back to article RAND report finds that, like fusion power and Half Life 3, quantum computing is still 15 years away

Quantum computers pose an "urgent but manageable" threat to the security of modern communications systems, according to a report published Thursday by influential US RAND Corporation. The non-profit think tank's report, "Securing Communications in the Quantum Computing Age: Managing the Risks to Encryption," urges the US …

Page:

  1. karlkarl Silver badge

    Even if an iPhone had a quantum processor, it would still be a locked down piece of crap. I would still only be able to consume apps of a tiny uninovative store and I would still need to rent provision (DRM) profiles from Apple to run my own code.

    Basically, quantum computing will be useless unless it is "open". I don't even particularly mean "open-source". More like the IBM PS/2 compatible with a BIOS where we can do what we need to.

    Because of this, I don't think quantum computers will ever be owned by us "plebs". We will probably only ever be able to access their "benefits" via a stupid web browser.

    1. Mage Silver badge
      Boffin

      Even if an iPhone . . .

      What have iPhones or Apples to do with it? Also Apple is for people with money to spare. Us plebs buy $99 android phones and $400 laptops, some of the laptops run Linux.

      The PS/2 didn't have an open BIOS and was a failure in the market place.

      Real Quantum computers may or may not happen. Currently they need cryogenics, so even a "small" tower cased sized one might need a room full of support.

      Also Quantum computers are not simply faster computers. They handle specialised sorts of problems, so even if really useable ones ever exist you'd probably not want one.

      -

      Fusion power might be closer, it's hard to say. Basically unless there is a sort of working prototype of something any forecast is total guess work.

      Prototypes can take 6 months to 10 years to be commercial:

      See Passenger Jet Aircraft (idea is from 1938!). LED is 1924 or 1962. Radio 1898, but home in 1922. Electronic TV proposed 1905, working 1935. LCDs.

      Lithium cells nearly 20 years before used in phones and laptops.

      1. karlkarl Silver badge

        Re: Even if an iPhone . . .

        I used iPhone as an umbrella for consumption hardware. You think Android is more open? You still can't do anything on them but browse silly websites.

        The PS/2 compatible (compatible with PC/AT/XT) had 2 bios. One was "open".

        What I guess I was trying to say is that the quantum processor isn't important if we aren't allowed to do anything productive with it other than viewing adverts.

        1. Mage Silver badge
          Facepalm

          Re: Even if an iPhone . . .

          The PS/2 used different I/O cards. MCA. It never had an open BIOS. Nor did the PC/XT/AT.

          The compatibles of the PC non-PS/2 systems didn't have an Open BIOS either. Compaq developed their own and the Phoenix BIOS a commercial clone supposedly created by a separate team reverse engineering the IBM PC BIOS (not PS/2) and issuing a spec. It was licenced. It was never open source at the time.

          https://en.wikipedia.org/wiki/Phoenix_Technologies#Cloning_the_IBM_PC_BIOS

          The IBM PS/2 was really the third generation IBM-PC. Unlike the first which used catalogue HW, IBM proprietary BIOS and MS's version of a reverse engineered CP/M-86, it had more proprietary HW and more IBM designing in it.

          "These models were in the strange position of being incompatible with the IBM-compatible hardware standards previously established by IBM and adopted in the PC industry. "

          The higher spec models were supposed to run OS/2, incompatible with DOS. The lower spec models were inferior spec to PC-AT clones at the same price, so despite about 3M corporate sales, the PS/2 was a failure. The PCjr even more so outside the USA.

          No PS/2 models could take the popular ISA expansion cards.

          Ironically the PS/2 came out the same year as the Archimedes running on ARM. 1987.

          PS/2 and OS/2 were too little, a couple of years too late.

          1. martinusher Silver badge

            Re: Even if an iPhone . . .

            The PC BIOS was indeed open because in keeping with IBM documentation policy both the schematic for the computer and the assembler source for the BIOS were in the user manual. This is really what set the whole clone thing on fire -- anyone who could throw a PC board together and had a supply of the chips to make it was in the clone business. These halcyon days came to an abrupt end when IBM noticed the competition and sent a politely worded nastygram to the various manufacturers giving them a month to replace the BIOS. This resulted in a mad scramble to quickly code up a replacement -- not that difficult except that these early machines used a bunch of hardware specific gotchas to implement things like system calls so the result was 'compatible' but not necessarily 'identical'.

            Subsquent models of the PC were required to be software backwards compatible so essentially the difference between a PC, XT, AT and so on what was in what was added to support things like hard disks. Things like the audio casette interface (and eventually, ROM based BASIC) gradually disappeared since they were less than useful -- there's no point in cluttering up the address space with a language if you're not going to use it.

            (Luckily for everyone IBM's lawyers hadn't come up with the API copyright wheeze or we would have really been screwed!)

        2. Doctor Syntax Silver badge

          Re: Even if an iPhone . . .

          "You still can't do anything on them but browse silly websites."

          This may come as a surprise to you, ut you can also use them to make phone calls.

          1. jake Silver badge

            Re: Even if an iPhone . . .

            Some of us only make and recieve telephone calls from our phones. A simple tool that does a simple job, but does it extremely well. (I'm an old UNIX hacker. So sue me.)

      2. werdsmith Silver badge

        Re: Even if an iPhone . . .

        Apple is for people with money to spare. Us plebs buy $99 android phones and $400 laptops, some of the laptops run Linux.

        This pleb doesn't have money to spare for phones so i buy ised iPhones for not very much money.

        To avoid Android.

        1. jake Silver badge

          Re: Even if an iPhone . . .

          Aren't all iFads ised, by definition, making "ised iPhones" redundant?

      3. TeeCee Gold badge
        Facepalm

        Re: Even if an iPhone . . .

        The PS/2 didn't have an open BIOS and was a failure in the market place.

        I have news for you. Shunned by sad nerds who only want the thing to run something else on isn't a failure. Sold like hot cakes for kids' games, which is what it was for. Ended up pretty much wiping out Nintendo and Sega, which was success beyond the wildest dreams of Sony when they introduced it.

        1. stuartnz

          Re: Even if an iPhone . . .

          Um, sorry but the OP was talking about, and here I quote, "the IBM PS/2" - not the Sony PlayStation 2 The IBM PS/2 did not wipe out either Nintendo or Sega, nor was it introduced by Sony.

  2. Brian Miller

    Quantum vs COBOL

    Ok, so the US just might be vulnerable to encryption being cracked by quantum computing. Maybe. If there is usable quantum computing...

    The thing is, the really sensitive government stuff is protected by encryption that is not public. The problem is one for the rest of us out here, who are actually far more vulnerable to a chair and rubber hoses than encryption being broken by quantum computers. And because some mainframe back there is running COBOL, doesn't mean that it has information that is sensitive enough to warrant the expense of being cracked using a quantum computer.

    Current encryption is far more vulnerable to math and GPU attacks than the alleged eventual arrival of quantum computers.

    1. DavCrav Silver badge

      Re: Quantum vs COBOL

      "The thing is, the really sensitive government stuff is protected by encryption that is not public."

      When designing a cryptosystem, you should assume that the enemy knows what it is. Security through obscurity is rarely a good idea.

    2. Martin Gregorie Silver badge

      Re: Quantum vs COBOL

      Quantum and COBOL is the obvious approach.

      Follow the path pioneered by CHAPS - all the complex message handling and protocol management was initially done on relatively slow Tandem NonStop fault-tolerant computers, which front-ended the member bank's mainframes. These undoubtedly did, and still do, run COBOL programs that handle all the bank's internal accounts and transactions. The Tandem machines used a very secure encription engine* which handled all outbound message encryption and inbound message decryption. The combination provides a secure, reliable gateway to the CHAPS financial network.

      This is a good way of handling security while maintaining network uptime and throughput. At the same time it keeps the internal systems well isolated from network nasties. Of course such a setup isn't cheap, but if manglement thinks its sensitive data doesn't justify the expense, then they deserve to carry the can if/when proved wrong.

      [*] This encrypt/decrypt engine could be quantum-based when, if ever, that technology achieves 99,99% uptime, something that Tandem NonStop systems and encryption engines achieved in the early '80s.

    3. Persona Silver badge

      Re: Quantum vs COBOL

      Current encryption is far more vulnerable to key stealingthan math and GPU attacks

      1. Mage Silver badge

        Re: Quantum vs COBOL

        XKCD. Hit him with the spanner till he gives the key

        1. Anonymous Coward
          Anonymous Coward

          Re: Quantum vs COBOL

          Citation needed:

          https://xkcd.com/538/

          1. Charles 9 Silver badge

            Re: Quantum vs COBOL

            But what happens when you're up against a masochist or a wimp? A masochist would ask for harder and a wimp would faint before you even got started.

            1. Alan Brown Silver badge

              Re: Quantum vs COBOL

              "But what happens when "

              Then you start on the family instead.

              1. Charles 9 Silver badge

                Re: Quantum vs COBOL

                Either orphaned or a black sheep. The former has no family to threaten and the latter could care less...or even provide their address. There IS such a thing as Too Kinky to Torture...

                1. jake Silver badge

                  Re: Quantum vs COBOL

                  Whatever. Show me one example of a dude (or dudette, of course!) with those attributes and a need for strong encryption. Absurd example is absurd.

                  As a thinking point, does anyone reading this really have a need for strong encryption? And by "need", I mean "might get tortured into releasing the key, it's that important to someone else". (Yes, I use it too ... but I'm not so vain as to think anyone but me actually cares that I use it.)

                  1. Doctor Syntax Silver badge

                    Re: Quantum vs COBOL

                    At that extreme, if they do, they're not going to tell you. But as a contra thinking point, do any of our everyday communications contain anything TPTB have any business being able to intercept?

                    1. jake Silver badge

                      Re: Quantum vs COBOL

                      "do any of our everyday communications contain anything TPTB have any business being able to intercept?"

                      No, of course not. That's why I encrypt them as I see fit, even if especially if there is absolutely no reason to do so :-)

                      1. You aint sin me, roit Silver badge

                        Re: Quantum vs COBOL

                        Most encryption will be applied automatically without you needing to provide a "key". The TLS handshake between your device and the server, and subsequent data encryption, will be transparent to the user. You might need to authenticate to some sites, so you'll need a "password ", and preferably a strong one...

                        Same applies to streaming services. A user might see no need for strong crypto, but a provider will!

    4. werdsmith Silver badge

      Re: Quantum vs COBOL

      I feel that a Quantum computing breakthrough might not be immediately revealed and spooks will operate it secretly like a modern Bletchley Park.

      1. jake Silver badge

        Re: Quantum vs COBOL

        I think that because most research (outside Universities) is being done in publicly traded companies, any and all breakthroughs will be announced immediately to drive up the price of their stock. As seems to be happening, if you follow the news.

        1. Charles 9 Silver badge

          Re: Quantum vs COBOL

          Really? What about black projects where the mere existence would be denied? Suppose the Utah data center is just a front for a black quantum computer project?

          1. jake Silver badge

            Re: Quantum vs COBOL

            Has anybody noticed a brain-drain, where any of the top Quantum researchers coming out of Uni just drop off the face of the Earth? Or do they all seem to wind up in public corporations (if they don't stay at Uni doing post-grad work, and teaching). From what I've seen, it's the latter, thus suggesting there are no major black quantum computing projects.

            Top boffins of this caliber are rather rare critters, and well known by all players, starting at a fairly young age. They don't appear out of thin air any more than they disappear into it.

            Sorry, my "Conspiracies Are Us" hat is in need of repair.

            1. Charles 9 Silver badge

              Re: Quantum vs COBOL

              Usually doesn't work that way. People in black projects tend to have cover jobs to explain their absences. At least, that's how I read things worked during the work on the likes of the F-117 (which was a black project as well, not to mention more conspicuous given they were working on something comparably large in physical size).

              1. jake Silver badge

                Re: Quantum vs COBOL

                Works for general engineering like F-117 (which had no really new tech in it ... the way it was put together was the new part) ... but for something like Quantum Computing, the high-end talent pool is several orders of magnitude smaller. People going missing from their "day job" would be noticed immediately.

                1. Charles 9 Silver badge

                  Re: Quantum vs COBOL

                  Point is, they would still have a day job in a related but mainstream field. They wouldn't vanish. And I don't see much difference between a quantum computer and a stealth fighter; it's the design that's the hard part.

                  1. jake Silver badge

                    Re: Quantum vs COBOL

                    The trouble is that the few people researching in this field don't have time for so-called "black" projects[0]. There quite simply aren't enough hours in the day. If any were skiving off for the government of your choice, the time spent doing that would stick out like a sore thumb.

                    [0] May I coin a phrase? "Black DevOps" ...

  3. Denarius Silver badge
    FAIL

    what is problem with COBOL ?

    who cares what language source code is written in ? As for underlying hardware, does it matter anymore outsiide of specialist cases like extreme conditions? What matters are human factors. PHB bean counters getting rewarded for cutting costs, not punished for not funding upgrades or determining if systems are still fit for purpose, secure and supportable. In short, a complete management and leadership failure. None of which matters has anything to do with languages. For the record, I like COBOL, verbose as it is but if properly written, it has three big benefits.

    (a) source code is documentation if variables are rationally named. Again, human factor.

    (b) Excellent control structures are mandated by language.. Again, code quality is created and maintained by humans.

    (c) The DATA section forces thought and understanding of the data types. This is where so much goes wrong.

    Would I use COBOL outside of business ? Probably not. C, Fortran or whatever is most suitable for problem, even if by some fluke a fashionable language might be a candidate.

    1. Yet Another Anonymous coward Silver badge

      Re: what is problem with COBOL ?

      >who cares what language source code is written in ?

      It does suggest that if your government IT is still running the same code from the 60s with layers of patches because it has always lacked the money, management and technical staff to update it.

      Then switching the security model overnight to a post Quantum Computing world might be "challenging"

    2. Anonymous Coward
      Anonymous Coward

      Re: what is problem with COBOL ?

      Over 30 years ago I was teaching old time COBOL programmers how to program in this new fangled language called 'C' (the K&R version).

      1. Steve 114

        Re: what is problem with COBOL ?

        As an 'old time COBOL programmer' I have been waiting for decades for something better. Good corporate experience with the 'Simplicity' skin for APL, still waiting for Object Orientation to deliver something remotely usable. Use-case: define the data/business problem unambiguously in words or diagrams that can be checked, then leave all the heiroglyphics to a compiler that will flag hidden ambiguities.

        1. bombastic bob Silver badge
          Devil

          Re: what is problem with COBOL ?

          migrating COBOL data storage systems to SQL, at the very least, might be worth while [but requires that time/money thing to do it]

          Trying to gerrymander "Object Oriented" into that mix is likely to result in HORRIBLY INFLATED contracts and plenty of job security later on... as in "It is SO Object Oriented that only the author understands it".

          Object Oriented = Highly Overrated [and I wonder if those who are responsible for 'those kinds of abominations' could properly DEFINE what 'an object' is, in a single sentence... [it's a nice job interview question that will weed out the "job security" types from your development team]

          1. Anonymous Coward
            Anonymous Coward

            Re: what is problem with COBOL ?

            Wanting someone to define a complex idea in a single sentence is a sign that you really do not understand how definitions work. (Try defining "electricity" in a single sentence. Three quarter page sentences like the ones in a Thomas Hardy novel don't count.)

            Object orientation actually flows naturally from the SQLmodel of data, data flow diagrams, and scoping of functions just like we did in the 1970s. The problem in my experience is that spaghetti code programmers don't get it, regardless of age, because the ability to visualise a problem as blocks of functionality and data joined together by hooks of data flow and decision making is somewhat different from coding. For one thing, I think a degree of visual sophistication is necessary to design and understand architectures, and it's obvious that many programmers are very non-visual.

            I have always distrusted any programmer who cannot express themselves with diagrams on paper.

            1. jake Silver badge

              Re: what is problem with COBOL ?

              "Wanting someone to define a complex idea in a single sentence is a sign that you really do not understand how definitions work."

              Thinking that the interviewer is expecting a one sentence answer to that complex question is a sign that you don't understand what the interviewer is actually looking for.

              It's not the answer, rather it's how it is answered.

              1. Anonymous Coward
                Anonymous Coward

                Re: what is problem with COBOL ?

                read BB's post to which I was responding. Jake, you can do better than these weak responses.

                1. jake Silver badge

                  Re: what is problem with COBOL ?

                  As usual, bob's commentardary was all over the place ... "multi-threaded", if you will. I was responding to the "interviewer" thread because it was the only portion that interested me at the time. I'm sorry if I upset your world view.

            2. jake Silver badge

              Re: what is problem with COBOL ?

              "I have always distrusted any programmer who cannot express themselves with diagrams on paper."

              One of the best programmers I have ever known was born blind. She doesn't do much drawing on paper.

              1. Anonymous Coward
                Anonymous Coward

                Re: what is problem with COBOL ?

                So, take extreme corner case and try to use it to disprove a generalisation.

                You'd certainly fail my interview tests.

                1. jake Silver badge

                  Re: what is problem with COBOL ?

                  Handling edge cases is a major part of engineering. Failure to take them into account in the initial spec is a rookie maneuver.

            3. Charles 9 Silver badge

              Re: what is problem with COBOL ?

              "Try defining "electricity" in a single sentence"

              Electricity (noun): A form of energy resulting from the existence of charged particles (such as electrons or protons), either statically as an accumulation of charge or dynamically as a current. (Source: Lexico (OED))

              There, now where's my five pounds?

              1. Anonymous Coward
                Anonymous Coward

                Re: what is problem with COBOL ?

                Nowhere. You have given no useful information at all. You have recited a dictionary definition of a noun,and haven't told anybody anything useful about what electricity actually is. Read Feynman's account of his visit to Brazil to see what I mean.

                1. Charles 9 Silver badge

                  Re: what is problem with COBOL ?

                  I don't care. I defined (by the definition of definition) the word you described within the specified limits.

                  You see, I tend to treat the English language like a computer language: with some degree of literal-mindedness. I tend to seek and (when I can) speak precisely. If someone wants a pack of cigarettes in a box, then that is what they get (a soft pack of cigarettes in a separate box). IOW, be careful what you ask for. What you really meant to say was that you didn't really want a definition but a thorough explanation. A thorough explanation in two sentences or less would be far-fetched, as would trying to explain a concept to someone who's never experienced it before (part of the plot of The Gods Must Be Crazy was how Bushmen dealt with something totally alien--a soda bottle--literally falling into their lives).

          2. Doctor Syntax Silver badge

            Re: what is problem with COBOL ?

            "Object Oriented = Highly Overrated"

            I spent ages being suspicious of OO until I eventually discovered about 75% of it was IT SOP: wrap new words round what you've been doing all along.

            If you're programming in C you have a struct to define some data. If you're using SQL you have DDL to describe a table. At a design level you might have entities in an ER diagram. The extra bit with OO is that instead of having your struct and then free-floating functions you combine the two. Once you've got used to the changed vocabulary and the extra functionality it gives you you've added the next 24%.

            All you have to avoid is the 1% of the OO religious fundamentalists with the "everything's an object" approach. Stay pragmatic and it's simply another way of thinking about data and how to handle it. Sometimes you reach out for the set-oriented view of RDBMS and SQL, sometimes OO is what fits the job.

            1. jake Silver badge

              Re: what is problem with COBOL ?

              Well said.

      2. Mage Silver badge

        Re: Over 30 years ago I was teaching

        In 1987 I was writing C++ tutorials That's 30+ years ago.

Page:

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2021