Silver Lining
End of XP is not all bad.
The Indian State of Tamil Nadu will solve its Windows XP problem by adopting Linux. Tamil Nadu is home to over 70 million people and its capital city is Chennai, a hub for India's business process outsourcing industry second only to Bangalore. The decision to move to Linux is outlined in this letter (PDF) from the State's …
"An obscure state in a third world country where labour to look after high mantenance software is cheap picks the no money up front high maintenance solution - who would have thought it?! it must be the year of Linux on the desktop"....
'Tamil Nadu possesses the second-largest economy (2011–12) among states in India after Maharashtra.`
...the third world country where Satya Nadella was born and brought up, and received his degree Electronics and Telecommunications from the Manipal Institute of Technology...
...the third world country whose population (now 1.26bn) will overtake China's (now 1.39bn) in our lifetime (in 2028 by current estimates) to become the largest in the world...
From the tone and content of your comment, I assume you are a Microsoft attack dog trying to Scroogle Linux but of course I could be wrong
I guess all of you downvoters buy into Cameron's drivel about India as a massive source of hi-tech excellence, rather than actually having worked with the kind of developers their outsourcing companies provide.
My comment is nothing about MS/Linux, it's about Indian IT skill levels. Companies like Tata have terrible reputations for sending over incompetent workers and abusing foreign work visas.
Tata's incompetent workers doesn't implicate Indian IT in general, just like CACI's or HP's don't implicate US IT in general. When the customer is happy to keep paying for people whose only job qualification is a pulse, why spend the extra money to hire people who know what they're doing?
AC I wish I could agree, especially given my long (and once proud) history of working as a successful Windows sysadmin as a day job. Here's the kicker though: after much frustration over years of volunteered support for small Windows-based networks for various organisations I figured there must be better ways of doing things for places that I'm not directly involved with for eight hours every week day.
As it turns out, migrating such networks to either to Linux-based (Ubuntu desktop/Zentyal server) or Mac desktops/notebooks and Qnap/Synology NAS has not just reduced but almost eliminated support overhead from my perspective. I get my time back and feedback from users is overwhelmingly positive. Stuff really just does work with little fuss and the computers don't mysteriously slow down after a few months use. A school I chose the latter path for was so amazed at the difference that they now discourage staff from using Windows devices, even though I was fully supportive of continuing Windows support for those who may have wanted to use it.
Now admittedly all-in-one solutions like Zentyal and Qnap don't scale so well to large networks, but Linux-based operating systems generally scale well to larger environments and there are some great FOSS solutions out there for SSO, policy and lifecycle management.
You really should get out some time and explore the *whole* world of tech a little - you might be pleasantly surprised!
@AC
"An obscure state in a third world country where labour to look after high mantenance software is cheap picks the no money up front high maintenance solution"
The BOSS Linux desktop client appears to be based on Debian Stable, so that BOSS Linux 5.0 seems to be Wheezy based (GS 3.4.1 along with KDE and XFCE available on the install DVD). Not sure that I would describe Debian Stable as 'high maintenance'.
Tamil Nadu is pretty big as well and not obscure to me or my neighbours.
I'm posting this from a default BOSS 5.0 linux installation on a Thinkpad X200s, more powerful than target machines I guess.
The i386 DVD includes a live session, and the installer.
What you get is Debian Wheezy with Gnome 3.4.1 but set to fallback mode by default, with effects switched off. This is enforced by setting a dconf key and by commenting out the gnome-shell option in /etc/gdm3/greeter.gsettings. Someone thought about that so the live session and default installation will run on machines with lower end graphics and won't get spannered when fingers start playing.
Customised theme, nice icons, looks like Gnome 2 with the top and bottom bars comes with a BOSS user manual, and a first run pop up with toll-free support number and Web addresses. Applications include both Iceweasel and Chromium, the full LibreOffice, GIMP, Banshee, VLC, mp3 codecs and Gnash flash all out of box. I tend to set HTML 5 settings for Youtube to reduce the processor load.
Debian text mode installer has been simplified, and the installer has to run in English, no other language choice. Just type and confirm username, password and then accept defauts and you get a sensible setup. I was able to install offline easily, the network autodetect just ran and failed and then the installation resumed. There is an /etc/apt/sources.list file set up that points to the BOSS repository in India. About 30Mb of updates when rebooting.
BOSS add ons include an 'easy' installer for XFCE4 that does not require admin. Support for a long list of Indian language settings is included. There is a promotional video which gives you more of an idea of the target market, chaps in a village popping the CD into a tower PC with a 15 inch LCD monitor. Children producing printouts &c.
Conclusion: stable, some thought about target users on XP PCs, recent stable Linux, update system working, support available. Looks viable to me.
Citations? Or you are simply talking bollocks. Again.
http://www.pcworld.com/article/2082460/moving-a-city-to-linux-needs-political-backing-says-munich-project-leader.html
"Munich city authority migrated around 14,800 of the 15,000 or so PCs"
"that migrating to LiMux instead of modernizing its existing Microsoft software would save it over €11 million"
"...or want a version of Office that actually works"
LibreOffice $works quite well for me as it happens. A couple of sample PDFs produced from .odt files...
http://sohcahtoa.org.uk/pages/files/maths_gcse-rules-and-tools-number.pdf
http://sohcahtoa.org.uk/pages/files/maths_gcse-rules-and-tools-data.pdf
In my case $works = 'can produce fairly long documents with tables, mathematical formulas, drawings and imported photos easily with little fuss on a recycled laptop of modest spec'
Can you specify your definition of $works?
"It is unlikely to be a need to access "MS software" but to legacy applications that only run on [legacy] Windows."
Well I'd be guessing unless you have specific info. , but the point is they have most of their desktops on Linux and even AC can't seriously have us believe that only 20% of their staff do real work
BTW
London borough to roll out Google Chromebooks to escape Microsoft's licensing costs
http://www.computing.co.uk/ctg/news/2337233/london-borough-to-roll-out-google-chromebooks-to-escape-microsofts-licensing-costs.
http://www.computerworlduk.com/news/public-sector/3509773/barking-dagenham-council-rolls-out-google-chromebooks-in-face-of-xp-deadline/
Brilliant! They can squeeze out some more years of work from existing hardware, support local software shops, that provide the distribution and if they go for a XP-like window manager like XFCE, they don't even retrain their people a lot.
The question is, if they can pull it off in such a short lead time? I think they are at least one year too late. Of course it depends on what they use their PCs for. If all applications are browser based, hosted on a central server, and perhaps a bit of local office suite usage for writing letters, it might even work.
Anyway - whatever OS they choose, they will be more or less in the same situation.
FOSS proponents have always noted that support for Windows will eventually run out and that if support runs out no organisation can continue using it for extended periods of time. If you have FOSS you can just continue to support it yourself, and you automatically pool your efforts with everybody else still using that software.
> f you have FOSS you can just continue to support it yourself, and you automatically pool your efforts with everybody else still using that software.
That rather assumes that your replacement software is FOSS too, not just proprietary on top of FOSS.
If you can't recompile, you'll run into the same issues as XP end of support.
I find it so inconvenient whenever I hit proprietary licenses - it pushes IT into all sorts of contortions to try to minimise costs. I've a project to run a couple of internet-facing sFTP Servers in an MS shop. Haha!, you want how much for a server OS and how much for patch management? Its really low throughput but we need dedicated HA through-out, so double the ftp servers, double the AD servers for authentication, network management servers. It's all for a function which could run a couple of atoms in terms of throughput. It's just madness!
I guess we will hear a lot of stories like this in the coming years. A refresh cycle of 3 years for hardware and software is not sustainable, except for the bigger players. Maybe 5-6 years with FOSS would possibly lead to big savings.
I took charge of a small infrastructure a few years back where the servers were 6-10 years old and they had a budget for one new server. I visualized the lot using a free virtualization offering on the new server and it worked like a charm.
Linux is truly achieving world domination at last
You don't need a 3-year refresh cycle with Windows. The fact people are still running XP on the machines that came with XP is testament to this. Business support for Windows7 runs till at least 2020 so that's a 6 year refresh cycle right there even if you didn't already upgrade - longer if you did like you should've done.
If you buy into the sales pitch to upgrade every version of the OS you might lose out, but business has no problem skipping versions of Windows.
> skipping multiple versions of Windows - Vista, 7, and now 8.
Seriously, the _only_ people who would need to upgrade more or less automatically are the developers of software that _must_ run on Microsoft's latest-and-greatest - and the maintainers of that software. The rest of the world can take their own sweet time.
There is a growing trend to employ clueless managers in the IT arena. They are good at being yes-men, pencil pushers & bean counters... and of course excel following stupid orders.
Any transitional issues, including Munichs headaches, can all be attributed to this management shortcomming. IT should be left to people in the know, not the clowns with titles.
The choice of OS is the least of the problems - Windows 8 upgrades would fall victim to the same managerial incompetencies.
There is a growing trend to employ managers that are absolutely clueless, but that are good at being yes-men and following stupid orders. Independant intelligent thought is not encouraged and profit is not only the bottom line, but the only line.
Any transitional issues, including Munichs headaches, can all be attributed to this blindspot in top level corporate management. The OS is the least of the problems - Windows 8 upgrades would fall victim to the same incompetencies.
Leave I.T to the people in the know !
@JDX and all
Does anyone know what these PCs are actually used for and what the 'backoffice' is?
If they are just stand alone PCs with local printers and perhaps Internet access to open standards compliant Web sites - like a community centre/drop in set up - a direct swap out might work especially given a one panel XFCE4 desktop.
If they are like the (fully maintained Win7) client PC I'm sitting at now, there is the SharePoint based Intranet, the 'business applications' that look like Delphi on a bad day and the AD authentication to replace along with the SQL Server student database. Not happening any decade soon!
Is anyone from India reading here? Want to let us know what kind of operations we are looking at?
In their haste to migrate their users off XP (W7 then, horrors, W8) MS may have made a significant strategic blunder.
I have "numerous" clients for whom XP was more than sufficient. There was NOTHING in 7 or 8 that added to the functionality that already existed in the workplace environment which was worth the expense in an SMB setting. Perhaps if MS included a "lite" version of Office in 7/8, the migration MIGHT make sense. But if you evaluate the needs of a typical office drone, using word, excel and a browser, migrating made no financial/business sense.
As Trevor Potts pointed out in an earlier article, MS could save face by offering a subscription service for XP upgrades. But then again, I don't recall MS ever eating crow....
> 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.
This probably makes a lot of sense for an Indian organisation, the relative cost of upgrading to another Windows version would be a lot higher than for, say, a UK company. Support skills in India will be cheaper, but again this is relative so maybe a similar cost to UK organisations.
For UK companies though, the cost of upgrading hardware/software is trivial; the person sitting in front of the PC is costing 15,000 GBP per-year minimum. The desk and chair the PC and user sit on, together, probably cost as much as a low-spec PC. The main issue is how much lost productivity the organisation has during the "learning what's different" phase.
With Windows 8, Microsoft made that last bit much higher than previously and this was a major error; It won't kill them, but after such a solid product as Windows 7 it makes corporate IT very unlikely to upgrade until something more work-oriented is delivered.
I dont agree with all your post, but I do agree that this is a great opportunity for countries like India to take advantage of a Linux based OS.
The loss of XP should create a vacuum that can be filled with Linux based operating systems.
In fact, I wouldnt be shocked to hear about Linux OS's being tailored specifically to suit municipal services etc.
"[...] but I do agree that this is a great opportunity for countries like India to take advantage of a Linux based OS."
And develop the skill set of large numbers of IT support staff. Especially with migration issues. I think a few businesses could emerge from this that might be getting customers in Europe and States.
As usual we are letting a new market slip by here in sunny old Britain...
What dropping support actually means is that Patch Tuesday (13th?) May 2014, will not support XP, so until that date, (unless there is a hideously bad vulnerability discovered) there is no difference from before.
This is an extra month to play with, (providing you are willing to take the small risk of serious vulnerability.)
My expectation is that when MS don't backtrack, then they will roll it out over the next few months, but they are hoping for a U turn, so they don't have to.
"so until that date, (unless there is a hideously bad vulnerability discovered) there is no difference from before."
Pffft. Even *I* have unpatched vulnerabilities for XP in my back pocket, and I'm not a professional black hat by any means. Please...XP is wide open and in a little under 48 hours from now killing it dead will become a bloody sporting event.
"which of the Linux distros will adopt the Patch Tuesday program."
Maybe I'm misreading what you said, but Linux commonly has a "Patch Everyday" program. Patches are available once security issues are fixed. You don't need to wait for anybody to bundle them up and decide for you which bits are important or optional updates. There's also absolutely no point in waiting for 6 days if a new flaw is found and fixed on a Wednesday.
This "Patch Tuesday" marketing bollocks really works, doesn't it? As if that was a desirable feature...
Predictable patching schedules are highly desirable for organisations / business - especially large enterprises. One of the major issues with using Linux is the vastly larger number of security patches that are released on a random schedule to be evaluated - compared to a Windows platform which has far fewer patches, that are almost always released to a known schedule - so that testing and deployment can be planned and resourced for.
Predictable patching schedules are highly desirable for organisations / business - especially large enterprises.
No! Security patches being available as soon as they are ready is highly desirable!
I'm opening RTF files now! I don't want to wait until next month to be fixed!
Without you knowing it, your fanboyism has automatically turned a flaw into a feature. Is your anti-virus update also once a month? Do you all sit around a table discussing whether a remote exploit in IE should be patched or to tell users not to view jpegs?
If you so desire, you can hold the updates, let them build up for a month and on Tuesday, pick through them.
One of the major issues with using Linux is the vastly larger number of security patches that are released on a random schedule to be evaluated
It's not at random - it's as soon as possible!
Windows platform which has far fewer patches
Interesting sly dig, but there are different levels of patches. Most a just features and general upgrades.
Also, on Windows there is far less to patch. Linux patches usually cover the entire system - not just the OS... so good luck with all your other Windows software - now that is a random schedule, and update method... assuming the vendor supports it.
Now, I'm a Windows developer, and we have to implement and deploy this functionality ourselves... and too many people dislike the "check for updates [aka phone-home]"
"Predictable patching schedules are highly desirable...One of the major issues with using Linux is the vastly larger number of security patches that are released on a random schedule to be evaluated"
A couple of very good and pertinent points.
To answer them;
1. the users machines can easily be pointed at the organisations own patch server and no other, and locked down so only approved and locally tested patches will be made available to users, just like a good Windows sysadmin will do.
2. The "large" number of patches relate to the entire ecosystem, not just the OS. ie all the 10's of 1000's of apps as well as the OS and desktop environment. A very large proportion of those patches will be irrelevant to the organisation and any modern update application will only show those the users system can actually make use of. Note also that in some cases there will be multiple patches, eg the recent SSH patch set for each of the libraries, clients and servers where MS would bundle that all into a single patch file (possibly even with other, unrelated patches) so the actual count of patch files is not relevant or comparable in any meaningful way.
You raise a good point. In my extensive experience with both sides I actually lean toward the Linux side because:
- Linux patches very rarely break things (and even for Windows it's pretty rare).
- Linux repositories typically include patches for non OS software too (e.g. Firefox and LibreOffice), and many third party applications are now provided via repositores now so can take advantage of the same built-in update mechanism (e.g. Apt/Yum).
- Apt and Yum repositories are far easier and less bulky to manage and maintain than WSUS.
"Predictable patching schedules are highly desirable for organisations / business - especially large enterprises."
Yes they are, which is why you can maintain multiple stages of package repository clones, depending on severity, target environments, etc
Then you can test whatever set of patches YOU decide to roll out before doing so.
And you know what? You then can roll out critical stuff within less than 24 hours (after testing) while leaving everything else on a regular schedule. And yes, that works for enterprises. And most of it, apart from some level of testing, can be fully automated.
Been doing exactly this for the last 4 years in a global company with ~30k employees. Thanks.
For some types of businesses it's completely unacceptable to be unable to turn around critical fixes quickly and everything else on their own schedule.
"Still, i cannot help wondering how long it will take to speculate Linux flaws and even more, which of the Linux distros will adopt the Patch Tuesday program."
@Kracula
1) Thanks for posting with your user name, so many anons when being critical.
2) Security patches are rolling into OS when needed. Remember the SSH vuln on GNU/Linux OSes [1] reported some weeks ago (just after the MacOS one)? An update arrived on my gNewSense 3.1 installation within 23 hours (it could have been as little as 16 hours as I'm in a different time zone to the maintainers and need my shut-eye). I have to say that gNewSense isn't exactly the biggest or most heavily supported Linux out there.
[1] https://www.debian.org/security/2014/dsa-2869
First Munich, now Tamil Nadu. (And a few more I could mention). Each one makes it harder to claim that you can't do without Windows or Office. As every Dutchman knows, a few small leaks soon becomes an all-consuming flood. (But apart from Microsoft employees, this is a flood that all should welcome).
It's taking a very, very long time for this crack to build up.
If it is such an obvious choice then I am flabbergasted as to why these vast corporations, which exist solely to give money to their owners don't all switch over.
How many would it take to prove it works, 10?, 20?
I have worked at many Blue chips, most, if not all, used Windows everywhere. I often developed on Linux but had a Windows laptop right there as well (to do everything else). In some places the Apple laptops supplied for *nix use had had Windows installed on them on top of OS X - most were not even dual boot.
So, why don't some of Sony, Samsung, BA, etc. do it?
Is it because there are not enough people capable enough to look after their systems? Not so, apparently, they would need fewer, thus saving even more money.
is it because they are afraid it would not work? Well, if these other people did it, why can't we?
Interesting that the citations are for non-corporate entities, saving money for the public not for shareholders.
Is it because they just couldn't manage the retraining etc? Apparently not, the replacements are all similar (how interesting it was to see a Start-like button on Ubuntu - no doubt disappeared now).
Is it because the software wouldn't work? Apparently not, it does everything one wishes, apart from stupid things that one shouldn't do anyway, like ActiveX or something.
Is it because they have weird apps that don't exist elsewhere? Maybe, but everyone? hardly. Keep a few old Windows systems around until the need disappears, as it surely must.
So, apathy? They change and restructure and acquire and revamp endlessly in an attempt to get ever richer so I doubt that would stop them.
Presumably Google prove that this is possible and Google are very, very successful and very, very rich so why is their success not sufficient reason to switch systems starting tomorrow?
So, what is the reason that a story whereby some outfit decides not to use MS somewhere in the world is big enough news to actually publish it here?
And, I believe the jury maybe still be out on Munich and it is definitely out on Tamil Nadu for some considerable time.
"is it because they are afraid it would not work? Well, if these other people did it, why can't we?"
Legacy software and/or industry-specific software combine with the back-breaking cost of VDI/VDA/App-V/Thin-App and any other variant on "remotely delivering legacy Windows-based software to non-Windows endpoints.
It takes time, money and expertise to move staff from Windows to Linux, even if you have like-for-like applications across the board. When you have industry-specific stuff dragging you down...
...look, Mainframes are still around for the same bloody reason. Some of them run a dozen layers of emulation so that they can keep an application written in the bloody 60s going, because all the business logic for the entire organization lies in that ancient code. Microsoft will be the same.
The world is moving on. We are moving to newer operating systems and to companies we trust ever-so-slightly more than Microsoft. One Chromebook, iPad, Android all-in-one, Linux desktop and SaaS application at a time.
What you aren't seeing is a wholesale move from Microsoft's Windows to another single platform. Instead, you are seeing a diversity of platforms being experimented with, chosen and carefully refined to meet the needs of the niche that embraces them.
There may never be another "general operating system" like Windows again. The time for a one-size-fits all monopoly is behind us. The future belongs to task-specific devices, operating systems and applications delivered in the manner that best suits the customer, not the developer.
Competition. It's occurring right now, and no matter how much some folks want to desperately deny it, the world has changed forever from the days of Redmondian supremacy.
Microsoft is culturally incapable of making the changes required to foster trust amongst its customer base. That will be its mortal wound. Maybe it will shrink back to a small cluster of die-hard fanboys like Apple, and then come through the looking glass punching above its weight. maybe.
But it won't have the same market conditions Apple did. There won't be just one or two major players to contend with. There will be an army of quality developers catering to every niche, each with a fiercely loyal userbase. It isn't just turning the ship around that's going to be a bitch: the hearts and minds already lost will spread dissatisfaction and affection for the enemy virally. Countering that may not be possible.
"According to C. Umashankar, managing director of Electronics Corporation of Tamil Nadu (Elcot), government departments across the state will switch from Windows desktops to Novell's Suse Linux and OpenOffice from this year. Elcot is Tamil Nadu's state-owned IT supplier."
This is what they said in 2007 when Microsoft asked $150 per copy of Windows XP (and Tamil Nadu countered with $7 per copy). Was that just a price-negotiation tactic then?