lunches are catered every day by a chef on staff. Not only you no-longer have to spend money on lunch, you get to eat some really good food. so i hear. they might even cover breakfast and whatnot. they also have naked chicks running around and giggling. err. scratch that last one.
... at least in FireFox. Not sure about Win/IE. Google/ig has a same concept, in that things reflow as you drag, but i personally find that to be a usability bug. It offends me. Google was more subtle in making you drag an outline, which i find a bit less confusing than the actual cloned Node you're dragging on the MSN counterpart. I'm pretty sure both google.com/ig and start are adopting this concept due to technical limitations. Doing drag and drop correctly is a sizable challenge. Note how neither MSN or google make you drag widgets that were below the fold. The reflowing 3-column layout you're looking at, is as such, i surmise, for many reasons not the least of which have to do with accurately calculating mouse and object position on a global document, rather than within the confines of a window object. Also dragging objects beyond the folds would ideally require automated scrolling. Not entirely trivial either to make it all work in Safari, Firefox and IE. But possible:) I find it easier to develop in Safari first by sticking to standards. Making various things work in Firefox and then IE, in that order, is typically a matter of minor incremental tweaks. In most cases, if it works in Safari, it works in FF and IE.
moderators: please mod this up. it's an important distinction. I don't believe a single methodology will fight spam. But a combination of methodologies will.
... can be found in the comments section of this blog post. Executive summary: "organizations" have exactly NOTHING intelligent to contribute because their premises are flawed, and fail to take into account the fact that economics of such projects vary vastly from one community to another.
All very valid points. Skype may very-well for now be very nicely established.
When Hybrid WiFi/GSM/3G handheld devices come to market with the capability of doing STUN+SIP+RTP(encrypted or not), things will likely change. Take a service like Vonage. They've recently released their own wireless handheld device that allows you to use your Vonage account over WiFi. It's a more restricted version of SIP. But now imagine various ISPs getting into this business and operating as MVNOs, giving you a smart handheld device that will place FREE SIP calls *for you* when you say "call this person", because it detected that the person has a SIP address (username@some_SIP_provider.com) in its address book entry. How did you get that address book entry in the first place? That person likely e-mailed you their vCard earlier, which added it to your computer's address book, which sync'ed it to your handheld device. Or they might have sent it to you over bluetooth or infrared. The nice thing about a SIP address is that it looks and works just like an email address: it is globally addressable, regardless of who the provider is. You don't even NEED to know whether a person is "online" to try to place a SIP call to them. If they're not there, you can leave a voicemail, which gets e-mailed as.wav file to the recipient. bickity-bam.
SIP opens a world of interoperability and global communications which Skype's closed protocol will NEVER come CLOSE to.
Firewall and NAT issues detractors of SIP complain about can all be worked around by having a SIP provider route RTP traffic for you. It's a less than ideal situation, but Skype does just about the exact same thing by having some poor sap on YOUR network, relay traffic for you. The STUN protocol addresses the vast majority of NAT issues, i've verified this myself on very exotic NAT setups at home, with various SIP providers such as EarthLink, SIPphone, and Free World Dialup. SIP should work fine for most regular home users. Skype doesn't go through ALL firewalls, stating otherwise is simply a factual inaccuracy. If a firewall is set-up in "Deny all" mode by default, with specific "Allow" rules for specific services, you bet your furry rear-end Skype won't go through, which is the case at my work.
There's no magic here. only networking. Where Skype works, SIP can technically be made to, it entirely depends on how motivated your SIP provider is, to make your call go through. From here it depends on cost-effectiveness. If too many SIP users are behind messed-up router/firewalls/networks, SIP providers might elect to relay RTP traffic for you for a fee.
But it's about delivery and time to market. It's very true, right now, Skype is King, and insanely convenient. I'm hopeful developers will get their act together and truly emphasize USABLE (usability is key) interoperability. Which was the reason behind the "Wish List" i posted in the story.
as a side-note, if you're a Mac OS X user, you already get this for free with just about all VoIP software out there. Mac OS X lets you register various audio input/output devices, whether they work over bluetooth or USB. Any OS X application developer can easily "surface" those devices to enable their users to pick where the sound should come from, and where it should go out. Pretty cool if ye ask me.
And the beauty of SIP is that no software vendor needs to define an API... because SIP is an open protocol anybody can write to.
The Skype API is Skype's limited answer to its closed proprietary protocol.
SIP is only a signaling protocol which allows two machines connected to an IP network to find each-other and exchange data in real-time. That data can in theory be anything, including encrypted UDP packets. It should be a simple matter of software writers to implement encryption. From here, the Session Description Protocol, which helps to SIP clients do handshaking to determine how they'll be talking to one-another, ought to help two clients exchange encrypted data.
you might consider reading-up on the STUN protocol, which came out in 2003, and solves the VAST MAJORITY of NAT traversal issues, albeit, not *all* possible issues. But this is where your SIP provider comes in, and offer to relay RTP (voice data essentially) traffic for you, which is a similar practice as Skype using people as "supernodes" without their knowledge or conscious consent. Except that in the SIP model, it's the SIP provider who spends the extra network resources, not some hapless random person Skype may have picked for you.
fair enough. my point is, i've successfully done SIP with other people through pretty extreme use cases. Chances are if it works for me, most people will be pleasantly surprised. Why shoot-down SIP/Gizmo without giving it a chance to work, especially when it means that in the end, you and i don't have to belong to the same VoIP system to chat with one-another?
It depends on who your SIP provider is. When all else fails, your SIP providers might relay RTP traffic between you and your party, which is similar to practices employed by Skype whereby they'll get around network limitations by routing calls through various members, without members' knowledge. Which I find insanely unethical. Do give Gizmo a try and see for yourself instances where you're unable to place or receive calls where Skype can. I have personally not run into such instance, but it'd be interesting to see people document those instances.
I've used various SIP clients (XLite/SJPhone/Gizmo) through 2 levels of NAT in my own home (192.0.0.* network plugged to 10.0.0.* network hooked to my DSL WAN), calling a free world dial-up user over his SIP URL, who sits behind asterisk at home, and one big NAT. It works just fine for me. I've SIPed countless users who sit behind your average linksys router, or behind your average 2wire home-network-in-a-box kinda thingie. Upon inspecting tcpdumps and SIP client logs, In no case did i ever see the various SIP providers actually routing the RTP traffic for me.
So again, SIP may not work everywhere, but damn, I've tried some really ODD shit with SIP working seamlessly.
well i can tell you that Gizmo works wherever I've used Skype before. So does SJPhone. So does XPro and XLite. I'm pretty sure all of'em do it through STUN, i know XPro/XLite do it through STUN for sure. And where Skype fails, Gizmo also fails, such as my work network with heavy firewalling. What we need is to document the precise instances where Gizmo and other SIP clients don't work where Skype actually does.
Again, Gizmo, like all other half-decent SIP clients DOES work through NAT, i've verified this myself. They very likely use the STUN protocol. See previous post.
STUN came out in 2003, while SIP's been around since 1995. The reason SIP wasn't practical until 2003 was indeed because of NAT. Now that we have STUN, SIP is practical. Gizmo works.
SIP by its very definition defines and enables P2P based on open standards. The SIP protocol is a signaling protocol, which enables two machines to find each-other on the internet and start exchanging data in real-time without going through a 3rd-party.
Voice quality is absolutely fantastic with Gizmo, but it entirely depends on which SIP client you use, and which SIP client the party you're calling uses. The SIP protocol, is only a signaling protocol. There are a buttload of other open protocols such as SDP (Session Description Protocol) that come into play and allow for infinite layers of interoperability based on users' computing and network resources. Through handshaking, two SIP clients can easily agree to the best codec to use.
Have you downloaded and tried Gizmo Project for Mac? I have reviewed it for theappleblog.com (currently down because slashdotted), and it's very very nice, albeit still in beta with rare glitches. It works flawlessly with all sound input and output devices you've registered with Mac OS X, including the bluetooth motorola h1810 headset i use.
Gizmo run on Mac OS X, Windows and there's a linux version due out in a month or two.
WRONG, NAT was SIP's weakness ONLY until the STUN protocol came out in 2003. Since then, all SIP clients use the STUN protocol to traverse NAT at least as well as Skype does. I've used EarthLink SIP, sipphone.com SIP, FWD through multiple layers of NATs without a glitch.
John Leger writes "Despite legal challenges faced by the rumor sites, The Apple Blog today is releasing what appear to be exclusive early concept models of Apple products." Interesting concept, and features a large disk drive, ipod dock, and wireless connectivity
Nothing in the above quote claims this actually came from Apple. "early concept models" can very-well be that the author woke-up one morning early, scratched his ass, and spent the next 72 hours relentlessly drawing-up those images.
This April Fool's joke would have been even cooler if TheAppleBlog had retained the same level of ambiguity on their site as the "submitter" of the story kept.
you said "what we need is more competition". How do you think this is going to happen?
Here's the problem: our country's broadband infrastructure is owned and operated by a couple of corporations who own all the pipes to people's homes. I don't have a problem with corporations building out infrastructure and seeking to make a very lucrative buck off of it. This is what they do. This is free enterprise. Free enterprise is a good thing. Making money is a good thing. However, and understandably enough, local governments just handed them out permits to dig in our streets to lay their pipes without any attempt to negotiate a future "pipe-sharing" plan with supporting infrastructure for competing businesses to offer content/services/data over those pipes, after they'd recouped their costs and made profits to the tune of $X amount, or after Y years of sole operation and ownership. At the time, we didn't really think in terms of data. Fair enough.
We are at a turning point in history, where we now have the ability to change this.
Contrary to what the incumbents would have us believe, municipality-driven broadband infrastructure would, in my opinion, become the ultimate enabler of free enterprise from the private sector in data, media, and communication SERVICES.
Municipality-driven WiFi is just ONE step in an overall encouraging direction.
Municipality-built broadband infrastructures, beyond providing the ability for said municipality to provide very basic connectivity for free or cheap to its constituents, also provides an opportunity to welcome the private sector to compete on an equal footing. The infrastructure must simply be allowed to evolve to allow for mostly automated ways to "share the pipe".
A WiFi system can be easily extended to enable such sharing. So could a fiber-optic network.
Consider today's "sharing" alternatives in the DSL field: it's bleak. My only real DSL alternative is my local Telco, Verizon. Thankfully, I'm able to get service from EarthLink at about the same price point as Verizon, and instead of getting mere connectivity with the insanely useless MSN premium package, i get stuff i actually find useful, such as Mac OS X Address Book synching with my earthlink online address book, which is tied into the challenge-response-based spam filtering. But here's the problem though, while EarthLink is competing on services, it can't compete with Verizon on speed, because they're only able to resell Verizon's DSL connectivity to me, and from what i've heard, we ain't looking at a big margin here.
I want hundreds of EarthLink's competing on both speed and services.
In the case of Muni WiFi, I could for example get free basic connectivity throttled at lower speeds from the City, with no-other services, and justify spending money with fine services such as knowspam.net to protect myself from spam, flickr.com for photo sharing, TypePad for blogging, Rojo.com for news reading, Prodigem.com for Torrents creating/seeding,.Mac for reliable WEBDAV hosting, some packaged-deal from EarthLink, and/or hundreds of cheap services which may be useful TO ME. There's a lot of innovation on the Internet, many of those innovators are struggling to find sustainable revenue models.
Such a broadband scene will also open the doors to triple-play packages: data, media, communications over a single pipe. Many competitors, the best few ones would win, the customer wins.
Right now, in my area, Verizon and Adelphia are the big winners. I, as a consumer, am not. As far as i'm concerned, these fsckers have no business offering internet services, what the fuck do they know abo
i don't mean to troll but remind me how you get digital still cameras, DV camcorders, bluetooth mobile devices, bluetooth headsets to work in Linux? Assuming you do... what do you do then? Where's the Linux device synchronization strategy, like, you know, iSync?
What does the digital music marketplace and music device market look like for Linux?
What's a decent document authoring tool in Linux? vi, emacs, and Open Office do not count.
the post author just might be interested in doing niftier things with his computer. you know, like when you have a girlfriend... then a wife... then kids... and you like to take pictures and videos, and perhaps easily author DVDs
sure... he could stick to Linux. Linux is nice. I've used it for a while. but today, why bother? OS X is hands-down a far superior operating system. And that's precisely because it not-only draws from the open-source community, but also adds a whole layer of innovations targeted at fulfilling the needs of, you know, mere mortals who don't spend their time compiling kernels all day long.
you've obviously got a lot of learning to do about OS X.
Slashdot Mirror
lunches are catered every day by a chef on staff. Not only you no-longer have to spend money on lunch, you get to eat some really good food. so i hear. they might even cover breakfast and whatnot. they also have naked chicks running around and giggling. err. scratch that last one.
overrated?
OVERRATED?
PSHAH!
... at least in FireFox. Not sure about Win/IE. Google/ig has a same concept, in that things reflow as you drag, but i personally find that to be a usability bug. It offends me. Google was more subtle in making you drag an outline, which i find a bit less confusing than the actual cloned Node you're dragging on the MSN counterpart. I'm pretty sure both google.com/ig and start are adopting this concept due to technical limitations. Doing drag and drop correctly is a sizable challenge. Note how neither MSN or google make you drag widgets that were below the fold. The reflowing 3-column layout you're looking at, is as such, i surmise, for many reasons not the least of which have to do with accurately calculating mouse and object position on a global document, rather than within the confines of a window object. Also dragging objects beyond the folds would ideally require automated scrolling. Not entirely trivial either to make it all work in Safari, Firefox and IE. But possible :) I find it easier to develop in Safari first by sticking to standards. Making various things work in Firefox and then IE, in that order, is typically a matter of minor incremental tweaks. In most cases, if it works in Safari, it works in FF and IE.
moderators: please mod this up. it's an important distinction. I don't believe a single methodology will fight spam. But a combination of methodologies will.
... can be found in the comments section of this blog post. Executive summary: "organizations" have exactly NOTHING intelligent to contribute because their premises are flawed, and fail to take into account the fact that economics of such projects vary vastly from one community to another.
All very valid points. Skype may very-well for now be very nicely established.
When Hybrid WiFi/GSM/3G handheld devices come to market with the capability of doing STUN+SIP+RTP(encrypted or not), things will likely change. Take a service like Vonage. They've recently released their own wireless handheld device that allows you to use your Vonage account over WiFi. It's a more restricted version of SIP. But now imagine various ISPs getting into this business and operating as MVNOs, giving you a smart handheld device that will place FREE SIP calls *for you* when you say "call this person", because it detected that the person has a SIP address (username@some_SIP_provider.com) in its address book entry. How did you get that address book entry in the first place? That person likely e-mailed you their vCard earlier, which added it to your computer's address book, which sync'ed it to your handheld device. Or they might have sent it to you over bluetooth or infrared. The nice thing about a SIP address is that it looks and works just like an email address: it is globally addressable, regardless of who the provider is. You don't even NEED to know whether a person is "online" to try to place a SIP call to them. If they're not there, you can leave a voicemail, which gets e-mailed as .wav file to the recipient. bickity-bam.
SIP opens a world of interoperability and global communications which Skype's closed protocol will NEVER come CLOSE to.
Firewall and NAT issues detractors of SIP complain about can all be worked around by having a SIP provider route RTP traffic for you. It's a less than ideal situation, but Skype does just about the exact same thing by having some poor sap on YOUR network, relay traffic for you. The STUN protocol addresses the vast majority of NAT issues, i've verified this myself on very exotic NAT setups at home, with various SIP providers such as EarthLink, SIPphone, and Free World Dialup. SIP should work fine for most regular home users. Skype doesn't go through ALL firewalls, stating otherwise is simply a factual inaccuracy. If a firewall is set-up in "Deny all" mode by default, with specific "Allow" rules for specific services, you bet your furry rear-end Skype won't go through, which is the case at my work.
There's no magic here. only networking. Where Skype works, SIP can technically be made to, it entirely depends on how motivated your SIP provider is, to make your call go through. From here it depends on cost-effectiveness. If too many SIP users are behind messed-up router/firewalls/networks, SIP providers might elect to relay RTP traffic for you for a fee.
But it's about delivery and time to market. It's very true, right now, Skype is King, and insanely convenient. I'm hopeful developers will get their act together and truly emphasize USABLE (usability is key) interoperability. Which was the reason behind the "Wish List" i posted in the story.
Wikipedia's coverage of the London Transport Explosions is very thorough and very-well organized.
Somebody mod parent up, it would appear Michael Robertson has graced us with his presence, hehe.
as a side-note, if you're a Mac OS X user, you already get this for free with just about all VoIP software out there. Mac OS X lets you register various audio input/output devices, whether they work over bluetooth or USB. Any OS X application developer can easily "surface" those devices to enable their users to pick where the sound should come from, and where it should go out. Pretty cool if ye ask me.
And the beauty of SIP is that no software vendor needs to define an API ... because SIP is an open protocol anybody can write to.
The Skype API is Skype's limited answer to its closed proprietary protocol.
SIP is only a signaling protocol which allows two machines connected to an IP network to find each-other and exchange data in real-time. That data can in theory be anything, including encrypted UDP packets. It should be a simple matter of software writers to implement encryption. From here, the Session Description Protocol, which helps to SIP clients do handshaking to determine how they'll be talking to one-another, ought to help two clients exchange encrypted data.
you might consider reading-up on the STUN protocol, which came out in 2003, and solves the VAST MAJORITY of NAT traversal issues, albeit, not *all* possible issues. But this is where your SIP provider comes in, and offer to relay RTP (voice data essentially) traffic for you, which is a similar practice as Skype using people as "supernodes" without their knowledge or conscious consent. Except that in the SIP model, it's the SIP provider who spends the extra network resources, not some hapless random person Skype may have picked for you.
fair enough. my point is, i've successfully done SIP with other people through pretty extreme use cases. Chances are if it works for me, most people will be pleasantly surprised. Why shoot-down SIP/Gizmo without giving it a chance to work, especially when it means that in the end, you and i don't have to belong to the same VoIP system to chat with one-another?
It depends on who your SIP provider is. When all else fails, your SIP providers might relay RTP traffic between you and your party, which is similar to practices employed by Skype whereby they'll get around network limitations by routing calls through various members, without members' knowledge. Which I find insanely unethical. Do give Gizmo a try and see for yourself instances where you're unable to place or receive calls where Skype can. I have personally not run into such instance, but it'd be interesting to see people document those instances.
I've used various SIP clients (XLite/SJPhone/Gizmo) through 2 levels of NAT in my own home (192.0.0.* network plugged to 10.0.0.* network hooked to my DSL WAN), calling a free world dial-up user over his SIP URL, who sits behind asterisk at home, and one big NAT. It works just fine for me. I've SIPed countless users who sit behind your average linksys router, or behind your average 2wire home-network-in-a-box kinda thingie. Upon inspecting tcpdumps and SIP client logs, In no case did i ever see the various SIP providers actually routing the RTP traffic for me.
So again, SIP may not work everywhere, but damn, I've tried some really ODD shit with SIP working seamlessly.
well i can tell you that Gizmo works wherever I've used Skype before. So does SJPhone. So does XPro and XLite. I'm pretty sure all of'em do it through STUN, i know XPro/XLite do it through STUN for sure. And where Skype fails, Gizmo also fails, such as my work network with heavy firewalling. What we need is to document the precise instances where Gizmo and other SIP clients don't work where Skype actually does.
Again, Gizmo, like all other half-decent SIP clients DOES work through NAT, i've verified this myself. They very likely use the STUN protocol. See previous post. STUN came out in 2003, while SIP's been around since 1995. The reason SIP wasn't practical until 2003 was indeed because of NAT. Now that we have STUN, SIP is practical. Gizmo works.
SIP by its very definition defines and enables P2P based on open standards. The SIP protocol is a signaling protocol, which enables two machines to find each-other on the internet and start exchanging data in real-time without going through a 3rd-party.
SIP opens the doors to far more than just real-time communications (text, voice, video), it also opens the doors to file sharing. It's a matter of someone writing a client that does file sharing over SIP. OH WAIT, someone already did as a proof of concept, and released the source code under a BSD-style license.
Voice quality is absolutely fantastic with Gizmo, but it entirely depends on which SIP client you use, and which SIP client the party you're calling uses. The SIP protocol, is only a signaling protocol. There are a buttload of other open protocols such as SDP (Session Description Protocol) that come into play and allow for infinite layers of interoperability based on users' computing and network resources. Through handshaking, two SIP clients can easily agree to the best codec to use.
Have you downloaded and tried Gizmo Project for Mac? I have reviewed it for theappleblog.com (currently down because slashdotted), and it's very very nice, albeit still in beta with rare glitches. It works flawlessly with all sound input and output devices you've registered with Mac OS X, including the bluetooth motorola h1810 headset i use.
Gizmo run on Mac OS X, Windows and there's a linux version due out in a month or two.
WRONG, NAT was SIP's weakness ONLY until the STUN protocol came out in 2003. Since then, all SIP clients use the STUN protocol to traverse NAT at least as well as Skype does. I've used EarthLink SIP, sipphone.com SIP, FWD through multiple layers of NATs without a glitch.
STUN is a major enabler of SIP
I've also posted a couple of tips and tricks on getting Gizmo working with other SIP systems, and also listing some other players in the SIP field.
Nothing in the above quote claims this actually came from Apple. "early concept models" can very-well be that the author woke-up one morning early, scratched his ass, and spent the next 72 hours relentlessly drawing-up those images.
This April Fool's joke would have been even cooler if TheAppleBlog had retained the same level of ambiguity on their site as the "submitter" of the story kept.
... welcome our new Apple Overlords in the Living Room
Imagine a Most Entertaining, Action-Packed Beowulf Cluster of these!
you said "what we need is more competition". How do you think this is going to happen?
Here's the problem: our country's broadband infrastructure is owned and operated by a couple of corporations who own all the pipes to people's homes. I don't have a problem with corporations building out infrastructure and seeking to make a very lucrative buck off of it. This is what they do. This is free enterprise. Free enterprise is a good thing. Making money is a good thing. However, and understandably enough, local governments just handed them out permits to dig in our streets to lay their pipes without any attempt to negotiate a future "pipe-sharing" plan with supporting infrastructure for competing businesses to offer content/services/data over those pipes, after they'd recouped their costs and made profits to the tune of $X amount, or after Y years of sole operation and ownership. At the time, we didn't really think in terms of data. Fair enough.
We are at a turning point in history, where we now have the ability to change this.
Contrary to what the incumbents would have us believe, municipality-driven broadband infrastructure would, in my opinion, become the ultimate enabler of free enterprise from the private sector in data, media, and communication SERVICES.
Municipality-driven WiFi is just ONE step in an overall encouraging direction.
Municipality-built broadband infrastructures, beyond providing the ability for said municipality to provide very basic connectivity for free or cheap to its constituents, also provides an opportunity to welcome the private sector to compete on an equal footing. The infrastructure must simply be allowed to evolve to allow for mostly automated ways to "share the pipe".
A WiFi system can be easily extended to enable such sharing. So could a fiber-optic network.
Consider today's "sharing" alternatives in the DSL field: it's bleak. My only real DSL alternative is my local Telco, Verizon. Thankfully, I'm able to get service from EarthLink at about the same price point as Verizon, and instead of getting mere connectivity with the insanely useless MSN premium package, i get stuff i actually find useful, such as Mac OS X Address Book synching with my earthlink online address book, which is tied into the challenge-response-based spam filtering. But here's the problem though, while EarthLink is competing on services, it can't compete with Verizon on speed, because they're only able to resell Verizon's DSL connectivity to me, and from what i've heard, we ain't looking at a big margin here.
I want hundreds of EarthLink's competing on both speed and services.
In the case of Muni WiFi, I could for example get free basic connectivity throttled at lower speeds from the City, with no-other services, and justify spending money with fine services such as knowspam.net to protect myself from spam, flickr.com for photo sharing, TypePad for blogging, Rojo.com for news reading, Prodigem.com for Torrents creating/seeding, .Mac for reliable WEBDAV hosting, some packaged-deal from EarthLink, and/or hundreds of cheap services which may be useful TO ME. There's a lot of innovation on the Internet, many of those innovators are struggling to find sustainable revenue models.
Such a broadband scene will also open the doors to triple-play packages: data, media, communications over a single pipe. Many competitors, the best few ones would win, the customer wins.
Right now, in my area, Verizon and Adelphia are the big winners. I, as a consumer, am not. As far as i'm concerned, these fsckers have no business offering internet services, what the fuck do they know abo
i don't mean to troll but remind me how you get digital still cameras, DV camcorders, bluetooth mobile devices, bluetooth headsets to work in Linux? Assuming you do ... what do you do then? Where's the Linux device synchronization strategy, like, you know, iSync?
What does the digital music marketplace and music device market look like for Linux?
What's a decent document authoring tool in Linux? vi, emacs, and Open Office do not count.
the post author just might be interested in doing niftier things with his computer. you know, like when you have a girlfriend ... then a wife ... then kids ... and you like to take pictures and videos, and perhaps easily author DVDs
sure ... he could stick to Linux. Linux is nice. I've used it for a while. but today, why bother? OS X is hands-down a far superior operating system. And that's precisely because it not-only draws from the open-source community, but also adds a whole layer of innovations targeted at fulfilling the needs of, you know, mere mortals who don't spend their time compiling kernels all day long.
you've obviously got a lot of learning to do about OS X.