The GPL is crafted to avoid any additional restrictions being imposed on the code. It does not try to evaluate if the additional restrictions are 'simple' or 'reasonable'. It errs on the side of caution by not allowing any restrictions.
Put yourself in the shoes of whoever wrote the GPL. Can you write a license that preserves code freedom while, in advance, allowing some types of restrictions? Based on what criteria?
It's not that the vanity clause in XF86 license is unreasonable. It's that it's one kind of restriction, and these are forbidden by the GPL.
Before you whine about this, you recomendded hiring "a DBA per hundred employees". You obviously knew what you were talking about, and are backpedaling. There's a huge difference between hiring a programmer to do some contract work and hiring a DBA.
You know there's more than one person commenting on slashdot, don't you? Open mouth, insert foot.
You did not address his point. A small business person does not have hundreds of employees by definition. Someone with maybe one or two employees cannot afford a dedicated DBA just so their database can be normalized.
No I did not address that particular case. Easily solved: go to rentacoder.com and get someone who knows what they're doing to design the database for you. Software design is a lot like auto mechanics: you could fix it yourself, it probably would work for some time, but will break more often. And you don't need a fleet of a hundred vehicles and a full-time mechanic to justify professional help, you can hire as you need.
No database. Track information on paper, as it was done 50yrs ago.
Use access, make a huge number of kludges that work, and then call in consultants to solve the whole mess, paying by the nose in consultant fees (I've seen this happen several times)
Hire a DBA per hundred employees, right out of college.
Robert: The core OS itself is closed source. This way I can make sure that no forks of SkyOS will be made, and that there is always only
a single SkyOS distribution. Furthermore, keeping the core OS closed source makes it possible for me to control and change any kernel function as fast as
possible without waiting for other developers to checkout/checkin due to different time zones and other considerations.
Forking, mystified as the big bad wolf in OSS projects occurs only when the project seems doomed. Furthermore, I have no example of a fork that caused a project to fail. I'd like to know the real reason behind not releasing the OS as open source. Surely, it can't be the huge market out there for Yet Another Operating System.
We are looking for some good software developers to join the SkyOS Software Development team. Their duty will be to port or write some good applications and drivers before our 5.0 release. In order to join the team you must have:
(...)
This is not a paid position. There are, however, some benefits that come with it:
(...blabber about access to dev s/w, duh!...)
WTF? I didn't know the market was so bad people took s/w positions in exchange for access to toy operating systems. Well it isn't. And you can't get people to work for free while you make all the money. Choose open source or closed source, but play by the rules...
From your description, I guess you'll like Gentoo. Take the time to go through the docs (very good quality documentation), namely the USE variable configuration. You should be able to have portage compile all but the most exquisite packages for you, configured just the way you want it.
Gentoo does have one problem. It'll make very easy to drive on the bleeding edge of libs/apps. Stay on the bleeding edge too long, and you'll see system breakage -- probabilities don't lie. My last one was on an upgrade from GCC minor versions, where I had to boot from rescue CD and recompile everything anew (laptop downtime of almost 24h, but recovery was unattended).
A good system would explicitly know about dependencies, and would then concurrently start/stop everything for which the dependencies are already satisfied. That's what multi-tasking is for. Each time a daemon reports a successful start, all the other services whose dependencies are satisfied by it, would immediately be concurrently launched.
You've just described Gentoo's init system, which builds a dependency graph, and can be configured to start/stop services concurrently (although that's not the default).
I would expect more from an established news source, such as BBC. In the wake of the David Kelly affair, which has profoundly stained BBC's reputation, here is a perfect example of a story based on pure speculation and hear-say.
There is no suggestion that MyDoom was written by a Linux zealot, much less evidence. For all we know, it could have been SCO doing it, so its case reaches the headlights, and the pump-n-dump scheme fairs better. SCO's case against IBM is a wreck, and will hit a wall very soon on its own.
I hope this letter produces editorial regulation inside BBC against unfounded stories. Don't compete with "The Sun".
The worm takes advantage of a common software bug called a buffer overflow. Buffers overflow when a data string is written into memory without its length being checked by the program. If the string is too long, the tail end of the data overwrites the program's own code.
From your previous post (my emphasis):
...data from your application spills across into another application's memory
Memory protection is guaranteed via hardware. The too common segmentation faults one sees when developing stuff in C are just the result of a stray pointer trying to write onto an address outside the memory area allocated to a process. The operating system usually kills the offending process. It is, therefore, impossible for a program to overwrite another programs memory, luckily... Current buffer overflows would pale against an attack that could target and overwrite a super user process.
Your suggestion fixes one problem with security, but it doesn't resolve others - how about buffer overflows where data from your application spills across into another application's memory?
Memory protection has been with us x86'ers since the 286...
Can someone explain why it's so bad? Because it sure seems like a great way to associate windows and tools together into one cohesive group. The fact that MS does something is never a good enough reason for me to do something, that they stop doing something won't make me stop either.
For me, it's the fact that MDI apps implement a windowmanager with much more limited features than whatever I choose for the regular desktop. I'm a windowmaker user for some years now. Taking gimp as an example, I've configured gimp to always open windows on the same virtual desktop, with a specific ordering of windows -- so that toolboxes are never covered -- with specific positions for some toolboxes. I can never get that kind of customizability with an MDI app, whose internal windowmanager is just a quick kludge to glue the app together.
All in all, I guess I use GIMP just as an MDI, by dedicating one virtual desktop to it, and then make full use of the excellent windowmanager running on my desktop.
Wrong on both accounts. Northern European countries, namely Sweden and the Netherlands have better quality of life than the USA. Even Canadians live better than US citizens. As for free trade, the US is about on par with the EU, but actively trying to protect itself more -- see current Euro<->Dollar exhange rates and recent steel import taxes.
You decided to become a device driver developer and handle client/server systems. Why should I (or any part of society) be held responsible for your choice? I'm sure when you went to school for it, you thought you'd make money, pay the bills, etc. You made a bad decision, but yet I'll have to pay out of my paycheck to keep you fed. Ludicrous.
In other words: Don't you ever ever spend 3-5 years (5% of your lifetime) specializing in anything. Specializing is a risk, and you are the one running that risk. Only you pick the profit of specialization, so only you carry the burden of failure if evolution makes your abilities obsolete. Let's have a nation of farmers and stone artisans. Those trades will never get obsolete.
Like what the other guys were saying, in essence, is that you're only right in the cases of luxury products (Ferrari cars, gucci clothing, prada bags). There's no such equivalent in software/IT/tech operations.
Luxury goods are one example of items where price is not a significant factor in the purchase decision. In tech products, the only factor that I can see may supercede price is (perceived) quality/brand name. For Joe Consumer, brand names mean a lot, and this alone may sustain a price higher than would be dictated by competition (see B&O products or Sony consumer goods).
I don't see how manufacturing costs can not be related to price in the long run. When we're discussing jobs moving to other countries, of course it starts with profit margin, but in the long run it is producing similar product quality for a cheaper price.
It's very easy for the manufacturing cost not to have a significant effect on price, even in the long run. You just have to make it for the price to be a minor decision factor for the consumer. This is especially true for luxury items, such as fashion clothing, luxury travel or collectible items. Quality, style, fashion, rarity are all factors that may supercede price when it comes to choose a product or service.
With the ability to get cheaper labor off short in the tech world, the prices for certain tech consumer goods (from software to DVDs to car computer brains) will fall, allowing prices to fall as well.
You're wrongly assuming that the price of goods is in any way dependant on the manufacturing cost. Price is dictated by market -- supply and demand. The only constraint is: if your manufacturing costs exceed the price, your profit margin is negative and your product is dead.
There are a whole lot of products that have gigantic profit margins, proving my statement. Think about the production cost of a Ferrari. Think about the production cost of a good Porto. Think about the production cost of designer clothing. Examples are everywhere...
You're under the illusion that prices are based on production costs rather than market conditions. Goods and services are always sold at the most efficient price supported by the market. When the most efficient price has negative efficiency, companies lose money and products/services die. When a company cuts costs, there's no pressure to deflate prices. The only possible pressure is dictated by buyers (the market).
If you measure your coffee in pots, you aren't drinking real coffee. Two real expressos in a row, and I'm jumping around like a jackhammer. After a pot of coffee I could die and my heart would continue beating on its own.
$15000 bucks is a little less what I make per year. It'd be much more cost effective if I built one of these myself. Not that I need one, anyway. VoA (Voice over Air), gesture controlled light switches and junk-made file/mp3/web server work perfectly well at my home.
I think you are confusing the kernel with the distro. There's not much in the linux kernel that makes it usable only in Unix boxes -- apart from single-rooted filesystem and the concept of 'everything is a file', I can't remember much else. It'd probably be easier to use Linux/*BSD underneath a masquerading layer to change unwanted behaviours.
So 20 years ago, back when Sun was inventing Corba, they weren't a Unix company?
How did the fact of being invented by a Unix company brand Corba as a Unix protocol? It's implemented and used in both Unix and non-Unix environments, and designed with interoperability in mind. Just the opposite of the family of protocols started by DDE which grew all the way to the current (D)COM(+). May I add just as a note, that I don't like Corba. I think it's overly resource greedy, and difficult to debug -- much like the Microsoft equivalent familiy of protocols.
I restate that Windows developers tend to blindly follow whatever is dictated by Microsoft and be impervious to other technologies. People working on more mixed environments tend to search and evaluate the best solution for any given job. This dictatorship goes well against my gut feelings (i.e. it smells).
I also restate that X11 is a specific application protocol. Not in the sense of a final user application, but in the sense of being applied to a single problem -- remotely displaying computer interfaces. It's not a generic glue protocol, like Corba, (D)COM(+) or piped plain text. It does not belong in this discussion.
Which Unix applications get the most use?
I'd wager it's the likes of ls, grep, sort, sed, awk, tar or uniq, not Apache, mozilla or any other big application. I don't have any numbers, however.
And just a note: Some of the apps you mentioned do very usefull work if stuffed in a pipeline via stdin/stdout, namely gcc or emacs (although for this purpose, sed or awk are easier to use), and I very regularly use ssh in a pipeline to do simple backups (e.g. ssh somehost "tar czvf - importantdir" > backup.`date --iso-8601`.tgz). Your lack of knowledge is probably the reason for the poor evaluation of the importance of common, simple, well documented interfaces on every application. Do yourself a favor. Go and read the book mentioned in the article.
Slashdot Mirror
Put yourself in the shoes of whoever wrote the GPL. Can you write a license that preserves code freedom while, in advance, allowing some types of restrictions? Based on what criteria?
It's not that the vanity clause in XF86 license is unreasonable. It's that it's one kind of restriction, and these are forbidden by the GPL.
- No database. Track information on paper, as it was done 50yrs ago.
- Use access, make a huge number of kludges that work, and then call in consultants to solve the whole mess, paying by the nose in consultant fees (I've seen this happen several times)
- Hire a DBA per hundred employees, right out of college.
2 doesn't look so good anymore.And yet, on their main page:
WTF? I didn't know the market was so bad people took s/w positions in exchange for access to toy operating systems. Well it isn't. And you can't get people to work for free while you make all the money. Choose open source or closed source, but play by the rules...Gentoo does have one problem. It'll make very easy to drive on the bleeding edge of libs/apps. Stay on the bleeding edge too long, and you'll see system breakage -- probabilities don't lie. My last one was on an upgrade from GCC minor versions, where I had to boot from rescue CD and recompile everything anew (laptop downtime of almost 24h, but recovery was unattended).
Cheap, simple, and it has a flashlight. Feature list: makes calls, does short messaging, and has a flashlight.
Did you steal that UID, kid?
All in all, I guess I use GIMP just as an MDI, by dedicating one virtual desktop to it, and then make full use of the excellent windowmanager running on my desktop.
Wrong on both accounts. Northern European countries, namely Sweden and the Netherlands have better quality of life than the USA. Even Canadians live better than US citizens. As for free trade, the US is about on par with the EU, but actively trying to protect itself more -- see current Euro<->Dollar exhange rates and recent steel import taxes.
Duh!
There are a whole lot of products that have gigantic profit margins, proving my statement. Think about the production cost of a Ferrari. Think about the production cost of a good Porto. Think about the production cost of designer clothing. Examples are everywhere...
You're under the illusion that prices are based on production costs rather than market conditions. Goods and services are always sold at the most efficient price supported by the market. When the most efficient price has negative efficiency, companies lose money and products/services die. When a company cuts costs, there's no pressure to deflate prices. The only possible pressure is dictated by buyers (the market).
If you measure your coffee in pots, you aren't drinking real coffee. Two real expressos in a row, and I'm jumping around like a jackhammer. After a pot of coffee I could die and my heart would continue beating on its own.
$15000 bucks is a little less what I make per year. It'd be much more cost effective if I built one of these myself. Not that I need one, anyway. VoA (Voice over Air), gesture controlled light switches and junk-made file/mp3/web server work perfectly well at my home.
Is there a porno version of Star Treck?! Where? Where?
I think you are confusing the kernel with the distro. There's not much in the linux kernel that makes it usable only in Unix boxes -- apart from single-rooted filesystem and the concept of 'everything is a file', I can't remember much else. It'd probably be easier to use Linux/*BSD underneath a masquerading layer to change unwanted behaviours.
I restate that Windows developers tend to blindly follow whatever is dictated by Microsoft and be impervious to other technologies. People working on more mixed environments tend to search and evaluate the best solution for any given job. This dictatorship goes well against my gut feelings (i.e. it smells).
I also restate that X11 is a specific application protocol. Not in the sense of a final user application, but in the sense of being applied to a single problem -- remotely displaying computer interfaces. It's not a generic glue protocol, like Corba, (D)COM(+) or piped plain text. It does not belong in this discussion.
I'd wager it's the likes of ls, grep, sort, sed, awk, tar or uniq, not Apache, mozilla or any other big application. I don't have any numbers, however.And just a note: Some of the apps you mentioned do very usefull work if stuffed in a pipeline via stdin/stdout, namely gcc or emacs (although for this purpose, sed or awk are easier to use), and I very regularly use ssh in a pipeline to do simple backups (e.g. ssh somehost "tar czvf - importantdir" > backup.`date --iso-8601`.tgz ). Your lack of knowledge is probably the reason for the poor evaluation of the importance of common, simple, well documented interfaces on every application. Do yourself a favor. Go and read the book mentioned in the article.
Sorry for the typo.