1101 posts • joined 27 Apr 2009
> Along with a 4inch tablet.
Nokia N770, N800, N810, N900
I still use my N800, it is a nice size.
> I think the current UK Windows Phone market share of over 11% ... will continue to grow rapidly ...
Do try and keep up with more current data :-
"""Overseas, things are worse. In the U.K, its market share in February, 2014 was 10.1%, down from 11.3% the month previous. Back in August, 2013, its market share there was 12%, and it's been on a downward trajectory ever since. In France, things look more dismal. In February it had an 8.3% market share. It was all the way up to 12.9% there in November. In China, things are bad as well. It had a 1% market share in February, slightly up from 0.7% in January, but well under the 3.5% share it had in October, 2013."""
Re: And now he is back-peddling...
"If I cannot make money on handsets, I will not be in the handset business."
Note the personal pronoun. If he fails to make money then he will be kicked out of the company.
Re: Here's what to do if you're stuck with it
> I think a prominent example given was a CnC machine driven by a custom ISA board;
Exactly. I know one one-man-band machine shop that had a 2nd hand flat-bed machine with a controller running OS/2. It was fixed it a couple of times by clearing the log files that had filled up the hard disk. Last year the OS/2 controller completely failed so the owner had to get a replacement controller (for $20,000 or so) and it runs XP.
"""Using Python scripts"""
The serpent from the tree of knowledge, no doubt.
Re: Took them long enough...
> That pushes the oldest APIs Microsoft still supports back to 1981
MS-DOS 1.x APIs were cloned from CP/M with only minor changes, so make that 1975.
Re: None of this changes anything
> As for ACL: There is some support for the older, withdrawn version that can be switched on/used. Under current Windows versions - they are on. Always.
That is because *nix always had a useful permissions system, along with multiple groups, sgid, and multiple file links (hard or symbolic) that could achieve the necessary levels of permissions and restrictions.
Windows never had anything like that so they had to graft on ACLs. Linux can have ACLs but they are an unnecessary complication in most situations. Windows _must_ have ACLs because there is bugger all else.
> (Try porting a Swing based UI to Android,
"""Now though CodenameOne allow you to create mobile apps using not just Java but even Swing. Best of all there's even a free version. """
Or AjaxSwing: http://www.creamtec.com/products/ajaxswing/solutions/java_swing_ui_on_ipad.html
Try porting _anything_ to Windows Phone.
Re: Is Universal Windows the equivalent of yet another API?
> I'm confused! It sounds to me as if Microsoft is busy developing yet *another* mini-Windows. A fourth?
Actually that would be the fifth unless we ignore WP7 that was dumped and replaced by the incompatible WP8.
Re: Too Little Too Late
> MySQL is on a par to the free version of SQL Server
Except it is not really free. Not only is it time limited but you need to buy CALs for the clients to access it.
"""SQL Server 2014 is available for download today as a 180-day free trial version through Microsoft's TechNet Evaluation Center here."""
PostgreSQL is better.
> Nokia's devices arm is now back in profit.
Another 'TheVogon unfact' !!
While Nokia as a whole made a profit ...
"""[Nokia] phone business, which reported an adjusted operating loss of 191 million euros in the fourth quarter """
Re: How will it make money?
> The license fees payable in the other direction overtook the value of the support payments some time ago....
I very much doubt that. Nokia made around 30million WP phones in 2013 - about twice as many as in 2012. One billion divided by 30 million is about $33 per phone.
It may be that $30-$35 is what Microsoft would like to have charged per licence and so it has just become in balance, but 'some time ago' is unlikely because that would have required > $40 per phone.
The problem that Nokia had was that the support payments ended this year and they were losing money already on every phone sold.
Now with 8.1 being free later this year or next year this will have the effect of shutting down WP8 production to save paying MS with the plan to start up again with 8.1 for free plus newer, cheaper, chips so that they might even make money at last.
Of course Nokia won't be paying the tax anyway so could undersell the OEMs. That will please them no end.
> That's actually one of its advantages.
That was the marketing talking. When WP7 could only handle single core they said: why would you need dual-core ? Of course WP7 was strictly single-tasking and 'tombstoned' apps in order to start another. The only 'background' tasks were like MS-DOS TSRs.
When WP8 required dual-core they said: 'why would anyone need quad-core'. That was because WP8 did not support quad-core (now apparently it will do with 8.1).
> the lower-specced hardware it can get away with places less demand on the battery
Actually the _reason_ for multiple cores is to cater for idling with less battery demand. ARM chips can turn off the cores when not required and each core (or the last one) can idle at reduced consumption. So instead of using, say, 30% power at idle with a single core, a quad core can run at 30% of one quarter of the full power, or less if it has asymmetric cores.
Newer SoCs can also have lower battery demand than one from 2012 because of having newer manufacturing technology. They often have more functionality in the chip and this can also reduce power demands and, with fewer components in the phone, can save manufacturing cost.
Being stuck on 2 year old designs meant that Nokia had to sell at a loss (even when MS bunged them a billion a year) and hope that volume would increase so that eventually there would be at least break-even. Maybe that will happen. Meanwhile Mozilla has said that the latest chip could make a $25 phone because, basically, it will be a single chip plus a screen.
Being _old_ and lower spec does not make it cheaper.
> There's no intrinsic reason why that should be the case. New chip comes out that MS want to support, they can just as easily allocate engineers as Google can.
Yes there is. Microsoft must do the software for WP. With Android _anyone_ can tune the system to a new SoC. For example: Intel are tuning Android to their newer chips while tuning the chips for Android.
Android, and other Linux based systems, could also run on MIPS based Dragon chips or Intel or indeed any newer architecture.
> About half a billion dollars last year I read here I recall.
Given that there were a total of nearly a billion phones sold last year, and more than half of them were Android, then it's not a big deal.
> but making a phone using it is not.
Quite. One of the problems with WP is that it only supports a limited set of specific SoCs, most from 2012. It seems that 8.1 will, later this year, support some additional ones, even some quad core.
Android can support newer SoCs much quicker, these can have more features integrated and thus reduce manufacturing costs, while WP has to wait for Microsoft to catch up.
So far Nokia has been making a loss at selling WP while many Android phone companies are making a profit.
Re: Inspired by cp/m
>>There may be many BASICs but there are only vague similarities between most of them.
I would suggest that you are rather limited in your knowledge of the many varieties of what are called 'BASIC's. Portability between them is not one of their strengths.
>>The "most portable programming language" for CP/M (and later MS-DOS) was COBOL,
I am not sure of what you calling 'wrong'. COBOL, at the time, was one of the most portable of languages. In many cases, such as with RM COBOL, it was not even necessary to recompile to move an application between completely different systems, just get the appropriate runtime, it was byte coded.
In the late 70s I used MicroFocus, RM and Microsoft COBOL on CP/M, MP/M, OaSys, DRX and later on MS-DOS, Xenix, Unix and others. Moving code between these was not a problem at all.
>>Tim Paterson worked
> Time Paterson wes making machines running MS BASIC. And by the time BSD started (with Pascal), BASIC was already well entrenched in business.
I suspect that you really meant UCSD Pascal.
SCP made machines, such as the Zebra CP/M and MP/M range, that could run Microsoft COBOL, Microsoft Pascal, and very many other languages.
While there were quite a number of applications written in various BASICs the serious business ones were mostly done in 'commercial BASICs' : Pick BASIC, cBASIC2. It was necessary to rewrite code between these and several other BASICs in common use.
> You shouldn't make such simple mistakes. It will confuse readers who weren't actually there at the time.
You seem to have had a very limited exposure to the range of products available at the time. Just MS-BASIC was it ?
> Please do your homework on what Windows 1 was ... it was a GUI for DOS and not a file manager.
While Windows 1 provided a graphics library that could be built into stand-alone programs or programs to run under Win1, when Windows itself was loaded it was basically a clock, a calculator and a file manager that was similar to, but not as good as, XTree. Both could run programs.
Re: Inspired by cp/m
> The $ is pronounced "string", and was a familiar idiom from what was, at the time, the most portable programming language available. Tim had previously worked in this area: the only software available for the hardware he built was a stand-alone (customized) version of that programming language.
You seem to be way off beam.
Tim Paterson worked with 8080/Z80 machines running CP/M, such as the SCP Zebra range.
The "most portable programming language" for CP/M (and later MS-DOS) was COBOL, available since 1978 from Microfocus and Microsoft with RM not far behind. Others available before 1980 were various Pascals, C and SmallC, Fortran, and many incompatible BASICs*.
None of those are why BDOS function 9 requires its output string to be terminated by a '$'.
(you seem to be confusing this with some BASICs marking string variable names with a '$' which was nothing to do with the question or the answer.)
* Note: There may be many BASICs but there are only vague similarities between most of them.
Re: It was known for being followed by 2.1 and 2.11
> Novell DOS 7 was waaay better than 6.22 (unless you tried to use the multitasking part, which was rather buggy).
Which was tragic. DRI (which Novell bought to get DR-DOS*) had great multiuser/multitasking back to MP/M in 1978. This went through MP/M-2, MP/M-86, Concurrent-CP/M-86 (1982) to Multiuser-DOS. DR-DOS was derived from the multiuser systems. Initially it was DOS-Plus which retained some multi-tasking then DR-DOS 3.4 and 4 which has source code from Concurrent-DOS. DR-DOS 5 from Multiuser-DOS. DR-DOS 6 added back a task switcher, TaskMax, so you could run multiple programs and switch them with one actually executing at a time.
Novell DOS 7 added multitasking but instead of using DRI's from Multiuser-DOS** they invented their own buggy and poorly performing one.
* When Bill Gates announced a new version of 'MS Advanced Server' (originally an IBM product) he also said that "The next version MS-DOS/Windows might not support Netware". This was to scare users into dropping Novell for MS. Novell responded by buying DRI and announcing that each Novell server would be supplied with DR-DOS for each client. They agreed that MS would support Netware and Novell would drop Novell-DOS.
** I used Multiuser-DOS as a desktop for many years as well as having many clients running it. It could run Windows 3.11 as one (or more) of its tasks, including in 386 mode).
> This is the one that IBM adopted into MS-DOS 4
IBM wrote this into PC-DOS because most of the OEMs already had this features. I had a machine with Wyse 3.21 that had a 80Mbyte partition. They passed the code back to MS.
"""Compaq DOS 3.31 and Wyse DOS 3.21 both support >32mb disk partitions in the same fashion as DOS 4.x. """
I recall the advertising for MS-DOS 4: "It _finally_ breaks the 32Mb barrier." Like they weren't a couple of years behind the curve.
> However, disks over 512MB used inefficient 8KiB clusters -
While large clusters may be wasteful when there are many small files they are more efficient when large files are being accessed. I had COBOL applications running on FAT discs (with DR-Multiuser-DOS and similar) using ISAM file systems. These DRI systems, including DR-DOS, allowed the cluster size to be chosen when formatting disks.
Changing the cluster size from 2Kb to 8Kb improved the random access performance three fold.
Primarily this speedup is because the only way to access a particular position in the file is to start at the directory entry and track down the FAT entries until the correct sector is found. The bigger the cluster the fewer FAT entries to be read.
FAT also had no mechanism to have 'sparse' files - another inefficiency compared to a real file system.
Re: What was 2.0 really known for?
> The big problem was that some MSDOS commands used / to introduce flags.
They all did. This was copied from CP/M which copied it from DEC's RSTS/E or similar.
While it says v1.1 source it is actually MS-DOS 1.25.
It is alledged that with PC-DOS 1.0 Gary Kildall was able to get it to produce a DRI copyright message. Both SCP and MS were full DRI OEMs and had everthing for CP/M that DRI would provide. SCP for the Zebra range of computers and MS for the Z80 Softcard.
At the time there were 'decompilers with notations' for specific versions of software. While the CP/M BDOS had been written in PL/M the decompiler produced Assembly code. The program had all the comments and notations built in but required the BDOS binary to trigger the output of these to avoid copyright. It is alledged that CP/M 1.3* was decompiled and then put through Intel's 8080 - 8086 translator to be the starting point of QDOS.
When the DRI copyright was demonstrated IBM settled by giving DRI money, rewriting code to replace DRI code, and giving DRI rights to use any mechanism in MS/PC-DOS**. This rewrite was PC-DOS 1.1 and passed back to MS who called it MS-DOS 1.2x.
* 1.3 because that version had a bug in the FCB handling on a file close that existed in MS/PC-DOS 1.x but was fixed in CP/M 1.4.
** This is why DRI was never sued over DOS-Plus and DR-DOS using FAT file file system and other stuff.
Re: Only half?
> A quick look doesn't show IBMIO.sys
IBMBIO.COM is only found in PC-DOS for the 1.x and 2.x versions. Other OEMs would have their own specific IO system. The source includes IO.ASM which, apparently, is for SCP's Zebra S-100 bus systems.
> maybe 20% that might need some access to MS software
It is unlikely to be a need to access "MS software" but to legacy applications that only run on [legacy] Windows.
Re: They already do for ELA customers.
> A couple of million is still cheaper than new machines or a migration.
But they _still_ need to do new machines _and_ a migration.
Extended support this year means they will pay even more (double) for extended support next year if they don't do a migration, which will likely require new machines, before then.
> putting it at around 72%
It seems that you didn't read the whole article:
"""the 1,000 people surveryed [sic]"""
"""The survey was carried out in May by TNS, on behalf of Deloitte, among people in the UK between the ages of 16 and 64."""
It may be that the 1000 were in large urban areas. As it was only taken on a restricted age range then it is incompetent to extrapolate to being a figure to use over the whole population.
> but if we're going with marketshare as being 'sales' of SmartPhones
You may find this instructive on the issue.
> I'd certainly suggest that making a statement like "It is unlikely that there are as many as 1million WP phones in the UK." with no clear evidence to support it, would fall into the category of something you /want'/ to be true as opposed to something you /know/ to be true.
While your fanciful speculative "good calculation" claim that there were 4.5 - 5 million was based on: complete ignorance of what the figures represented.
> Here in the UK, the market share of Windows Phone is actually just above 10% ... and growing. Given that about 75% of people in the UK have a smartphone, we can probably make a good calculation that the number of Windows Phone users in the UK is between 4.5million and 5million people.
One can see you oozing enthusiasm for the product, almost in a cultist way. Unfortunately you have made several gross errors.
> the market share ... just above 10%
Market share refers to sales over the counter. It will take a couple of years to transform that into a similar figure for ownership.
> Given that about 75% of people in the UK have a smartphone
"""Presently around 53% of the UK mobile-using population of 60m has a smartphone,"""
Note that the 60million is the total population. Other estimates are that around 50% of adults have a smartphone. So the total smartphone ownership is likely to be around 20million.
It is unlikely that there are as many as 1million WP phones in the UK.
> The sold a lot of 520s because you could buy one without contract for around $70 (less than 45£)
Which is probably less than the manufacturing cost before even adding on the WP licence fee.
Re: @Martijn Bakker
> If you can point to one case where they've hired outsiders to write their software ...
Applesoft BASIC. They hired Microsoft to implement MS BASIC on Apple II.
Mac Office. Microsoft implemented it for Apple.
> So let's be clear 1/10the number means MS have gained a toe hold in the market
It wasn't 1/10.
"""3.6 million Surface devices globally last year. [..] Apple sold nearly 31 million units of the iPad... in Q4."""
"""Sales of tablet computers surged to 195.4 million last year"""
So that would be less than 1/50.
Re: A slice off the top
> Of which Firefox gobbles 1¾" -
Try "Side Tabs" add-on.
It save vertical space and the side space can be adjusted.
Re: Hey Satya, free tip..
> they had "hotmail" as a strong brand,
Hotmail was a strong brand before Microsoft bought it.
Re: MS was a Cult...
> Who in an office can work with Linux?
You are confused. Office workers don't use 'Windows', they use applications. HR uses payroll and personnel applications. Office drones use purchasing and order systems, transport and despatch systems. They may type up the odd memo. Accountants may fiddle with spreadsheets, but in general the OS is irrelevant except as a launcher for the real work programs.
Re: Hmmm..... Let's see....
> get on with what you're good at- making things that let other people make things.
The problem is that people want to make things for Android, iPhone, RaspberryPi, and such, and Microsoft won't build the tools for those.
Re: basic issue is the different API's
> Lack of integration with home entertainment
Interestingly MS had a survey taken back in the early 90s that showed most households had their PC and TV in the same room. Bill concluded that this was because they want to watch TV and use the PC at the same time so he 'innovated' a device that was both by including a broadcast tuner controlled by software.
The actual reason they had both in one room is that, unlike Bill, they didn't have a 22 room mansion.
Re: chess playing
> eventually deployed enough brute-force to outwit Kasparov
I thought Kasparov used 'brute force' (ie evaluating every move to some depth) when playing his games.
Re: " all intelligent devices - notebook, slabbies and smartphones"
> there's the dial home that most Linux installs do when installed, there is the use of update web sites.
While some distros, such as RHEL, do have registrations for the support and thus can be counted, other distros do not. I have several machines here with CentOS (4), Fedora, Mint and Ubuntu. I use CentOS as a desktop (and server) and Fedora as a server (and desktop). Plus I have clients with CentOS servers, and others with RHEL servers they built themselves.
None of these boxes are in the 'server market statistics', tough possibly some of the RHEL revenue for the OS alone might be.
As for the 'dial home', while that may indicate 'usage' it does not get included in the stats which only counts _cost_ (or sales revenue).
Re: " all intelligent devices - notebook, slabbies and smartphones"
> There are also plenty of ways to tell if someone has installed a Linux server, there's the licence that you pay for your support,
No. You are wrong, yet again.
The _license_ for Linux is free. The support agreement is _not_ a license, it is a contract. Many Linux server distros are not tied to support contracts (ie they are optional), eg CentOS, Scientific. Most Linux distros, even when not specifically servers, include server applications and can be used as anything from personal servers, SMB servers, to dedicated server clusters. And then there are various BSDs.
None of these would be in the 'server market' as a cost (or revenue). Note that _support_ costs are separate from 'the server market'.
Certainly if IBM sells an AIX box or a box with RHEL installed that would be included, but most will buy bare servers boxes and do their own install even if that is RHEL or SUSE.
> I don't understand why my initial post attracted so many downvotes.
Because, as usual, you are _wrong_. Market share in dollars is _not_ 'usage'. It is _cost_.
Re: " all intelligent devices - notebook, slabbies and smartphones"
> Most sold units.
Most _cost_ of sold units.
The 'server market' is measured in dollars, not units. A bare box with CentOS installed counts as $0 (because it isn't even included), machines built in-house (Google, Amazon, etc) aren't counted, one Windows server counts as $thousands.
But AC 'The Vogan' will still peddle his misleading stats because he doesn't understand them.
The Fall 1984 issue of byte gives prices:
64Kb PCJr entry model $599
128Kb PCJr extended 1 floppy $999
256Kb PC 1 floppy $1995
256Kb PC 2 floppy $2420
256Kb PC XT 1 HD 1 floppy $4395
In all case the monitor is extra
Color Monitor $680
Re: Floppy drives
> A 5 1/4" double-density disk had a 360KB formatted capacity per side. If your drive read only 360KB, then it was a single-sided drive. If it read 720KB, it was a double-sided drive.
No. IBM PC and XT used 5-1/4" 40 track drives. Originally they were full height 40 track single sided 8 sector giving 160Kb. Double sided gave 320Kb. An updated controller could run at 9 sector to give 180Kb or 360Kb.
Other manufacturers did use the later 5-1/4" DD 80 track drives to give 640Kb, 720Kb or even 800Kb on double sided double density.
The IBM PC AT had HD 80 track DS to give 1.2Mbyte.
If it was 5-1/4" and had 720Kb it is likely it wasn't from IBM. (though they did put 3-1/2" 720Kb drives in some models IIRC).
I had a copy of PCAlien that could read and write dozens of different manufacturer's formats (as long as an appropriate drive was available).
Re: Java on the desktop is dead
> the Dalvik engine, which is Google's implementation of a Java VM.
Yes, except it is not 'Google's'* nor is it a Java VM**.
ART*** is Google's new Dalvik VM.
* """It was originally written by Dan Bornstein"""
** """Unlike Java VMs, which are stack machines, the Dalvik VM uses a register-based architecture."""
*** """a replacement, called Android Runtime (ART), that should improve the performance of Android apps by a huge margin."""
Re: Just a question: "utilize"?
> and I have seen similar uses from other authors.
Given that their writing would be more than transitional, surely that should be:
"and I have seen similar utilisations from other authors."
(actually it should be "usages")
> When a product requires a JRE just to install, you know it's a pile of crap underneath....
You mean like having to install a .NET framework.
Re: Perhaps missing the point?
>. "...Thus the statistics are heavily skewed to unsophisticated users, ie typical Widows users (and Mac).
> Er, I run Ghostery, Privoxy, a customized hosts file, and a couple of other goodies. Obviously, I am an unsophisticated user as I run them on a Mac...
As you are thus unlikely to feature in the statistics, and are not typical, then you do not fit in the category 'unsophisticated users'.
Re: @Matt: reasons to switch away from Windows
> Total bollocks. Windows is the most frequently used server (75% market share), desktop and laptop (89% market share) platform on the planet
Yes, what you say is "Total bollocks".
Windows may have '75% market share' but that is a measure of the _cost_ (or revenue to the OEMs). It says _nothing_ about usage. The thousands of servers at Google, Amazon, and many other huge sites are not counted, corporates and individuals building their own CentOS (or many other) servers are not counted. In any case one could buy several Linux based servers for the cost of one windows server.
It seems that you have no clues about the things you talk of. Frequency of usage is not the same as cost.
Re: Perhaps missing the point?
> These are Web Access Statistics.
They are rather limited web access statistics as 1) it measures only those sites that subscribe to the particular trackers 2) it measures hits not machines or users 3) it only measures those who do not avoid tracking (NoScript, Ghostery, many others).
Thus the statistics are heavily skewed to unsophisticated users, ie typical Widows users (and Mac).
Re: Never mind
> I am fairly sure that none of my Linux machines has ever been counted
Exactly. I run NoScript and Ghostery. Those counters do not include my machines.
Re: Excuse my Ignorance but
> What is "Google" in this table? The others relate to a web server offering but what (if any) is Googles web server offering?
Re: no surprise here
> as dual-booting with a weaker competitor only helps the competitor.
Also secure boot inhibits dual boot even after the customer owns the machine.
> From the customer's point of view, dualboot or no is not a major issue,
Except that it is likely to cause the phone to be more expensive. For two reasons: the license fee for WP (which may temporarily become 'free') and having to make the phone with hardware that supports both OS (WP has limited support for modern SoCs).
A couple or 4 years ago there were some 'x86 Windows tablets' offered for sale here. They dual booted with Windows 7 and Android. They had Office on board and were only a couple of hundred dollars more than an equivalent ARM Android. It turns out in the fine print that Windows and Office were time limited 'Trial' versions. After a couple of months the cost doubled - or reverted to being an expensive Android.
Re: no surprise here
> the three remaining rusted-on weirdos who blindly adulate every Microsoft product ever made no matter what.
> ^ 19 v 3
Yes, those 3.
- Mounties always get their man: Heartbleed 'hacker', 19, CUFFED
- Analysis Oh no, Joe: WinPhone users already griping over 8.1 mega-update
- Leaked pics show EMBIGGENED iPhone 6 screen
- Opportunity selfie: Martian winds have given the spunky ol' rover a spring cleaning
- OK, we get the message, Microsoft: Windows Defender splats 1000s of WinXP, Server 2k3 PCs