Posts by Peter Gathercole

Re: So why'd you do it then? @Rich 2

You've missed the point.

When IR35 was first mooted, and increasingly as it has morphed into what we have today, the premise was that the 'companies' and 'small businesses' were really not companies at all, but a vehicle to be employed by a company with some significant financial benefits. So they introduced the term 'disguised employees' and then deemed that their Personal Service Company (PSC) was not really a business at all, so did not fall into the category of either businesses or companies, and thus are not in scope for his statement.

In addition, for tax purposes, self-employed and employed by a PSC are completely different things, even now. So they can very easily claim that they are supporting the self-employed while taxing the hell out of a PSC.

The problem they have is differentiating a PSC from a larger company with more employees. As far as corporate law is concerned, there is not a lot of difference between a PSC and a small employer/SMB. There are some accounting and audit differences once you get over a certain level of revenue, but most of the tax/PAYE/dividend rules are largely the same for joe_blogs_it_systems Ltd, and the domestic larger IT consultancies like Capita, Deloitte, NTT, PWC, Earnst and Young etc.

So they want to put rules in place that hit PSCs, but do not affect the larger PLCs. And so far, they've been vary poor at it. Now with the wave of a legislative wand, the treasury want to put laws in place that threaten the engaging companies rather than the PSCs, hoping that the large companies will just turf over the yax and NI while still engaging the contractors. But actually, the large companies are so risk averse that they will just stop employing contractors and taking the people onto the books, but at permie rates not contractor rates.

Of course, if this happened, the tax take would actually go down, so one wonders whether this may end up being an own goal!

Alternatively. we may actually see a rise in the number of IT professionals engaged through the large consulting firms, which would probably benefit nobody other than the shareholders of those firms!

Re: Smoke And Mirrors @DontFeedTheTrolls

Actually, your statement about assessment, although true, is not the only thing it changes.

It also changes who the obligation to pay tax and national insurance sits with.

If a contractor is deemed inside of IR35, then the engaging client, or one of the intermediaries between the client and the contractor's company has to pay the PAYE and NI to HMRC, not the contractors company. This is a huge change. This obligation also exists if HMRC challenge the IR35 determination for a contractor. If the contractor was deemed outside, and HMRC challenges it, it falls to the company that performed the IR35 determination to pay for the back tax and NI HMRC deemed is owed.

It is the risk of the IR35 status being overruled by HMRC that has the client companies running scared, as if a client with a significant number of contractors suddenly has to pony up tens of thousands of pounds per contractor, the financial risk becomes very significant, even to banks and insurance companies.

And it is then the case that if it wanted to challenge this in the court, the court costs will also be significant.

I expected some of the larger companies to try to defend their position against HMRC, but it seems like HMRC have persuaded them that the risk is too great to take the chance. So it's really HMRC sowing FUD with engaging companies, with no real idea of whether they will follow through with their threats!

From the contractors side, it has been argued that if a contractor stays in the same role, and accepts that they are inside IR35 when they had previously self assessed themselves as outside, that gives HMRC a lever to chase back tax and NI for the whole of the time the contractor has been in the role as the previous assessment was obviously incorrect. This has contractors running very, very scared, as very few of them are likely to have the money to pay this readily available. This is why contractors are looking to leave their current contracts before the end of March next year.

And this is all happening before we even know that the legislation will hit the statute books (it's still draft legislation!)

Re: Ah the good old days

Hmm. have to go back to my documentation. I don't remember setting links on the system board itself, I'm pretty certain it was on the keyboard for the model B, but you may be right about it not being the normal switch block, but links on the other side of the keyboard. I'm pretty certain you could put another 8 switch block, allowing you to set addresses from 1 to 254. 0 and 255 were special, and the documentation talked about having a special router node, numbered as 255 IIRC, probably a system 3 or 4, allowing multiple Econets to be linked together.

I only set up one network with about two dozen stations, back in the early '80s. My network was mostly model B's with a couple of Master 128s added at a later time, and I believe one of the first Econet Level 3 fileservers (with a very early version of ANFS) sold, with a 10MB hard disk, and another model B acting as a print server. The fileserver was so new, there was no printed documentation on it!

I'd take my BEEB apart if I had it with me, but I'm away from home at the moment.

Re: XWindows

This could be prevented with the correct use of xhost to close down access to the X server, and if you had a shared filesystem could also be controlled quite easily using Magic Cookies.

And of course, if you are really careful, you use X protocol tunneling in ssh, although there are ways of subvirting even this.

Re: No one has mentioned Elite?

The ultimate controller IMHO was the BitStik high precision joystick. The throttle was a twist on the joystick, and the extra buttons could be used to change the view.

Re: Ah the good old days

* represeented an OS or sideways ROM command, not just Econet.

Basically, when the CLI parser saw a line starting with a *, it offered it to each sideways ROM in the system in the order of priority (I think it went from 15 down to 0), and if none of the sideways ROMs claimed the command, the OS checked against itself, and then looked at the currently active filesystem to see whether there was a command that matched in the current directory.

Econet was implemented as just another sideways ROM, and also provided a filesystem type.

Acorn did a really excellent job of designing the BBC Micro operating system, something it started in the System 2 and Atom days, and continued on into Arthur and RiscOS on the Archimedes and RiscPC.

The only real problem with the BBC and Econet is that it was completely impossible to secure the network.

There was the concept of a privileged workstation that had higher capabilities than the rest of the network. It was identified as having a station number of 0 (the station number was set either by soldering links on a location on the keyboard or fitting an 8 switch DIL switch in the same location (we used to call them DIP switches, can't remember why). This was read into a memory location when the BEEB was turned on, but unfortunately, the BEEB having no protected hardware, allowed the station number to be overwritten by whoever was on the system. This gave anybody the capability of becoming station 0.

There was a similar byte that could be overwritten with the current user ID, which identified you to the network, allowing you to masquerade as anybody on the network!

So Econet was good in principal, but unfortunately not so good in practice.

Re: Disposal cost

IBM Hursley is/was a development lab. and thus is not under the IBM UK company, but reports directly to the US holding company. The full company name is IBM UK Laboratories Ltd.

The Hursley site is not full any way you look at it. Some space is leased out to bits of IBM UK (for example, the remains of the UK Software Centre and parts of the TSS Hardware Front Office) is in Hursley. This has happened as other IBM locations in Basingstoke and the South East have been closed.

UK CICS support and training, and some other Z Series support and training is done at Hursley, although I believe that the main CICS development was transferred to Toronto some time back. I think there is also some Storwize/Spectrum Virtualize development still done there, but there is now very little in development work done anywhere in IBM in the UK nowadays.

I also think that Hursley may be one of IBM's UK cloud centres.

Re: Disposal cost

You nay have a point, but as there are very few operating System 360 or 370 systems left in the world, the chance of obtaining a near complete example of either is a chance to be jumped at.

Unlike your example of Concord, which would be almost impossible to return to operational status (see the trials of getting Vulcan XH558 to fly again), there is a chance that one or both of these systems may live again.

Both 360 and 370 were hugely influential, both in architectural and construction terms for the whole of the computing industry, introducing the concept of documented near-open architectures, plug compatible peripherals, solid-state electronics, common instruction set across all models, and introducing the concept of backward compatibility that even extends in some degree to the latest z15 systems today.

You could argue that this is all available under emulation, but until you are in the presence of one of these things quietly rumbling away in a machine room, you don't really appreciate how big they were!

What I'm surprised about is that IBM Germany of UK have not themselves provided any resource, as even they don't have a 360 or 370 in their Hursley (UK) computing museum (not sure about Germany). Mind you, even to them these machines are ancient history, as I doubt that any hardware engineers from this era are still working for IBM. I only have vague memories of working on these types of machines from my time at Durham University, using both an System 360/65 and a System 370/168, and I'm getting towards the end of my working life.

Re: Same process = no problem @Traveller

Whilst in general I agree, it is now the case that within a browser, each tab is run in a separate thread, with protection between the threads (there was an article about this in Firefox a few months ago).

As each tab could be running client side code downloaded from different servers, each in their own thread, it is necessary to have some degree of protection to prevent code in one tab getting information from another tab (imagine if you are doing some online banking in one tab, which could be spied on from a tab running some 4Chan or LOLcats pages in another).

So it is not safe to have hyperthreading turned on even inside a single process.

In addition to this, as most OSes schedule at a thread level, not a process level, it is practically impossible to have hyperthreading turned on in one process and not in another. Maybe on a core-by-core basis, but then you have all sorts of core affinity problems, and that does not solve any problems!

Re: Fixing the symptom…

On a UNIX-like system, the i-node order matters not one jot. Each directory entry contains a file-to-inode mapping, and the directory file is what is accessed when you look at a directory contents, and the i-nodes appear in any order.

On almost all UNIX-like systems, when you look at a file in a directory, the named entry is looked up, and then if more than just the name is required, the i-node number is fetched from the directory entry, and the i-node read to get the rest of the required information (which is pretty much everything about the file other than it's name).

If your OS has it, you may be able to use the ncheck (ancient UNIX command predating fsck, along with icheck and dcheck) utility to read the inodes in sequence.

Re: Fixing the symptom…

The order of files in a directory is not always the sequence of creation. Several filesystem types keep the filenames in something other than a date ordered list. In this case, it will be the order that the directory structure is traversed.

JFS2 on AIX keeps the entries in the directory structure sorted according to a particular collating sequence that is set by the base language that the system has (and re-sorts it every time a file is created). GPFS uses something like a B-tree structure in a meta-database for the directory entries. I have seen some other things, but can't remember any more off the top of my head.

And please note that some collating sequences are not [A-Z][a-z} sequence like ASCII. I remember when iso8859 became common, and the collating sequence for this is something like AaBbCc.. etc, so when doing sorts and particular transliterations (using tr) you get some very unexpected effects.

Re: You know it was mostly created by guys...just saying @Esme

My first job in 1981 was in the DP department of our local government.

In the team was a Manager, who was a man, two Data Analysts, both women, four Programmers, one of whom was a woman, and a Data Prep team, who also acted as the operators, all six of whom were women.

So out of 13 people, the men were a minority 9 to 4. And I think that most of the Data Prep. team were paid more than me and the other junior programmer (a man), so the salaries weren't skewed towards the men either (with the obvious exception of Manager, of whom there would only ever be one).

Re: CBM64 - COBOL compiler

I have to ask who actually thought COBOL on a C64 was a good idea!

I know that there were versions of Pascal for various 8-bit micros, but putting a compiler on a micro that does not really have the ability to do concurrent input and output to a filesystem.

The best I found on an 8-bit micro that the Acornsoft Pascal that was sideways-ROM based, so you could edit in memory, then compile in memory without either the editor or the compiler occupying any of your (rather precious) memory. It would also allow you to read the source from a file so that you could maximize the size of the program, but doing this from ECONET was terribly slow, as it would read it one character at a time with Acorn NFS (ECONET Level 1 or 2) as shipped with the original Model B with Econet. ANFS worked better, but raised PAGE by several pages.

I wrote a buffering system that used the RS423 and cassette buffers to read 256 bytes off the net at a time which speeded things up hugely!

Actually made our lab. of BBC micros a useful teaching tool for Pascal as well as structured BASIC and it's main reason for being, as a Computer Appreciation lab.

Legacy items on your CV

I have one line on my CV that mentions that I had learned APL when at University between 1978 and 1981, and I recently (less that six months ago) got a phone call from an agent who was desperate to find someone who even knew what it was for a role in Sweden (I think it was for Saab Defence).

He said that he would put me forward if I wanted, as the client did not want to train someone from scratch, and that the rate was very good.

I declined, as I was not looking for a contract abroad and even though I loved working in APL, it was 40 years ago dammit!. But it does show that sometimes keeping legacy things on your CV can throw up opportunities.

Re: My first programs were in COBOL

Long turnaround of card decks is why I used to check the RC of the compile in the JCL. I would speculatively put a test run in the same deck as the compile, so that if it compiled successfully, I would get at least one cycle's worth of debugging.

And if it didn't compile, the test was never run, preventing me wasting any machine time. The only thing that was wasted is the minuscule amount of energy moving a slightly larger card deck around.

I'm sure the managers didn't understand conditional execution in a JCL stack, because they kept checking that I wasn't wasting system time.

Re: First language @big_D

I echo your comments on RPG/II.

I left Uni. in 1981, having been formally taught PL/1, APL and 6502 assembler, and having taught myself C on UNIX editions 6 and 7 on the PDP11/34 that the School of Maths (computing was still considered a mathematics discipline at the time at Durham) operated for fun and research projects.

I fell in to a job at the Borough Council at my parents home town (allowing me to move back with them), and was put through a crash self learning course for RPG/II on a Sperry Univac 9/30, and initially found the language confusing (the implied loop cycle, something I appreciated later when I started using awk), limiting, and very difficult to code anything complicated other than read-a-record, do some calculations, write a record (either to a disk file, or to the printer).

A year later, having been told that I did not merit a pay rise, even though I had explored many of the more obscure features, nothing had changed my basic opinion, and I remember calling RPG/II a jumped-up macro assembler (which, in hindsight was actually flattering to it) to the Borough Treasurer when he was trying to claim that it was actually a very high-level language (his claim was that the compiler did all the work so the programmer did not have to). I pointed out that the compiler had very little work to do, as the limited set of operators were not that different from the machine instructions, and the condition codes mapped into the processor condition registers almost exactly. All it needed to do was keep track of variable names and any exception processing.

I was so angry that I actually said something along the lines of the fact that he wouldn't recognize a real high-level language if he saw it, but he at least did not fire me straight away, but let me find a new job and work out my notice.

Re: Good encapsulation, Dr S

IBM learned about registries all the way back in the 1990's.

AIX implemented a registry called the ODM with AIX 3.1 on the RS/6000. It went down like a lead balloon with almost every technical user of AIX criticizing it (including many people in IBM itself!)

Nowadays, although it's still there for some things (like device configuration and software inventory tracking), almost everything else is using flat files.

This is what experience shows us works.

Re: Complements the RISC @Michael Wojcik

I have a copy of that document. I was quoting it from memory, as it is at home, and I was away when I wrote the post.

Re: Complements the RISC @martinusher

It's quite debatable whether POWER is really RISC, although IBM came up with a couple of really bad bacronyms - "Performance Optimization With Enhanced RISC", and "Reduced Instruction Set Cycle-time" (this one is seriously bad, and goes back to the RISC System/6000 days, when IBM actually thought it could redefine RISC).

The POWER ISA has more instruction types and addressing modes than many CISC designs, although it does run instructions at or near the clock speed (or faster when you consider super-scalar execution), and does have a large register set, which are, or were, some of the defining concepts for a RISC processor.

Re: Any key

Mine is "Keyboard not present. Press F1 to continue"