Posts by Stephen Channell 328 publicly visible posts • joined 28 Sep 2007
The first recorded email-virus was a Christmas Tree REXX script that rendered a twinkling tree on 3270 terminals, with the addition that it sent itself to everybody on you Profs contact list.. seized-up IBM for weeks and only purged by an emergency fix that that filtered-out Christmas.
On MVS JES enabled you to run a job on another mainframe (e.g. cataloguing a tape for another job to use), and running a job that spawned another was as simple as writing to a DD mapped to the intrdr.
Who needed a USB drive when you could use ftp & tcp/ip to move files around and optionally submit jobs through ftp with a “quote address intrdr”.. and yes those TN3270 telnet sessions had clear-text passwords.. and there was always RJE for those without TCP/IP FEP.
No, the mainframe was no more secure, but it was : secure-by-obscurity; what security there was, was on by default; everything was audited; the price for hacking was terminal.
You mean the processor architecture that Intel has been trying to kill for the last thirty five years!
Long before the Itanic adventure, and even before the IBM PC, Intel’s cunning plan was a new ADA machine iAPX 432. If Intel thought there was a future in x86 chip series they never would have licensed it to AMD and others.
What dammed fool allowed a BETA of Skype out into the wild; with its integration into the People-tab, phone & messaging.. you’re not Apple (where people will blame themselves); you’re not Google (where people will assume it was the battery); you got to be as solid as a dropped iPhone (but without the cracked screen).
Having all my {phone, fleecebook, linkedin, twitter, skppe} contacts in the people tile is only a feature if it’s solid, otherwise you’re better off with half-a-dozen apps. Making Skype calls from the phone app is neat for long-distance but has to be rock solid ‘cus the phone app is safety critical
It’s a documented fact that in the 1991 USSR coup all data-lines (mainly Telex) were cut, but modems still worked because the KGB thought the bit-stream was faults. With no secure Telex lines, the British Ambassador drove to Berlin to brief Whitehall (which was on alert for possible war). Meanwhile print journalists were filling ‘live’ updates from Moscow using Tandy kit & acoustic couplers.
At the time I thought it was quaint that our primitive government didn’t have the kind of kit I’d played with in the local Tandy store… but now understand why. “Why” because millions had been spent developing a home-grown version after evaluating and rejecting the Tandy with its Microsoft software.
Forgive me for defending Microsoft (I don’t work for them), but yes in this particular case not buying Microsoft very nearly got us all killed!
The Tandy TRS 80 model 100 that the Civil Servants “evaluated” revolutionised journalism. The “foreign correspondent” did not need to dictate a story over the phone to typist, but could type it up before hand and transmit at a whopping 30 baud with an acoustics coupler.
The journalist use-case is the same as the Civil Servant.
The article mentions the NEC & Olivetti versions of the Tandy that was manufactured by Kyocera, but neglects to mention why different firms where able to make essentially the same kit.. because Kyocera did not make the software.. a little Seattle outfit did that bit.
The little Seattle outfit would have chewed Terry’s arm of for the customisation contract because it hadn’t made it big by then, and the programmer still goes on about “information at your fingertips” today.
Sure, the Liberator looks neat; but if they’d just bought and customised a few thousand off-the-shelf standard boxes, they’d have saved a stack of money & the Moscow embassy wouldn’t have been cut-off when Gorbachev was deposed (telex lines were cut in the USSR coup).
For the historically curious, the programmer was Bill Gates and the software was customisable with Microsoft BASIC
time was when financial accounts were considered too confidential for computer systems, then a decade later it was considered too confidential for shared (switched) cable, then a decade later too confidential for outsourced data-centres and contract DBA.. today we know better.. We know how to match confidentiality with controls and Chinese walls.
I've loaded my companies source code into the tfspreview.com, and have less worries than I had with my SVN hosted with Webfusion. MS pays all the cost of Chinese walls to stop me suing them for stealing my code.
I'm impressed with TFS and like the combination of all continuous integration components together.. And may even use the continuous deployment with VM too test-case management
Clearly somebody didn't read the "I put in on the touch-screen computer " before commenting "you're not really the target audience".. because.. err.. actually yes, my touch-screen IS the target audience!
Sure I could do the flick from the right of the screen but my PC's got a border round the touch screen so I have to jam my finger into the border before flicking, except that don't work unless you've got no nails (1cm long & it don't work) and the patience of Sinclair ZX81 owner.. so utter, utter shite.
her's a thing, my PC screen is twenty-five times the size of my (great) WinPhone screen, but the phone's the one that makes space for a home icon on the screen. not since the paper-clip has Microsoft done such a stupid thing with the PC UI.. utter utter shite
I have a Microsoft Partner licence that included Windows-8, so I put in on the touch-screen computer in the kitchen. My kids moved the wireless keyboard because “it was in the way”.. but that’s fine ‘cus Windows-8 is touch oriented, and Windows-7 worked really quiet well without mouse or keyboard, but…
This utter, utter, shite software is the first Microsoft OS for 25 years that actually requires a keyboard! Accidentally open a Metro-app on a touch-screen PC without a keyboard and your only option is to hit the power button (or go and hunt-down a keyboard).
I thought the biggest change with Windows-8 was that my kids would play Angry Birds, but no, the biggest change is having to tape spare batteries to the back of the screen to be sure that the ESSENTIAL keyboard will always work.
I have a Windows7 phone, I've installed Windows Server 2012 and configured the shiny new box.. I'm even coming round to the dumb ribbon.
I like my phone & the apps on it, and I like the snappy W8 UI for remote access.. and I thought Vista was fine if you had enough memory.. but Windows8 is utter utter shite.
I accept simple apps on my phone because I know its limits, but a shite apps takes me back to Windows2.. But that's not why W8 is shite..
The start screen was so familiar to my phone, that I found myself looking for the start icon because there's one on my phone (despite weeks of WinServ2012).. along the way I discovered something about the way my brain works.. I'm either in visual mode or code-mode & need somewhere click.
Windows8 is not the new Vista.. it is the worst desktop OS Microsoft have EVER produced.
Really what a dumb comment.. We're talking about phones.. You know that thing you dial 911 with in an emergency, or arrange to meet.. What exactly is the criteria for calling the "OS" shit?
Could it be phone function:no, could it be battery life:no, is there anything else a phone can do that couldn't be better done with some other kind of device:no..
There is something a little absurd about turning crops into Ethanol to water-down petrol, just for it to be belched out of the back of a huge car that makes a WWII Sherman tank look like an eco-vehicle.
The solution is very simple, America need to start taxing fuel and encouraging people to buy smaller cars.
Bedouin is much more succinct than Parallel Generational-Copying Garbage Collection with a Block-Structured Heap
Where Java oriented Garbage Collection takes a Ghetto approach of squashing objects into low memory to minimize working-set, Haskell (specifically, but also other functional languages) take a "bedouin" approach to cleaning-up immutable objects.. the process memory foot-print is larger, but the working set remains small.
Haskell doesn’t get much choice about how it uses memory; but the kicker is that you have lower contention and can take advantages of Transactional Memory which the Java –Ghetto cannot.
By the time TSX is mainstream we’ll be using 1Tb DIMMs and won’t need to pack unrelated objects into the same memory pages.. Bedouin Memory Management
The AVX2 (long long) integer operations will be in there for cryptographic processing.
The really interesting bit is the transactional memory TSX extensions (IBM’s is already well along the curve). TSX should be a big kicker for TP and HPC, but writing software to take advantage of it is going to take a big paradigm shift away from Garbage-collection to Bedouin memory management
With Windows 9 Microsoft will enable “Modern UI” to be Windowed, re-sized and docked on a desktop computers and big-screen panels. The task-bar will be replaced with a Start bar that are active tiles instead of icons.. many will call it a return to Win7, but MS will describe it as WinPhone experience for large displays, and those without MS keyboards (that replace function-keys with a touch panel). There will be no return to the “start” button, but there will be a “Home” app that is more like a WinRT version of Win3 Program-Manager intended as a launch-pad for Windows-Store-Azure-Apps. Windowed WinRT apps will look pretty indistinguishable to WPF apps (which will gain traction after Microsoft open-sources .NET).
We’d wonder what all the fuss was about, but will be told it was all needed because the Windows-8 touch interface didn’t fit with Windows-9’s gesture interface & kinetic keyboards.
As for programming… even MFC will still be supported!
I know you can get Maven to convert “customer.ContactCount += 1” to “customer.setContactCount(customer.getContactCount()+1)”, but lack of property methods is the one area where VisualBasic4 is better than Java.. a hacky language that was already deprecated when Java was first written.
I waited six years for GJ to become Generics in Java; I’ll have to wait six years for lambda functions and six years for properties.. that’s a whole load of 6-6-6 when (as the author observed) there are more fun things to do than write boilerplate code.
What is wrong with you ? Can't you smell fishy cases or what ? no, actually no.
If we (ordinary folk) go to Sweden and get carried away with a childhood fantasy about Swedish au-pair, there’s a good chance we’ll just get kicked out because “consent” pretty much puts the burden of proof on the woman (unless she’s also been beaten-up).. rape conviction rates are woeful the world-over.
Judges/prosecutors like to make examples of the rich/powerful/famous, because “doing” rich/powerful/famous people sends a deterrent message that nobody is above the law… Mr Arseange is powerful/famous and fits the profile.
Mr Arseange may well be an encrypted crusader fighting for the truth and downtrodden everywhere, violating the secrecy of evil governments that abuse their citizens and partners; but two downtrodden women in Sweden have the right to know that the law will fight for them and expose the truth of whether they were or were not violated though raped/assault.
Rape ruins the lives of millions of women, and is the most common abuse of power: Mr Arseange should reflect on the damage he does to his mission, and comply with those laws created to protect individuals from abuse by the powerful. We like to say “the innocent have nothing to fear”, but with burden of proof in a just society he might well “get off” even if he’s guilty.
Mr Arseange is right in one respect though, his life (as a convicted rapist) would be ruined.. the truth’s like that.. tough.
offloading code generation, garbage collection and runtimes is what you do to “threads” running on different “cpu cores” not thrown out to a massively parallel GPU.
Where GPU does come into play is in graphical rendering, which is where OpenGL comes into play.. but how you would get a JFC/Swing app with all its Java event handlers to talk OpenGL sounds like magic. Microsoft went through this with the move from WinForms (Win32/GDI) to WPF (Direct-X/GPU) – everything changes.
The other use of GPU is for parallel-compute.. which can work just dandy in .NET with a pinned array of struct, but you need to go IBM X10 to get anything like it in the Java world
HP trimming down on the spec for “entry level” is nothing to do with expanding the base for Tandem NonStop systems, but all about providing development & test servers to try and get new applications written for existing MPP systems.. that will feed into bigger system upgrades.
NonStop is no spring-chicken, the answer is not to port the Kernel to yet another processor but to port the TMF & messaging system to a different OS (think CICS over Encina).. and integrate with platform management & infrastructure virtualisation.. NonStop on OpenStack could provide much needed brand differentiation for the HP/EDS cloud offering.
They might have retired/”best-shored™" all their engineering expertise in Fault-Tolerant TP.. but “NonStop Cloud” is just the conceptualisation that the HP brand needs to transform the vision of its core mission statement/value proposition.
Tiered storage is nothing new, Fujitsu and other mainframe vendors have been doing Hierarchical storage for decades. "cloud gateway" storage might replace tape storage, but it does not address the big-picture of making location transparent storage, which is more a traditional data-centre issue than a "cloudy" one.
Given the success of Scala in a relatively short time with little money compared to a ten-year Sun project that sucked up DARPA funds; it is fair to say that the project simply failed.
Ok so they worked out quite quickly that the Java VM is not that great for parallel algorithms, but they couldn’t fix the JVM for political reasons and couldn’t use a different VM again for political reasons.. so why spend years flogging a dead horse.
Java’s lack of structs (value types) is the parallel handicap (& reason why IBM forked the X10 VM).. if you can’t do something as simple as an array of complex numbers without scattering over the heap, you’re not going to be able to feed a GPGPU (witness the differences between .NET & Java binding to NAG libraries).
The main reason for switching to a more modern version like Office 2003 is that it comes on CD, and is often bundled with a new PC.
After five years, I'm coming round to the idea that the Ribbon is better than toolbars, but if they're honest dumping the menus was about nixing the competition.
Repetitive strain injury (RSI) is the big problem with tables, not lack of apps.. proper applications just highlight the need for a keyboard and mouse-like device.
If Apple can continue to improve Siri and Windows and Android voice remains wowfull, then enterprise apps will start to be rewritten.
Economics is called “the dismal science” because understanding is incomplete and predictions come with huge margins for error driven by human behaviour.. if you used understanding/prediction to ban derivatives, you’d ban a lot more as well and damage the real economy.
Derivatives were first created to finance mid-west farming; Forex-options for global trade; interest-rate swaps for price stability and CDO’s for pension-funds facing a drought of AAA debt to invest in. The evil is not derivatives or bankers especially, but the asset-bubbles fuelled by greed, gullible fools; demands for double-digit grown with flat inflation and quite a few spivs… but those spivs are mostly not bankers.
At the height of the battle of Britain, parliament debated at length the legality of bombing munitions factories because they were private property and not covered by the declaration of war.. the world had not woken up to total-war, and some MP’s were distracted by the details.
Fast forward seventy years and the credit crunch threatened financial Armageddon more serious than “simply” the interbank lending market, demanding unprecedented coordinated action to flood the market with liquidity and print money. While the Bank of England continues to provide unlimited “free money” to banks, LIBOR pretty much a notional measure. Fretting over LIBOR in total-economics is like fretting over property in total-war.
At the height of the crisis LIBOR rates were so high relative to bond-futures that the money-markets froze and yield-curves (with LIBOR on the short end and futures on the long) pointed negative breaking financial models for every wholesale financial product including equities. In the circumstances it’s incredulous that politicians did not lean on Bankers (just as incredulous as Jeremy Hunt being the first to use an advisor as a back-channel), but if you’re going to mention a short rope at Blackfriars, you’re going to use a clean-line.
Always looks good, and sometimes actually is.
A “responsible” manager under stress will be looking for a scapegoat and fixate on the first goat facsimile; making suggestions increases the probability that the selected goat is actually the right one.
If it’s is your fault, make cunning suggestions like “when did the DB2 DBA last check the CA-7 backup”.. they may not know that the plan is actually in a VSAM file, plus lots and lots of options gives you a chance to fix it while problem managers are looking elsewhere.
.. but on a serious note: “Or at the next place to get hit” might be an observation of stochastic probability, but these kind of disasters are now less likey to happen because the FSA is going to put a price on the disaster. Full weekend regression testing and seasoned lags starts to look like very good value for money.
There are good and bad people everywhere, a Scottish graduate is just (if not more) likely to cock-it-up as an Indian. The issue is about cost, experience, orientation, training and commitment.
Mainframe schedulers like CA-7 & OPC/A have the concept of “current-plan” that is a pre-calculated optimisation to avoid resource contention and release jobs very quickly irrespective of the size of the schedule. If you’ve got 18k jobs “Current-plan” is vastly better than the ponderous dispatch you get from Unix schedulers that do things dynamically.
My hunch would be that it’s the “current-plan” that was lost; and the extended delay was caused by some of the jobs being run before the problem was found. CA saying the problem was unique to RBS, is probably not a polite version of “they’ve got an idiot there”, but simply that CA-7 had local customisation (perhaps additional resource dimension of running three banks off one schedule).
My guess is that they’ve replaced older mainframe-era staff with younger staff lacking the orientation to understand architectural differences; younger staff looking for a stepping stone; managers without the relevant experience.. but the real problem is cost.
Scheduling an upgrade mid-week is always about cost; Mr Hester should take it on the chin and not hide behind some poor soul, but state what he’s going to do better to deserve the trust we place in him.
a junior operator might have been the root-cause, but to trash the database and all backups sounds more like half the batch had run when the backout decision was taken and the queue needed to be reset to work from restore. bitch all you like about ITIL, but a senior change manager would have taken the backout decision, not an operator.
Say what you like about Fred Goodwin, but this would not have happened on his watch..
.. and why was a software upgrade happening mid-week, if not to save on overtime costs for Sunday?
For sure the premium iPhone market in the developed world is saturated, but while trendy coffee-shops are stuffed with iPhone strokers, laptops still outnumber tablets for three simple factors [1] RSI [2] productivity [3] flexibility (useful work apps like wordprocessors, Spreadsheets).
Sure, Apple “owns” the tablet market, but it’s not a saturated market because of the coffee-shop scope for grown (but maybe the very-very-big-iPhone market is). The market for a capable hybrid tablet/ultrabook is probably as big as the current iPad market, and if RSI cases are significant, the pure tablet-bigger-than-a-paperback model might have peeked.
The latest and next generation of apps are not more Angry-Birds or ever better shopping-lists; the future for successful “apps” is with web-enriched services and information; that is where HTML5/ECMAScript has advantages over {Objective-C, Java (ME/Davlik), .NET}. Mattie’s statistics actually point to web-technology hegemony not Android.
In a HTML5 world, you’ll get just as Angry-Birds as you got with iOS/Android/Mango/Symbian, but with one code-base, and no device lock-in.. Mattie’s “Surface plan means the world belongs to Android” is irrelevant twaddle.
Apple pretty much own the tablet market with huge economies of scale and scope to halve the price and still make a profit; but they have a weakness: iPad “is” just software and marketing with the hardware done by Foxconn in China.
Too take-on Apple at its weak point, you need to be able to prime the actual hardware manufacture with a compelling prototype and a huge order.. and this is where MS Surface comes in: If Foxconn take the MS prototype and make it into a real product, it will have all the quality and finish of an iPad, but different software.. with plenty of scope for OEM to undercut on price, or simply stick a different label on the box.. in short “the Nexus play”.
Too conclude that “Surface” -> “software dead” is both absurd (Apple hasn’t been a hardware company for a long time), and misguided: “Open Source” doesn’t mean there isn’t any money in software anymore, but the days sitting back on fat licence fees are gone.
‘nice access to testing tools’ kinda hits the nail on the head for systematic code-coverage because the language needs to provide rich metadata and a reflection interface for code-coverage analysis. You can’t do code coverage with a dynamic scripting code (C++ code-coverage tools are pretty woeful too). When you can’t even see what other scripts are dependent on the one you’ve just changed.. you’re bound to hit the occasional “it didn’t do that before”. JVM/CLR provided the metadata and reflection interfaces.
Functional testing covers the full range of test scenarios from unit-testing to end-to-end acceptance testing (does what it is functionally required to do)
You gotta love statistics, take a look at the query for {PHP; Python; Scala;F#} from the same site at PHP, Python, Scala, f#
When you take Scala & F# into the picture PHP and Python start to look like the legacy stuff.
All these recruitment statistics hide some of the fundamentals, like the need to constantly recruit script developers for those “oh, it never did that before.. better get someone in” scenarios where inadequately tested code falls over. Systematically tested C#/Java code with high code-coverage during functional & performance testing is less likely to fall-over and need constant maintenance.
Functional programming on the other hand has the productivity of a scripting shell, the performance of C#/Java and the architecture for parallel programming on all those cores to rival C++ for raw throughput.
The article actually laments the perception of IT presented by piss-poor writes/producers who think a Commodore PET in the IT Crowd or an XPS laptop in Sheldon makes them look knowledgeable.
Truth is IT is young dynamic profession that is much more interesting than Law (or accountancy); who’s practitioners are bereft of the emotional need to write idealised fiction to justify their years of toil. Maybe as IT matures, some technical authors will get so bored they will get round to writing a cracking screen-play, until then we’ve got Father Ted, Father Dougal & Mrs Doyle re-envisioned as Roy, Moss & Jen.. very funny.. but not exactly representative.
Read the article, consider being a girl-geek mentor.. you might even get…
The fact that the arduplane has to talk to the Pi over Ethernet points to a bit of a design flaw: the Pi does not have Arduino compatible GIO pins, and there is no miniport Linux driver for applications to talk to sensors.
The wonderful thing about Arduino is that the open-source design has built (over years) a rich ecosystem of sensors and controls that span smart-clothing to smart droids.. Raspberry Pi is neat, but it would have been better if it followed the netduino approach and tried to be a clever microcontroller before trying to be a cheap PC
I can’t see this being any more irritating that the pissie office ribbon that hides my document when I hit “File”, just so “they” can say that it’s not really a menu.. trashing the “Start” is consistent with other daft decision that we’ve already accepted... except maybe…
A tablet combines {processor, screen, keyboard, pointer}, but a tablet is now considerably cheaper than the original IBM PC Keyboard.. why should the tablet metaphor be applied to the screen and not to the keyboard/mouse?.. look at a Bloomberg keyboard, maybe the “Search”, “News”, “People” buttons are a simple example of how a metro UI should be applied to a business computer.
Maybe the question should not be about the Ctrl-Esc/Start key, but the SysRq and other legacy keys that could be replaced by a metro-tab..
From a software development perspective integrating SCVMM with TFS so that my automated MSBUILD jobs for compilation and unit-testing can also be deployed is something nothing else does.
TFS was already a neat one-stop integration that replaces {SVN, JIRA, Hudson/Jenkins/TeamCity, etc}, adding the automated tagging of a VM with a build & associated {bugs, stories, task} saves a lot of manual effort.
Given that {system; regression; integration; performance; operational; acceptance} testing is a big driver for VM management, the TFS/SCVVM combination is very appealing to big dev-shops
If MS can provide something that is as flexible, reliable, scaleable, manageable, adaptable, open and cheap as any of the alternatives then hay why not?
Windows HPC is a case-in-point: they bundle a bunch of tools and slash per-server licence to compete, overtake Sun in the TOP-500 ranking.. and don’t get to antsy if the WinHPC cluster is actually running Hadoop or DataSynapse.
Are they on some “Parallel Universe trip”? not really, the bulk of Enterprise Servers run Windows and the next hardware refresh of physical servers will see Hyper-V replacing some VMWare ‘cus its cheaper.. add-in simper integrated security for hybrid private/public cloud and all-of-a-sudden the “IT-Managers” (that were the target for the pitch) start seeing a cheaper way to get Disaster-Recovery.
If MS want to slash cost to win.. well let them!.. just make sure it’s a big slash..
DB journaling will write the data twice for recovery, but the old journal will be dropped after a DB backup (or mirrored checkpoint), so the journal is never as big as the data… unless you never checkpoint or backup… in which case it will grow to 100%+ (but that’s academic ‘cus a recovery would fail anyway)
33% decrease in latency is also iffy because journal-write has almost no seek latency & a decent SAN will have NV write cache.. SSD will win big on random read.. but the rate depends workload.
For any workload more important than the musings of a spotty teenager, you’d want the journal anyway for replication.
If the first world war, they cut down iron railings to “make tanks”, in the second world war they collected “pots & pans” to “make aircraft”.. in neither case did any of this effort contribute to the engineering of tanks or aircraft.. it was all about changing perceptions, feeling involved and motivating people.. not a single pot or pan was transformed into aluminium, but productivity was higher, and more planes got built.
In the same way today: the issue is not whether out carefully separated recyclables actually end up in landfill, but whether we cut down on packaging in supermarkets; not whether hybrid car help the environment, but whether we start switching engines off when stationary.
Maybe foot-fall generators can cover their costs, but I’d guess there more use for tracking people movements around shopping centres and transport hubs.. and we’re about to see the mother of all stress tests.
Don’t underestimate the importance of the politics: there is an environmental LIE (Limitation, Innovation, or Extermination)
Doing a sort on any other key than the Hadoop distribution key is always going to be slower because the data must be re-distributed in one step then aggregated/merged in another and the second step must wait for the first to finish. Dryad on the other hand is pipelined and can work on all steps concurrently.. in much the same way as Teradata did twenty years ago.
In the big-data space you can’t slosh the data into multiple stores, you have to pick one and that means Microsoft’s FDS must work with Microsoft’s version of Hadoop (Daytona) and preferably full Apache Hadoop.. just like Teradata does.
The question then is whether FDS is faster than HDFS and whether Daytona is faster than Apache Hadoop..
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
