Posts by Paul Crawford 5667 publicly visible posts • joined 15 Mar 2007
Go for VM use. Unless you have specific hardware needs, or are dedicated to gaming on a bare-metal installation of Windows, running in a VM has so many advantages: Never-changing hardware, ease of creating a copy/snapshot if you want to monkey with it, can be moved across hardware and host OS, and often malware won't run under virtualisation to protect its secrets so another bonus!
"with proper convergence in to NT in w7 we almost got there"
What are you talking about? The 16-bit DOS era kernels ended (badly) with Windows ME. With the relese of XP MS dropped 16-bit kernels and moved the "consumer" market to the 32-bit path started with NT.
XP was the direct successor to W2000 in terms of code/release, and that was the direct successor to NT4. You might argue about the goals of NT being better reached by Win7, but that has absolutely nothing to do with code convergence.
"Stable, AD, direct x, good driver support, backwards compatibility, etc etc"
In my case the only difference I saw was USB support. I had less stability issues under w2k, never used AD anyway, and never had driver problems or PnP issues on any of the machines I installed w2k upon. Maybe XP was more stable for some users/program combinations, but for me the only advantage was USB (plus longer support for patches, of course)
It depended to some degree on which branch you reached XP by:
1) From Windows 95->98->ME as a consumer
2) From NT3.5->NT4->W2000 as a professional
In the former case you lost quite a lot of DOS games and win16 support, but gained much better stability and security (yes, I know pre SP3 XP was hardly great, but compared to 16-bit?!)
In the latter case you got...few more devices supported and a Fisher-Price interface? Oh yes, and "product activation". But at least you could go for classic look and be back like w2k (as I did). Having said that, all that w2k effectively gave me over NT4 was USB support really.
Due to product activation, and some other reasons, w2k was last Windows I bought, XP came as work system. Now I am using Linux almost exclusively and my old w2k & XP machines run as VMs on top. Critically the license for them allows that, something the consumer versions of Fista & Win7, etc, do not.
What, you mean that people actually used the space we said was "unlimited" like there was no limit?
We all know that unlimited is a stupidly impractical thing to offer, but MS deserve a good PR kicking over this for the sheer stupidity of offering this and not expecting many to use it.
It also is a timely reminder of how putting your key data in the 'cloud' is basically giving someone else the power to change T&C and boot you out if they don't like you. This time MS appear to be giving folk a year to mend their ways, but in the future?
Paris - as she is smarter than MS marketing bods it seems.
Similar to my experiences in recent years, Linux installs with less pain than a typical machine without the correct pre-configured Windows OME image to install from (which users almost never saved when they had the chance). Less dicking around with updates as well.
And yes, I have suffered the exasperation of re-installing Vista to help a friend and it was crushingly slow to get and apply updates. After 3 hours I went home and told him to reboot it in the morning.
To be fair, installing an enterprise copy of Win 7 on recent hardware was no major trouble, but still slower than typical Linux install and try as I might, I could not get it to go from a USB stick. So it had to be a DVD written and temporary DVD reader to get it to boot and then install 7.
Maybe Win 10 has solved the USB boot and install process? Perhaps I shall never know care...
I guess in a lot of cases they don't have much choice, they have a good enough job and that company often wants them to keep doing what they need done. No offers of new projects or training on things as they come along.
I'm as guilty as any. I have not pushed myself to change job as life has been OK-enough here, and the steps in my relevant knowledge have come not by planned progression but by projects coming along and I end up doing them. Hence learning a new skill, like how to write "C code" in python...
Have an up-vote :) I wish I could give to 100 votes for the "universal macro assembler aka 'the 'C' programming language'" though!
Also share some of your views on FORTRAN, great for scientific work due to its built-in support for maths and complex numbers, extensive libraries (NAG & IMSL, etc) but had some horrible attributes as well (implicit typing, joys of GOTO being used far too often, being able to enter a function at multiple places, etc)
(1) "how users disable or never install updates" Maybe because they break things, e.g. removing media centre?
(2) "no idea why users never install antivirus" Maybe because AV is mostly crap and an on-going fee or incessant nagging? (OK must at least give MS a vote for providing a low overhead free choice here).
(3) "Microsoft finally listened, then made you redundant by taking all those little jobs out of your hands" Good for that! So never again will I have to support some friend/relative who has, yet again, trashed their system and/or got it infected?
(4) "Linux is not ready...sacrificial chickens and chalk pentangles" Good to see you have recent experience of both Windows and Linux in terms of ease of installing and sorting out problems. Never had to registry edit I presume? Never has to get a driver from some web site and side-step the scams, bloatware (looking at you printer manufacturers, WTF does a driver need to be > 100MB for?) and shitty toolbars that come with the territory?
People want and expect bug-fixes, in particular for glaring security holes.
They do not want changes that break basic functionality (e.g. removal of media centre) or require re-training to use (have you ever had to give telephone support to an elderly relative?). Android is a basket-case in this respect, but Windows has a long history of keeping those two aspects separate, until this W10 cock-down-throat push.
A while ago I had a motherboard with stupid fan speeds and ran the pwmconfig script to set it up. Might have had to set it to run on init though. Also you may have to have installed the sensors package first, as that gives you the readout of the speeds and voltages.
Or do you mean they have broken that?
""Of course transfers can always be made on the basis of an individual’s consent"
No that should not be the case, as that is asking someone to sign away their rights because they need gas or electricity, etc. Deciding not to deal with a given company because they are going to send my data to the US is often not an option, as you may only have one or two suppliers and enough do it to make competition on that basis impractical.
If I decide to deal with a US company that is one thing, but any company claiming to operate in the EU should not be allowed to break basic rights in return for slightly cheaper IT back-end supply.
How many of you have pulled the Big Red Knob on the master switch to the building/campus to see what really happens when the mains fails for more than a second?
Do the UPS hold up the machines but not the A/C systems?
Is there enough emergency lighting and torches (in working order) to get around and do stuff like check the power outage is not one of your own breakers tripping on a now-cleared fault?
For certain user groups Chrome OS is pretty good. Given that a lot of folk only really need web & email, plus some basic calendar, document edit & spreadsheet support its got them and locked down so you have to try spectacularly hard to screw it up.
Of course, the Google spying is not nice[1], and if you want much else its kind of barren, but for the price and security its hard to beat.
1. Given our glorious leaders want to spy on our every on-line activity anyway, having Google whore you from advertiser to advertiser is probably less of a risk if you don't conform to the norms of the day/party in charge :(
At first I thought he was some sort of complete idiot for having important stuff on an AOL mail account, but the suggestion that talking to Iran to try and sort things out is an unexpected breath of sanity in this world.
Still, we all have nothing to hide, so nothing to fear from our emails. Right?
@J J Carter
The simple economics you give are only a small part of the picture. As well as reduced costs from year 4-ish onwards, you also have a number of other factors:
1) Less likely to be p0wnd by script kiddies (only slightly less for nation state, though) as no macros/VB
2) No need for Windows for the OS, so some flexibility and possible cost savings there.
3) More pressure to have open formats for data exchange.
4) A very good sick to beat MS with for pricing and licence terms as your gonads are no longer quite so tightly in their vice.
Sure there are a number of cases when MS products are the only or best option, but anything that gives them cause to sit up and listen to the user base instead of screwing them for more money and/or personal data is a good thing.
Always have a separate /home partition. Ideally also a spare 20GB of unpartitoned HDD space.
Then worst case you can either re-install the OS over a badly borked one, or install a new one along side it, then boot it up, edit /etc/fstab to mount your old home partition and Robert is your little grandfather.
Now if only Windows always separated the OS partition form all user data and settings...and didn't bitch about activation keys, etc, if you do have to reinstall.
A long time ago (like 1998) I was involved in a project that decided to go with Windows NT instead of VMS because it was going to support multiple hardware platforms and be flexible and secure. All of the things that MS promised.
Alas, after buying some Alpha workstation for this (not cheap, but super-computer like speeds then) MS announced the death of all non-x86 platforms. Our customer (who was quite technical) was far from pleased and although the project was completed fine (and better than some other partner's work) the change in MS' support was a major blow.
Fast forward almost 20 years and I can see MS blow this way and that, and I am might pleased not to be dependent on them for any of my work (other than the odd Windows VM to run CAD software).
If could vote, I would turf every one of the current board out and appoint a new lot. Ones more interested in the Internet's general well-being and less on their personal fiefdoms.
In addition, I would make it a rule the no board member can stay on longer than, say, 4 years, and all have to have at least some real and recent computer science background (e.g. degree) or experience (e.g. successful management of software-heavy project).
Oh, and a personal unicorn would be nice while I'm at it...
Actually anything that marks the death-knell of "premium numbers" and stupidly over-priced foreign calls is a good thing!
It can't be beyond the wit of the telcos to have a reasonable model for data based on some monthly minimum and some reasonable extra for large amounts of data that will keep the lights on. All we need is some honesty in advertising and a regulator willing to beat them until the comply.
"a better way would be to use UTF-16 everywhere in a Windows application"
No a much better way would be some thin compatibility layer for Windows that allows UTF-8 to be used in Windows in place of UTF-16.
UTF-16 is horrible and breaks all of the native C/C++ string handling and all legacy text applications. At least UTF-8 is usable, even if you have the unpleasantness of off characters in old editors and variable length strings for a fixed number of "characters" when outside of the ASCII Latin alphabet.
Also remember that the speed of development is due to simply bolting together off-the-shelf cubesat bits and not having to design for long life and no single-point-of-failure (since its so cheap, and then they don't care if it fails soon).
The long term consequences of a vast number of short-lived and then (or even by design) uncontrollable small satellites is a serious one. Really, those things should only ever be put in a very low orbit so they will de-orbit all by themselves in a couple of years at most.
Thanks for reminding me of valgrind. Yes, it is not quick but it is a useful tool!
My comment about the efecne library is it has some minor performance hit on the allocation/freeing, but once you have an array it is pretty much full speed and not having to check array indexes on every access as the chip's VM unit will alert on out-of-bounds access. How much that impacts on a program depends on the relative amount of malloc'ing versus amount of array access.
If you are using malloc & free then you can run the code using the electric fence library (or similar) that uses the system's VM manager hardware to enforce bound checking and will trigger a segmentation fault and thus a core dump for debugging the code. This has very little performance penalty and requires no code change other than linking with the efance library.
What is much more of a pain is the abuse of stack-allocated arrays as they are much more likely to lead to code injection, and often confuse the debugger if the function context (return address) gets trashed.
Anyone know of a simple way to debug that? I.e. some automated way of using an electric fence style of check on stack arrays without a massive code change?
Also it is worth noting that a number of tools like Coverity are quite pedantic about array use from a static analysis point of view and will help find such problems even before you run the code. Not always of course, but use all the tools you have...
"lack of security for IoT deices results in a negative externality, where a cost is imposed by one party (or parties) on other parties"
OK, simple solution - make IoT vendors liable for the consequences of security breaches if any identified flaw is not automatically fixed within 30 days, maybe forcing them to have some insurance policy to cover it. That liability and/or how the premiums are calculated might just focus the idiots design and marketing minds of having a proper development, testing and support process.
What, then IoT is too expensive?
Oh dear, how sad, never mind! </Windsor Davies>
The only solution to the latency issue is to have you processing "machine" on the cloud-provider's infrastructure. At that point you surrender any security as that machine would necessarily have the key(s) to decrypt your data.
Otherwise you can use cloud for secure storage so long as it is encrypted at your end using a key not known to the storage provider, which is a good options for some situations (e.g. off-site backup).
You have a valid point, that someone has to support it.
Sadly, often the paid-for support is only a little better than what a popular (e.g. FreeNAS) forum has. I'm guessing you can also get paid support for open source solutions like FreeNAS, so its not an either-or option.
Can we have a straw poll on which major storage vendors really provide good support?
It is a fair point, that with several MLOC and a closed environment for a few dozen folk to review the code, you have very little chance of finding anything.
If, and that is a hypothetical "if", the TLA have had backdoors planted you can be damn sure they are not so dumb as to have obvious code and matching comments to draw attention to it. Most likely it would be some apparent 'typo' that allows an exploit to be deliverer, or it would be some obscure cryptographic flaw (or blind use of closed hardware support) that makes it easy for them and hard for others to exploit.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
