Interesting findings on the internet

[bd139]

The "take big risks" was the mantra of HPLabs, of course. I know of one engineer who was dinged because his work that was transferred to a division wasn't a big enough success.

"Take big risks" is a common mantra in startup culture. Along with "move fast and break things". I've always treated this mentality as naive, bordering on idiotic. It stems from the idea of creating new markets for products or services. It makes me feel dirty seeing the amount of this garbage I've had to sit through over the years by people with a track record of having one shot coincidental successes and trying to apply it to new things or existing solely to instil their apparent virtues on others.

My personal mantra in that space is "solve a need". If you aren't approaching a problem from the perspective of a well-defined need then you are taking an incalculable risk or doing something pointless and that is stupid. The art is really finding and understanding a need and that's a very much lost one when someone will throw millions at a bunch of stupid ideas and hope one of them hits the jackpot one day.

[tggzzz]

The MBAs crawled out when Bill and Dave were dead. They wanted a "50:50" company, i.e. 50k employees, $50bn. They didn't like the expense of HP Labs, thinking they could simply and more cheaply buy successful technologies. You know, things like Autonomy and WebOS.

My, didn't that work well.

(Oh, one CEO was allegedly screwing more than the company into the ground; rooting in the Aussie sense)

[tggzzz]

The "take big risks" was the mantra of HPLabs, of course. I know of one engineer who was dinged because his work that was transferred to a division wasn't a big enough success.

"Take big risks" is a common mantra in startup culture. Along with "move fast and break things". I've always treated this mentality as naive, bordering on idiotic. It stems from the idea of creating new markets for products or services. It makes me feel dirty seeing the amount of this garbage I've had to sit through over the years by people with a track record of having one shot coincidental successes and trying to apply it to new things or existing solely to instil their apparent virtues on others.

I've been in a couple of startups. They were technically risky, but not financially nor career-wise. The path of the second has been described by Herman Hauser; it didn't surprise me.

As for "move fast and break things", that has its niche. But not, please, autonomous cars.

My personal mantra in that space is "solve a need". If you aren't approaching a problem from the perspective of a well-defined need then you are taking an incalculable risk or doing something pointless and that is stupid. The art is really finding and understanding a need and that's a very much lost one when someone will throw millions at a bunch of stupid ideas and hope one of them hits the jackpot one day.

There are many examples where that isn't sufficient. If you ask 99.9% of customers what they want next, they say smaller faster cheaper. You miss the gamechangers.

You also miss the new concepts/products where people don't conceive there is a problem.

[Zenith]

Interesting. I wonder where/how I picked up that Ely was the response rather than the cause.

The "take big risks" was the mantra of HPLabs, of course. I know of one engineer who was dinged because his work that was transferred to a division wasn't a big enough success.

There can come to be a load of politics and personalities involved in these things. The leadership can shape the organisation for better or worse. I would have thought HPLabs should have been a big R, small d lab, taking on projects which not the remit of any existing division or were too venturesome for the division which might be assumed to develop and make them.

HPLabs in Palo Alto came to have a reputation as a retirement home for divisional managers, whose performance indicated it was a time in their career when they should make an individual contribution.

[Specmaster]

The "take big risks" was the mantra of HPLabs, of course. I know of one engineer who was dinged because his work that was transferred to a division wasn't a big enough success.

"Take big risks" is a common mantra in startup culture. Along with "move fast and break things". I've always treated this mentality as naive, bordering on idiotic. It stems from the idea of creating new markets for products or services. It makes me feel dirty seeing the amount of this garbage I've had to sit through over the years by people with a track record of having one shot coincidental successes and trying to apply it to new things or existing solely to instil their apparent virtues on others.

My personal mantra in that space is "solve a need". If you aren't approaching a problem from the perspective of a well-defined need then you are taking an incalculable risk or doing something pointless and that is stupid. The art is really finding and understanding a need and that's a very much lost one when someone will throw millions at a bunch of stupid ideas and hope one of them hits the jackpot one day.

I couldn't agree more, I had my bellyful of that crap in my last job. My mantra was to try and discover what problems a potential new customer was having, and then looking to see if I could offer a service that would solve that issue for them. Then, if I was able to do that I had a repeat customer for a long time to come and by keeping close to them and looking at ways to further improve our service, I'd be one or two jumps ahead of my competitors.

[bd139]

I've been in a couple of startups. They were technically risky, but not financially nor career-wise. The path of the second has been described by Herman Hauser; it didn't surprise me.

As for "move fast and break things", that has its niche. But not, please, autonomous cars.

Startups are fine. Technically risky is not what I'm talking about really. I'm talking about commercial risks i.e. market creation, poor market appraisal and finding the market has no customers.

As for autonomous cars, that will turn into a very large rant if I get into that. I agree entirely.

There are many examples where that isn't sufficient. If you ask 99.9% of customers what they want next, they say smaller faster cheaper. You miss the gamechangers.

Actually not really. They don't say that most of the time. I've spent a long time actually talking to customers about what they want and they don't know what they want initially. You have to sit down with them and get them to show you what they are doing. Never a suggestion needs to be made as they will almost always list off the things that they need as they are working. Eventually you can turn that into a product that is based on a need and a market gap. I've done that myself a couple of times and in the third case I'm still mostly working for the company I helped get off the ground in 2007 which exists on actually listening to people and bridging a market gap.

You also miss the new concepts/products where people don't conceive there is a problem.

Those are only viable if you have the capital to waste to take a significant risk and a brand with enough of a reputation to sink a tanking product or service. On top of that the success rate is incredibly low. Most businesses trying to enter conceptually new areas fail. And if they don't they are mostly limping corpses. But we hear all the romanticised success stories rather than the reality so we assume that we can and should do that when we probably should run away.

[Zenith]

The MBAs crawled out when Bill and Dave were dead. They wanted a "50:50" company, i.e. 50k employees, $50bn. They didn't like the expense of HP Labs, thinking they could simply and more cheaply buy successful technologies. You know, things like Autonomy and WebOS.

My, didn't that work well.

(Oh, one CEO was allegedly screwing more than the company into the ground; rooting in the Aussie sense)

I thought they just wanted to get rid of him, and although that was a definite transgression, in other circumstances it would have been quietly ignored.

Autonomy was pure incompetence, a basic lack of due diligence, but on an epic scale. I recall they paid $9 billion for it. It had been previously offered to Oracle for $3.5 billion and Ellison looked at it and said it wasn't worth anything. You can argue they were defrauded but they look very much like suckers. There are a whole string of HP acquisitions which were disastrous. One that springs to mind is Bluestone bought for half a billion. Within 6 months the products had disappeared and the employees had left or were distributed through HP. I'm not familiar with WebOS. The BoD of HP was ridiculous, with the spying scandal and all sorts of silly things.

There are a few recent examples of large companies whose BoDs lost contact with reality, who their customers were and weren't, what they wanted to buy, how to present it and generally what their business was about. Victoria's Secret and Budweiser come to mind.

[tggzzz]

Interesting. I wonder where/how I picked up that Ely was the response rather than the cause.

The "take big risks" was the mantra of HPLabs, of course. I know of one engineer who was dinged because his work that was transferred to a division wasn't a big enough success.

There can come to be a load of politics and personalities involved in these things. The leadership can shape the organisation for better or worse. I would have thought HPLabs should have been a big R, small d lab, taking on projects which not the remit of any existing division or were too venturesome for the division which might be assumed to develop and make them.

A completely new product line isn't a bad achievement.

Towards the end it became ridiculous. The almost whole of HPLabs' projects were rammed together in an effort to be seen to be big game changers. I wasn't surprised when it failed.

HPLabs in Palo Alto came to have a reputation as a retirement home for divisional managers, whose performance indicated it was a time in their career when they should make an individual contribution.

That worked both ways. I knew the person who developed HP's first decent digitisers. When I told him the name of the local manufacturing division's manager, he was gobsmacked, and repeated he "never would have believed it".

[vk6zgo]

The MBAs crawled out when Bill and Dave were dead. They wanted a "50:50" company, i.e. 50k employees, $50bn. They didn't like the expense of HP Labs, thinking they could simply and more cheaply buy successful technologies. You know, things like Autonomy and WebOS.

My, didn't that work well.

(Oh, one CEO was allegedly screwing more than the company into the ground; rooting in the Aussie sense)

I thought they just wanted to get rid of him, and although that was a definite transgression, in other circumstances it would have been quietly ignored.

Autonomy was pure incompetence, a basic lack of due diligence, but on an epic scale. I recall they paid $9 billion for it. It had been previously offered to Oracle for $3.5 billion and Ellison looked at it and said it wasn't worth anything. You can argue they were defrauded but they look very much like suckers. There are a whole string of HP acquisitions which were disastrous. One that springs to mind is Bluestone bought for half a billion. Within 6 months the products had disappeared and the employees had left or were distributed through HP. I'm not familiar with WebOS. The BoD of HP was ridiculous, with the spying scandal and all sorts of silly things.

There are a few recent examples of large companies whose BoDs lost contact with reality, who their customers were and weren't, what they wanted to buy, how to present it and generally what their business was about. Victoria's Secret and Budweiser come to mind.

I always say Victoria has many "secrets", but they aren't sexy lingerie, they are mouldering corpses buried in the Dandenong Ranges, but that's off topic!

In my opinion, both HP & Tektronix started "circling the gurgler" when their sales staff would demonstrate the totally inadequate early versions of DSOs & argue with us when we pointed out their shortcomings.
They tried to argue that a horrible mess of aliasing was an equally valid way of looking at a waveform.
Amazingly, telling your customers they are stupid is not a good sales technique.

[tggzzz]

There are many examples where that isn't sufficient. If you ask 99.9% of customers what they want next, they say smaller faster cheaper. You miss the gamechangers.

Actually not really. They don't say that most of the time. I've spent a long time actually talking to customers about what they want and they don't know what they want initially. You have to sit down with them and get them to show you what they are doing. Never a suggestion needs to be made as they will almost always list off the things that they need as they are working. Eventually you can turn that into a product that is based on a need and a market gap. I've done that myself a couple of times and in the third case I'm still mostly working for the company I helped get off the ground in 2007 which exists on actually listening to people and bridging a market gap.

I'd argue that isn't "them telling you".

It is more a case of eliciting their pain points (which they may not have recognised), followed by some concept validation.

You also miss the new concepts/products where people don't conceive there is a problem.

Those are only viable if you have the capital to waste to take a significant risk and a brand with enough of a reputation to sink a tanking product or service. On top of that the success rate is incredibly low. Most businesses trying to enter conceptually new areas fail. And if they don't they are mostly limping corpses. But we hear all the romanticised success stories rather than the reality so we assume that we can and should do that when we probably should run away.

HP Labs was expected to fail most of the time. That sometimes avoided the divisions going down blind alleys, e.g. the Apple Newton in the early 90s. It also sometimes lead to new markets, e.g. PARISC machines becoming a standard for telecom SS7 servers.

[tggzzz]

The MBAs crawled out when Bill and Dave were dead. They wanted a "50:50" company, i.e. 50k employees, $50bn. They didn't like the expense of HP Labs, thinking they could simply and more cheaply buy successful technologies. You know, things like Autonomy and WebOS.

My, didn't that work well.

(Oh, one CEO was allegedly screwing more than the company into the ground; rooting in the Aussie sense)

I thought they just wanted to get rid of him, and although that was a definite transgression, in other circumstances it would have been quietly ignored.

Autonomy was pure incompetence, a basic lack of due diligence, but on an epic scale. I recall they paid $9 billion for it. It had been previously offered to Oracle for $3.5 billion and Ellison looked at it and said it wasn't worth anything. You can argue they were defrauded but they look very much like suckers. There are a whole string of HP acquisitions which were disastrous. One that springs to mind is Bluestone bought for half a billion. Within 6 months the products had disappeared and the employees had left or were distributed through HP. I'm not familiar with WebOS. The BoD of HP was ridiculous, with the spying scandal and all sorts of silly things.

Never heard of Bluestone; too far after my time.

The Autonomy purchase was as ridiculous as the purchase which destroyed GEC-Marconi.

The BoD was a bunch of stuffed shirts/skirts.

There are a few recent examples of large companies whose BoDs lost contact with reality, who their customers were and weren't, what they wanted to buy, how to present it and generally what their business was about. Victoria's Secret and Budweiser come to mind.

Boeing anybody?

[Zenith]

A completely new product line isn't a bad achievement.

Towards the end it became ridiculous. The almost whole of HPLabs' projects were rammed together in an effort to be seen to be big game changers. I wasn't surprised when it failed.

On the face of it, it appears ridiculous, but I don't know anything about it and it's probably inappropriate to discuss it in public.

Just before I left I saw strange things happen, such as people praised with great praise, for no apparent reason, as soon as they walked in the door. Enquiring as to why was "negative". Other people, sometimes who'd been well thought of for years, were persona non grata for equally obscure reasons. I came to the conclusion it proceeded on its own logic, impenetrable to me. I thought it was all going to implode and the only surprise was that it took so long.

I only saw HPLabs as an outsider, and a lot of things didn't make any sense for years.

If you like, we can talk about it next time we meet up.

[mansaxel]

HP Labs was expected to fail most of the time. That sometimes avoided the divisions going down blind alleys, e.g. the Apple Newton in the early 90s. It also sometimes lead to new markets, e.g. PARISC machines becoming a standard for telecom SS7 servers.

PA-RISC succeeded in telco partly thanks to EHPT, the Ericsson - HP partnership on billing, which was the key application. Also, telcos, being the inventor of made-up prices (like mobile roaming charges) have no idea what things cost in reality, which fit the insane pricing of nice machines with a sub-par operating system. (A SDH STM-64 line card for a Cisco router cost about 6 times the price of a 10Gig Eth line card for the same platform. The line rate is the same, the components probably are 85% same, but the SDH card was bought by telcos, the Ethernet card sold to people who had a business to run...)

I have no love for the HP UNIX(tm) business. The machines were very reliable, very well built, albeit a bit slow, but the OS was pretty uninspiring and a hassle to work with. Adding insult to injury, they had a much nicer Unix variant in Tru64, coupled with a much better hardware platform in the Alpha CPU, both which came in as Compaq had bought DEC. But that was axed in favour of H-PUKES on Intel Titanic. Yeah, that's a winning team. Not.

[tggzzz]

HP Labs was expected to fail most of the time. That sometimes avoided the divisions going down blind alleys, e.g. the Apple Newton in the early 90s. It also sometimes lead to new markets, e.g. PARISC machines becoming a standard for telecom SS7 servers.

PA-RISC succeeded in telco partly thanks to EHPT, the Ericsson - HP partnership on billing, which was the key application. Also, telcos, being the inventor of made-up prices (like mobile roaming charges) have no idea what things cost in reality, which fit the insane pricing of nice machines with a sub-par operating system. (A SDH STM-64 line card for a Cisco router cost about 6 times the price of a 10Gig Eth line card for the same platform. The line rate is the same, the components probably are 85% same, but the SDH card was bought by telcos, the Ethernet card sold to people who had a business to run...)

The whole telecom industry is incestuous, possibly necessarily.

SS7 cards always seemed to be exorbitantly priced.

I have no love for the HP UNIX(tm) business. The machines were very reliable, very well built, albeit a bit slow, but the OS was pretty uninspiring and a hassle to work with. Adding insult to injury, they had a much nicer Unix variant in Tru64, coupled with a much better hardware platform in the Alpha CPU, both which came in as Compaq had bought DEC. But that was axed in favour of H-PUKES on Intel Titanic. Yeah, that's a winning team. Not.

Alpha was an interesting exercise in minimalism: what's the least you have to specify. Unfortunately they had to backtrack with byte addresses and operations.

I have little opinion on the relative merits of operating systems; it is swings and roundabouts all the way down. (Exception: operating systems that randomly change process priorities in order to disguise deadlocks). OTOH, I have a loathing of mode-sensitive editors, e.g. those where "delete-word" will delete the preceding or following word depending on what happened sometime in the past. Yes, EDT, I'm looking at you. And vi(le), for similar reasons.

As for Itanic... I watched a talk in ~96 where the speed was obtained by hand-crafting the assembly instructions. Change the implementation slightly, and start again. I ignorantly presumed compiler technology was up to the job, and that recompilation when moving to a new microarchitecture was acceptable. I asked about context switch times with the (thousands of) hidden registers, and was suspicious of the response. Then I noticed that power consumption was becoming a key bottleneck, and that Itanic gained speed by performing many irrelevant operations.

Still, Itanic was successful in one way: it kept PA-RISC customers on-board as the end of PA-RISC evolution approached. (And Intel became tarred with the public Itanic implementation fallout)

[tggzzz]

A completely new product line isn't a bad achievement.

Towards the end it became ridiculous. The almost whole of HPLabs' projects were rammed together in an effort to be seen to be big game changers. I wasn't surprised when it failed.

On the face of it, it appears ridiculous, but I don't know anything about it and it's probably inappropriate to discuss it in public.

Just before I left I saw strange things happen, such as people praised with great praise, for no apparent reason, as soon as they walked in the door. Enquiring as to why was "negative". Other people, sometimes who'd been well thought of for years, were persona non grata for equally obscure reasons. I came to the conclusion it proceeded on its own logic, impenetrable to me. I thought it was all going to implode and the only surprise was that it took so long.

I only saw HPLabs as an outsider, and a lot of things didn't make any sense for years.

If you like, we can talk about it next time we meet up.

I missed such "personality cults", having decided to self-defenestrate after enduring a Princess Fiorina performance. Sounds like the march of time made things worse.

Sounds like there is a topic for chin-wagging!

[BU508A]

Robots are gaining ground in agriculture.

https://www.reddit.com/r/oddlysatisfyin ... _cow_barn/

[BU508A]

"The front fell off!" - new case at Boeing:

https://www.theguardian.com/business/20 ... -falls-off

[tggzzz]

"The front fell off!" - new case at Boeing:

https://www.theguardian.com/business/20 ... -falls-off

I remember when Boeing used to overengineer things...

[Zenith]

"The front fell off!" - new case at Boeing:

https://www.theguardian.com/business/20 ... -falls-off

I remember when Boeing used to overengineer things...

I recall there was a time in the late 80s when Boeing had a string of failures. They seemed to sort it out and it was nothing like the current mess.

[Cubdriver]

"The front fell off!" - new case at Boeing:

https://www.theguardian.com/business/20 ... -falls-off

This one can't really be laid at Boeing's feet seeing as the last 757 rolled off the assembly line in 2004. That said, they REALLY need to get their heads out of their asses with this 737 debacle - merging with and being taken over by McDonnell Douglas was their big mistake. The wrong culture prevailed.

-Pat

[bd139]

Crusty the Mac that wouldn't die: https://geocities.ws/rehasoft/crusty.html

[bd139]

And a hilarious video about crypto...

[BU508A]

Sometimes you can't eat as much as you want to throw up.

This morning in the Heise newsticker (German)

"Personalised surveillance instead of advertising: mobile phone data analysed and sold

The Patternz case provides new insights into the state-industrial surveillance complex. The technology can also identify a target's children and routes." (translation via DeepL)

Source:
https://www.heise.de/news/Gezielte-Werb ... 09259.html

Some links taken from the article:
https://www.iccl.ie/wp-content/uploads/ ... crisis.pdf
https://www.404media.co/inside-global-p ... me-bidding

[bd139]

Now I suppose the question is if the EU are going to regulate against it or add it to their surveillance prospectus

Really it's best to do what you do in full view of this. And things you don't want to be seen fully outside of it in side channels. And generally avoid social media entirely because that's the principal data collection engine now.

Edit: the elephant in the room here is mostly Google who's entire business is built on this and most of their revenue depends on. I would suggest if you really care about privacy, then the first step is to stop giving them money and data and that means dumping Google, Gmail, Android, Chrome etc.

[Zenith]

Don't forget, "If you've got nothing to hide, you've got nothing to fear", a very naive statement in my view.

Blair and Hague have been pushing for the sale of NHS records. I seem to recall in the 90s, when the MMR vaccine controversy was at its height, Blair absolutely refused to say whether his own young child had been given it. He said it was a private medical matter and people had no right to ask. So probably not.