Slashdot Mirror

It's about how hosts hardcodes avoid redirects (hosts can do that)!

I.E. -> You can also avoid DNS servers being unpatched for 1/2 a decade now (even though a fix exists, worst of all, @ the ISP level), AND likely this redirect problem for facebook technically as well:

---

5 Years After Major DNS Flaw Found, Few US Companies Have Deployed Long-term Fix:

http://it.slashdot.org/story/13/01/29/1859257/5-years-after-major-dns-flaw-found-few-us-companies-have-deployed-long-term-fix

---

Again - Since hardcodes in hosts determine the host-domain name resolution 1st:

---

Microsoft TCP/IP Host Name Resolution Order:

http://support.microsoft.com/kb/172218

PERTINENT QUOTE/EXCERPT:

"The client checks to see if the name queried is its own.

The client then searches a local Hosts file, a list of IP address and names stored on the local computer.

A sample hosts file, Hosts.sam, is installed with the TCP/IP protocol showing the proper format.

Domain Name System (DNS) servers are queried.

If the name is still not resolved, NetBIOS name resolution sequence is used as a backup. This order can be changed by configuring the NetBIOS node type of the client."

---

Loaded @ OS startup too, by the IP stack itself (since host are TIGHTLY INTEGRATED as part of it also) running in Ring 0/RPL 0/kernelmode (vs. far, Far, FAR SLOWER usermode/Ring 3/RPL 3 & slower code in addons (ala AdBlock's python/javascript code) + browser addons SLOW DOWN WEBBROWSERS (known issue in FireFox, stack up a few & see for yourself)...

AND

Hosts are referred to by default FIRST by ANY WEBBOUND APP:

Including webbrowsers & their addons (which CANNOT resolve ip addresses "by themselves") which makes AdBlock or Ghostery, redundant (especially since they're advertiser)...

APK

P.S.=>

Please inform your company IT department about this. I'm not making this up: About Licensing

That has nothing to do with using downgrade rights. You have to get downgrade media from the OEM, however.

http://www.microsoft.com/oem/en/licensing/sblicensing/pages/downgrade_rights.aspx

seamless:
Perfectly consistent: a seamless plot in the novel.

Exchange: The most widely used email/scheduling/office management server software on the planet: http://www.microsoft.com/exchange/en-us/default.aspx

access: verb (used with object):
To make contact with or gain access to; be able to reach, approach, enter, etc.: ie: Bank customers can access their checking accounts instantly through the new electronic system.

Nothing like being repeatedly sold fluffware (i.e. software designed for retards), being constantly treated like a brain-dead toddler

BAHAHAHAHAHHAAHAHAHAHAHAHHAHA...oh holy fuck that's a good one, let me introduce you to Microsoft Windows, it's got big colorful tiles that make it easy to hit regardless of your spasticity, meanwhile I'll stick to UNIX thank you very much.

Well considering that even MSFT ended support for SP2 over 2 years ago I think a minimum of SP3 would be sensible, and since we are talking AV here and we all know how script kiddies use the patches to reverse engineer new hacks I think expecting the OS to be kept up to date really wouldn't be out of line.

Of course this is all ignoring the elephant in the room when it comes to XP, which is the rampant piracy of that OS. Damned near every out of date XP system that comes into my shop is running "XP Pro Corp Razr1911 Edition" with updates turned off by the guy that installed the hot copy of Windows. But I don't think its up to the AV companies to condone and support piracy so they should expect the OS to have the latest security patches and warn the user if it does not. If they choose to ignore that warning then its on them and like with any other unsupported piece of software if it works it works and if it don't it don't.

No, this is the template for all Windows products. OEMs must conform to the attached specifications, but otherwise have freedom in their designs. The fact that Microsoft has patented many aspects of the Surface such as the vents and type cover show it is not mean to be a "template" or hardware reference.

Surface Pro is just an example that shows Windows doesn't have to conform to stanard laptop/desktop form factors. Many other OEMs have taken the hint, and in some ways delivered better products like the Lenovo Yoga and Helix, Dell XPS 12, or the Samsung 700T. I don't think we would have ever seen innovative hardware like this if Microsoft never released the Surface.

All products fill a niche. Surface isn't the best table, and isn't the best laptop, but its a hell of a better laptop than a pure tablet, and vice vesa. Those looking for a pure tablet or a pure laptop will not find the Surface suitable. Those looking for a hybrid might. This is the niche Surface fills. You can tell us all you want how your Nexus 7 is a better tablet, and that's great. I'm happy for you. But you can keep it because it's useless to me.

I wasn't talking about servers. I was talking about something like the XP-mode virtualization on the desktop versions of Windows 7 (Pro/Ultimate/Enterprise). Granted a desktop still likely has more horsepower than a tablet, but I wouldn't expect it to be unbearably slow.

Anyone who says "no compromises" design is displaying a fundamental lack of understanding of what design is. Design IS compromise. EVERY design is trying to solve a problem -- but there's never just one problem. So you have a tablet OS and a traditional OS in the same package? Great. You also have more complexity than either alone and more disk space used. And more heat and worse battery life than a tablet, and lower performance than a dedicated laptop.

It's a red flag: if you hear a designer say "no compromise!" then walk away, because you're dealing with someone who doesn't know shit about what they're supposed to be doing. Or possibly someone who does know better but is in a position where they must say what the CEO wants them to say, which is just as bad. The point is that someone, somewhere, is not being honest at all.

A lot of people I know see low-end dell laptops as a good choice to buy, and with all the talk of 'It's up to the OEMs to decide if to allow the bootloader to be unlocked' you can see where this might be going.

For Windows 8 certification, on x86 hardware, the user must have the option of disabling Secure Boot. See the official requiements (Windows 8 System Requirements PDF), page 121. A couple of select quotes:

17. Mandatory. On non-ARM systems, the platform MUST implement the ability for a physically present user to select between two Secure Boot modes in firmware setup: "Custom" and "Standard". Custom Mode allows for more flexibility as specified in the following:

18. Mandatory. Enable/Disable Secure Boot. On non-ARM systems, it is required to implement the ability to disable Secure Boot via firmware setup. A physically present user must be allowed to disable Secure Boot via firmware setup without possession of PKpriv.

Obviously, this is the requirements for Windows 8, and there's no guarantee Microsoft won't remove that for the next iteration - and it's only for x86, whereas ARM must be locked down. We should definitely be wary & weary, but let's still stick to the facts, right?

Who said anything about safe CIL?

PCs don't require safe CIL unless the system administrator has changed the policy to require safe CIL. Windows Phone 7 and Xbox Live Indie Games require safe CIL, and there's no way for an end user to change this policy. So if GNOME were using Mono, and you wanted a Windows Phone app and a GNOME app to share any code, you'd have to stick to safe CIL.

And standard C++ certainly does compile down to full CIL.

Standard C++ doesn't compile at all when you turn on /clr:safe. The syntax for code that compiles with /clr:safe differs for classes, pointers, references, and arrays.

You may laugh, but originally that was the only way for Windows NT to get C2 certification. :-)

* http://support.microsoft.com/kb/93362
" Microsoft has opted not to include certain components of Windows NT in the evaluation process, ... It may be enough to consider networking to be another subsystem, ... "

Direct2D and Direct3D are not being abandoned, they've been moved into the Windows 8 SDK along with other APIs that have replaced components of DirectX such as XAudio2. Microsoft is just retiring obsolete, deprecated, and unsupported APIs like DirectSound (replaced by XAudio2), DirectMusic, and DirectInput. More information is available at Where is the DirectX SDK?.

Try not to look like a retarded neckbeard stuck in 2001 when you reply.

they didn't. 2 of the 3 pie's fat slices involve the OS.

From http://www.microsoft.com/investor/CompanyInfo/SegmentInfo/ServerAndTools/Overview.aspx
"Server and Tools product and service offerings include Windows Server, Microsoft SQL Server, Windows Azure, Visual Studio, System Center products, Windows Embedded device platforms, and Enterprise Services. Enterprise Services comprise Premier product support services and Microsoft Consulting Services."

To be fair, you're both going semantic and offtopic: rerun this argument like the monty python spanish inquisition script, and say that OS and business and servers are the cash cow yadda yadda...

Bloatwarepad is subsidized, just like Zune was.

Incorrect. If you add your corporate Exchange account to your iphone, you are giving your Exchange admins the ability to wipe your device. I know this because the it happened to a friend of mine - the IT function at her company apparently had a spazz attack and remote wiped ALL the phones.

iOS remote wipe from exchange

But, the Surface Pro isn't just a storage device. It would be more like renting a furnished apartment as 1200sqft and getting complaints that all the furnishings take up 700sqft. It might be excessive, but it would be installed anyway if the user were building their own device.

I'm guessing the controversy stems from the marketing decision to call it "storage" rather than "hard disk size". On one hand, more people understand "storage" vs "HDD", but on the other hand "storage" implies available storage space. Guess they should come up with a new term.

At least the Surface specifications page has an asterisk next to "storage" explaining the issue. They even have a full page dedicated to it.

Grandma's wondering why her grandson was so cheap he only got her a 32 Gig Surface with only 16 Gigs of space (on the RT model)...

Oh wait, did you get grandma the yet-unreleased Surface Pro? What exactly is Grandma's use case that supports the i5-based Surface Pro YET leave her incapable of attaching a USB drive to her tablet?

so a hard drive is no longer a device that stores date? what is it then? also, http://www.microsoft.com/Surface/en-US/surface-with-windows-rt/help-me-choose You see the little number 3 next to the storage size? Yea, that's a disclaimer. Since you seem to have forgotten that a hard drive IS a device, I'm assuming you've forgotten how to read.

The 32 GB Surface with Windows RT has aprox. 16 Gig of storage space for user content, the 64 GB Surface with Windows RT has about 45 GB of storage space for user content.

The 32 GB Surface with Windows RT has aprox. 16 Gig of storage space for user content, the 64 GB Surface with Windows RT has about 45 GB of storage space for user content.

The Surface costs less than a comparable Apple product (MacBook Air).

The specifications for the Surface Pro are fine - it's an i5 processor, 4 Gigs of RAM, 64 or 128 Gig of storage space, JUST LIKE THE BASE MacBook Air!

Who ever told you the Surface would make a suitable desktop replacement? And please, define "butchered version of Windows"? It will run ANY Windows 7 application that can fit within it's RAM and storage limitations...

I love it. Modded insightful already, yet the post is clueless. We already know how big win Rt is: http://www.microsoft.com/Surface/en-US/support/surface-with-windows-RT/files-folders-and-online-storage/surface-disk-space-faq

win rt + office + apps is 8 GB, so win rt is about 3 GB. This neglects to mention this story had *nothing* to do with win rt.

Yes, most of -- this means . What has age to do with this? .NET 1.0 was released in 2002, and it was a very early version, nowhere close to being mature, so it had issues that needed fixing. v1.1 fixed a lot, but there were design flaws in the system that couldn't just be patched on top of the existing VM -- something that Sun/Oracle decided to do with Java, because god forbid making the new code not work in my old cellphone (although that finally happened in Java 7, IIRC).

This is a guide to exactly what breaking changes were done between .NET 1.1 and 2.0: Breaking Changes in .NET Framework 2.0. Almost all the changes are either fixes for design flaws, or problems with the implementation that couldn't be fixed without breaking compatibility. So yes, MOST of the code works, and whatever code may not work, it's probably because it was using/abusing features that were broken.

Also, Microsoft is just the main developer of the .NET technologies. If Microsoft decided to suddenly drop all support for .NET and stop all development, we could still continue to use it and work on improving the standard through the Mono project, some fork of Mono, or a whole new implementation.

Ms has always tried to get popular FOSS applications running decent on their platform in a futile attempt to negate the need to run GNU/Linux for those said apps. Then when Linux became the killer app Ms went out of their way to accomodate Linux on their hyper-v system. This is not because they want Linux or FOSS around in the marketplace. They know that if they do not accomodate FOSS their system will become more and more marginalized by emerging tech.

MS only started that strategy when those FOSS applications on Linux really started eating their lunch - back around 2005. Prior to that, they were utilizing the Unix compatiblity layer to get people to run their applications on Windows in order to get Windows in the shop, and then force them to get performance by converting to Win32 from POSIX. Problem with applications like PHP and Apache was they already had those Win32 layers and very good performance, but the performance under Linux was still extremely better. IIS couldn't compete with Apache - still can't.

The FAQ does not say it requires internet access to actually use the products in their original form (only to upgrade, manage account, save to SkyDrive, etc).

http://office.microsoft.com/en-us/home-premium/#FAQs

1) From the documents you linked:

When a subscription expires, the subscription enters a brief grace period during which administrators receive notification email messages and see alerts, when they log in to the Office 365 portal, that warn that the subscription will soon be disabled.

...

If you do not renew the subscription, the subscription will soon be disabled; user accounts assigned to the expired subscription are disabled, and users are unable to access the expired subscription. However, administrators can still access the service.

More FAQs here.

I would expect Office 365 simply won't allow you to save new files to your SkyDrive account if it fills up. It doesn't seem it won't allow you to save locally though.

After actually reading the articles, I am still unclear about two things when your subscription expires: 1) How long will I have access to my documents? According to current documentation for enterprises and small business:

When a subscription is removed, all data is permanently lost.

http://onlinehelp.microsoft.com/En-ca/office365-enterprises/hh143495.aspx http://onlinehelp.microsoft.com/en-ca/office365-smallbusinesses/hh143522.aspx 2) Subscribers get an additional 20GB in Skydrive. What happens to my documents if I am using 100% of Skydrive (including the additional 20GB)? Is there a grace period? They don't make it easy to find the information to these questions. The answers are likely the same for any other cloud service that provides a free and paid offering but why do we have to guess.

From: http://www.zdnet.com/microsofts-office-365-home-premium-what-happens-when-subscriptions-expire-7000010498/ "For users with an Office 365 Home Premium subscription, as the expiration date of that subscription approaches, users will receive notifications inside the Office applications and via e-mail to remind/nag users about the approaching expiration date. Once the subscription expires, the Office apps will enter a "read-only reduced functionality mode." This means users will be able to view or print documents, but won't be able to create any new documents or edit existing documents. Users who want to regain their full Office capabilities will be able to purchase a new subscription (via Office.com) or a set of predesignated retailers. Users also will have the choice of simply using older, existing versions of Office or to just use the free Office Web Apps on SkyDrive for basic editing. If a user has stored documents created/edited with Office 365 Home Premium in their SkyDrives, these documents will still be downloadable once subscriptions expire. Users can save SkyDrive documents to another computer or drive at any time, according to Microsoft. (With Office 365 Home Premium, users get an additional 20 GB of storage on top of their existing SkyDrive quotas.)"

After actually reading the articles, I am still unclear about two things when your subscription expires: 1) How long will I have access to my documents? According to current documentation for enterprises and small business:

When a subscription is removed, all data is permanently lost.

http://onlinehelp.microsoft.com/En-ca/office365-enterprises/hh143495.aspx http://onlinehelp.microsoft.com/en-ca/office365-smallbusinesses/hh143522.aspx 2) Subscribers get an additional 20GB in Skydrive. What happens to my documents if I am using 100% of Skydrive (including the additional 20GB)? Is there a grace period? They don't make it easy to find the information to these questions. The answers are likely the same for any other cloud service that provides a free and paid offering but why do we have to guess.

From: http://www.zdnet.com/microsofts-office-365-home-premium-what-happens-when-subscriptions-expire-7000010498/ "For users with an Office 365 Home Premium subscription, as the expiration date of that subscription approaches, users will receive notifications inside the Office applications and via e-mail to remind/nag users about the approaching expiration date. Once the subscription expires, the Office apps will enter a "read-only reduced functionality mode." This means users will be able to view or print documents, but won't be able to create any new documents or edit existing documents. Users who want to regain their full Office capabilities will be able to purchase a new subscription (via Office.com) or a set of predesignated retailers. Users also will have the choice of simply using older, existing versions of Office or to just use the free Office Web Apps on SkyDrive for basic editing. If a user has stored documents created/edited with Office 365 Home Premium in their SkyDrives, these documents will still be downloadable once subscriptions expire. Users can save SkyDrive documents to another computer or drive at any time, according to Microsoft. (With Office 365 Home Premium, users get an additional 20 GB of storage on top of their existing SkyDrive quotas.)"

It's important to remember that there are 2 ways of buying Office 2013 (at least for home use): Office 2013 and Office 365. MS has a nice simple comparison here. The $99/year gets you 5 computers while the other SKUs only let you install on 1 computer.

One important change for the stand-alone SKUs is the # of computers you can install on. In Office 2010, there were SKUs that let you install on 3 PCs for "Home & Student" edition or 2 PCs for "Home and Business" edition. While Office 2013 is 1PC for all editions of the stand-alone. I'm guessing this is MS trying to push Office 365 (the subscription).

If I was installing on 5 PCs, the subscription may be worth it, but I'm not sure I like the idea of my software license expiring and possibly losing data.

When a subscription is removed, all data is permanently lost.

http://onlinehelp.microsoft.com/En-ca/office365-enterprises/hh143495.aspx
http://onlinehelp.microsoft.com/en-ca/office365-smallbusinesses/hh143522.aspx

2) Subscribers get an additional 20GB in Skydrive. What happens to my documents if I am using 100% of Skydrive (including the additional 20GB)? Is there a grace period?

They don't make it easy to find the information to these questions. The answers are likely the same for any other cloud service that provides a free and paid offering but why do we have to guess.

When a subscription is removed, all data is permanently lost.

http://onlinehelp.microsoft.com/En-ca/office365-enterprises/hh143495.aspx
http://onlinehelp.microsoft.com/en-ca/office365-smallbusinesses/hh143522.aspx

2) Subscribers get an additional 20GB in Skydrive. What happens to my documents if I am using 100% of Skydrive (including the additional 20GB)? Is there a grace period?

They don't make it easy to find the information to these questions. The answers are likely the same for any other cloud service that provides a free and paid offering but why do we have to guess.

"TCP/IP has been with us since the late 1970's. Was the primary protocol in use on the Internet from the early 1980's on"

As you so rightly pointed out, TCP/IP is a communication protocol, insecurity was provided at a higher level.

"Microsoft Internet Commerce Strategy"

"Server Foundation. A server and tools commerce foundation based on .. Microsoft Windows NT Server security".

Perhaps I should have been clearer:

Microsoft have already mandated that systems with ARM platforms MUST NOT have an option to disable Secure Boot, in order to qualify for Windows 8 hardware certification.

Source:
Mandatory. Enable/Disable Secure Boot. On non-ARM systems, it is required to implement the ability to disable Secure Boot via firmware setup. A physically present user must be allowed to disable Secure Boot via firmware setup without possession of PKpriv. A Windows Server may also disable Secure Boot remotely using a strongly authenticated (preferably public-key based) out-of-band management connection, such as to a baseboard management controller or service processor. Programmatic disabling of Secure Boot either during Boot Services or after exiting EFI Boot Services MUST NOT be possible. Disabling Secure Boot must not be possible on ARM systems.

Windows 3.11 is still available for technet subscribers. Simple!

I've been told that a lot of new laptops sold with Windows 8 have Wi-Fi chips with no Windows 7 driver.

May have been a slick sales pitch. Windows 8 software should run just fine on Windows 7. (Metro apps won't run under Windows 7.)

"If your PC is running Windows 7, your files, apps, and settings will easily transfer to Windows 8."
"Programs that run on Windows 7 will run on Windows 8."
http://windows.microsoft.com/en-US/windows-8/meet

"The gateway to get to Windows 8 is Windows 7, and we will have backward compatibility with Windows 7 embedded into Windows 8. That's something that we're very committed to. But that's a really important first pillar" Kevin Turner - Microsoft's COO said at the Worldwide Partner Conference 2011.
Despite having a slick new tile interface, the desktop underneath for Windows 8 appears to closely resemble that from Windows 7.
http://www.tomshardware.com/news/windows-win8-win7-software-compatibility,13085.html

I have used Vista 64 bit drivers on Windows 7 before, and they have worked just fine. Just because there aren't "official" drivers for Windows 7 doesn't meant the drivers that are out won't run on it.

No, Microsoft needed to split with the past APIs (.NET, win32, COM, etc) and build a single one to replace them all.

If you're referring to WinRT, it's just another facade over Win32 (like .NET is, not that .NET is going anywhere) that is leaking ugly Win32 compatibility warts like MAX_PATH. The API layer beneath Win32, the native kernel syscall API, does not have have that restriction. FYI, WinRT uses COM to define WinRT, so that's not going anywhere either. If they're building their shiny new API on top of a different existing middle layer (Win32), I don't see it becoming independent or a split with the past.

Windows 7 is stable, usable, and a sufficient progress over windows XP. WIndows XP dominated the last 10 years, and my prediction is that 7 will dominate on PCs in businesses the next 10 years.

Windows 7 can't dominate that long, and it really shouldn't. I'd go more insane if it did.

The painfully prolonged dominance of Windows XP is simply due to the failure of Vista. That had Microsoft committing to support XP until the middle of 2014, almost 13 years after it shipped. Windows 8, though, is not such a terrible failure (at least it shipped on time), so Microsoft is only committing to support Windows 7 until 2020, 10 years after release.

As with any version of Windows, some companies will use XP and 7 long after it's prudent to do so. But the majority of desktops should switch long before the end of the extended support period.

Just use some chutzpah. Contact them, and specifically ask about this project:

http://research.microsoft.com/en-us/jobs/default.aspx

Microsoft contributes to more than just Linux, including opening a lot of their own (previously closed) products. They even brag about it on their website if you want to read what they have to say about it. It's not just a one-off thing.

Ms has always tried to get popular FOSS applications running decent on their platform in a futile attempt to negate the need to run GNU/Linux for those said apps. Then when Linux became the killer app Ms went out of their way to accomodate Linux on their hyper-v system. This is not because they want Linux or FOSS around in the marketplace. They know that if they do not accomodate FOSS their system will become more and more marginalized by emerging tech.

Letting a DB server out on the internet is moronic by itself, but not having installed a patch that was available 6 months before the worm started spreading, well, that's even worse.

The worst thing of all, however, is that Microsoft *itself* had unpatched instances of SQL Server out on the net and they themselves got pwned.

How I really like the Windows registry: Go to HKFOOBARZZZ, click on FOOBAR, Local User, Applications, Windows, Internet Explorer, Version Xxxx, 50 clicks and more

Let me introduce you to reg.exe.

See, that's exactly my point: once you know what tools are available and where to find out more about them, everything is easy.

Now you can just google (or bing) it.

Of course I do that. But that one works in Fedora, this one in Ubuntu, that in Debian (deliberatly naming distros here, not saying that's exactly that) ... it wasn't easy.

The graphic have only 80x25 resolution (the classical DOS resolution), but it's still a GUI...

Sure, it's a GUI ... but with the command line always present, where I con do my net use, arp or whatever I need to do. And - that's the most important part for me - I need no mouse at all.

Don't get me wrong - I'm not saying "Windows is better". I stay away from those cultistic flame wars. I was just describing what's difficult for me to get the hang for Linux. If I would have started out using Linux, I'd most likely say the same about Windows.

The Surface Pro isn't good at anything.

Surface Pro is the best tablet that can run x86 applications. Show me another machine that is as thin, light, and powerful with a stylus. You can't.

I'll also save you the trouble of reading two replies to your posts....

It's too heavy, expensive and crappy on battery power to be a good tablet

You might say a tablet is defined by being thin and light and having all day battery life, but that's different from what I want in a tablet. I want a stylus for writing. Surface Pro has this, 90% of tablets today don't. I want ports for USB drives, portable harddrives, video out. Surface Pro has this, 90% of tablets today don't. I want an SD card slot for expansion and swapping cards. Surface Pro has this, 90% of tablets today don't. Hell even Google's Nexus tablets don't have this. So in my book, Surface Pro is the only real tablet out there. It all depends on the user's needs. The Surface pro is a little thicker and heavier than iPad (.5 lbs and .5") but the tradeoff is more power with an i5 and compatibility with millions of applications and devices. So great for you if your iPad or Android tablet is thinner and lighter and lasts all day. I don't care because they are as useless as rocks for my needs, and I'll gladly pay $899 for a tablet that does what I want.

and it doesn't have a real keyboard or adequate screen size to make it a good laptop.

What about this isn't a "real" keyboard? Or the infinite variety of bluetooth keyboards or wired USB keyboards that can be plugged in? I also fail to see how 10.6" is not an adequate screen size but 11.6" on the Macbook Air somehow is. That extra 1" is the threshold? I might similarly say the Macbook Air has inadequate resolution (1366×768 vs 1920x1080 on Surface Pro). Or that Macbook Air is too heavy (2.38lbs vs 2.01lbs). Or too thick (0.68" vs 0.5"). Or too expensive ($999 vs $899).

time_t has also been 64 bit for a number of years now

...on most (all?) 64-bit platforms and on those 32-bit platforms that have gone with 64-bit time_t .

At least as I read the Visual Studio 2012 documentation, time_t is 64-bit by default; you have to define _USE_32BIT_TIME_T to get a 32-bit time_t. This goes back as far as Visual Studio 2005; in Visual Studio .NET 2003, it's a "long integer", which is 32 bits in Microsoftland.

A quick check of my Solaris 11 (virtual) machine's /usr/include indicates that it defines time_t as a long, so it's 64 bits only for 64-bit code (can't break binary compatibility...), so only 64-bit code avoids the Y2.038K problem on Solaris. A quick check of that machine's host, running OS X Mountain Lion, reveals the same thing (again, can't break binary compatibility...), so only 64-bit code avoids the Y2.038K problem on OS X.

I'm not certain what headers are used when compiling 32-bit on, for example, my (64-bit) Ubuntu 12.04 (virtual) machine, but, if I compile a program that prints sizeof(time_t) with -m32, it prints 4.

So, on two fairly significant UN*Xes, time_t appears to be 64-bit only for 64-bit code. Now, it may well be that, well before 2038, for desktop and server computing, and possibly even smartphone/tablet computing, all libraries and applications will be 64-bit.

However...

So it's not a real problem even on embedded systems

...I wouldn't necessarily assume that. The good news is that truly embedded systems (I don't consider smartphones etc. to be embedded systems, as, even if they don't support development tools running on them, they can run third-party applications, and thus might be subject to binary-compatibility issues) don't have the sort of binary-compatibility issues that would force them to keep time_t 32-bit. However, if they're running a software platform that, by default, makes time_t 32-bit on 32-bit processors, unless they've made an effort to override that default, time_t is 32-bit. Hopefully the lifetime of those embedded systems is sufficiently short (and the developers of those systems sufficiently clueful) that anything with 32-bit time_t gets replaced in time.

All that over and above...

However, even those compiled with 64-bit time_t aren't necessarily safe - you'd probably find most of them assume it's still a 32-bit quantity and end up storing it as such - ignoring the compiler warnings or casting to get rid of them. So even programs of today with 64-bit time_t's won't necessarily make it past 2038 either.

And if stuff like that is done, recompiling does diddly - the bug will still strike despite a 64-bit everything. Hell, someone may have decided during the conversio nto increase the timestamp size from 32 to 64 bit, but didn't realize someone squashed it down to 32-bit upstream.

...what you've already noted, in addition to file formats with 32-bit time stamps (e.g., pcap format; fortunately, pcap-ng format can, with time stamps with nanosecond resolution, last until 2554).

time_t has also been 64 bit for a number of years now

...on most (all?) 64-bit platforms and on those 32-bit platforms that have gone with 64-bit time_t .

At least as I read the Visual Studio 2012 documentation, time_t is 64-bit by default; you have to define _USE_32BIT_TIME_T to get a 32-bit time_t. This goes back as far as Visual Studio 2005; in Visual Studio .NET 2003, it's a "long integer", which is 32 bits in Microsoftland.

A quick check of my Solaris 11 (virtual) machine's /usr/include indicates that it defines time_t as a long, so it's 64 bits only for 64-bit code (can't break binary compatibility...), so only 64-bit code avoids the Y2.038K problem on Solaris. A quick check of that machine's host, running OS X Mountain Lion, reveals the same thing (again, can't break binary compatibility...), so only 64-bit code avoids the Y2.038K problem on OS X.

I'm not certain what headers are used when compiling 32-bit on, for example, my (64-bit) Ubuntu 12.04 (virtual) machine, but, if I compile a program that prints sizeof(time_t) with -m32, it prints 4.

So, on two fairly significant UN*Xes, time_t appears to be 64-bit only for 64-bit code. Now, it may well be that, well before 2038, for desktop and server computing, and possibly even smartphone/tablet computing, all libraries and applications will be 64-bit.

However...

So it's not a real problem even on embedded systems

...I wouldn't necessarily assume that. The good news is that truly embedded systems (I don't consider smartphones etc. to be embedded systems, as, even if they don't support development tools running on them, they can run third-party applications, and thus might be subject to binary-compatibility issues) don't have the sort of binary-compatibility issues that would force them to keep time_t 32-bit. However, if they're running a software platform that, by default, makes time_t 32-bit on 32-bit processors, unless they've made an effort to override that default, time_t is 32-bit. Hopefully the lifetime of those embedded systems is sufficiently short (and the developers of those systems sufficiently clueful) that anything with 32-bit time_t gets replaced in time.

All that over and above...

However, even those compiled with 64-bit time_t aren't necessarily safe - you'd probably find most of them assume it's still a 32-bit quantity and end up storing it as such - ignoring the compiler warnings or casting to get rid of them. So even programs of today with 64-bit time_t's won't necessarily make it past 2038 either.

And if stuff like that is done, recompiling does diddly - the bug will still strike despite a 64-bit everything. Hell, someone may have decided during the conversio nto increase the timestamp size from 32 to 64 bit, but didn't realize someone squashed it down to 32-bit upstream.

...what you've already noted, in addition to file formats with 32-bit time stamps (e.g., pcap format; fortunately, pcap-ng format can, with time stamps with nanosecond resolution, last until 2554).

time_t has also been 64 bit for a number of years now

...on most (all?) 64-bit platforms and on those 32-bit platforms that have gone with 64-bit time_t .

At least as I read the Visual Studio 2012 documentation, time_t is 64-bit by default; you have to define _USE_32BIT_TIME_T to get a 32-bit time_t. This goes back as far as Visual Studio 2005; in Visual Studio .NET 2003, it's a "long integer", which is 32 bits in Microsoftland.

A quick check of my Solaris 11 (virtual) machine's /usr/include indicates that it defines time_t as a long, so it's 64 bits only for 64-bit code (can't break binary compatibility...), so only 64-bit code avoids the Y2.038K problem on Solaris. A quick check of that machine's host, running OS X Mountain Lion, reveals the same thing (again, can't break binary compatibility...), so only 64-bit code avoids the Y2.038K problem on OS X.

I'm not certain what headers are used when compiling 32-bit on, for example, my (64-bit) Ubuntu 12.04 (virtual) machine, but, if I compile a program that prints sizeof(time_t) with -m32, it prints 4.

So, on two fairly significant UN*Xes, time_t appears to be 64-bit only for 64-bit code. Now, it may well be that, well before 2038, for desktop and server computing, and possibly even smartphone/tablet computing, all libraries and applications will be 64-bit.

However...

So it's not a real problem even on embedded systems

...I wouldn't necessarily assume that. The good news is that truly embedded systems (I don't consider smartphones etc. to be embedded systems, as, even if they don't support development tools running on them, they can run third-party applications, and thus might be subject to binary-compatibility issues) don't have the sort of binary-compatibility issues that would force them to keep time_t 32-bit. However, if they're running a software platform that, by default, makes time_t 32-bit on 32-bit processors, unless they've made an effort to override that default, time_t is 32-bit. Hopefully the lifetime of those embedded systems is sufficiently short (and the developers of those systems sufficiently clueful) that anything with 32-bit time_t gets replaced in time.

All that over and above...

However, even those compiled with 64-bit time_t aren't necessarily safe - you'd probably find most of them assume it's still a 32-bit quantity and end up storing it as such - ignoring the compiler warnings or casting to get rid of them. So even programs of today with 64-bit time_t's won't necessarily make it past 2038 either.

And if stuff like that is done, recompiling does diddly - the bug will still strike despite a 64-bit everything. Hell, someone may have decided during the conversio nto increase the timestamp size from 32 to 64 bit, but didn't realize someone squashed it down to 32-bit upstream.

...what you've already noted, in addition to file formats with 32-bit time stamps (e.g., pcap format; fortunately, pcap-ng format can, with time stamps with nanosecond resolution, last until 2554).

PowerShell is built on top of .NET, so objects are .NET objects. And it can be any object at all, not some specific predefined ones. When you access their properties (or invoke methods - which you can also do, although it's relatively rare in PS scripts), it uses reflection to discover what the object actually supports. So, yes, in practice you can pass any data, organized in any way you want (since you can always have a stream of one object, and said object can have properties that return other objects - so really it can be an arbitrary rooted graph). Generally speaking, standard commands try to produce instances of standard framework classes - e.g. 'ls' produces streams of FileInfo and DirectoryInfo.

The way this interacts with traditional text streams is pretty simple: strings are also objects in .NET (System.String), and every .NET object has a ToString method - so PS can always treat an object stream as text, or text stream as object, as needed. That's pretty much what it does when you have a regular program anywhere in the pipeline, or for the last step in it (the results of which you want to print out).

In practice, it's somewhat more complicated, because ToString does not necessary return a pretty-printed human-readable representation of the object, so it also has its own formatting layer there - the various Format-* cmdlets, like Format-Table (aliased as ft).So you can do things like:

ls | ft name,length -g extension

i.e. print out entries one per line, in two columns representing file name and file length, and grouping them by file extension. If your last step in the pipeline is not ft (or something else that produces a text stream), it tries to guess formatting depending on object type.

The 'where' command in my example does not actually do any parsing. It just takes a predicate function as a parameter, applies it to its input stream, and only outputs those objects that match the predicate. The {} syntax defines a single-parameter lambda function, where $_ is the name of that parameter in the body of the lambda - it's all built-in PS syntax, 'where' doesn't know anything about it.

PowerShell is built on top of .NET, so objects are .NET objects. And it can be any object at all, not some specific predefined ones. When you access their properties (or invoke methods - which you can also do, although it's relatively rare in PS scripts), it uses reflection to discover what the object actually supports. So, yes, in practice you can pass any data, organized in any way you want (since you can always have a stream of one object, and said object can have properties that return other objects - so really it can be an arbitrary rooted graph). Generally speaking, standard commands try to produce instances of standard framework classes - e.g. 'ls' produces streams of FileInfo and DirectoryInfo.

The way this interacts with traditional text streams is pretty simple: strings are also objects in .NET (System.String), and every .NET object has a ToString method - so PS can always treat an object stream as text, or text stream as object, as needed. That's pretty much what it does when you have a regular program anywhere in the pipeline, or for the last step in it (the results of which you want to print out).

In practice, it's somewhat more complicated, because ToString does not necessary return a pretty-printed human-readable representation of the object, so it also has its own formatting layer there - the various Format-* cmdlets, like Format-Table (aliased as ft).So you can do things like:

ls | ft name,length -g extension

i.e. print out entries one per line, in two columns representing file name and file length, and grouping them by file extension. If your last step in the pipeline is not ft (or something else that produces a text stream), it tries to guess formatting depending on object type.

The 'where' command in my example does not actually do any parsing. It just takes a predicate function as a parameter, applies it to its input stream, and only outputs those objects that match the predicate. The {} syntax defines a single-parameter lambda function, where $_ is the name of that parameter in the body of the lambda - it's all built-in PS syntax, 'where' doesn't know anything about it.

PowerShell is built on top of .NET, so objects are .NET objects. And it can be any object at all, not some specific predefined ones. When you access their properties (or invoke methods - which you can also do, although it's relatively rare in PS scripts), it uses reflection to discover what the object actually supports. So, yes, in practice you can pass any data, organized in any way you want (since you can always have a stream of one object, and said object can have properties that return other objects - so really it can be an arbitrary rooted graph). Generally speaking, standard commands try to produce instances of standard framework classes - e.g. 'ls' produces streams of FileInfo and DirectoryInfo.

The way this interacts with traditional text streams is pretty simple: strings are also objects in .NET (System.String), and every .NET object has a ToString method - so PS can always treat an object stream as text, or text stream as object, as needed. That's pretty much what it does when you have a regular program anywhere in the pipeline, or for the last step in it (the results of which you want to print out).

In practice, it's somewhat more complicated, because ToString does not necessary return a pretty-printed human-readable representation of the object, so it also has its own formatting layer there - the various Format-* cmdlets, like Format-Table (aliased as ft).So you can do things like:

ls | ft name,length -g extension

i.e. print out entries one per line, in two columns representing file name and file length, and grouping them by file extension. If your last step in the pipeline is not ft (or something else that produces a text stream), it tries to guess formatting depending on object type.

The 'where' command in my example does not actually do any parsing. It just takes a predicate function as a parameter, applies it to its input stream, and only outputs those objects that match the predicate. The {} syntax defines a single-parameter lambda function, where $_ is the name of that parameter in the body of the lambda - it's all built-in PS syntax, 'where' doesn't know anything about it.