An honest question: what did Apple contributed back ? I know about Rendezvous/zeroconf, and I suppose they participate actively in KHTML (for Safari), but beside that ? There is obviously Darwin, but it is not being used very much outside of OS X AFAIK.
Also, Apple produce quite a bit of hardware. How are they doing about providing drivers for Open-Source OS, or documentation for these to be written ? For example, are their WiFi adaptor supported in Linux/PPC ?
An even better story will be when folks realize that it is OK for the whole world not to agree with them on philosophy. Especially when those philosophies have economic ramifications.
I wish it was only philosophical issues. Unfortunately, closed-source drivers pose a lot of technical and legal problems, too.
Attachments disappear or become garbled, mail disappears into the great beyond, or arrives hours after it has been sent, even within the same ISP.
I suggest you either switch ISP, or host your own mail server with a competent sysadmin.
Personnally, I never used email as much as I do today. The volume of spam suck, and the time wasted pampering SpamAssassin and other spam counter-measure is insane, but so far the gain still outweight the effort for the vast majority of people apparently.
Can I slap an OC-48 card in a PC? Unlikely. OC-192? How about 10GigE? Again, unlikely. For
some of us, PCs can't compete with dedicated routing hardware.
The keyword here is "some" of us. Personnaly, I work full-time in IT and know of no one using these kind of pipe. So, yes, if you are a telco, a large ISP or run an extremely big network, Cisco is what you need. The rest of us (all 99% of it) can probably get by using software routing on PC hardware.
Ok, could you explain what does email encryption, ssh and VPN have in common (beside encryption) ? Why should they be administered within a single tools instead of many specialized (and probably better suited) tools ?
Personnally, I hate monolithic tools. You are stuck using sub-par components of the suite instead of picking best-of-breed specialized tools.
I am too lazy to check myelf, so I'll ask the collective : does GFS support locking and mmap() ? I am asking because this is a sine qua non condition to run my favorite mail server, Cyrus imapd. Redundant high-availability servers are one of the most asked for scenario. And no, Cyrus Murder don't cut it (it solve a different problem, that is scalability).
For "ground zero" year 1, this is true. But, for year 2, year 3, and so on, the migration and retraining costs are minimal (assuming they stay on linux), and there's *no* cost for software licenses for whatever new versions they *choose* to upgrade to.
More important than the recurrent upgrade licensing cost is the growth cost. You may buy 1000 licenses of MS Office, but if your company grow to 1100 users, you will need to buy 100 more licenses. OO.org ? No problem, bring'em on.
Also, assuming you don't want to upgrade your 1000 users base (because of licensing cost or, gasp!, migration and retraining cost associated with the new version), you will have to either buy more license of the version you use (if it is still available), or have your userbase using different version.
And we don't even talk about the management of your software licenses inventory. And the legal risk associated with error in this inventory (too many users, too few licenses). For some people, closed-source software seem like a no-brainer but I personnally prefer to avoid that bullshit and use OSS instead.
And if you wanted a GUI front-end for package managers, apt-get has a Synaptic,
and I'm sure there are many others for RPM etc.
As good as Synaptic is, it still don't really cut it for "normal" users. Instead of being package-centric, a nice software management GUI should be application centric. It should display available application by name instead of package, and hide non-application package (libs, core OS package, etc) to simplify the presentation. I think Click'N Run from Lindows is something along this line, but I never bothered to try Lindows. Other niceties, such as screenshot, could be integrated in the application installation GUI too.
I think package management, if coupled with such a user-friendly GUI, could be Linux killer apps.
I'm the kind of person that linux is going to have to deal with. I browse the web from root! Try explaining to your boss that, no, infact he cant do X because it violates the security policy, or that he has to change users to install stuff.
Multi-users operating systems, one of the nice computing innovation of the 70s (or was it done in the 60s ?), come with that notion of variable privileges level. I know this is an abstraction that might be hard for the common men to grasp, but computing is already full of abstraction (like hierarchical file system) that we expect normal users to understand. I think it is about time, we IT people, start hammering the notion of privileges in the mind of our users. I suggest you start with your boss.
This trend of having normal user running with administrator privileges to shield them from the annoyance of understanding the reasonning behind privileges level is pathetic. If you are not part of the solution, I can only conclude you are taking part in perpetuating the problem.
- Much of the
cost saving of Linux over Unix comes from hardware - i.e. using Intel over mainframe/AIX/zSeries etc.
Wrong. Go buy a license for 100 mail users, or 50 MSSQL user licenses. OUCH. Now compare Postgres/MySQL or Sendmail/Qmail/god-knows-what-free-email-servers
I have to argue that, despite MS's other claims, I agree that TCO will be higher, primarily because most linux programs require a lot more user support than your average windows program, installed and patched with "software wizards".
Could you explain to us how using "software wizards" instead of package manager (yum, apt, urpmi, whatever your distro is based on) and related GUI for software installation and patching could translate into TCO saving ? Remember, most software installation and patching is not being done by the end-user anyway, but by their IT departement.
I spent four hours writing up a macro-enabled, data-validity-using spreadsheet for my company's linux users, while the identical spreadsheet in Excel took me about 45 minutes, and the linux version just didn't compare, and I'm not even a spreadsheet power user.
If you would be equally proficient with both MS Office and OO.org, that would be telling something. However, I am pretty sure you are not. With that assumption, the only conclusion I can draw ATM is that using software you are not familiar with take more time, especially for advanced stuff like data validation. Duh.
Just checked, and indeed the most interesting bit are covered by the Alladin Free Public License. Some part are covered by the LGPL, though.
As for the patches stuff, I would be interested in knowing what they did and did not contribute back. I kind of recall having read the WineHQ implementation of DirectX contain no Transgaming code, but that they contributed some other stuff. I guess "a lot" is relative.
You fail to take into account that fabulous force called "inertia". Even if all reasons holding back Linux on the desktop (LotD) are solved and it achieve feature/quality parity, people will still continue to use Windows massively. Something about the comfort zone. For people to switch, you would need bring some pretty substantial advantages, not just parity. If people choosed software on quality and features, the world would have switched from IE to Mozilla/Firewhatever a loooooong time ago.
Slashdot Mirror
It does. It's called eval. It is also use for exception handling.
I don't see the point. If you don't execute outside input to your program, how could it be a security issue ?
Oups, I mean 13 GB (damn NumLock never on when we need it !).
Something close to that happened to me. It was a GB Fireball AS. Could it have been a similar model ?
An honest question: what did Apple contributed back ? I know about Rendezvous/zeroconf, and I suppose they participate actively in KHTML (for Safari), but beside that ? There is obviously Darwin, but it is not being used very much outside of OS X AFAIK.
Also, Apple produce quite a bit of hardware. How are they doing about providing drivers for Open-Source OS, or documentation for these to be written ? For example, are their WiFi adaptor supported in Linux/PPC ?
I wish it was only philosophical issues. Unfortunately, closed-source drivers pose a lot of technical and legal problems, too.
Be able to upgrade kernel without worrying if it will break their video card drivers. See 4KSTACK.
Faraday Cage
Now, what is the wavelength of these ultra-wideband thingy ?
What else ?
... Film at 11.
I suggest you either switch ISP, or host your own mail server with a competent sysadmin.
Personnally, I never used email as much as I do today. The volume of spam suck, and the time wasted pampering SpamAssassin and other spam counter-measure is insane, but so far the gain still outweight the effort for the vast majority of people apparently.
The keyword here is "some" of us. Personnaly, I work full-time in IT and know of no one using these kind of pipe. So, yes, if you are a telco, a large ISP or run an extremely big network, Cisco is what you need. The rest of us (all 99% of it) can probably get by using software routing on PC hardware.
Ok, could you explain what does email encryption, ssh and VPN have in common (beside encryption) ? Why should they be administered within a single tools instead of many specialized (and probably better suited) tools ?
Personnally, I hate monolithic tools. You are stuck using sub-par components of the suite instead of picking best-of-breed specialized tools.
I am too lazy to check myelf, so I'll ask the collective : does GFS support locking and mmap() ? I am asking because this is a sine qua non condition to run my favorite mail server, Cyrus imapd. Redundant high-availability servers are one of the most asked for scenario. And no, Cyrus Murder don't cut it (it solve a different problem, that is scalability).
More important than the recurrent upgrade licensing cost is the growth cost. You may buy 1000 licenses of MS Office, but if your company grow to 1100 users, you will need to buy 100 more licenses. OO.org ? No problem, bring'em on.
Also, assuming you don't want to upgrade your 1000 users base (because of licensing cost or, gasp!, migration and retraining cost associated with the new version), you will have to either buy more license of the version you use (if it is still available), or have your userbase using different version.
And we don't even talk about the management of your software licenses inventory. And the legal risk associated with error in this inventory (too many users, too few licenses). For some people, closed-source software seem like a no-brainer but I personnally prefer to avoid that bullshit and use OSS instead.
Hint: I'm a card carrying Linux zealot (TM), and I cut down my posting in the BF a lot (I prefer the more productive LKF now).
...
I'm changing my Tribu right now to leave subtle hint
As good as Synaptic is, it still don't really cut it for "normal" users. Instead of being package-centric, a nice software management GUI should be application centric. It should display available application by name instead of package, and hide non-application package (libs, core OS package, etc) to simplify the presentation. I think Click'N Run from Lindows is something along this line, but I never bothered to try Lindows. Other niceties, such as screenshot, could be integrated in the application installation GUI too.
I think package management, if coupled with such a user-friendly GUI, could be Linux killer apps.
Multi-users operating systems, one of the nice computing innovation of the 70s (or was it done in the 60s ?), come with that notion of variable privileges level. I know this is an abstraction that might be hard for the common men to grasp, but computing is already full of abstraction (like hierarchical file system) that we expect normal users to understand. I think it is about time, we IT people, start hammering the notion of privileges in the mind of our users. I suggest you start with your boss.
This trend of having normal user running with administrator privileges to shield them from the annoyance of understanding the reasonning behind privileges level is pathetic. If you are not part of the solution, I can only conclude you are taking part in perpetuating the problem.
Once you get your hand on it, post about it in the BattleFront.
;)
Nope, I don't use the same handle here and there fellow.
Apple to orange, maybe ?
My opinion exactly, but I'll add the desktop status is somewhat debatable.
And I am a card carying OSS Zealot (TM).
Could you explain to us how using "software wizards" instead of package manager (yum, apt, urpmi, whatever your distro is based on) and related GUI for software installation and patching could translate into TCO saving ? Remember, most software installation and patching is not being done by the end-user anyway, but by their IT departement.
If you would be equally proficient with both MS Office and OO.org, that would be telling something. However, I am pretty sure you are not. With that assumption, the only conclusion I can draw ATM is that using software you are not familiar with take more time, especially for advanced stuff like data validation. Duh.
Off course there's none, but we may ask.
As a side note, which card do people recommend to watch HDTV with Linux ?
I don't agree here. I would buy their packages just for convenience (5$/month ... peanuts) even if the whole thing was LGPLed.
Gotta think about this license thing a little more before I take out the credit card, though ...
Just checked, and indeed the most interesting bit are covered by the Alladin Free Public License. Some part are covered by the LGPL, though.
As for the patches stuff, I would be interested in knowing what they did and did not contribute back. I kind of recall having read the WineHQ implementation of DirectX contain no Transgaming code, but that they contributed some other stuff. I guess "a lot" is relative.
Thanks for the clarification.
You fail to take into account that fabulous force called "inertia". Even if all reasons holding back Linux on the desktop (LotD) are solved and it achieve feature/quality parity, people will still continue to use Windows massively. Something about the comfort zone. For people to switch, you would need bring some pretty substantial advantages, not just parity. If people choosed software on quality and features, the world would have switched from IE to Mozilla/Firewhatever a loooooong time ago.