My original point was: if there is any live-DVD I wanna get it to see how it has been made. I am a Gentoo guy. But at this time it doesn't matter - any distro LiveDVD (means bootable DVD) will help me to understand how to build it on my own.
Well, I've got a lot of "theoretical" answers, perhaps too theoretical. So, I am keep looking for it:)
A thousand lines of C will certainly "do more" than the same number of lines of Sparc assembler. Likewise (though arguably not on the same scale), a thousand lines of Java will certainly "do more" than the same number of lines of C.
Why do you stop? Keep going:
A thousand lines of Python will certainly "do more" than the same number of lines of Java. Likewise (though arguably not on the same scale), a thousand lines of Lisp will certainly "do more" than the same number of lines of Python.
With even higher level languages, entire classes of mistakes cannot even exist, but I don't have time to go into the advantages of such languages like SML or Haskell.
Ok, let's just put ML and Haskell right next to Python and Lisp in the thousand line formula above:)
Programmers no matter how experienced, are going to make mistakes. What matters is how costly those mistakes are. And they cost a whole lot more in C than in Java.
The cost in a well orginized process is majorly defined by amount of debugging. Two thing worth to mention:
the amount of low-level mistakes is lower in a higher-level programming language due to... well there is no low-level programming there:)
the refactoryng of a high-level language coded program is faster than of a low-level one;
But all this logic doesn't work in real life, where most of corporate decisions are based on a marketing bullshit, rather than on real benefits. That's why Java is everywhere, especially in placases where it's not supposed to be from the first place: network server-side applications and GUI.
I think one of the biggest keys created by Apple (and killed by Apple too) was OpenDoc as an DOM precessor, and based on it CyberDog - what Mozilla is trying to be today, but at time when Netscape and IE could barely run longer than 10 minutes without being crashed.
Where was that Scully when the technology was closed? Why wasn't it at least open-sourced?
So many stupidy-based decisions were, are and will be driving Apple.
Microsoft was (and still is) using lots of legal AND illegal methods to take (and keep) customers to Windows.
Linux... Wait a minute, there is no any equal Linux vendor to put here that would be appropriate to compare to microsoft and Sun regarding the marketing and selling strategies. In fact, there IS NO marketing strategies in Linux! Of course there are few commercial vendors with their distros, there are some support companies and integrators. And of course they have marketing strategies about their business. But that has nothing to do with development of Linux kernel (and a dozen of independent distros). Sometimes businesses like IBM contribute some good code, but the decisions about CVS commits usually are done by independent people (sometimes even working in a competiting company. That's the nature of open source.
So, it is not correct to say that Linux is eating Sun's customers. No, it is another way around: Sun is letting customers (existing and potential ones) to choose Linux by keeping the high prices for Sun products, inappropriately high compare to offered features AND features of competing system.
I think that Sun still has chances to save itself. Somebody indisde Sun (in fact - on atop of it) must stop covering own ass trying to prove the importance of having done investments to Solaris, must dedicate some efforts to move the rest of still superior Solaris features to Linux (especially Linux/Sparc) and must abandon Solaris in a favor of Linux/Sparc. They will save on cutting the cost of OS developemtn and support. And the customers will be more cofortable to save on skill investments.
B/c make is so unfamous by its two decade obsolte design that many people has already created many projects dealing with dependencies in a much better way without any harsh and using much better (and more programmable) syntax.
As for BSD and make - it's pretty balanced combination, where everything old-and-well known (even when it's too obsolete already) is considered as good.
All systems come with bash. Why not use it? Python is not needed for this.
I still remember systems where bash scripts did not work as bash was broken or bash accidently was disabled for the user running the scripts. That time (3-5 years ago) there were many flamewars bash-vs-sh. What's happen? Bash is everywhere. Not precisely everywhere, but more systems have bash today.
I think the evolution of unix-like operating systems (especially Linux ones) is moved far forward enough to begin flamewars python-vs-bash. Yes, Python is not everywhere, but so isn't bash, Python has just a little bit smaller % of system with no Python, but still many of them have it. Especially if it will be required.
Why Python? Several years ago Bash won a simple sh just b/c it has a little bit more convinient programming extensions. Even with that Bash is still not exactly general programming language and if you try anything complex on it - you know what I mean. Python is much better for scripting complex tasks. Dependency checking is one of reasons to use it. String parsing, regeneration of config files - another ones. With bash you don't have a good luck to do it.
Evolution of system tools wasn't frozen these years. Get over it. It's time for real tools to come to the scene. If you don't like it - choose an OS more conservative than Linux, for example BSD - sure they will stisk to a simple sh for a couple of more decades. No wonder it's dea... sorry...:)
Well, thanks to the project like this, and Gentoo as well, in few years someone will point on this flamewar and say: "remember some conservators did not want to use Python? Now it's time for..." Who knows, will it be Erlang or Haskell?:)
Yes it's it tru. In both projects the amount of concurrent authenticated sessions was bellow 2000 per cluster. The amount of concurrent requests (I hope you understand the difference) was bellow 200 per server (which is a pretty reasonable limit for many considered HTTP standalone listeners).
We talked about Java vs Python and EJB vs Zope here. Tests I've made shoed that on the same hardware (P3-4x500MHz + 2GB RAM) both JSP-no-EJB (Tomcat) and JSP+EJB (Tomcat+JBoss) are actually slower than Python+Zope. Partially - b/c the server was tow busy doing lots of swaping trying to feed a memory-hungry JVM.
Of course you can drop a cluster of super-computers to the project to feed a JVM up. But I have a feeling that even in THAT case Python+zope will work not significantly slower than EJB.
Finally I saw DocBook filter and thought it's time to be happy with OOo. Not so fast - it's broken. OOo is freezing each time it tries to use that form. Too bad:(
Well, one or two jokes per first 10-20 comments is ok. But when ALL first 10 comments are the same stupid joke and my comment is the first one which is not - that's bad.
There is spam in email, SMS and even NNTP but the society is somehow begin fighting against it using filters and laws. As for today the laws do not work, so we are filtering. The laws (karma one) do not work on/. either. So we need a filter for it.
Oh, God! 10 comments and ALL 10 are silly jokes about Beowulf clusters. If we need a Beowulf cluster then that would be to scan/. traffic and filter such stupid jokes out.
Can we at least install on/. some neural-network scanners that would mod all such obsolete jokes down?
The parent certanly does not have any experience (or a motivation) of doing right the refactoryng and migration form Java to Python.
I've done it with two projects, one was heavily overbloated with EJB, another one was a typical JSP thing. In both cases I've moved to Python+Zope and it was done pretty quickly and smoothly.
Well, I admit, I've done it without Jython, as I've found there was no need for old/new code temporary integration aside of transparent authentification (which was simple - through LDAP). And I've made sure that in the middle of the transiotion no need to share any session objects.
Performance has been improved (shut-up about that common myth that "Zope is slow"), and so has been a memory usage.
So, I know on practice - it's doable.
By the way, I've never found the situation, when you think about re-writing some Python function to C to accelerate your web-server AND Java was fast enough with the same (logically same) function. In a well designed web-system (including templates and database) a web server has no potentially bad issues. Plus, you can always cache something. And that is the same with Python and Java.
Xfree86 in Linux eats memory too. But somehow it doesn;t affect (slow down) anything. I think it's because Linux has the most optimal (as for today) schema of swapping. If X gets too much of fonts for caching, but rarely use most of them - the kernel will dump them to the swap untill they will be required. IANALKH (I Am Not A Linux Kernel hacker), so it's just my home theory.
By the way, Linux eats a free memory also - for disk caching, but it doesn't slow down anything (oppositely - it accelerates a lot!) and it releases it immidiately when more memory is required for real process needs.
I've migrated from 10.1 -> YDL 2.1 -> Gentoo 1.4 -> 10.2 on my Powerbook. I can confirm: Mac OS X is the slowest os (after Mac OS 9 of course) that was running there. As soon as I'll change my scanner to work with SANE I'll go back to Gentoo, which is the fastest OS for Mac hardware today I ever tested.
In lousy designed system many drivers might be lousy too.
I had recently been suffering from BSOD each time I've tried to access a specific file from the shared drive, which happaned to have some hardware errors. By the way, the drive was SCSI. NTFS over IDE by some reasons is more stable.
It must log the disk error message, not fail catastrophically!!!
It is Windows design that must be blamed, not anything else.
P.S. After reinstalling Linux on that PC everything works fine, including Ext3 on that SCSI drive and Samba sharing of the same set of files from the same drive. Go figure...
1. Does it include RBAC (Role-based Access control)?
2. in 2.4 ALSA worked only on 40% boxes for me (60% among x86). I am looking forward to have ALSA on PPC.
3. Are you saying it will be different (more general?) than CONFIG_BLK_DEV_NBD (Network block device support), which is already in 2.4?
Although I am not sure I understand you further example. But thanks anyway - I already see 4 different reasons to try it on production systems (adding the responsivity of desktops answered in another comment).
So, now Apple's fruit is bitten (is Bill Gates the one who did bite the side of that apple fruit originally?) AND cracked (is Linus Torvalds the one who eventually will help to split software and hardware business of this forever-troublesome company?)
Well, splitting Apple's hardware and software businesses might be a not bad idea for end-users - we will see stores selling Macs with pre-installed Linux (and BeOS too?) as well as OSX working on x86 and everything else (Pocket OSX someday?).
If Apple demands the split of Microsoft OS, Internet and Office busineses, then why not try the same medicine on Apple?
Is there anything in Europe, Australia or Russia you could use? Sure some of their bands have nothing to do with RIAA. They may have their own contracts with local labels, but those contracts could be have enough flexibility ot leave a chance for you. Take a look at the Orchestra of Oleg Lungstrem (Russian) - in their economy they may want to have one more channel of audittory and money.
By the way, if you are looking forward to open the "Blues" section, then you may want to look at Yury Naumov's music, which he is selling independently from his own site. And I remember he was (may be still) very popular in Russia. And still popular among russian immigrants in North America.
One more idea: Flamenco music teachers (from Spain, but not only) have own web-sites with their records, some still not bound to any labels.
Also I am glad to find choral music in your store. Do you accept a choral music from churches?
You have wrong logic in your sig leading to wrong conclusions. So do you in your comment.
"An open source startup" in a context of the original article means "a startup with a business model which is not under NDA". It doesn't mean "a startup selling open source software".
By the way, arguments "vi vs Emacs" are not useless in the industry where the usability is still a common problem. That's why Emacs, with its vim-mode, is the best:)
Slashdot Mirror
Well, I've got a lot of "theoretical" answers, perhaps too theoretical. So, I am keep looking for it :)
Hey, jokes aside, THAT's the vehicle NASA should use: Soyuz. It safes money and lifes. Hmm can be used in a slogan :)
Not exactly true counting the fact that Apple has even supported MkLinux those days.
Any URL proving that you are not trolling?
Why do you stop? Keep going:
A thousand lines of Python will certainly "do more" than the same number of lines of Java. Likewise (though arguably not on the same scale), a thousand lines of Lisp will certainly "do more" than the same number of lines of Python.
With even higher level languages, entire classes of mistakes cannot even exist, but I don't have time to go into the advantages of such languages like SML or Haskell.
Ok, let's just put ML and Haskell right next to Python and Lisp in the thousand line formula above :)
Programmers no matter how experienced, are going to make mistakes. What matters is how costly those mistakes are. And they cost a whole lot more in C than in Java.
The cost in a well orginized process is majorly defined by amount of debugging. Two thing worth to mention:
- the amount of low-level mistakes is lower in a higher-level programming language due to... well there is no low-level programming there
:)
- the refactoryng of a high-level language coded program is faster than of a low-level one;
But all this logic doesn't work in real life, where most of corporate decisions are based on a marketing bullshit, rather than on real benefits. That's why Java is everywhere, especially in placases where it's not supposed to be from the first place: network server-side applications and GUI.Where was that Scully when the technology was closed? Why wasn't it at least open-sourced?
So many stupidy-based decisions were, are and will be driving Apple.
What is the URL I can download Gentoo LiveDVD image?
Any URL to download the image of LiveDVD?
Don't forget also Steve Jobs, the guy who is trying to kill Linux/PPC.
True about windows, but not Linux.
Microsoft was (and still is) using lots of legal AND illegal methods to take (and keep) customers to Windows.
Linux... Wait a minute, there is no any equal Linux vendor to put here that would be appropriate to compare to microsoft and Sun regarding the marketing and selling strategies. In fact, there IS NO marketing strategies in Linux! Of course there are few commercial vendors with their distros, there are some support companies and integrators. And of course they have marketing strategies about their business. But that has nothing to do with development of Linux kernel (and a dozen of independent distros). Sometimes businesses like IBM contribute some good code, but the decisions about CVS commits usually are done by independent people (sometimes even working in a competiting company. That's the nature of open source.
So, it is not correct to say that Linux is eating Sun's customers. No, it is another way around: Sun is letting customers (existing and potential ones) to choose Linux by keeping the high prices for Sun products, inappropriately high compare to offered features AND features of competing system.
I think that Sun still has chances to save itself. Somebody indisde Sun (in fact - on atop of it) must stop covering own ass trying to prove the importance of having done investments to Solaris, must dedicate some efforts to move the rest of still superior Solaris features to Linux (especially Linux/Sparc) and must abandon Solaris in a favor of Linux/Sparc. They will save on cutting the cost of OS developemtn and support. And the customers will be more cofortable to save on skill investments.
As for BSD and make - it's pretty balanced combination, where everything old-and-well known (even when it's too obsolete already) is considered as good.
I still remember systems where bash scripts did not work as bash was broken or bash accidently was disabled for the user running the scripts. That time (3-5 years ago) there were many flamewars bash-vs-sh. What's happen? Bash is everywhere. Not precisely everywhere, but more systems have bash today.
I think the evolution of unix-like operating systems (especially Linux ones) is moved far forward enough to begin flamewars python-vs-bash. Yes, Python is not everywhere, but so isn't bash, Python has just a little bit smaller % of system with no Python, but still many of them have it. Especially if it will be required.
Why Python? Several years ago Bash won a simple sh just b/c it has a little bit more convinient programming extensions. Even with that Bash is still not exactly general programming language and if you try anything complex on it - you know what I mean. Python is much better for scripting complex tasks. Dependency checking is one of reasons to use it. String parsing, regeneration of config files - another ones. With bash you don't have a good luck to do it.
Evolution of system tools wasn't frozen these years. Get over it. It's time for real tools to come to the scene. If you don't like it - choose an OS more conservative than Linux, for example BSD - sure they will stisk to a simple sh for a couple of more decades. No wonder it's dea... sorry... :)
Well, thanks to the project like this, and Gentoo as well, in few years someone will point on this flamewar and say: "remember some conservators did not want to use Python? Now it's time for ..." Who knows, will it be Erlang or Haskell? :)
We talked about Java vs Python and EJB vs Zope here. Tests I've made shoed that on the same hardware (P3-4x500MHz + 2GB RAM) both JSP-no-EJB (Tomcat) and JSP+EJB (Tomcat+JBoss) are actually slower than Python+Zope. Partially - b/c the server was tow busy doing lots of swaping trying to feed a memory-hungry JVM.
Of course you can drop a cluster of super-computers to the project to feed a JVM up. But I have a feeling that even in THAT case Python+zope will work not significantly slower than EJB.
Finally I saw DocBook filter and thought it's time to be happy with OOo. Not so fast - it's broken. OOo is freezing each time it tries to use that form. Too bad :(
There is spam in email, SMS and even NNTP but the society is somehow begin fighting against it using filters and laws. As for today the laws do not work, so we are filtering. The laws (karma one) do not work on /. either. So we need a filter for it.
Can we at least install on /. some neural-network scanners that would mod all such obsolete jokes down?
I've done it with two projects, one was heavily overbloated with EJB, another one was a typical JSP thing. In both cases I've moved to Python+Zope and it was done pretty quickly and smoothly.
Well, I admit, I've done it without Jython, as I've found there was no need for old/new code temporary integration aside of transparent authentification (which was simple - through LDAP). And I've made sure that in the middle of the transiotion no need to share any session objects.
Performance has been improved (shut-up about that common myth that "Zope is slow"), and so has been a memory usage.
So, I know on practice - it's doable.
By the way, I've never found the situation, when you think about re-writing some Python function to C to accelerate your web-server AND Java was fast enough with the same (logically same) function. In a well designed web-system (including templates and database) a web server has no potentially bad issues. Plus, you can always cache something. And that is the same with Python and Java.
Do you mean S-exps? I agree.
By the way, Linux eats a free memory also - for disk caching, but it doesn't slow down anything (oppositely - it accelerates a lot!) and it releases it immidiately when more memory is required for real process needs.
I've migrated from 10.1 -> YDL 2.1 -> Gentoo 1.4 -> 10.2 on my Powerbook. I can confirm: Mac OS X is the slowest os (after Mac OS 9 of course) that was running there. As soon as I'll change my scanner to work with SANE I'll go back to Gentoo, which is the fastest OS for Mac hardware today I ever tested.
In lousy designed system many drivers might be lousy too.
I had recently been suffering from BSOD each time I've tried to access a specific file from the shared drive, which happaned to have some hardware errors. By the way, the drive was SCSI. NTFS over IDE by some reasons is more stable.
It must log the disk error message, not fail catastrophically!!!
It is Windows design that must be blamed, not anything else.
P.S. After reinstalling Linux on that PC everything works fine, including Ext3 on that SCSI drive and Samba sharing of the same set of files from the same drive. Go figure...
if SElinux is being ported to BSD, why not to port TrustedBSD to Linux? Just to keep the balance and the choice :)
2. in 2.4 ALSA worked only on 40% boxes for me (60% among x86). I am looking forward to have ALSA on PPC.
3. Are you saying it will be different (more general?) than CONFIG_BLK_DEV_NBD (Network block device support), which is already in 2.4?
Although I am not sure I understand you further example. But thanks anyway - I already see 4 different reasons to try it on production systems (adding the responsivity of desktops answered in another comment).
Well, splitting Apple's hardware and software businesses might be a not bad idea for end-users - we will see stores selling Macs with pre-installed Linux (and BeOS too?) as well as OSX working on x86 and everything else (Pocket OSX someday?).
If Apple demands the split of Microsoft OS, Internet and Office busineses, then why not try the same medicine on Apple?
By the way, if you are looking forward to open the "Blues" section, then you may want to look at Yury Naumov's music, which he is selling independently from his own site. And I remember he was (may be still) very popular in Russia. And still popular among russian immigrants in North America.
One more idea: Flamenco music teachers (from Spain, but not only) have own web-sites with their records, some still not bound to any labels.
Also I am glad to find choral music in your store. Do you accept a choral music from churches?
"An open source startup" in a context of the original article means "a startup with a business model which is not under NDA". It doesn't mean "a startup selling open source software".
By the way, arguments "vi vs Emacs" are not useless in the industry where the usability is still a common problem. That's why Emacs, with its vim-mode, is the best :)