* Posts by nerdbert

77 publicly visible posts • joined 12 Oct 2012

Page:

Twitter begs some staff to come back, says they were laid off accidentally

nerdbert

Re: (seeing the writing on the wall)

I've been through almost exactly what the Tweeps are going through when IBM nuked my business line here in the US, so when I say the Tweeps won't win, I mean it. Notice the timeline. WARN says 60 days notice, which is what I got. They've been given 90 days in which they are technically still employees, they're just not allowed on site or in to work (same happened to me). To see that they're still technically employees, notice that the company specifically told them that they are bound by the employee handbook. (Depending on what the handbook says, that may mean that they technically can't get employment elsewhere for 90 days if Elon is feeling vindictive, although in CA I doubt he could make it stick, nor would it be worth his time/money to deal with the lawyers.)

America's chip land has another potential shortage: Electronics engineers

nerdbert

Re: I can see the pattern

When I graduated with the doctorate, I had a choice: take it in Solid State Physics with few job prospects, or at least double the starting salary and an order of magnitude more prospects in EE.

Didn't spend a whole lot of time making that decision.

nerdbert

Re: It's pay, stability and pay

One thing I've always told new hires. "This company has a job. You have a career. If you're smart, you'll tend to your career and not the job, since jobs will always come and go."

The smart ones will leave if the company pigeonholes them into something that won't help their career. The others will be laid off at 40 with very poor job prospects. IBM was famous for that kind of crap since they had their own proprietary tools that didn't translate into industry standard design flows. And we all know how IBM views age from all the discovery details that have leaked from their recent suits.

nerdbert

Re: Means of production

It's worse than you expect.

SoCs are becoming the norm. Everyone wants the efficiency and cost savings. That means that only very large companies can afford to have the teams to supply all the various functions. (Yeah, buy the IP, sure, and then spend more time trying to integrate it into your design flow and foundry. You're usually better off doing it yourself by the time you've done all that work.)

But the chips are large, thus expensive to fab as well as design, and big companies are generally run by the spreadsheet folks and they refuse to take risks. So you get just a few companies, making very large, very "safe" designs that are just incremental improvements on what's been provided before. So this means that chip guys are just expected to churn out variations of what's been done before. Hardly an exciting prospect.

Then, look at where all the work has gone. For decades the chip companies have been trying to shift jobs to China, India, and the like. Hell, back in the late 90s to about 2010 you couldn't get a VC to even look at a proposal unless you had a plan for a significant number of folks in China or India. They've wised up a bit now, but for several decades anyone considering EE had to look at a shrinking job market with fewer opportunities in the West.

So you're going to do a job that's far tougher than software engineering, that pays less, and that is much, much harder? I'm one of those EEs, and an analog EE at that, so I've done all the Ph.D. level EE and solid state physics stuff. I've also got a degree in software (undergrad), and even I say you'd be nuts to go EE if you find even a modicum of satisfaction writing code.

But I'd also tell you that unless you choose your specialty very, very carefully you'll find a ton of age discrimination in both industries. Neither one likes gray hair unless you're in a field where if it doesn't at least function first time, every time you'll have a tough time finding a job past 40.

Bipolar transistors made from organic materials for the first time

nerdbert

Re: Totally of subject

The troll density is high today...

Graphical desktop system X Window just turned 38

nerdbert

Ah, the olden times

For the longest time I preferred X10 to X11, especially on the Suns we were running at the time. X10 had animated icons that X11 lacked, and for many years X11 was a buggy piece of cr*p that burned far too much memory in an era when memory was a very precious resource.

If you fire someone, don't let them hang around a month to finish code

nerdbert

Re: Not asking for a handover

"Unless the contractors were already in contract elsewhere that's not very professional. The professional approach is "Don't get mad, get even. But payment upfront this time.""

If you leave the job on someone else's initiative, your new full-time job is to find another job. But personally, I would realize that even if it is cash upfront, the odds of contracting with a company like this being beneficial and lucrative are low. You already know that the company, in this case, won't honor T&Cs, so engaging with them for *any* reason is dangerous for you and could lead to more legal entanglements than you want to deal with -- they're already floundering and desperate, so why put yourself in their crosshairs voluntarily? They've just given you about the most trouble-free excuse to leave from what is obviously a bad situation with less than ethical folks. Seriously, in this situation you should see all sorts of red lights and klaxons going off, so smile, demure, and run like hell for the exit.

Don't wrestle with the pig -- you'll only get dirty and the pig will enjoy it.

Sealed, confidential IBM files in age-discrimination case now public to all

nerdbert

"HR people see a 'strong CV' and hire these on the spot and don't bother to think that a person who hasn't ever been even 3 years in a company, won't stay longer than that in yours either."

I have to say that early job hopping isn't bad. You get interesting projects and more diverse experience, letting you find what you like and are good at. You don't put new folks into supporting legacy projects, and if you have no history with a company you don't have a string of past projects you're expected to support. And in good times you generally get paid pretty well. Chip guys who are job shopping these days can pretty much write their own ticket.

Japan solves 5G airliner conundrum: Keep mobe masts 200m from airport approach paths. That's it

nerdbert

One other minor issue in the US relative to the world

The US FCC has authorized those same frequencies - but has allowed its carriers in those bands (Verizon and AT&T) to use double the power of the rest of the world. Yeah, even if you guardband with the same mask the rest of the more sane world does, there's more spillover into adjacent frequencies if you double the power.

Never attribute to malice what could be attributed to rank stupidity, ESPECIALLY if the government is involved. Actually, if the government is involved, stupidity is usually tied with corruption/regulatory capture as an explanation, but that's for a longer rant. The really sorry thing is that the FCC actually looks competent when compared to the FAA.

Fired credit union employee admits: I wiped 21GB of files from company's shared drive in retaliation

nerdbert
FAIL

Re: Several levels here…

<blockquote>How long did the help desk take to restore from a shadow copy or backup? </blockquote>

I'm guessing it actually could take quite a while to restore the state. Assume nightly backups as most of the commenters here are suggesting. That means you could lose a day's work.

How many loan applications could get approved/rejected in a day?

How many accounts could be opened/closed?

How many supporting documents could be added to accounts/applications?

And how much is the financial business done in a day related to their shared drive?

A $10K bill for restoring that day's data is probably a drop in the bucket to what they lost by this deplorable lack of control and proper backups. In this sort of business I'd expect hourly incremental backups at the very least, even for relatively less valuable data like supporting data that could be re-requested.

And this is even without contemplating the issues of possible data modification before deletion (corrupting the files you would try to restore), trying to audit any possible access by this employee, etc. This is an IT clusterfsck in a business where some real money is at stake. Shame on the CU for their IT practices.

The Linux box that runs the exec carpark gate is down! A chance for PostgreSQL Man to show his quality

nerdbert

Re: At Jenny, re: config files.

Back in the days when Linux wasn't even a wet dream and when emacs meant "eight megs and constantly swapping" (i.e. when even workstations with 4 MB of RAM were considered top-of-the-line), emacs was the first program I installed on a workstation simply because it automatically made a backup of any file you edited with it. It saved my butt more times than I can count. Before that, the tape drive and OS tapes where my frequent companions when installing the latest workstation coming down the line.

Uncle Sam courting Intel, TSMC to build advanced chip fabs on home soil – report

nerdbert

There are costs, and there are related costs

People tend to think of only immediate costs. This is a mistake. For example, as Andy Grove will tell you, there is tremendous value in knowing how things are actually manufactured when you are designing things to be manufactured, even if most of the work is done overseas. Trying to figure out if the idea you have will scale into something big requires you to consider something beyond the mere invention of a thing, and to figure out how to scale that thing to production. I know in semiconductors that the US has lost a tremendous advantage in having an intimate knowledge of how things are done as most of the fabs have gone overseas or to foreign investors.

And that is without considering just what a shooting war in Taiwan would do to semiconductor supplies worldwide, and the impact it would have on the US's technology dependence. It seems to me that tariffs and other costs would be the equivalent of an insurance policy against rogue nations. After all, China not too subtly threatened the US drug supply when the US criticized the CCP's handling of the recent virus.

Call us immediately if your child uses Kali Linux, squawks West Mids Police

nerdbert
Facepalm

Re: Be a government informer! Betray your family and friends! Fabulous prizes to be won!

Actually, my son has nearly every one of those tools on his computer, I know about it, and commend him for it.

He's actually taking courses at the local community college on network security and computer administration, as well as being a devout gamer so if he didn't have those on his computer, I'd be upset.

Hate Verilog? Detest VHDL? You're not the only one. Xilinx rolls out easier-to-use free FPGA programming tools after developer outcry

nerdbert

Re: "Programming"

Verilog looks a little like C, except it's not a programming language, it's a HARDWARE description language. More obvious with VHDL.

I think that that concept is the hardest to communicate to folks coming in from a traditional software framework.

For an HDL, you get the concept that, unless you specify otherwise, everything runs in parallel, which is hard to communicate to a non-hardware person. And in VHDL in particular you can blow out the size of your design if you're not careful to specify how you create your states and wind up generating far more hardware than you expect. It was one of my bigger complaints about VHDL versus Verilog. In Verilog you are much closer to the hardware and it tends to be a bit harder to be stupid about what you're doing in terms of hardware utilization. At the system level Verilog is harder to deal with, and VDHL is a better choice, but neither of them fit the traditional programming paradigm at all.

nerdbert

Where life is on the line - the medical biz - you'd hope that somewhat more care was taken than for some social media "killer app" (could be a vain hope of course).

Not so vain. I've got friends in the medical device biz, and coworkers who used to work at the biggest names in the biz, and I can say that the amount of care they take is both amazing and frustrating. From architectural chip sims where even changing the slightest thing requires approval from 3 levels of management, to simulations where even 9-sigma design margins have to pass, to years and years of measuring lead wire designs in various configurations, aging conditions, etc are required.

We'll turn a new chip in 9 months in a new technology and get something to work and release it a couple of months after that.

It'll take 7+ years to make the smallest change to a pacemaker, much longer if there's any change to something like the silicon inside it. And you know, I think I like it that way.

But there are times that it's frustrating to see how conservative the FDA is. There are biohackers who are taking old Medtronic insulin pumps, rooting them, then connecting them to blood sugar monitor watches and making an "artificial pancreas" where the monitoring a correction of blood sugar is automatic and far superior to the way things are done now (not controlling blood sugar well leads to damage that accumulates over time, and the closed loop feedback minimizes time that blood sugar gets too high, as well as eliminates the risk of it going too low and the person going into shock or a coma, which is even more dangerous) . In fact, one of the board members of the JDRF, the main Type I diabetes research foundation, does just that and is extremely pleased with the results. But Medtronic and the other manufacturers are moving very, very slowly on that path because of the FDA and concerns that they would be "practicing medicine" if they do more than provide information and be far more liable if something went wrong. The engineers would like to get a prototype system out there for those willing to take a chance, but the regulatory system and management are pretty firmly against it. Management is mostly doctors, and the regulators want well structured test groups, double blind studies, constant monitoring, human subjects regulations followed, etc. In other words, it'll be a long, long time before the public gets something through them, rather than engaging in the hacking subculture now going on in the diabetes community.

Four more years! Four more years! Svelte Linux desktop Xfce gets first big update since 2015

nerdbert

Not one to follow fashion

Why would you use Xfce? The main reason would be either because it runs better on old or low-end hardware than systems like GNOME Desktop (used by the main Ubuntu distribution) and KDE, or because you prefer its performance and minimalist approach.

I run xfce on the highest end of high end of Xeon processors, on machines not equipped with less than 1T of RAM, and it's not for any of those reasons.

I use a chip design program that's regularly updated to whatever is a reasonably recent version of CentOS. That means I get updates at least annually, and have been for the last 20 years. I've seen window managers come and go, but it's been more painful to see them evolve. That's one of the reasons I like xfce. While I can't be sure how the latest version of KDE or GNOME will affect me, xfce doesn't change. So all I'm left is figuring out how the blasted CAD tool has screwed things up rather than trying to figure out if it's some new interaction between the desktop and the program.

So in summary, I'm an engineer and I use my machines to get work done, not for eye candy. And if a desktop makes it so I can get my job done easier, I'll use it.

That's a sticky Siemens situation: Former coder blows his logic bomb guilty plea deal in court

nerdbert

[...]evidence that Tinley added code to the complex spreadsheets that "had no functional value, other than to randomly crash the program,"

Sounds like 99.99% of all code submitted by first time contributors to a FOSS project.

Only somewhat more seriously, if this is a felony then nearly any "legacy" software system could populate a prison.

Black Hat USA axes anti-abortion congressman as keynote speaker after outcry – and more news from infosec land

nerdbert

Re: There are now answers to this that will make everyone happy.

Opposing planned parenthood isn't a 'flaw' though. It's an outright attack on 50% of the population.

Why are we even arguing this? It's totally tangential to the issue at hand, and if you're willing to drive away an ally on this issue because of his stance on another completely unrelated issue, you're risking your ability to make progress on this one, and to stand any chance of persuading him on the other. You risk becoming the very pure, but exceptionally powerless, Libertarian Party by narrowing your focus to only those who share your beliefs on all issues.

And yes, this is how you get demagogues like Trump, who are completely unafraid and unashamed at upsetting convention and consensus. When you, yourselves, become demagogues you give rise to such behavior in your opponents.

Buy, buy this American PCIe, drove my PC on the Wi-Fi so the Wi-Fi would fly

nerdbert

How to leak information...unintentionally

No idea if the 10 bit word is still true or not

That was dropped in PCIe v3. Neither v4 nor v5 have that 10 bit run length.

'Software delivered to Boeing' now blamed for 737 Max warning fiasco

nerdbert

Re: And now we get to the ...

I don't know about your company, but for my products I'm much happier when senior management is not involved in the detailed engineering decisions, since, you know, I'd like the product to actually work. Andy Grove was a great engineer in his time. but by the time he was a C-level creature even he knew that he had to stay away from a lot of the actual design.

Middle management can be a crapshoot. Sometimes you get that rare individual who can find his arse with both hands, but often you can't. It's director level and above that forgets everything they've ever learned about engineering since they haven't been allowed to touch the tools for so long that they're only allowed to play company politics and sell their wares anymore. It's the middle management that's trying to transition to Director and above who's the most dangerous link in the chain.

nerdbert

Re: 'Software delivered to Boeing'

You're right: Boeing is definitely not off the hook. In fact, from what little I know I'd suspect that they're the most exposed, although they will do their best to try and blame their subcontractor (who probably will fight back pretty hard from what little I've heard).

MCAS was delivered by a supplier, yes. From the western hemisphere, though. DAL B, at best. I've heard rumblings that even the subcontractor wasn't pleased about the system they were designing and had internal battles about what they were expected to deliver and at the quality level. Apparently the coders were thinking DAL A was more appropriate and that the system they were told to deliver wasn't what they wanted to deliver, but that's not how Boeing wrote the requirements.

Boeing is the systems integrator. They define how the systems behave, not the subcontractors. The contractors who write the software have specs from the system integrators that they follow, and they write their software as they're told. Boeing in this case has two AoA sensors, one per side. In the NG planes they were independent, one for the pilot and one for the co-pilot, and it was pretty obvious if one disagreed with the other since the pilots could compare results and ignore the one that disagreed with their eyes. But in the MAX the MCAS is tied to one side only, and Boeing seems to have hidden the AoA and override functions. The subcontractor probably suspected that this was a shortcut and suboptimal, but even if they did, they don't know how the other systems that Boeing was installing would interact with MCAS, how the override would function, etc. Unless it was clearly written in the contract that MCAS needed to monitor both sides I doubt that the subcontractor will get a whole lot of the blame since they really couldn't see the overall system.

(A good systems engineer is a rare thing. It's a lot more fun to work on the building blocks. I did system engineering for a decade, got great praise and reward for the stuff I did, and I got an excellent feeling why in NASA it's hard to fly anything new, even on test satellites. Then I quit and went back to designing the subsystems. It was a lot more relaxing and fun working on the subsystems where I could really figure out what was going on and optimize what was delivered without wondering if the fact I changed some minor thing was going to bring the whole system crashing down.)

International Bullying Machine? Big Blue seeks exposure of corporate canary

nerdbert

Re: IBM

If the lawyers are so good, how come they let Big Blue get into this mess in the first place?

I've dealt with their IT lawyers quite a bit. They were very good every time I dealt with them. You'd expect that, given that their IT lawyers generate $1.5B+ every year. They're passed around from division to division depending on who is management's golden haired boy.

But the IT lawyers form their own entity inside IBM. HR is its own realm and is a cost center rather than a revenue generator. They're tasked with carrying out management's orders and trying to pass off their actions as legal. Sometimes it's not all that easy to do, as in this case.

Dead LAN's hand: IT staff 'locked out' of data center's core switch after the only bloke who could log into it dies

nerdbert
Pint

Re: BOB

Exactly this.

Manager: So we need to document this flow. And we to fix these things ASAP.

Me: You can have one or the other.

Manager: Why? Why can't we have both?

Me: Because you insisted on a three hour meeting, face to face in a conference room to go over this. Now I have time to do one or the other.

Manager: Hmmmmm... Ok, just fix the things. Then we'll have to have another meeting to go over the plan to document everything.

Me: *sigh* Sure thing, boss. Just put the meeting notice on my calendar. (Screaming inwardly to myself, "And you've still learned NOTHING!")

Just a lightly edited extract of a real situation. Much of the problem is that for managers their work product is meetings, so they think things are getting done if you have a meeting about a subject rather than actually working on the problem.

Pay row latest: We aren't biased against Big Tech, says Uncle Sam as it rolls eyes at Oracle

nerdbert

New Oracle HR activity: generating an infinite number of new job titles

"Such concerns about the DoL's requests for data are said to have been raised at the meeting, as well as the department's reliance on job titles and roles to assess pay disparities."

I can foresee a new activity for HR drones: generating new titles for every employee. I can foresee things like "Reversion Testing Level 2 Manager", "Individual Contributor for Scripting Checkers of Product XYZ Data Entry." Things that are totally meaningless for making comparisons.

(We had similar silly things in IBM when the DoJ tried their antitrust suit. We had to keep every slip of paper for anything we did in product development and submit a copy of it to the lawyers, who promptly buried the DoJ in an avalanche of paper that was completely meaningless for them, but had to be provided given their broadly framed requests.)

Three no Trump: China says it's 'open' to giving Qualcomm-NXP wedding its blessing

nerdbert
Big Brother

Re: Qualcomm to acquire Dutch semiconductor maker NXP

Qualcomm is a special case for US regulators. They do quite a bit of *cough* "special" *cough* work for various three-letter-agencies of the US government, so those regulators are "guided" in their decisions by forces other than consumer interest. That much a fairly open secret in the industry.

As for the NXP deal, that's a different question. The US probably likes that as a way of getting more fabs that can handle the various special projects that Qualcomm does for the US government. There aren't many of those fabs these days, especially with IBM paying Global Foundries to take its fabs (note all the special qualifications on GF keeping some of the fabs that do DoD work that were embedded in that deal). If Qualcomm could make a go of older fabs for some of their work (and older fabs are often better for things like RF) and use them for various special projects, then we should have no doubt that the US regulators would be told by various agencies that the deal needs to go through no matter what.

What a meth: Woman held for 3 months after cops mistake candy floss for hard drugs

nerdbert

Re: The best and the brightest fighting the war on drugs again!

The cops can pull you over for any reason in practice, but the key for them is to get permission to keep you. If you say no to the request for a search (and you always should to prevent cases like this), and ask if you are free to go there's a clock that starts ticking. If the police decide to delay you beyond the reason for the initial stop, and you keep asking if you're free to go, then anything they might find becomes inadmissible.

In essence, by asking if you're being detained you force the officer to come up with and articular probable cause or to release you. The SCOTUS has made quite clear that detaining you after the purpose of the stop is impermissible without probable cause.

The fact that you have to be knowledgeable, quite clear, emphatic in enforcing your rights is absurd, but that's the way lawyers like it.

Supreme Court raises eyebrows at Google's cozy $8.5m legal deal

nerdbert
Mushroom

Re: Original judge

We can fix this situation very quickly. Make the lawyers the *last* to be paid, not the first, when damages are calculated. If the damages were only $0.04/person, then lawyers should make no more than what each damaged person did. Oh, if we feel particularly generous, let's let the lawyers count a reasonable number of support staff in these suits, too. The firm in this case might make $1.20 for bringing this!

IBM talks 'emerging, high value segments' – so you know the Q3 numbers aren't great

nerdbert
FAIL

How to leak information...unintentionally

IBM has SG&A expenses that are almost 3x R&D expenses.

Intel has SG&A expenses that are less than half their R&D expenses.

And you wonder why IBM is so screwed up?

What's Big and Blue – and makes its veteran staff sue? Yep, it's IBM

nerdbert
FAIL

And more recently, over the last five years IBM shares lost 21% of their value (even before inflation), whereas the NASDAQ gained 109%.

May I point out that over those same 5 years IBM bought back almost 20% of their shares, too? Without their stock buyback plan, the stock would have performed even more abysmally.

I know that stock buyback plans are popular with investors, but they're also an admission by management that they really don't know how to use the money they have on hand to drum up new business. In IBM's case, that little bit of financial engineering has become their bread and butter since the corporate finance guys booted the tech guys out of the company leadership in the early 80s. Is it any wonder the company has gone downhill since then?

Sysadmin trained his offshore replacements, sat back, watched ex-employer's world burn

nerdbert
Mushroom

Re: This smells like BS.

Also, how the fuck can you "shut down every data centre in every location in the world"

I work for a large chip company. We use a certain company's software to manage versioning in our chip databases. My company is also notoriously cheap. A decade back or so the team managing the software was reduced to one person (call him 'Chip'). One of the chip designers (call him 'Dave') was a very senior designer and a big proponent of that crappy piece of software given the role of helping out when Chip went on holiday.

Now Chip was a nice enough guy and a wizard at the versioning software, but he wasn't too knowledgeable on remote protocols, and we were using the software on multiple sites under multiple OSes (HPUX, Solaris, Linux, etc) even though it wasn't designed to be networked (think something like rcs vs. git). I'm a chip designer, but I've also had IT administrator experience, so I had written a whole bunch of scripts that Chip used to mirror commands across the company, but having an intense dislike for our crappy commercial versioning software I never learned most of the more dangerous administrative commands and had no checks in the scripts. I had merely handled the networking aspects of the software and told Chip that he was responsible for any checking other than to see that the remote command worked since I had no idea what he needed to do, and he'd never done that since he was too busy doing the work that had been done by six people to implement checks.

Of course, Chip went off on holiday and Dave took over. Dave innocently went to completely reset a private library in the versioning software (delete the old private library and all past versions and recreate a blank one). Only what he thought was his own private library was not only wildcarded, he was doing it with the networked command environment, and included all the production libraries for that silicon generation. And of course he did this at midnight before going to bed since even for his smaller private library the procedure would take some time (I told you it was crappy software).

By early morning, about 800 chip designers all around the globe were howling about being unable to work because the all the production databases had disappeared or were in the process of disappearing. Of course, the Unix backups were only performed at one site because of the sheer volume of data (remember that I said the company was cheap?). And of course, being cheap, backups had not been tested during all the downsizing and they weren't working. All told, the cutting had gone beyond the meat and well into the bone, and now the cost was a company's worth of highly paid designers being idle for two weeks as the databases were recovered. In my case I wasn't affected too much since I wasn't following official procedure and had been working in unmanaged private libraries anyway (remember I said I hated that piece of software?).

I rather suspect that that incident was why Chip was given two guys to help him, and Dave was removed from the administer's group on the software.

So yes, I've seen entire companies' sites all go down at once because someone did something with tools he didn't understand completely.

Intel confirms it’ll release GPUs in 2020

nerdbert

Re: Always good to have competition to rein in that nVidia/AMD duopoly

Google isn't doing graphics. Look at the papers on the Tensor chips they've been doing and you can see that while the architectures are similar (SIMD machines with massive high bandwidth memory access), there are distinct differences between a Tensor machine and a graphics card. But from just the papers Google has published you can get an estimate of what their Tensor chips run, and a reasonable estimate is that those chips alone, not counting the HBM, assembly, and all else, run much more than a maxed out 1080 Ti card. Google may be large as companies go, but they're still not large enough to get the massive discounts you get from volume Si production.

Internet engineers tear into United Nations' plan to move us all to IPv6

nerdbert

Re: And... this is where it gets political

Whatever their internal disagreements, the IP community have a Unix-like disdain of outsiders, so I wouldn't assume automatically that the ITU's ideas are necessarily any more fanciful than have at times emerged from those of the true faith.

Obviously you've never had to deal with the ITU at the spec level if you don't understand the IP community's dislike of dealing with the ITU. Not that it's pleasant dealing with spec making at most times, but dealing with the ITU in particular makes no anesthesia dental work look like a holiday in comparison. Calling the ITU more political than technical is the mildest of the complaints that can be made.

US Congress finally emits all 3,000 Russian 'troll' Facebook ads. Let's take a look at some

nerdbert

Re: The poor English reminds me of the 419 scam.

Don't fix the candidates. Fix the system that filters out better ones.Don't fix the candidates. Fix the system that filters out better ones.

On the Democrat side, the system was fixed. Just ask Bernie about how the DNC behaved and how the super-delegates system works. Yet still I can't believe the system nominated someone who, had she been anyone else, would have been behind bars in any non-politicized justice system. I know I'd have been in Leavenworth if I'd mishandled classified materials like that.

Strangely enough, the Republicans actually had the more democratic (little d) nominating system. For the GOP the problem were there were so many similar "mainstream candidates" splitting the votes in a winner-takes-all system that the outlier was the one who survived to win. Trump may well have won because the highly partisan media gave him far more exposure than he deserved, and there's a fair bit of this country that dislikes the media. There's evidence Clinton and the DNC conspired to promote his candidacy, and in the most perverse sense they may well have nominated the only candidate who touched issues that allowed him to peel off states like Ohio and Wisconsin that were more sensitive to trade and immigration issues. Absent that manipulation, Clinton should have beaten Cruz handily, for example.

I'm not sure that making the GOP system less democratic and more like the Democrat's system is the best idea as the GOP establishment is at least as corrupt as the Democrat's. But I would like it if the media were more trusted on political matters and wasn't so partisan that it is dismissed by half the country. That would be the biggest and best improvement you could make in the system at present, but the odds of that happening are minuscule.

Now Europe is getting jitters over Broadcom's Qualcomm takeover bid

nerdbert

Broadcom's nationality?

I thought that Broadcom was moving its headquarters to become an American based company: https://www.usatoday.com/story/tech/news/2017/11/02/trump-announces-broadcom-moving-legal-hq-u-s/825600001/

That said, I'm massively confused by the statement "We are concerned about the possibility of a European company handling sensitive data of EU citizens falling in the hands of a company that is based in Singapore, where data protection standards are lower than in the EU." What PII will Qualcomm have on EU citizens other than their own employees?

And the speculation of possible data exfiltration in their chips seems highly unlikely. While the company may be legally domiciled in Singapore, the vast bulk of Broadcom's design efforts and corporate decision making centers are US based so it would be more likely that any exfiltration implementation would come at the behest of the US rather than Singapore. And in purely practical terms, getting exfiltration implemented in a quiet manner in an SoC seems to me like it would be massively difficult given how freely the folks in Silicon Valley talk and the huge number of folks involved in the design, implementation, and testing of an SoC.

Honestly folks, if you're talking about the possibility of exfiltration you really should be talking about software rather than hardware. It's far easier to hide those nefarious programs than it is to design them into the hardware.

But given that Hock Tan's modus operandii is to buy a company and spit out the parts he doesn't want makes Niels Annen's concerns pretty valid about the fate of the Hamburg plant. I'll certainly give him that.

Desktop PC shipments dip below 100m/year

nerdbert

Re: Prebuilt-vs-components

I haven't met anyone who's bought a prebuilt desktop in years. If you care about performance and durability you either build or get a custom built system.

Moore's Law and Dennard scaling (base semiconductor scaling) are dead. We've hit the practical Amdahl limit, too. So if you look at the practical speed of a general purpose CPU for the last 5 years you'll see it's almost flat. Why upgrade your system if your CPU and RAM are basically unchanged in speed?

The only thing that's still scaling up at a respectable pace are GPUs. I've upgraded my GPU regularly, but I haven't seen enough of an improvement to even consider a CPU update, meaning that my desktop system with high quality components has served me well for years, and probably will for more years to come.

H-1B visa hopefuls, green card holders are feeling the wrath of 'America first' Trump

nerdbert

"It has the most Nobel prizes precisely because most of the people who have won them have immigrated from other countries to be there."

Somehow I don't think Einstein would have had any trouble getting through a merit-based immigration system, nor would he be disqualified under the "moral turpitude" clause that's hitting that Polish-born doctor. He's being reviewed for deportation (and it's not a certainty) because of convictions for receiving stolen goods years ago. But as the article noted, had he gone and applied for citizenship as many green card holders have done, this wouldn't be an issue now for him.

To me, the more telling comment is the lawyer who said that H1-B applications would have more chance of being approved if they didn't pay the absolute minimum wage required. Isn't that exactly how the program was sold? That H1-B was for those cases where jobs couldn't be filled by Americans?

Black & Blue: IBM hires Bain to cut costs, up productivity

nerdbert

Re: Why bother?

Already done. Last I heard the patent portfolio brought in north of $1.5B/year. They pass it around to the latest fair-haired division to make that VP look good. Terribly political.

BTW, in IBM-ese RA is not redundancy action, it's "Resource Action", although the effect is still the same.

1 in 5 STEM bros whinge they can't catch a break in tech world they run

nerdbert
Flame

Re: Isn't it a small minority

I also used to work at NASA. That is until the HR drone and my manager gave a presentation that flatly stated that no white male would get a promotion until there was "equality" in our department. There were 20 engineers in our department, all "white males" as far as HR was concerned (i.e. we were all males of Caucasian or Asian decent). We did the math and very soon the department was down to 5 engineers and still shrinking last I heard. Why stick around in those situations?

Intel, Microsoft confess: Meltdown, Spectre may slow your servers

nerdbert

Re: "8th Generation Core platforms..."

There are a fair number of ways that Intel can fix the Meltdown issues cleanly since AMD already does that. (Yes, let's acknowledge that Intel chose the more risky architecture for speed reasons.) TLB isolation or mirroring, change the order of the execution, etc.

Spectre will be a touch harder to fix. Right now it's almost secure on AMD, while it's a gaping hole on Intel's processors. Again, there are fixes, but what ones will impact performance the least? That's probably a big unknown, even inside Intel.

As a practical estimate, look what it takes Intel to design a new processor. Their ping-pong strategy using 2 design groups should tell you that it takes probably 18 months to make each fairly large change in their processor, and this is likely to be a fairly large change in an area that's notoriously finicky (branch prediction is an art). As a rough estimate, I'd say that pushing either the Spectre or Meltdown fixes through the process is probably at least 6 man months worth of effort (new architecture required with performance optimizations, RTL implementation and checking, new P&R, lab validation, etc).

The timing couldn't be worse for Intel. They typically announce desktop processors in the fall. That means that they're probably in the testing and finalization stages of validating next fall's announcement now. Any attempt to put a fix in there will likely hit fall's announcement hard. You're talking designing a fix, implementing it, 2 months to turn around the design in the fab, and then testing the fix. Maybe it's doable, but there are going to be a lot of sleepless Intel engineers if they hope to keep the schedule. My personal bet is that they'll have to slip the schedule AND rob the engineers blind on overtime.

Memo man Damore is back – with lawyers: Now Google sued for 'punishing' white men

nerdbert
Alien

How to leak information...unintentionally

I'm certainly no fan of politics in the workplace. That said, this suit was filed in California, and as with many things, California is an oddball. And remember, this suit is filed in California, in California courts, with California law controlling.

Google has a policy of actively attacking those who don't hew to one particular viewpoint and firing them as Damore found out. Under California employment law, that's illegal. Remember that Damore didn't discriminate against anybody according to the reason the CEO gave for canning him, it was because he promoted "badthink". In most right to work states that's actually legal, but not in California where there are protections against firing for political reasons. Given the storm of accusations and leaked documents showing managers who would blackball employees based on their political beliefs, Google's really behind the 8-ball in this suit.

The suit would have had to be structured differently if it were filed in Federal courts, since firing for political reasons is allowed under Federal law. The discrimination and hostile workplace claims could still be made, but they're harder to prove. Frankly, the ability to add the political angle probably made filing in the California state courts preferable given the statements that Google made when they fired Damore. It will take some pretty fancy gymnastics on the part of Google's lawyers to make this go away.

Qualcomm sues Apple for allegedly blabbing smartphone chip secrets in emails CC'd to Intel

nerdbert
Thumb Up

Re: re: "We were told to ignore intellectual property rights when designing the modem."

Exactly right. When you are doing development you are nearly always told that you should *NOT* do searches for IP related subjects. Every company I've ever worked for will tell its engineers the same thing: NO IP SEARCHES! Willful violations of patents are *extremely* expensive (see CMU vs Marvell for an example), while inadvertent or unknowing ones are far less expensive.

Besides, once the product is out there and making money any IP violation becomes a patent war, and very, very few companies want to go up against some of the titans of patenting. Although in this case, Qualcomm vs. Intel would be an epic battle. Both have some very fundamental patents in some key but very different niches. The negotiations would be epic.

And that points out a fundamental difference between Qualcomm vs. Apple and Qualcomm vs. Intel. Apple doesn't have anywhere near the depth of IP related to chips that Intel does, so Qualcomm has a heck of a better chance of strong arming Apple than it does Intel. It's not like Apple's patents have much overlap with Qualcomm's business, but since Apple is a consumer of Qualcomm and competitor's products, they're more vulnerable to Qualcomm's IP threats.

IBM offloads Notes and Domino to India's HCL Technologies

nerdbert
Stop

Re: Jesus, is Notes still going?

Oh IBM used Notes, and even had it ported to their POWER machines in the 90s.

I have to say, Notes got me my first laptop. The POWER workstation port was so bloated it made the PC version look svelte and quick in comparison (and it was slow on the PC, but that's another story). Even though we were R&D we were expected to use it, but it was so big and slow that I loaded it up twice a day: once when I first got in and once before I left. I the usual course of events, my 3rd line needed something right away and asked me for it, but I didn't see it until well after he wanted it. When asked why, I explained how I couldn't do both my work and run Blotus Notes at the same time. It wasn't a week later when my manager dropped by with a laptop so that I wouldn't miss another management meltdown. And this was a time when laptops were nearly unheard of outside of sales and management.

Hash of the Titan: How Google bakes security all the way into silicon

nerdbert
Thumb Down

How to leak information...unintentionally

You don't do chips, do you?

TSMC has several modes of operation. One where you pass off RTL to them and they do the synthesis, P&R, etc. This is the "handholding for newbie startups" mode. TSMC could, if it desired, change your logic and hide it from you since they also design the test patterns.

The other mode is where they take GDS2 (geometric trapezoids) and hand you back silicon. This is the one that serious companies use. In this case TSMC is practically locked out since they'd have to decompile the GDS, make changes, and hope like h*ll that they didn't change the test patterns you've already generated. The odds of this are infinitesimal on any practically sized SoC.

I expect Google is a serious company, with serious money to spend given that they are going to this extent for security so TSMC isn't a practical attack vector. Your better bet would be to corrupt one of Google's IP suppliers and try to inject a vulnerability there. I seriously doubt Google is designing the microcontroller, for example, so that's where I'd start if I wanted to corrupt this sequence, although you could do it on any of several IP blocks they use.

Surfacegate: Microsoft execs 'misled Nadella', claims report

nerdbert
FAIL

Re: I would guess MS is to blame

May I point out that MS had the same inability to read a thermal spec issue with the Xbox 360 and the RRoD? Which is why they had to go to IBM's more expensive SOI process to move the processor temperature down. (Not that it was totally MS's fault, but they should have done a better transition to lead-less solder like most other companies did.)

I would suggest there may be a lack of attention to thermal engineering in the MS hardware department and too-slavish deference to design engineering. A failure in one flagship product is understandable, but not learning from it speaks volumes about the culture in MS's hardware division.

Indian Business Machines? One-third of Big Blue staff based there and Bangladesh

nerdbert
FAIL

How to leak information...unintentionally

It's not like anybody didn't suspect the numbers, it's just interesting how those numbers got out.

IBM is moving their key work to low wage locations.

IBM sales haven't grown in 20 quarters.

Not that management will see any connection. So, so, so glad I left long ago when they went from techie management to cookie management. Any company that's been pumping its stock price with buybacks for way more than a decade has been telling stockholders that it has no idea how to be a tech business because they can't find anything useful to do with the money they're making, and IBM shows what that sort of clueless management leads to.

Pence v Clinton: Both used private email for work, one hacked, one accused of hypocrisy

nerdbert

Re: Apples and Oranges

You also need to consider that it was *illegal* for Pence to use his state government email for personal email. That there was some spillover between political and state emails is almost guaranteed in such situations. Even if Pence obeyed all the rules, there's the case where a donor might email the governor's personal email for some assistance with bureaucratic red tape down the road.

And to compare that to Clinton's situation is laughable. We already know that Clinton was required by law to turn over *all* work related emails, yet she culled numerous emails that were obviously work related from the cache she turned over and that were only recovered by the FBI getting them from other sources not inside the government. She was actively trying to make sure that none of her political emails were subject to FOIA by running her own filter, and she violated the law by not turning over all the email that she was legally obliged to do. Pence has no such similar intentional violation of Indiana laws.

Why software engineers should ditch Silicon Valley for Austin

nerdbert
Happy

Re: Ummm ... have you ever spent a summer in Austin, TX?

Austin is nothing compared to Houston. And as for raw heat, you might as well live in an oven as in Phoenix. But I'd take either of those places over SV.

They've tried to get me out in SV many times, but I always laugh at the recruiter and ask them if they can match my quality of life. Sure, there's a shot at making millions (I've had friends who've done that in SV), but the odds of that are low enough that it's not worth the pain of living in SV. Unless you're at a good startup you're not going to hit the lottery. Working for Google or Facebook in SV is a losing proposition unless you just want them on your resume, and for that you just want to get in and get out while you're young because those are no places to make a career.

IT team sent dirt file to Police as they all bailed from abusive workplace

nerdbert
Unhappy

Re: Uhhhmmmmm

Why accept the abuse for years on end before collectively resigning? Why not start by collectively speaking up?

Because it rarely works?

My version goes like this. We had a group in a big, TLA company back in its gold plated era. A productive group, we never missed a tapeout by anything other than the expected amount (i.e. for this company we never made the fab wait more than what they were running behind). It got our manager a promotion by 2 levels, and we got a new, freshly minted manager from a different area that was chronically late and buggy.

Step 1: 3 months later 25 people scheduled a group meeting with the former manager's and his subordinate who was our manager's manager, presented how badly he was treating us and how badly he was managing the schedule. "All proper concern" was expressed to our complaint, and "corrective action would be taken."

Step 2: after 6 months of nothing being done, two individuals were chosen to begin feeling out opportunities with external companies for the entire group. We had 3 bids and ...

Step 3: after 3 months of searching, 24 people put 24 15-minute appointments on our former manager's calendar to hand in our resignations. The new company was certainly better than the TLA with the infamous bureaucracy and paid better, but the important factor that going to work was fun again, and we actually got to do work rather than spend all our time avoiding management abuse. The TLA got out of that business line soon after that, as they never were able to put out another competitive chip in that market segment.

Tech Trump: Silicon Valley steps into the valley of unhappiness

nerdbert
Pint

Re: Bad for investors maybe. But is it even worse than now?

Using every resource to crush someone who has insulted you is SOP right now, just ask Lois Lerner.

One good thing about Trump being President is that all of a sudden there are a lot of people who will worry about too much government power and be a lot more interested in exposing government abuses.

Four reasons Pixel turns flagship Android mobe makers into roadkill

nerdbert
FAIL

The Pixel is dim

Pixel is too much money for too little return in my book. I've been a Nexus user from the start, and when I look at my 5x and compare the Pixel I'm massively underwhelmed. $400+ for that?! What I still want is a replaceable battery and a uSD slot and possibly water resistance (although that's not as big a deal), none of which come in the Pixel.

If Google's going to make an iPhone clone, it's got to do better than an iPhone because Google already comes to table with a big minus in that it lives by stripping away my privacy in more ways than Apple does. At this point, I'd do the iPhone before I'd do a Pixel if I had to choose. And I dislike my daughter's iPhone and I absolutely despise iTunes with white-hot hate.

I keep meaning to try Cyanogen and the Pixel might be what drives me to it. Now where'd I leave my wife's old LG G3?

Page: