execute on new ideas that address the declining revenue?
Scapegoating everything on OS + Viya is getting so tired. A competent crew surely could have cleaned the mess up by now!
Alot of talk about OS and Viya.
Viya has proven to be impotent at moving the revenue needle upwards. Good software addresses business needs that are big enough to drive revenue upward. In that regard, Viya is a costly failure.
Oliver was just another nerd who thought he knew best and had no business background what so ever.
His business acumen was so deficient that he shunned customer feedback that could have guided him towards recognizing important business needs. Rookie mistake. In this regard the notion of "build it and they will come" certainly does NOT apply.
Put the two together, and you get a strong understanding of why SAS is for sale...and has remained unsold.
I'm glad OS is in a happier place. But he showed that he did not care about my feelings, so it’s hard to care much about his. I care more about what he did that affected my life and others.
At some point, it must have become clear that Viya was competing against open source, and was not what the market wanted. I’ll be interested to hear the OP’s view on why SAS did not change course.
Do you think JG gives a flip about “emotional intelligence“ or put by effort into cultivating it himself?
I think JG knows that you can’t treat your workforce like sh1t and expect them to do good work.
In my considerable personal experience,Oliver@SAS was a driven individual who did not suffer fools or care much if they got butt hurt
Oh, god, it‘s the butt hurt guy. Yeah, yeah, we all know that you know have societal permission to be an a55hole. Just because you can doesn’t mean you should. Y’know, because it just makes you an a55hole. You and OS both.
You're presenting a false dilemma. You can be both emotionally intelligent and a good leader.
The "kinder / gentler" persona arises after a period of narcissistic mortification. Humpty Dumpty had a great fall and all that.
Various Tech CTOs don’t have the “leadership skills” that you are describing. To many they can appear to be aloof, socially insensitive or even downright arrogant. They are often people are very few words. Their job isn’t to exemplify modern corporate “emotional intelligence“ indoctrination. That said, some of us who knew him post-SAS, witnessed a generally kinder and gentler Oliver.