Granted. Just figuring you understand some of the implications, both ways.
Let's say I want to duplicate Gmail. I'd start with some sort of IMAP server. I'd put it on something like RAID (or better, ZFS), then replicate that setup on at least one other machine -- probably via DRBD. I'd put both of them on a UPS, and I'd still take regular backups, in case people delete something they don't mean to.
Then, I'd add webmail by picking one of several open source projects (the first that comes to mind is Squirrelmail, but I'm sure there are better), which can use IMAP as a storage system.
Then, I'd setup a webserver with webdav enabled, and train people to use Sunbird (or something similar) to share calendars. And a Jabber server, and of course, an SMTP server (probably Postfix).
Then, I'd still be short some of the groupware functionality (how do I send an invitation that someone can click on from their email, that will automatically add the event to their calendar?), and the webmail likely wouldn't be as good, though some users will appreciate the ability to use standard IMAP clients.
All the while, I'd be billing my own hours, and the company would be paying for all the hardware involved -- half of which would necessarily be sitting idle, and possibly more. I don't know what Google charges for corporate-level Google Apps, but I doubt it's more than my salary.
So, much as I'd like that job, this is something I think it makes sense to outsource. Unless you're large enough to run your own datacenter, you'll be outsourcing other things, anyway -- hosting, for example. Is a company based on Amazon Web Services "surrendering all its data"?
(To any recruiters reading this: I actually wouldn't mind that job, even if I don't agree that it's the best approach.)
If the issue is that Gmail might go away, there are backup scripts available, and IMAP access makes it easy to write more -- and if you've got a domain (cheap), you can migrate off Gmail later without having to switch email addresses or lose mail. If the issue is that Internet access might go away, Gmail has an offline feature -- I believe it uses Google Gears. If the issue is security, use https://mail.google.com/ -- you could even block access to it at the corporate firewall/proxy, if you have one.
Now, look at my own email address -- I already have a server I keep in my house, to play with. I put email on it, and bought a domain (when I was 15 -- it seemed cool at the time). I'd rather Google doesn't have my data, and I'd rather have the freedom to add whatever features I can write. So I do see your point.
I realize Gmail has IMAP, didn't mean to imply otherwise. TFA is talking about IMAP support as one of the items that was on the checklist of features Gmail has to have before it can lose the Beta tag.
Sysadmins doing backups is only part of the problem, and convenience is pretty valuable.
I had a longer post written, but then I realized you've got a gmail.com address obfuscated up there. Clearly you think the benefits are worth any real or imagined loss of privacy.
Aside from the fact that it's skinnable, I find the UI actually quite usable. It has most of the same keyboard shortcuts you'd expect in a mail client.
All of those things could easily be done without the beta label. The only way they'd really have much obligation to you is if you were paying them already.
As it is, the IMAP access means that worst case, you could be backing up your data right now and preparing for that worst case.
Frankly, I'd much rather have Outlook be gone, for several reasons:
- Gmail is pretty solidly technically superior, in most of the ways we care about. Example: It doesn't fall over if you put several hundred thousand emails in the same "folder".
- Gmail moves the data off of the end-user's computer. Far, far too many Outlook setups (especially in small businesses) store everything locally, with no backup -- one hard drive crash away from all that archived email gone.
- Gmail is platform-agnostic. It's actually annoyingly browser-aware, but all browsers are supported somewhat, and among the fully-supported browsers are Firefox and Safari, and Gecko and Webkit both exist for every platform I care about. That's one baby-step closer to Linux on the corporate desktop.
- Google actually seems to support open standards -- for example, Gmail includes GTalk, which operates over Jabber. Email is available via IMAP, and calendars via caldav. Contrast to Outlook/Exchange -- the Halloween documents show that Microsoft deliberately chose proprietary protocols, as well as proprietary extensions/perversions of existing protocols.
Now, I'd still prefer we all start improving the existing open implementations, and get to where this is entirely open standard, commodity stuff, just like IMAP and SMTP is today. But Gmail would be a marked improvement over Outlook, in many ways.
Right. Contrast this to something like Google Search, which, on the very few (like, three?) occasions it's ever been down, everyone assumed that it was their own Internet that was at fault.
That's what I would assume the criteria would be -- Gmail will come out of beta when it's as stable as Google's other services that are out of beta.
Of course, TFA seems to be operating under a different definition of "beta". IMAP is certainly a feature I would demand from a service like Gmail, but it really isn't a measure of stability.
I personally don't agree with that ruling, nor do I see how a reasonable person could.
More importantly, it never made it to the supreme court -- in fact, it was settled out of court.
For comparison, the KKK was allowed to have their own TV show. The rules are still very simple: If you don't like a message, speak out against it. If you feel a particular piece of information encourages a particular activity, speak out against that information, and take action against the activity if it's actually illegal.
None of these require actual censorship. Nor is that book necessarily anything other than a work of fiction, meant to entertain, not to encourage you to actually follow that life.
It's not so much the unique number, it's the hassle of typing it. That argument ultimately comes down to: Why are you making me do extra work to "activate" something I already bought?
Slashdot simply assigned me a unique ID. Furthermore, it was free, and it actually provides some value -- it allows me to gain a reputation (good or bad), track relationships (Slashdot was kind of a social network before there was such a thing), moderate occasionally, and customize how I view the site.
The unique ID with the game allows me to... uh... play the game. But wait, didn't I already pay my $50 for that very right -- to play the game?
See, while I may bitch about it, I actually use Steam, and consider it to be a fair trade. That's far more restrictive -- not only is there a unique ID, but I must be online at all times in order to use it.
However, there's something I get in return -- I get to buy the game once online, and then re-download as many times as I want. I get pretty close to no restrictions on copying it -- in fact, this is enabled, and Steam will actually create CD or DVD-sized backup files -- so long as I only run one at the same time.
EA is offering CD keys, with no added value other than "We're not completely raping you with SecuROM anymore! Please stop protesting... we'll be good, we promise..."
They don't care about being spied on, they find the idea that any pleb with an Internet connection can look at their house without their knowledge distasteful.
As opposed to anyone in the police, or anyone with a court order, or anyone who somehow gains access (legitimately or otherwise) to those recordings?
If the government made the results of this (I must say ambiguously defined) spying public in the way Google plans too, expect an uproar.
Because an uproar will solve everything, I'm sure.
The way you avoid that isn't to cause an uproar after the fact. The way you avoid that is to prevent the cameras from being installed in the first place.
Where I faulted IE in the past was when they were dictating de facto web standards. I don't care what they do in a corporate environment.
They still are dictating defacto standards, and it's largely because of that corporate environment. The more people who use IE -- whatever they use it for -- the less chance I have of saying "fuck IE" and just going with standards.
As long as I can't do that -- as long as I need to spend a good 20% of my time on any given project porting it back from Firefox to IE -- it continues to be a bane of my existence.
Even if I wanted to, it's not a win going IE-only -- Firefox has Firebug.
In fact, I understand the need for the integration that IE offers in a corporate environment because I've developed and deployed in such environments.
What does it offer that others don't, for that environment? What does "integration" mean in that context, anyway?
So what do you propose MS do? Open their document format up?
Uh... yes?
Who even knows how possible this is, given possible outside licensing issues and such.
Well, they did provide some amount of documentation for the older formats. The problem remains that they're willing to provide a "WorksLikeFoo" tag, but documenting how it actually works is "beyond the scope of that document". Oh, and there's the fact that their specs are several thousand pages long, as opposed to several hundred.
What might be easier is for them to start using the real standard -- OpenDocument -- and if it really can't provide the features they're whining about, work to improve that one, instead of ditching it entirely for their own "let's serialize the memory structures of Office as XML" format.
The fact remains, however, that word supports open formats.
Like what?
OOXML? We just covered that.
RDF? Word actually manages to screw that up quite a lot. I don't think there's a reasonable expectation, these days, of being able to take an RDF out of Word and opening it in TextEdit on OS X, or in OpenOffice.
Just because they solved it does not by any means make it ideal.
No, but it does mean it's not the end-of-the-world scenario that many like to suggest. Nor is it even necessarily worse than Windows -- I've already shown it's no worse than OS X. How much stuff did Vista break again?
Then there's the nice ability to actually maintain an older version. If you build a full stack -- say you've got some custom embedded system -- you can actually stick to a 2.4, even a 2.2 kernel, so long as you're willing to patch it yourself. I suspect that this is what Microsoft's "shared source" project is about.
PS3: technically superior, more open, better media capability.
Arguable.
"More open"? I don't think being able to run Linux in a hypervisor really counts for much. "Better media capability"? Only because Blu-Ray won.
Xbox 360: easier to develop for.
And let's not forget: Cheaper, and on the market sooner. At one point, you could almost get a 360 and a Wii for the same price as a PS3.
The PS3 would have to be pretty damned compelling to make up for that -- but that technical superiority would have to manifest itself in actual games, this being a console. And there really weren't any exclusives that anyone cared about.
The latter trumps all.
Then how do you explain the success of the Wii?
I would explain it much more simply: Price. The Wii is much cheaper, and provides something compelling and new. The 360 is more expensive, and provides a good HD experience -- plus it ties into Netflix now, I think. The PS3 is even more expensive, and has... well... it can run Linux in a VM, and it can play Blu-Ray discs.
Is the ability to play Blu-Ray discs worth an extra $200? For most people,
Bull. To keep their jobs they have to deliver the same margins, or the investors will punish them.
That assumes that there are investors, and that the same margins are being applied over the same number of people.
That means that if they offer something that costs more for a lower price, jobs have to go and/or salaries have to go down. They aren't going to make it up in volume.
They are if the volume goes away -- if company B provides a more costly service than company A, at a lower price, company B's investors will be patting them on the back as customers flock to them, and company A's investors will want to know why they're losing customers.
Eventually, company A will be forced to adapt or die. If they can't adapt, well, poor buggy-whip manufacturers. Boo fucking hoo.
Landline internet providers struggle with this,
Bullshit. Landline providers "struggle" with how to squeeze the most they can out of what they have, rather than expanding. The ones that actually do expand find that things like net neutrality really aren't an issue.
Information may want to be free, but copper and silicon most assuredly don't.
No, but they do want to get cheaper all the time -- which is also basic economics.
No, there's no such thing as a free lunch. So why should an operator of an antiquated cell network expect a free lunch?
Would you willingly take a paycut doing the same work?? No, then why the fuck do you think they are going to.
No one is suggesting that they (the people) have to take a pay cut. What we are suggesting is that they should catch up with the times, and provide more service in the first place.
Put another way: Working at a local ISP, you're probably doing about the same work for about the same amount of money now as you were five or ten years ago. And in that time, you've seen the bandwidth go from 56k to 100 mbits. Does it mean you're doing more work? No, you're doing less work and providing more value.
That's called productivity.
And sure, everyone would try to do less for the same amount of money if they could. That's called being lazy. I don't think it's a smart move in the long run, though -- as soon as someone is able to provide a dumb pipe at the same or a cheaper price than their charge-by-the-text-message pipes, they'll be forced to adapt or die. That's called competition.
sure, we wan't it to be different, but it isn't and they won't be for as long as they can.
"wan't"? Really?
But no, it's very much the same. If I'd been programming a couple decades ago, I'd probably be doing assembly, or a shiny new high-level language called C. Now, I work in Ruby, and consider C to be too low-level for most tasks.
Is that a pay cut? No, that's a productivity boost.
Sure, I could stick to some old app I wrote in assembly, and I could jack up the price and milk it for all I could -- charging thousands of dollars for each tiny little tweak. And it would work for awhile. Until all my customers went away to the guy down the block who rewrote it in a modern language, charged reasonably for real changes, and added config options so the clients didn't have to call him over every little thing.
Is the guy down the block taking a paycut, in that scenario? Not really. He's probably able to handle twice as many customers as a crusty old assembly guy is -- plus, the old assembler guy just lost all his business.
This doesn't always happen -- see COBOL. But it does happen often enough that it's not unreasonable to expect a similar change to be forced on cell providers.
Adapt or die. It's that simple.
You writing in all caps isn't going to change that one little bit.
You saying fuck every other sentence isn't going to change that one little bit.
Mod me down all you want fuckers, but this is the truth.
Translation: By posting "I know I'll get modded down", you hope to use reverse psychology on modders. What you are really saying is "I know I'm a troll, but please don't mod me down for it!"
With the exception of company intranets, this has been a non-issue for years. Stop clinging to outdated arguments.
So an argument which only applies to corporate America -- that is, to a huge portion of the computers that exist in the US -- is an outdated argument?
OOO has been able to read Word files for YEARS. They just can't render them consistently.
Yes, because being able to read a format is all that's required for an open standard.
Yeah, remember the whole OOXML debacle? The fact that the format can be read, and that we know it's supposed to be "SpacedLikeWordPerfect3.5.1" or whatever, is not the end of the story. It means we still have to pick up a copy of that specific version of Word, or WordPerfect, or Works, or whatever, and reverse engineer that.
Even if this were a non-issue, Word and the Office suite still blows OOO away.
Granted. But the fact is, OOo is a non-starter until this is resolved. I wonder how much development time is spent on this problem of compatibility, instead of actually building a better office suite?
It's about the target environment, which with Linux, you never know exactly what the target is.
Again: This is a solved problem.
How do I know? Because Adobe solved it. Id Software solved it. Epic Games solved it. 2DBoy solved it. Autodesk solved it.
Worst case, you piss off the tiny amount of Linux users who aren't using an x86 architecture. Even there, if you end up caring, it's likely just a recompile.
Do I really need to go on? Stop using outdated arguments.
Then why does Opera have to release a new installation package for each Ubuntu release?
I just installed the latest version in their repository, last modified June 2008, on Intrepid, which was released October of 2008, and which I've been updating since then. It was even the shared version, not the statically linked version, meaning it was using shared libraries that have likely updated since then.
It works pretty flawlessly.
Why don't you ask them why they do release a new version every Ubuntu? They certainly don't have to.
This makes no sense to me whatsoever. Are you implying that these are mutually exclusive??
Pretty much, in a given context. If it's completely outdated technology, like the 1980s-era tools I've pointed to, clearly there is some pretty powerful backwards compatibility happening. In fact, many of us still use tar (designed to work with tapes) to backup.
If, as you imply, backwards compatibility is completely broken, these old tools would stop working.
Otherwise, you're probably talking about some specific place where compatibility is broken. If so, please be explicit -- your one example already fell apart pretty spectacularly.
My point is that Objective-C only makes you marketable in the OS X/iPhone world, whereas Java, C, C++, Python, ruby are entirely agnostic.
That makes sense. It's also a different argument -- "no use" is quite different than "no marketability". Lisp is pretty hard to market, but those who use it seem to find quite a lot of utility.
But that is in no way "lock-in". The fact that others aren't using a clearly open technology in no way makes it "lock-in". The lock-in would be the IP they actually own, like Cocoa, or the little magnetic plug on the Macbook, or the gestures on the iPhone, or the fact that the iPhone must be cracked in order to run apps not blessed by Apple.
Certainly, apple has plenty of lock-in, and tends to be more proprietary than Microsoft. But you picked about the worst example of that -- especially when it compiles just fine on Linux, which is what this discussion was actually about.
If you ignore what the people who write the software tell you, it's your fault, not theirs.
And whose fault is it that it wasn't called 4.0 beta? Serously, how fucking hard is it to put a "beta" tag on it? Not that it was even beta quality, at that point...
Never mind that by 4.1, it's supposed to be end-user ready, and was shipped with Kubuntu, yet is still missing large chunks of functionality. 4.0 was just the most blatant form of that.
Another thing I noticed was that whenever I clicked something in that horrible Dolphin manager, Amarok would start up and start playing music for no good reason.
KDE defaults to a single click to open things. So if you were clicking a music file, that makes sense. If you wanted to do something other than open it, right-click.
I never got used to not being able to have icons on the desktop
There was always the ability to put a file manager widget on the desktop, and it does default to showing you the same Desktop folder GNOME does. It is now possible to use that as the background, though I'm not sure if you can put widgets on top of it.
in general, it was just very slow.
On older hardware, turn off compositing. On newer hardware, I haven't seen an issue.
But then, this one is hard to argue. If it's slow for you, it's slow.
Well, firstly, they called it 4.0. A dot-oh release generally means stable. Why couldn't they call it 3.9, or 4.0 alpha?
And second, 4.1 sucked quite a lot as well -- yet that is what Kubuntu Intrepid shipped with. And that's not "for developers" -- there was no KDE3 version of Kubuntu to download.
I don't care much about which is used -- arguably, a revolution was needed. I have only two observations here:
First, KDE4 changed quite a lot, but things could still segfault. There are languages which don't segfault, and I'd really prefer a rock solid GUI in something like Ruby or Python than a lightning-fast, unstable GUI in C++.
And second, KDE4 managed to completely fuck up the accepted convention of stable/unstable version numbers, and generally fuck up the transition period. Users were confused about which version they should use -- KDE3 was hopelessly out of date, but KDE4 wasn't nearly ready, and the same could be said for just about every KDE app, especially Amarok.
Seriously, the Linux 2.4 to 2.6 kernel upgrade was less painful. Why? Because the features people really needed from 2.6 were backported to 2.4, so no one felt "forced" to upgrade. And because they waited until 2.6 was done before releasing it -- they kept it as the clear development version (2.5) until then.
The rest of your comment is kind of rambling and kind of irrelevant. My windows is pretty solid, though not because of antivirus. But you'd have to pay me a lot to force me to use it as a primary OS.
Incidentally, one of the apps that just does not have a Linux equivalent is Quicken.
There's Gnucash and KMyMoney, but you're right. It's just not there yet.
Last time I looked, this app doesn't run in Wine.
When was that? Last I looked, it worked flawlessly in Wine.
But then, Quicken isn't the kind of thing you need to run all the time. I could live with booting Windows once a week, and living in Linux the rest of the time.
Can't say IE, because Firefox is available elsewhere.
IE is shipped out of the box, and some websites are IE-only. That is finally shifting, recently.
Can't say Microsoft Office, simply because there is nothing better out there.
You might argue that OpenOffice isn't better, but right now, it simply doesn't have a chance, when so much development time is spent reverse engineering old Windows formats.
Linux doesn't care about developers
I would think a developer would care about being able to completely customize their environment, in pretty much any way they want.
There is also the fact developing for Linux is a pain in the ass since there is no standard development stack.
Why does there have to be? Your basic components are there -- if your app links against glibc and runs on X, it'll work.
Oh, you want to talk about GNOME/KDE? Why does that matter? KDE apps run on GNOME, and vice versa. That's not a pain in the ass, that's choice, and you have the same kind of choice with Windows -- you're not forced to use MFC.
Or distribution.
Ubuntu.
An ISV makes a program, they don't want it breaking with every cleverly named alliteration release of your operating system.
And yet, somehow, most of the commercial programs which have been released haven't broken because of that.
In fact, in one breath, you whine that developers are forced to use "hard-to-learn and outdated technology", and in the other, you whine that backwards compatibility is broken. Which is it? I'll bet not a line of the code for the 'cp' command has changed in decades, so why should an ISV be worried?
On the other hand, most OS X programs at least require Tiger, if not Leopard. New OS X releases frequently break tons of apps.
Even Microsoft's own programs can break from one version of Windows to another. Example: Remember HD-DVD? Vista was already out when I started working on it, but HDiSim, Microsoft's HD-DVD software simulator, only ran on XP. Not 2k, not Vista, had to be XP. Oh, and it only worked on a specific version of Windows Media Player -- I believe it was 10. Not 9, not 11.
And this was pretty big, by the way -- HD-DVD was arguably a flagship product. If even Microsoft can't maintain backwards compatibility on Windows, yet all my old Linux apps still work, I don't know WTF you're whining about.
And, in Apples case, you're made to learn a crappy language that has absolutely no use outside Mac OS and Iphone development.
With how coherent the rest of your comment is, I don't know if I'd call it "crappy" or not, but it certainly works on Windows. If this particular whine is about Cocoa not being portable, well, neither is Win32, unless you're counting Wine -- yet, surprisingly, it's possible to write portable apps in C, even C++.
First, I hear that Linux is easy to install and that anyone can jump into it at will.
Linux is by far the easiest to install OS. However, installing an OS is not an easy task.
Then I hear from guys like you that people DON'T have a choice and that they aren't jumping to Linux because of the evil monopoly.
The preinstallation is only part of that. The other part is the tangled web of proprietary technologies that Linux would have to duplicate, perfectly, including the Windows API.
Let me put it this way: How is it Linux' fault that Quickbooks only exists for Windows and OS X?
Now, anyone can still jump in and try it out. For most people, Linux is 90% of the way there. The problem is, the last 10% is different for everyone. If it's something like Quickbooks, and it doesn't work flawlessly under Wine, then you need Windows, at least in a virtual machine. If it's something like Outlook, there's always the chance you could get people to switch to Kmail or Thunderbird.
For me, that last 10% is gaming -- I need Windows, because I'll never be able to count on Wine to give me as good an experience as XP. But I only need to boot Windows when playing a game that won't run on Linux, so that's good enough.
If you're offering a completely free operating system and people STILL aren't biting,
That's the other problem: It's seen as exactly as free as Windows. Highly technical people might tend to pirate Windows. Less technical people will buy a computer with Windows preloaded -- chances are, they didn't see an option to remove Windows and save themselves $100 or so -- so it's a hidden cost.
Which brings us back to that "contradictory" view -- Linux will only be seen as "free" when you have a company selling the same laptop with Linux and with Windows, and the Linux version is cheaper. And even when Dell does that, it's not the same laptop -- I'm talking about putting Linux as an option on the configuration screen. That is, buy this laptop with Vista Home Premium, Vista Ultimate, Ubuntu, or FreeDOS.
And all that does is really demonstrate to people that it's free. It still has to be shown as much better -- being "just as good" isn't enough.
Vendors who entered into exclusivity agreement with Microsoft did so of their own accord, they were not forced by anything other than market conditions.
Market conditions which Microsoft loves to support, of course, through means legal and otherwise.
The fact that it was caused by market conditions does not make it alright. Even if you make a good case that it's not Microsoft's fault, that does not make it alright.
Many many people do not have any aptitude at all for such things regardless of how simple they are.
That's mostly because they get hand-held all the time -- because we continue to support that some people are incapable of grasping computers, or at the very least, that they shouldn't have to, and that everyone should have their own personal IT department.
Obligatory car analogy: I don't know how to rebuild an engine, but I do at least know that I need to change my oil periodically, and I can find out how to do it myself if I really care.
This is the biggest lie Linux supporters tell. If Linux needs no fixing,
I don't think the point is that Linux needs no fixing. I think the point is that for those of us who are much more familiar with Linux, it is easier to fix a Linux problem than a Windows problem. I would argue that this is generally true, but I don't know enough about Windows to make that statement.
I do know that I am not willing to service Windows computers for free.
However, if we're playing anecdotes, here's mine:
I bought a Dell with Ubuntu on it. I tried to install XP.
Now, as I'm sure yo know, the usual way to install Windows drivers is to go to a vendor's website, download the driver, and install. This can take an hour or so under the best of conditions, due to the sheer amount of hardware that isn't supported out of the box by XP. Often, I find myself booting a Linux CD, so I can run lspci, which will at least tell me the brand of the hardware that I have -- the Windows device manager won't do that -- so I can then boot Windows, visit those manufacturers' websites, and download drivers.
However, for some reason, Dell laptops are worse. They've entered into a deal with nVidia that Dell is the only one who gets to distribute the nVidia drivers for this particular laptop. On Linux, I just install the normal nvidia drivers. On Windows, I have to get them from Dell.
Only problem is, XP isn't supported on this laptop. Vista is.
So I had to contact Dell tech support, which gave me all the links to drivers known to work with XP on this laptop. Most of them were from other laptops in the same series, a few were for other OSes, and it generally had a hacked-together feel, but it worked.
Let's compare to Linux. I did wipe the factory-installed Ubuntu with a 64-bit version. I vaguely remember having exactly one problem, with the touchpad, which a few minutes on Google corrected. Other than that, It Just Worked.
So yes, Linux needed fixing. So did Windows. And Linux needed a lot less fixing than Windows.
Slashdot Mirror
Granted. Just figuring you understand some of the implications, both ways.
Let's say I want to duplicate Gmail. I'd start with some sort of IMAP server. I'd put it on something like RAID (or better, ZFS), then replicate that setup on at least one other machine -- probably via DRBD. I'd put both of them on a UPS, and I'd still take regular backups, in case people delete something they don't mean to.
Then, I'd add webmail by picking one of several open source projects (the first that comes to mind is Squirrelmail, but I'm sure there are better), which can use IMAP as a storage system.
Then, I'd setup a webserver with webdav enabled, and train people to use Sunbird (or something similar) to share calendars. And a Jabber server, and of course, an SMTP server (probably Postfix).
Then, I'd still be short some of the groupware functionality (how do I send an invitation that someone can click on from their email, that will automatically add the event to their calendar?), and the webmail likely wouldn't be as good, though some users will appreciate the ability to use standard IMAP clients.
All the while, I'd be billing my own hours, and the company would be paying for all the hardware involved -- half of which would necessarily be sitting idle, and possibly more. I don't know what Google charges for corporate-level Google Apps, but I doubt it's more than my salary.
So, much as I'd like that job, this is something I think it makes sense to outsource. Unless you're large enough to run your own datacenter, you'll be outsourcing other things, anyway -- hosting, for example. Is a company based on Amazon Web Services "surrendering all its data"?
(To any recruiters reading this: I actually wouldn't mind that job, even if I don't agree that it's the best approach.)
If the issue is that Gmail might go away, there are backup scripts available, and IMAP access makes it easy to write more -- and if you've got a domain (cheap), you can migrate off Gmail later without having to switch email addresses or lose mail. If the issue is that Internet access might go away, Gmail has an offline feature -- I believe it uses Google Gears. If the issue is security, use https://mail.google.com/ -- you could even block access to it at the corporate firewall/proxy, if you have one.
Now, look at my own email address -- I already have a server I keep in my house, to play with. I put email on it, and bought a domain (when I was 15 -- it seemed cool at the time). I'd rather Google doesn't have my data, and I'd rather have the freedom to add whatever features I can write. So I do see your point.
I realize Gmail has IMAP, didn't mean to imply otherwise. TFA is talking about IMAP support as one of the items that was on the checklist of features Gmail has to have before it can lose the Beta tag.
Sysadmins doing backups is only part of the problem, and convenience is pretty valuable.
I had a longer post written, but then I realized you've got a gmail.com address obfuscated up there. Clearly you think the benefits are worth any real or imagined loss of privacy.
Aside from the fact that it's skinnable, I find the UI actually quite usable. It has most of the same keyboard shortcuts you'd expect in a mail client.
All of those things could easily be done without the beta label. The only way they'd really have much obligation to you is if you were paying them already.
As it is, the IMAP access means that worst case, you could be backing up your data right now and preparing for that worst case.
Frankly, I'd much rather have Outlook be gone, for several reasons:
- Gmail is pretty solidly technically superior, in most of the ways we care about. Example: It doesn't fall over if you put several hundred thousand emails in the same "folder".
- Gmail moves the data off of the end-user's computer. Far, far too many Outlook setups (especially in small businesses) store everything locally, with no backup -- one hard drive crash away from all that archived email gone.
- Gmail is platform-agnostic. It's actually annoyingly browser-aware, but all browsers are supported somewhat, and among the fully-supported browsers are Firefox and Safari, and Gecko and Webkit both exist for every platform I care about. That's one baby-step closer to Linux on the corporate desktop.
- Google actually seems to support open standards -- for example, Gmail includes GTalk, which operates over Jabber. Email is available via IMAP, and calendars via caldav. Contrast to Outlook/Exchange -- the Halloween documents show that Microsoft deliberately chose proprietary protocols, as well as proprietary extensions/perversions of existing protocols.
Now, I'd still prefer we all start improving the existing open implementations, and get to where this is entirely open standard, commodity stuff, just like IMAP and SMTP is today. But Gmail would be a marked improvement over Outlook, in many ways.
Right. Contrast this to something like Google Search, which, on the very few (like, three?) occasions it's ever been down, everyone assumed that it was their own Internet that was at fault.
That's what I would assume the criteria would be -- Gmail will come out of beta when it's as stable as Google's other services that are out of beta.
Of course, TFA seems to be operating under a different definition of "beta". IMAP is certainly a feature I would demand from a service like Gmail, but it really isn't a measure of stability.
I personally don't agree with that ruling, nor do I see how a reasonable person could.
More importantly, it never made it to the supreme court -- in fact, it was settled out of court.
For comparison, the KKK was allowed to have their own TV show. The rules are still very simple: If you don't like a message, speak out against it. If you feel a particular piece of information encourages a particular activity, speak out against that information, and take action against the activity if it's actually illegal.
None of these require actual censorship. Nor is that book necessarily anything other than a work of fiction, meant to entertain, not to encourage you to actually follow that life.
It's not so much the unique number, it's the hassle of typing it. That argument ultimately comes down to: Why are you making me do extra work to "activate" something I already bought?
Slashdot simply assigned me a unique ID. Furthermore, it was free, and it actually provides some value -- it allows me to gain a reputation (good or bad), track relationships (Slashdot was kind of a social network before there was such a thing), moderate occasionally, and customize how I view the site.
The unique ID with the game allows me to... uh... play the game. But wait, didn't I already pay my $50 for that very right -- to play the game?
See, while I may bitch about it, I actually use Steam, and consider it to be a fair trade. That's far more restrictive -- not only is there a unique ID, but I must be online at all times in order to use it.
However, there's something I get in return -- I get to buy the game once online, and then re-download as many times as I want. I get pretty close to no restrictions on copying it -- in fact, this is enabled, and Steam will actually create CD or DVD-sized backup files -- so long as I only run one at the same time.
EA is offering CD keys, with no added value other than "We're not completely raping you with SecuROM anymore! Please stop protesting... we'll be good, we promise..."
They don't care about being spied on, they find the idea that any pleb with an Internet connection can look at their house without their knowledge distasteful.
As opposed to anyone in the police, or anyone with a court order, or anyone who somehow gains access (legitimately or otherwise) to those recordings?
If the government made the results of this (I must say ambiguously defined) spying public in the way Google plans too, expect an uproar.
Because an uproar will solve everything, I'm sure.
The way you avoid that isn't to cause an uproar after the fact. The way you avoid that is to prevent the cameras from being installed in the first place.
Well, you're expecting them to be consistent in their paranoia.
After all, their government has been spying on them considerably more than Google, and it's Google they run out of town?
Where I faulted IE in the past was when they were dictating de facto web standards. I don't care what they do in a corporate environment.
They still are dictating defacto standards, and it's largely because of that corporate environment. The more people who use IE -- whatever they use it for -- the less chance I have of saying "fuck IE" and just going with standards.
As long as I can't do that -- as long as I need to spend a good 20% of my time on any given project porting it back from Firefox to IE -- it continues to be a bane of my existence.
Even if I wanted to, it's not a win going IE-only -- Firefox has Firebug.
In fact, I understand the need for the integration that IE offers in a corporate environment because I've developed and deployed in such environments.
What does it offer that others don't, for that environment? What does "integration" mean in that context, anyway?
So what do you propose MS do? Open their document format up?
Uh... yes?
Who even knows how possible this is, given possible outside licensing issues and such.
Well, they did provide some amount of documentation for the older formats. The problem remains that they're willing to provide a "WorksLikeFoo" tag, but documenting how it actually works is "beyond the scope of that document". Oh, and there's the fact that their specs are several thousand pages long, as opposed to several hundred.
What might be easier is for them to start using the real standard -- OpenDocument -- and if it really can't provide the features they're whining about, work to improve that one, instead of ditching it entirely for their own "let's serialize the memory structures of Office as XML" format.
The fact remains, however, that word supports open formats.
Like what?
OOXML? We just covered that.
RDF? Word actually manages to screw that up quite a lot. I don't think there's a reasonable expectation, these days, of being able to take an RDF out of Word and opening it in TextEdit on OS X, or in OpenOffice.
Just because they solved it does not by any means make it ideal.
No, but it does mean it's not the end-of-the-world scenario that many like to suggest. Nor is it even necessarily worse than Windows -- I've already shown it's no worse than OS X. How much stuff did Vista break again?
Then there's the nice ability to actually maintain an older version. If you build a full stack -- say you've got some custom embedded system -- you can actually stick to a 2.4, even a 2.2 kernel, so long as you're willing to patch it yourself. I suspect that this is what Microsoft's "shared source" project is about.
PS3: technically superior, more open, better media capability.
Arguable.
"More open"? I don't think being able to run Linux in a hypervisor really counts for much. "Better media capability"? Only because Blu-Ray won.
Xbox 360: easier to develop for.
And let's not forget: Cheaper, and on the market sooner. At one point, you could almost get a 360 and a Wii for the same price as a PS3.
The PS3 would have to be pretty damned compelling to make up for that -- but that technical superiority would have to manifest itself in actual games, this being a console. And there really weren't any exclusives that anyone cared about.
The latter trumps all.
Then how do you explain the success of the Wii?
I would explain it much more simply: Price. The Wii is much cheaper, and provides something compelling and new. The 360 is more expensive, and provides a good HD experience -- plus it ties into Netflix now, I think. The PS3 is even more expensive, and has... well... it can run Linux in a VM, and it can play Blu-Ray discs.
Is the ability to play Blu-Ray discs worth an extra $200? For most people,
Bull. To keep their jobs they have to deliver the same margins, or the investors will punish them.
That assumes that there are investors, and that the same margins are being applied over the same number of people.
That means that if they offer something that costs more for a lower price, jobs have to go and/or salaries have to go down. They aren't going to make it up in volume.
They are if the volume goes away -- if company B provides a more costly service than company A, at a lower price, company B's investors will be patting them on the back as customers flock to them, and company A's investors will want to know why they're losing customers.
Eventually, company A will be forced to adapt or die. If they can't adapt, well, poor buggy-whip manufacturers. Boo fucking hoo.
Landline internet providers struggle with this,
Bullshit. Landline providers "struggle" with how to squeeze the most they can out of what they have, rather than expanding. The ones that actually do expand find that things like net neutrality really aren't an issue.
Information may want to be free, but copper and silicon most assuredly don't.
No, but they do want to get cheaper all the time -- which is also basic economics.
No, there's no such thing as a free lunch. So why should an operator of an antiquated cell network expect a free lunch?
Would you willingly take a paycut doing the same work?? No, then why the fuck do you think they are going to.
No one is suggesting that they (the people) have to take a pay cut. What we are suggesting is that they should catch up with the times, and provide more service in the first place.
Put another way: Working at a local ISP, you're probably doing about the same work for about the same amount of money now as you were five or ten years ago. And in that time, you've seen the bandwidth go from 56k to 100 mbits. Does it mean you're doing more work? No, you're doing less work and providing more value.
That's called productivity.
And sure, everyone would try to do less for the same amount of money if they could. That's called being lazy. I don't think it's a smart move in the long run, though -- as soon as someone is able to provide a dumb pipe at the same or a cheaper price than their charge-by-the-text-message pipes, they'll be forced to adapt or die. That's called competition.
sure, we wan't it to be different, but it isn't and they won't be for as long as they can.
"wan't"? Really?
But no, it's very much the same. If I'd been programming a couple decades ago, I'd probably be doing assembly, or a shiny new high-level language called C. Now, I work in Ruby, and consider C to be too low-level for most tasks.
Is that a pay cut? No, that's a productivity boost.
Sure, I could stick to some old app I wrote in assembly, and I could jack up the price and milk it for all I could -- charging thousands of dollars for each tiny little tweak. And it would work for awhile. Until all my customers went away to the guy down the block who rewrote it in a modern language, charged reasonably for real changes, and added config options so the clients didn't have to call him over every little thing.
Is the guy down the block taking a paycut, in that scenario? Not really. He's probably able to handle twice as many customers as a crusty old assembly guy is -- plus, the old assembler guy just lost all his business.
This doesn't always happen -- see COBOL. But it does happen often enough that it's not unreasonable to expect a similar change to be forced on cell providers.
Adapt or die. It's that simple.
You writing in all caps isn't going to change that one little bit.
You saying fuck every other sentence isn't going to change that one little bit.
Mod me down all you want fuckers, but this is the truth.
Translation: By posting "I know I'll get modded down", you hope to use reverse psychology on modders. What you are really saying is "I know I'm a troll, but please don't mod me down for it!"
With the exception of company intranets, this has been a non-issue for years. Stop clinging to outdated arguments.
So an argument which only applies to corporate America -- that is, to a huge portion of the computers that exist in the US -- is an outdated argument?
OOO has been able to read Word files for YEARS. They just can't render them consistently.
Yes, because being able to read a format is all that's required for an open standard.
Yeah, remember the whole OOXML debacle? The fact that the format can be read, and that we know it's supposed to be "SpacedLikeWordPerfect3.5.1" or whatever, is not the end of the story. It means we still have to pick up a copy of that specific version of Word, or WordPerfect, or Works, or whatever, and reverse engineer that.
Even if this were a non-issue, Word and the Office suite still blows OOO away.
Granted. But the fact is, OOo is a non-starter until this is resolved. I wonder how much development time is spent on this problem of compatibility, instead of actually building a better office suite?
It's about the target environment, which with Linux, you never know exactly what the target is.
Again: This is a solved problem.
How do I know? Because Adobe solved it. Id Software solved it. Epic Games solved it. 2DBoy solved it. Autodesk solved it.
Worst case, you piss off the tiny amount of Linux users who aren't using an x86 architecture. Even there, if you end up caring, it's likely just a recompile.
Do I really need to go on? Stop using outdated arguments.
Then why does Opera have to release a new installation package for each Ubuntu release?
I just installed the latest version in their repository, last modified June 2008, on Intrepid, which was released October of 2008, and which I've been updating since then. It was even the shared version, not the statically linked version, meaning it was using shared libraries that have likely updated since then.
It works pretty flawlessly.
Why don't you ask them why they do release a new version every Ubuntu? They certainly don't have to.
This makes no sense to me whatsoever. Are you implying that these are mutually exclusive??
Pretty much, in a given context. If it's completely outdated technology, like the 1980s-era tools I've pointed to, clearly there is some pretty powerful backwards compatibility happening. In fact, many of us still use tar (designed to work with tapes) to backup.
If, as you imply, backwards compatibility is completely broken, these old tools would stop working.
Otherwise, you're probably talking about some specific place where compatibility is broken. If so, please be explicit -- your one example already fell apart pretty spectacularly.
My point is that Objective-C only makes you marketable in the OS X/iPhone world, whereas Java, C, C++, Python, ruby are entirely agnostic.
That makes sense. It's also a different argument -- "no use" is quite different than "no marketability". Lisp is pretty hard to market, but those who use it seem to find quite a lot of utility.
But that is in no way "lock-in". The fact that others aren't using a clearly open technology in no way makes it "lock-in". The lock-in would be the IP they actually own, like Cocoa, or the little magnetic plug on the Macbook, or the gestures on the iPhone, or the fact that the iPhone must be cracked in order to run apps not blessed by Apple.
Certainly, apple has plenty of lock-in, and tends to be more proprietary than Microsoft. But you picked about the worst example of that -- especially when it compiles just fine on Linux, which is what this discussion was actually about.
If you ignore what the people who write the software tell you, it's your fault, not theirs.
And whose fault is it that it wasn't called 4.0 beta? Serously, how fucking hard is it to put a "beta" tag on it? Not that it was even beta quality, at that point...
Never mind that by 4.1, it's supposed to be end-user ready, and was shipped with Kubuntu, yet is still missing large chunks of functionality. 4.0 was just the most blatant form of that.
Another thing I noticed was that whenever I clicked something in that horrible Dolphin manager, Amarok would start up and start playing music for no good reason.
KDE defaults to a single click to open things. So if you were clicking a music file, that makes sense. If you wanted to do something other than open it, right-click.
I never got used to not being able to have icons on the desktop
There was always the ability to put a file manager widget on the desktop, and it does default to showing you the same Desktop folder GNOME does. It is now possible to use that as the background, though I'm not sure if you can put widgets on top of it.
in general, it was just very slow.
On older hardware, turn off compositing. On newer hardware, I haven't seen an issue.
But then, this one is hard to argue. If it's slow for you, it's slow.
Well, firstly, they called it 4.0. A dot-oh release generally means stable. Why couldn't they call it 3.9, or 4.0 alpha?
And second, 4.1 sucked quite a lot as well -- yet that is what Kubuntu Intrepid shipped with. And that's not "for developers" -- there was no KDE3 version of Kubuntu to download.
So, was there anything frustrating you other than this one feature, which xfce lacks anyway?
I don't care much about which is used -- arguably, a revolution was needed. I have only two observations here:
First, KDE4 changed quite a lot, but things could still segfault. There are languages which don't segfault, and I'd really prefer a rock solid GUI in something like Ruby or Python than a lightning-fast, unstable GUI in C++.
And second, KDE4 managed to completely fuck up the accepted convention of stable/unstable version numbers, and generally fuck up the transition period. Users were confused about which version they should use -- KDE3 was hopelessly out of date, but KDE4 wasn't nearly ready, and the same could be said for just about every KDE app, especially Amarok.
Seriously, the Linux 2.4 to 2.6 kernel upgrade was less painful. Why? Because the features people really needed from 2.6 were backported to 2.4, so no one felt "forced" to upgrade. And because they waited until 2.6 was done before releasing it -- they kept it as the clear development version (2.5) until then.
Laugh if you want, bur IE does qualify as professional software that does not run under wine.
Ok, I'll laugh.
Yes, you can run IE6 perhaps, but not IE7 or 8.
Actually, you can.
The rest of your comment is kind of rambling and kind of irrelevant. My windows is pretty solid, though not because of antivirus. But you'd have to pay me a lot to force me to use it as a primary OS.
Incidentally, one of the apps that just does not have a Linux equivalent is Quicken.
There's Gnucash and KMyMoney, but you're right. It's just not there yet.
Last time I looked, this app doesn't run in Wine.
When was that? Last I looked, it worked flawlessly in Wine.
But then, Quicken isn't the kind of thing you need to run all the time. I could live with booting Windows once a week, and living in Linux the rest of the time.
Can't say IE, because Firefox is available elsewhere.
IE is shipped out of the box, and some websites are IE-only. That is finally shifting, recently.
Can't say Microsoft Office, simply because there is nothing better out there.
You might argue that OpenOffice isn't better, but right now, it simply doesn't have a chance, when so much development time is spent reverse engineering old Windows formats.
Linux doesn't care about developers
I would think a developer would care about being able to completely customize their environment, in pretty much any way they want.
There is also the fact developing for Linux is a pain in the ass since there is no standard development stack.
Why does there have to be? Your basic components are there -- if your app links against glibc and runs on X, it'll work.
Oh, you want to talk about GNOME/KDE? Why does that matter? KDE apps run on GNOME, and vice versa. That's not a pain in the ass, that's choice, and you have the same kind of choice with Windows -- you're not forced to use MFC.
Or distribution.
Ubuntu.
An ISV makes a program, they don't want it breaking with every cleverly named alliteration release of your operating system.
And yet, somehow, most of the commercial programs which have been released haven't broken because of that.
In fact, in one breath, you whine that developers are forced to use "hard-to-learn and outdated technology", and in the other, you whine that backwards compatibility is broken. Which is it? I'll bet not a line of the code for the 'cp' command has changed in decades, so why should an ISV be worried?
On the other hand, most OS X programs at least require Tiger, if not Leopard. New OS X releases frequently break tons of apps.
Even Microsoft's own programs can break from one version of Windows to another. Example: Remember HD-DVD? Vista was already out when I started working on it, but HDiSim, Microsoft's HD-DVD software simulator, only ran on XP. Not 2k, not Vista, had to be XP. Oh, and it only worked on a specific version of Windows Media Player -- I believe it was 10. Not 9, not 11.
And this was pretty big, by the way -- HD-DVD was arguably a flagship product. If even Microsoft can't maintain backwards compatibility on Windows, yet all my old Linux apps still work, I don't know WTF you're whining about.
And, in Apples case, you're made to learn a crappy language that has absolutely no use outside Mac OS and Iphone development.
With how coherent the rest of your comment is, I don't know if I'd call it "crappy" or not, but it certainly works on Windows. If this particular whine is about Cocoa not being portable, well, neither is Win32, unless you're counting Wine -- yet, surprisingly, it's possible to write portable apps in C, even C++.
First, I hear that Linux is easy to install and that anyone can jump into it at will.
Linux is by far the easiest to install OS. However, installing an OS is not an easy task.
Then I hear from guys like you that people DON'T have a choice and that they aren't jumping to Linux because of the evil monopoly.
The preinstallation is only part of that. The other part is the tangled web of proprietary technologies that Linux would have to duplicate, perfectly, including the Windows API.
Let me put it this way: How is it Linux' fault that Quickbooks only exists for Windows and OS X?
Now, anyone can still jump in and try it out. For most people, Linux is 90% of the way there. The problem is, the last 10% is different for everyone. If it's something like Quickbooks, and it doesn't work flawlessly under Wine, then you need Windows, at least in a virtual machine. If it's something like Outlook, there's always the chance you could get people to switch to Kmail or Thunderbird.
For me, that last 10% is gaming -- I need Windows, because I'll never be able to count on Wine to give me as good an experience as XP. But I only need to boot Windows when playing a game that won't run on Linux, so that's good enough.
If you're offering a completely free operating system and people STILL aren't biting,
That's the other problem: It's seen as exactly as free as Windows. Highly technical people might tend to pirate Windows. Less technical people will buy a computer with Windows preloaded -- chances are, they didn't see an option to remove Windows and save themselves $100 or so -- so it's a hidden cost.
Which brings us back to that "contradictory" view -- Linux will only be seen as "free" when you have a company selling the same laptop with Linux and with Windows, and the Linux version is cheaper. And even when Dell does that, it's not the same laptop -- I'm talking about putting Linux as an option on the configuration screen. That is, buy this laptop with Vista Home Premium, Vista Ultimate, Ubuntu, or FreeDOS.
And all that does is really demonstrate to people that it's free. It still has to be shown as much better -- being "just as good" isn't enough.
Vendors who entered into exclusivity agreement with Microsoft did so of their own accord, they were not forced by anything other than market conditions.
Market conditions which Microsoft loves to support, of course, through means legal and otherwise.
The fact that it was caused by market conditions does not make it alright. Even if you make a good case that it's not Microsoft's fault, that does not make it alright.
Many many people do not have any aptitude at all for such things regardless of how simple they are.
That's mostly because they get hand-held all the time -- because we continue to support that some people are incapable of grasping computers, or at the very least, that they shouldn't have to, and that everyone should have their own personal IT department.
Obligatory car analogy: I don't know how to rebuild an engine, but I do at least know that I need to change my oil periodically, and I can find out how to do it myself if I really care.
This is the biggest lie Linux supporters tell. If Linux needs no fixing,
I don't think the point is that Linux needs no fixing. I think the point is that for those of us who are much more familiar with Linux, it is easier to fix a Linux problem than a Windows problem. I would argue that this is generally true, but I don't know enough about Windows to make that statement.
I do know that I am not willing to service Windows computers for free.
However, if we're playing anecdotes, here's mine:
I bought a Dell with Ubuntu on it. I tried to install XP.
Now, as I'm sure yo know, the usual way to install Windows drivers is to go to a vendor's website, download the driver, and install. This can take an hour or so under the best of conditions, due to the sheer amount of hardware that isn't supported out of the box by XP. Often, I find myself booting a Linux CD, so I can run lspci, which will at least tell me the brand of the hardware that I have -- the Windows device manager won't do that -- so I can then boot Windows, visit those manufacturers' websites, and download drivers.
However, for some reason, Dell laptops are worse. They've entered into a deal with nVidia that Dell is the only one who gets to distribute the nVidia drivers for this particular laptop. On Linux, I just install the normal nvidia drivers. On Windows, I have to get them from Dell.
Only problem is, XP isn't supported on this laptop. Vista is.
So I had to contact Dell tech support, which gave me all the links to drivers known to work with XP on this laptop. Most of them were from other laptops in the same series, a few were for other OSes, and it generally had a hacked-together feel, but it worked.
Let's compare to Linux. I did wipe the factory-installed Ubuntu with a 64-bit version. I vaguely remember having exactly one problem, with the touchpad, which a few minutes on Google corrected. Other than that, It Just Worked.
So yes, Linux needed fixing. So did Windows. And Linux needed a lot less fixing than Windows.