It was true 5 years ago. At the time of network routers and "static-file" web servers. Today, at the time of Java-based application servers, databases and SOHO desktops it not true anymore.
BSD kernels are outdated in new hardware support, in new filesystems support. Most of Open Source Software is compiled on Linux straight forward, while on BSD BSD-specific patches are to be applied.
Linux helps you to grow, to get smarter. microsoft and apple addict you with their look-n-feel. Starring at GUI and dragging-n-dropping things improves your "professional mouse clicking" skills. But if you want to start thinking then Linux is your right choice.
Linux helps you to understand the system architecture, network protocols and even programming concepts. In a lot of cases of Linux usage scenario you don't have to program or administrate your system by yourself. But the good news that such chance is not hidden from you - it's around the korner.
Try to think about it (if you are still capable to think). Do you want to live the life of "professional mouse clicker"?
Look at all these OS-X-addicted users - it's alive picture of microsoft-addiced users back a decade ago. When people don't have brains they demand look-n-feel. And then write in a resume - "professional mouse clicker".
Actually I am not against look-n-feel, when it's not a goal, when it just helps, rather than disturbs.
I don't want to sit. I want to work. Linux is a perfect system to work - development, productivity, even home office - to work, what I want. OS X is a perfect system to sit and enjoy the view, what is not exactly I would want to limit myself. So, I run Linux (YDL) and work. And enjoy looking at how other people enjoy looking at results of my work.
The real consumer of that power would be : Mozila + Evolution + Open Office + Java on the same computer at the same time in environment of multi X11-terminals:)
I hope they will support Linux. Not only Linux/x86, but Linux in general, including x86, PPC, m68k, mips, Sparc, S390 - all official (at least non-experimental) architectures in the current Linux kernel source tree.
It might be a budget issue, so Macromedia might use the experience of Sun and Blackdown. It would be not big deal to create a team of developers porting Macromedia plugin and authoring tools from Linux x86 to the other architectures. I sure that developers from other linux platforms and from Mozilla team would help.
Would it be available for Linux in general I would support Flash rather than Java raw applets. I like its design and it's easier to create new very Flashy applications extending plain web sites and making web more attractive for customers.
Why Linux in general? Think more - it's about Flash plugin in Mozilla for embedded systems (i.e. kiosks and terminals). Also I would like to run Macromedia Flash authoring tools on my Linux/PPC rather than on that stupid Mac OS.
Linux is a very good OS for software development. Macromedia must understand it.
Whatever they decide, but they must use ONLY Open Source Software on the Lunar base! It's a neutral territory and it should stay free from any proprietary software.
GNU Public License will work just perfect as it is more protected from proprietary codejacking.
Maybe even we should declare Moon as a territory of GNU Public License! Let's vote RPM as a Moon President!
existing dominating Unix config standard format is historically closed to primitive shell scripts. It's not structured, badly verified and not designed for dependency control. That's because shell script language is too primitive.
I would recommend to use XML (namely RDF) format: the structuring would be much better, it would be under very good verification and it would not a problem to check configuration dependencies. At least it would be a foundation to create a standard for configuration dependency control.
parse all configs to/proc in XML format (or even better RDF) - let's stop Babilon Tower at last.
Re:I like Slackware's .tgz
on
Is RPM Doomed?
·
· Score: 1
Besides package dependency checking the installation needs configuration dependency checking: ports opened, apache folder aliases, sendmail user aliases and so on. Some of them is easy to check by awk/grep, some of them would take some scripts (are DB tables created and filled out?). RPM potentially can handle some of such issues. But how about PostgreSQL or Xemacs or Perl or Python which could be build after configuring with a broad range of possible options (see configure --help)?
The reality proves: tar.gz + natural brains cannot be substituted by any package dependency control:)
After years of experience with Slackware and RedHat, I've been convinced by my friends to try FreeBSD and to try BSD ports. The result did no meet my expectations. Everything related to Java did not work without intensive handwork. In fact, I've been hacking more than I've done it in Linux.
Well, I use RPMs only upto the point I've finished OS installation. After that point - I use only tarballs. Source, if possible, or binary, if no source code (i.e. Java)../configure --[option] - that's the tool you cannot compensate neither in BSD ports nor in
RPMs.
It doesn't mean I dislike BSD ports. It means that the worth of Linux benefits (better hardware and software support) is greater than the small benefits of BSD ports.
Conclusion: if you are a real sysadmin, smart and not lazy, Linux is your right choice.
better 3rd party intergation through Apache modules (.Net?)
can easily be a front-end for back-office applications
better compliance with open standards
Apache, drawbacks
no focus on COM, although there are some extensions for IIS
no focus on C#, although no problem to support C# Vm
too young for win32
PostgreSQL can be considered as an alternative option to MS Access or to MS SQL or to both.
PostgreSQL, benefits
less cost of production
better extensibility
better scalability both up and down
can easily be a DB for back-office applications
better compliance with open standards
PostgreSQL, drawbacks
no focus on COM, although it's not needed - there is ODBC
no native win32 - cygwin is required
no easy-to-use GUI-based DB-admin, although it's easy to fix with MS Jet.
Mozilla can substitute IE or co-exist with it.
Mozilla, benefits
less cost of production
better extensibility
Gecko can be embedded to OS same as IE was. But even better because of XUL
better compliance with open standards
Mozilla, drawbacks
no focus on COM, although some XPCOM mimics it
no focus on.Net, although it's easy to add.
it is slow, although it can be fixed with pre-loaded binaries.
Jabber could an alternative option to NetMeeting/MSMessanger.
Jabber, benefits
less cost of production
more compatibility with competitors
better compliance with open standards
Jabber, drawbacks
no focus on COM
GNOME could be ported finally to win32 (or at least to cygwin. It could be an alternative (to win32/VDM) option at the boot time, like in XDM you choose a desktop manager.
GNOME, benefits
less cost of production
better designed
more isolated from the kernel layer
better compliance with open standards
better designed with X11, although it is slow
GNOME, drawbacks
no focus on COM, although, is it needed?
GPL requires a careful approach. Will RMS approve a new set of exceptions from his rules?
not stable enough
slow with X11, although it might be improved with direct rendering
Linux, as a set of the kernel and utils could be used benith propretary GUI, like BSD under Aqua in Mac OS X
Linux, benefits
less cost of production
more stable
more isolated from the graphic layer
better compliance with open standards
better scalable up and down
less memory footprint
Linux, drawbacks
no focus on COM, although, is it needed at all?
GPL requires a careful approach. Will RMS approve a new set of exceptions from his rules?
Generally, all components listed above can either substitute their MS rivals or co-exist with them.
There is a very interesting question: what will be left if MS choose ALL components from the list above? You got it right - just yet another Linux distro, competing with RedHat and Debian. It would be more difficult time for RedHat and Debian. It would be less expensive production for Microsoft. And it would be more interesting time to live for all of us:)
XML is a format to share between different humans and different machines. People, speaking on the same language, prefer a natural lang like English. Otherwise they translated. Machines of the same platform can use same binary format, like ELF. Otherwise they use scripts. XML is a similar format of inter-lingua.
From the other point XML is not an inter-lingua by itself - it doesn't care about semantic. You need RDF, DAML, OIL, XML-ized Prolog and so on for it. XML is like ethernet in networks, if you will, it's on low level of system design.
One more point - binary format still requires format tagging. What's wrong with XML tags? Are they really big? I thing XML tags are same big for inter-lingua tagging as TCP packet headers. Should we cut or comprtess TCP packet headers?
Assembler is even better, isn't it?
- It works perfectly with relational algebra;
- It allows meta-programming;
- Predicates are the perfect fit for WHERE expressions;
- Database concept is already in Prolog.
The only problem is the lack of Prolog skills in Oracle CorporationIt was true 5 years ago. At the time of network routers and "static-file" web servers. Today, at the time of Java-based application servers, databases and SOHO desktops it not true anymore.
BSD kernels are outdated in new hardware support, in new filesystems support. Most of Open Source Software is compiled on Linux straight forward, while on BSD BSD-specific patches are to be applied.
Today Linux is everything BSD wants to be.
Linux helps you to understand the system architecture, network protocols and even programming concepts. In a lot of cases of Linux usage scenario you don't have to program or administrate your system by yourself. But the good news that such chance is not hidden from you - it's around the korner.
Try to think about it (if you are still capable to think). Do you want to live the life of "professional mouse clicker"?
Actually I am not against look-n-feel, when it's not a goal, when it just helps, rather than disturbs.
I don't want to sit. I want to work. Linux is a perfect system to work - development, productivity, even home office - to work, what I want. OS X is a perfect system to sit and enjoy the view, what is not exactly I would want to limit myself. So, I run Linux (YDL) and work. And enjoy looking at how other people enjoy looking at results of my work.
Does it make any sense?
How about desktop as a personal web-portal? It should have configurable/editable portlets in customizable layout.
Of course I would like to have an access to backend of such portals and therefore XUL/RDF of Mozilla is the way to go.
Some of portlets will be gnome-panels displayed in any customized locations (not olnly top/bottom/left/right).
The real consumer of that power would be : Mozila + Evolution + Open Office + Java on the same computer at the same time in environment of multi X11-terminals :)
It might be a budget issue, so Macromedia might use the experience of Sun and Blackdown. It would be not big deal to create a team of developers porting Macromedia plugin and authoring tools from Linux x86 to the other architectures. I sure that developers from other linux platforms and from Mozilla team would help.
Would it be available for Linux in general I would support Flash rather than Java raw applets. I like its design and it's easier to create new very Flashy applications extending plain web sites and making web more attractive for customers.
Why Linux in general? Think more - it's about Flash plugin in Mozilla for embedded systems (i.e. kiosks and terminals). Also I would like to run Macromedia Flash authoring tools on my Linux/PPC rather than on that stupid Mac OS.
Linux is a very good OS for software development. Macromedia must understand it.
It's not very strange. It just means that options should be organized/categorized/foldered better.
It might be useful to combine the best experience from M$ Win2k Administrator's console and IBM AIX Smith.
Also, it would be nice to have typical profiles. Hide some (or lot) of options in "non-super-user" profiles.
I hope GNOME 2.1 options will be more smart organized than now. It may take knowledge management patterns but it will worth.
GNU Public License will work just perfect as it is more protected from proprietary codejacking.
Maybe even we should declare Moon as a territory of GNU Public License! Let's vote RPM as a Moon President!
XML (especially DOM) is another way to get all my resources utilized. Although it takes it for good reasons.
Photoshop was another big demander of 32-bit. With 3D animation studios 64-bit will be a much better choice.
Also I expect a new wave (generation) of AI applications: in games, in financial applications and in search engines. And probably in speech interface.
Finally, it's not far away I will have 32GB of RAM on my home server. 32-bit is not good to access it :)
- not lazzy
- eager to learn
- having time to learn
Newbies in Wallmart should take Mandrake or RedHat.Or better Yellow Dog Linux :))
I would recommend to use XML (namely RDF) format: the structuring would be much better, it would be under very good verification and it would not a problem to check configuration dependencies. At least it would be a foundation to create a standard for configuration dependency control.
parse all configs to /proc in XML format (or even better RDF) - let's stop Babilon Tower at last.
The reality proves: tar.gz + natural brains cannot be substituted by any package dependency control :)
Well, I use RPMs only upto the point I've finished OS installation. After that point - I use only tarballs. Source, if possible, or binary, if no source code (i.e. Java). ./configure --[option] - that's the tool you cannot compensate neither in BSD ports nor in
RPMs.
It doesn't mean I dislike BSD ports. It means that the worth of Linux benefits (better hardware and software support) is greater than the small benefits of BSD ports.
Conclusion: if you are a real sysadmin, smart and not lazy, Linux is your right choice.
- PostgreSQL and XML (Xalan, 4Suite)
- Java and PostgreSQL (JDBC, Struts, JBoss)
- PostgreSQL and Zope
- Web Warehouse with PostgreSQL
If you have any ideas, proposals and suggestions - my email is axxackall@yahoo.comI just wish there was more Open Source stuff out there for Java related stuff...
Try Cocoon (if you love functional programming on XSL) or Struts (if you are poisoned by EJB) from Apache. Both work nice with PostgreSQL.
Or better try Zope and Python if you love OOP and hate Java-monster at the same time.
Apache is an obvious alternative option to IIS.
PostgreSQL can be considered as an alternative option to MS Access or to MS SQL or to both.
Mozilla can substitute IE or co-exist with it.
Jabber could an alternative option to NetMeeting/MSMessanger.
GNOME could be ported finally to win32 (or at least to cygwin. It could be an alternative (to win32/VDM) option at the boot time, like in XDM you choose a desktop manager.
Linux, as a set of the kernel and utils could be used benith propretary GUI, like BSD under Aqua in Mac OS X
Generally, all components listed above can either substitute their MS rivals or co-exist with them.
There is a very interesting question: what will be left if MS choose ALL components from the list above? You got it right - just yet another Linux distro, competing with RedHat and Debian. It would be more difficult time for RedHat and Debian. It would be less expensive production for Microsoft. And it would be more interesting time to live for all of us :)
Tired from semantical hard-coding? Try RDF-schema and ontologies.
From the other point XML is not an inter-lingua by itself - it doesn't care about semantic. You need RDF, DAML, OIL, XML-ized Prolog and so on for it. XML is like ethernet in networks, if you will, it's on low level of system design.
One more point - binary format still requires format tagging. What's wrong with XML tags? Are they really big? I thing XML tags are same big for inter-lingua tagging as TCP packet headers. Should we cut or comprtess TCP packet headers?
Here is the link to read at your free time: javascript