GIMP has color management and 48 bit color in its short term roadmap. Adobe should be worried of competition IMO (just like Microsoft is worried about Linux). This will be competition between a $650 piece of software and another, $0 one that you can download off the Internets.
BTW, you've seriously overpaid for your copy of Photoshop. If you have PSE, you can get photoshop for $299, which is how I bought it.
That's why no one is switching. It is perfectly possible to successfully compete with a graphics editor that sells for $650 a pop. And yeah, I do own a copy of Photoshop CS for Mac. My wallet still hurts. I wish it had competition at the time I bought it.
and it will never have it. As a photographer, I absolutely NEED color management. I can't do without it. Not having color management is a deal breaker for me, and for a lot of other folks. Besides, PSE's support of 48 bit color is castrated in order to not affect Photoshop sales.
So the key to world dominance is: 1. 48 bit color everywhere, without limitations 2. End-to-end color management capabilities 3. Decent, stable Windows version
Once GIMP people implement 48bit color and color management, they'll have a potential to take away a large portion of Adobe clientele - web designers and photographers (i.e. people in no way related to prepress and CMYK). When two products have equal capabilities in relation to your tasks, but one is $650 and one is free, the choice becomes really simple.
Right now GIMP is not yet there, but this doesn't mean it'll never be.
I like tabbed browsing actually, and I like tabs in VS.Net. If tabs were easily detachable, I would like this even more. Mac windowing is a horrible mess. They pay too much attention to transparency and shiny buttons, which in itself isn't a bad thing, IF the basics are nailed down. Whic, unfortunately, they aren't.
As things stand it's nearly impossible to have two instances of XCode running in parallel. You simply get lost in a dozen windows, because OS X doesn't offer anything to logically group them. No tabs, no MDI, no nothing. You can't even hide one set of project windows all at once. You have to either hide ALL XCode windows, or go through them and hide every god damn window manually.
This is fucked up, IMO. I'm literally 10 times more productive in VS.NET, and that's what I do - I write code. There's no competition to VS.NET right now as far as coding productivity is concerned.
Managed code is the future. XUL is great and shit, but some things you just can't do in XUL. You'll have to download extensions of some kind. These extensions better be sandboxed and have a verifiable security model..NET has all this, and it's also an ECMA standard.
Why not Java you might ask? Simply because it's a non-standard language specs to which are controlled by Sun. Plus, compared to C#/.NET Java looks like a bunch of amateur bullshit.
>> The days may be long gone, but they haunt >> my memories and have me running XP.
ACPI is not ready for realistic laptop use at this point, and all kinds of forums are littered with posts from users who had some major grief from setting it up. I'd predict that 95% of people who attempt to use Linux on their laptops revert to Windows XP/2000 sooner or later.
Driver support for wifi is kinda there (with ndiswrapper), but setting it up is _well_ beyond the capabilities of a Linux newbie, especially if this newbie wants proper WAP security.
Make it easy to transition back and forth between categories IF academic objectives are in line with the requirements. Make it _not cool_ to be in #3. Call them losers or something, because that's what they really are. If this hurts kids "psychologically", fuck psychology. The world is a harsh place, if you don't want to study (and your parents don't care), get ready to be a loser.
Finally, make it challenging but cool to remain in bucket #1.
The reason why kids don't give a crap about studying is because it's "cool" to not study. Change that, and ungrateful cocksuckers will study like you wouldn't believe.
Here's what I have to say on the topic: 1. Do not optimize _code_, optimize _algorithms_ , architectures and SQL schemas, and do so only in the case when you're 100% clear that there will be a sizable benefit. 2. If you do #1, don't do any optimizations until you've done at least preliminary perf testing and identified the bottlenecks. It's important for perf test to mimic real workload as close as possibly. Half-assed test scenarios are about as good as no testing, and sometimes worse because they give you the false sense of confidence, when in fact you should have none. 3. If you're making changes in an existing product, measure improvements under _varied_ workloads. Collect and study the trend data. 4. Don't optimize stuff which does not get called/accessed often or where you know the benefits of optimization will be minuscule. I've seen some programmers optimize error logging to avoid a fucking _function call_ (for some reason they thought it was expensive) and then turn around and do horrible performance decisions in SQL schema. 5. Don't optimize without first talking to your perf/stress person (if you have one). If you don't have a perf/stress guy, hire one. Some developers think they know a lot about perf and optimization, while in fact all they know are urban legends. So they spend a bunch of time on urban legends, and the real perf issues remain in the product. Your perf guy, if he's good, KNOWS what's broken and how to fix it. Listen to what he has to say a little more. 6. Pay the most attention to shared resources and threading issues around that. In this day and age, processors are blazingly fast, so the most common source of perf problems is locking on a shared resource. Do code reviews on code that's heavily multithreaded and employs a lot of locking. If you use MS SQL server, run Index Optimizer and see what it has to say. Also, run Microsoft PSS "Blocker" script under a real world workload to see if you're having locking issues.
Asus Terminator C3 - $115. VIA C3 processor, running at 800MHz. Comes with processor, floppy and CD-ROM. Put in your old PC2100 RAM, a couple SATA hard drives, install your favorite Linux/*BSD distro, and you have a perfect home server. It even has a PCI slot, FireWire, USB2 and TV OUT.
It's a mystery to me that Linux hasn't yet taken over both of these markets completely, end to end. It's free. It comes with full source code which you can tweak to your heart's content, it can be trimmed down to work in a frickin' wristwatch, it runs all the software you could possibly need, and if you need something extra, guess what, development tools are also free.
Yet PDA makers insist on paying the dough to MSFT instead of hiring a dozen Linux hackers to do "spit & polish" on their distro of choice.
I guess this is because PDA market is not yet cost driven, and PDAs are still perceived as useless geeky toys.
How about reducing the frikkin' power dissipation? Not to Pentium-M levels, but to Athlon 64 levels at least. People used to joke about Athlons, and now look at them, AMD fixed the issues without running them at 300MHz (that's the speed of Pentium-M processor in my notebook as I write this).
Sure there are crappy American programmers (albeit the percentage is dramatically lower), but there's an important cultural difference. A halfway decent American programmer is more likely to tell his boss to fuck off when he puts forth an unrealistic schedule or requirements. An Indian will not say anything. He will just copy&paste shit all over the place and call the resulting spaghetti code a product.
Yet over here, all Indians I've worked with do just barely enough to get by. Copy & paste is their preferred coding method, and architecture and proper design are nowhere to be found.
There was one exception in my 5 years in the industry - the guy was extremely, unbelievably bright. He wrote a book on ATL. He typed so fast I couldn't see his fingers, and he was the best programmer in 50 people team. But he's just an exception that it seems reinforces the rule.
Just three days ago, I bought this VIA C3 based system (processor dissipates 7W at idle), and they roll out sub 1W chips. Oh, well. Hard drive is gonna take 9W of power anyway.
So far it's just the usual bullshit. _IF_ we did that, we _could_, blah blah blah, yadda, yadda, yadda. I'm so sick and tired of empty promises. How's cut&paste doing, folks? Can I cut & paste between _everything_ yet or it's another 5 years away?
Slashdot Mirror
GIMP has color management and 48 bit color in its short term roadmap. Adobe should be worried of competition IMO (just like Microsoft is worried about Linux). This will be competition between a $650 piece of software and another, $0 one that you can download off the Internets.
BTW, you've seriously overpaid for your copy of Photoshop. If you have PSE, you can get photoshop for $299, which is how I bought it.
1. No polished apps
2. Poor hardware support
That's why no one is switching. It is perfectly possible to successfully compete with a graphics editor that sells for $650 a pop. And yeah, I do own a copy of Photoshop CS for Mac. My wallet still hurts. I wish it had competition at the time I bought it.
and it will never have it. As a photographer, I absolutely NEED color management. I can't do without it. Not having color management is a deal breaker for me, and for a lot of other folks. Besides, PSE's support of 48 bit color is castrated in order to not affect Photoshop sales.
So the key to world dominance is:
1. 48 bit color everywhere, without limitations
2. End-to-end color management capabilities
3. Decent, stable Windows version
Are we still in the 80s? I realize this is not an IDE, but I was talking about IDEs.
As a text editor it looks promising, though. Trouble is, text editor built into Mac OS X is perfectly adequate for my _text_ needs.
Once GIMP people implement 48bit color and color management, they'll have a potential to take away a large portion of Adobe clientele - web designers and photographers (i.e. people in no way related to prepress and CMYK). When two products have equal capabilities in relation to your tasks, but one is $650 and one is free, the choice becomes really simple.
Right now GIMP is not yet there, but this doesn't mean it'll never be.
I like tabbed browsing actually, and I like tabs in VS.Net. If tabs were easily detachable, I would like this even more. Mac windowing is a horrible mess. They pay too much attention to transparency and shiny buttons, which in itself isn't a bad thing, IF the basics are nailed down. Whic, unfortunately, they aren't.
As things stand it's nearly impossible to have two instances of XCode running in parallel. You simply get lost in a dozen windows, because OS X doesn't offer anything to logically group them. No tabs, no MDI, no nothing. You can't even hide one set of project windows all at once. You have to either hide ALL XCode windows, or go through them and hide every god damn window manually.
This is fucked up, IMO. I'm literally 10 times more productive in VS.NET, and that's what I do - I write code. There's no competition to VS.NET right now as far as coding productivity is concerned.
Managed code is the future. XUL is great and shit, but some things you just can't do in XUL. You'll have to download extensions of some kind. These extensions better be sandboxed and have a verifiable security model. .NET has all this, and it's also an ECMA standard.
Why not Java you might ask? Simply because it's a non-standard language specs to which are controlled by Sun. Plus, compared to C#/.NET Java looks like a bunch of amateur bullshit.
to be my second wife. Come on, we can "consider" all kinds of shit. Guess what, nobody cares what you "consider" until you actually make it.
If they put their base on the moon, that will be news. Their "considering" is not news.
You can buy a used CD on half.com for $8-10, delivered to your goddamn door. That's uncompressed, un-DRMd audio, and the booklet is included, too.
They should sell the stuff they're smoking instead, it's apparently very strong.
>> The days may be long gone, but they haunt
>> my memories and have me running XP.
ACPI is not ready for realistic laptop use at this point, and all kinds of forums are littered with posts from users who had some major grief from setting it up. I'd predict that 95% of people who attempt to use Linux on their laptops revert to Windows XP/2000 sooner or later.
Driver support for wifi is kinda there (with ndiswrapper), but setting it up is _well_ beyond the capabilities of a Linux newbie, especially if this newbie wants proper WAP security.
Divide classes into three categories:
1. Smart
2. Average
3. Below average
Make it easy to transition back and forth between categories IF academic objectives are in line with the requirements. Make it _not cool_ to be in #3. Call them losers or something, because that's what they really are. If this hurts kids "psychologically", fuck psychology. The world is a harsh place, if you don't want to study (and your parents don't care), get ready to be a loser.
Finally, make it challenging but cool to remain in bucket #1.
The reason why kids don't give a crap about studying is because it's "cool" to not study. Change that, and ungrateful cocksuckers will study like you wouldn't believe.
That's what pr0n looked like back then. It is widely known that the Internets were invented to transfer pr0n.
Here's what I have to say on the topic:
1. Do not optimize _code_, optimize _algorithms_ , architectures and SQL schemas, and do so only in the case when you're 100% clear that there will be a sizable benefit.
2. If you do #1, don't do any optimizations until you've done at least preliminary perf testing and identified the bottlenecks. It's important for perf test to mimic real workload as close as possibly. Half-assed test scenarios are about as good as no testing, and sometimes worse because they give you the false sense of confidence, when in fact you should have none.
3. If you're making changes in an existing product, measure improvements under _varied_ workloads. Collect and study the trend data.
4. Don't optimize stuff which does not get called/accessed often or where you know the benefits of optimization will be minuscule. I've seen some programmers optimize error logging to avoid a fucking _function call_ (for some reason they thought it was expensive) and then turn around and do horrible performance decisions in SQL schema.
5. Don't optimize without first talking to your perf/stress person (if you have one). If you don't have a perf/stress guy, hire one. Some developers think they know a lot about perf and optimization, while in fact all they know are urban legends. So they spend a bunch of time on urban legends, and the real perf issues remain in the product. Your perf guy, if he's good, KNOWS what's broken and how to fix it. Listen to what he has to say a little more.
6. Pay the most attention to shared resources and threading issues around that. In this day and age, processors are blazingly fast, so the most common source of perf problems is locking on a shared resource. Do code reviews on code that's heavily multithreaded and employs a lot of locking. If you use MS SQL server, run Index Optimizer and see what it has to say. Also, run Microsoft PSS "Blocker" script under a real world workload to see if you're having locking issues.
BillG is paying for it. Or more precisely The Bill and Melinda Gates Foundation. They have free internet access and LCD screens in every library, too.
http://www.newegg.com/app/viewproductdesc.asp?desc ription=56-110-030&DEPA=0
Asus Terminator C3 - $115. VIA C3 processor, running at 800MHz. Comes with processor, floppy and CD-ROM. Put in your old PC2100 RAM, a couple SATA hard drives, install your favorite Linux/*BSD distro, and you have a perfect home server. It even has a PCI slot, FireWire, USB2 and TV OUT.
I watch movies for free by using our local library (Washington, King County library system). One less monthly bill, yaw!
It's a mystery to me that Linux hasn't yet taken over both of these markets completely, end to end. It's free. It comes with full source code which you can tweak to your heart's content, it can be trimmed down to work in a frickin' wristwatch, it runs all the software you could possibly need, and if you need something extra, guess what, development tools are also free.
Yet PDA makers insist on paying the dough to MSFT instead of hiring a dozen Linux hackers to do "spit & polish" on their distro of choice.
I guess this is because PDA market is not yet cost driven, and PDAs are still perceived as useless geeky toys.
How about reducing the frikkin' power dissipation? Not to Pentium-M levels, but to Athlon 64 levels at least. People used to joke about Athlons, and now look at them, AMD fixed the issues without running them at 300MHz (that's the speed of Pentium-M processor in my notebook as I write this).
Sure there are crappy American programmers (albeit the percentage is dramatically lower), but there's an important cultural difference. A halfway decent American programmer is more likely to tell his boss to fuck off when he puts forth an unrealistic schedule or requirements. An Indian will not say anything. He will just copy&paste shit all over the place and call the resulting spaghetti code a product.
Yet over here, all Indians I've worked with do just barely enough to get by. Copy & paste is their preferred coding method, and architecture and proper design are nowhere to be found.
There was one exception in my 5 years in the industry - the guy was extremely, unbelievably bright. He wrote a book on ATL. He typed so fast I couldn't see his fingers, and he was the best programmer in 50 people team. But he's just an exception that it seems reinforces the rule.
Quantity doesn't matter. Quality does.
Kind of like "unbiased slashdot discussion" or "Microsoft-loving slashdotter".
According to TFA this shit happened 50000 years ago. Is this some kind of slashdot record, posting news that mattered 50K years ago?
Just three days ago, I bought this VIA C3 based system (processor dissipates 7W at idle), and they roll out sub 1W chips. Oh, well. Hard drive is gonna take 9W of power anyway.
So far it's just the usual bullshit. _IF_ we did that, we _could_, blah blah blah, yadda, yadda, yadda. I'm so sick and tired of empty promises. How's cut&paste doing, folks? Can I cut & paste between _everything_ yet or it's another 5 years away?