HowTo SMTP-TLS-AUTH w/ qmail [..and courier-imap, anti-virus & webmail]

Following Bruno’s post about qmail, smtp-auth with TLS support I googled for an old article I wrote some years ago (around 2002) on how to build a complete mail server using qmail, courier-imap (pop3-ssl, imap-ssl) and webmail (squirrelmail).  Here it is! Sorry no anti-spam related stuff there..

Btw, Bruno how can I comment your posts? :-S


Google maps the Universe

And so the Mountain View fellows did it again! The Sky is the final frontier..

Get the latest Google Earth and click in the “Sky” button. Explore the Universe from your chair. I’m feeling God already!

shot1.jpg shot2.jpg

Why Sun “open-sourced” UltraSPARC T2? [take two]

After this Hugo’s post over my last blog entry I feel I still have some more things to say about this matter.

UltraSPARC T2 (codename Niagara) is an high-volume microprocessor. The target market is not the average Joe, unlike the Intel/AMD commodity CPUs but the largest businnesses around the world like oil or multimedia companies (BP, Disney, Pixar, etc). Most of these companies are tied to mainframe manufacturers (i.e. IBM) for producing the biggest and powerful computing-grids in existence and the software to run over it.

Sun is now opening the doors to these same companies, both mainframe vendors and costumers . One is free to build and sell it. The other is free to develop on top of it.

I really don’t believe that there will be semi-conductor factories other than Sun’s deploying this kind of processing unit, still the option is available. What I trully believe is that some of these entities, which aren’t as strong as IBM/Sun, will embrace this opportunity and focus their efforts into selling and developing solutions while Sun provides the R&D 😉

And this is IMHO the new market that Sun is reaching: selling its Research & Development.

There will be no need for lock-in contracts. You [large company] can simply buy your mainframe-grid (for instance to HP) and have other company engineers building the tools for the task in hands.

Decoupling is healthy. It brings more quality and less costs. Isn’t that so?

Now, Hugo is right when he points out: “How much would it cost to build a processor like this without a dedicated factory? Emulators, maybe? FPGA? I’ll wait to see how many hackers will emulate this SPARC in order to improve it and find bugs. Not that it wouldn’t be cool, I just think it isn’t practical at this time.

How many architectures are supported by Linux/BSD? Have you ever asked how the hell a Linux hacker had access to an IBM s/390 mainframe in order to build a suitable kernel for it? The answer is simple: it’s the large companies that have been hiring these same hackers! So do you still think that it’ll be hard to get hands on this baby?

Why Sun “open-sourced” UltraSPARC T2?

Some of you will ask! The answer is not that hard..

Sun is not only opening this microprocessor specs. Sun is adapting the concept of OpenSource Software to the microelectronics field! And IMHO why it will be a tremendous success? I’ll tell you why..

  1. There’s no market competition – you’ll think that I’m nuts (I am!) but think of how many high-volume microprocessors you know that are available to other mainframe manufacturers than the producer of those same microprocessors? For instance, have you ever seen a SPARC-based microprocessor running on anything besides Sun’s machines? A Blue Gene running outside IBM’s? Guess not.. Well in this case you will most probably see Ultra-SPARC T2 bundled in other than Sun servers. And yes, that’s the main purpose of Sun, to feed a greater market than the existing (internal) one!
  2. More eyes, less bugs – I guess that the OSS world has been showing us the truth in this sentence. I’m already thinking about Linux, BSD and OpenSolaris hackers looking at the specs and the test-suites and finding some strange things that Sun engineers did not..
  3. Faster market adoption – The faster there are OSes and compilers supporting this architecture, the faster there will be a wide software repository available to it! And I’m still thinking on those hackers..
  4. Boost other Sun products adoption – OpenSolaris, ringing any bell?
  5. The “do-good-instead-of-evil” market effect – We’ve seen this on Google and we’re getting the same feeling about Sun (Java, OpenSolaris, Glassfish, Indiana Project..). Aren’t we people? 😉

I must admit that Jonathan Schwartz has done it again! Gratz go to Sun.

OOXML voted “Yes, with comments”

Hi fellow geeks,

In a predictable (and sad) move Microsoft got what it wanted: OOXML was accepted with a “Yes, with comments”!

I’ve posted the reasons before, and unfortunately things didn’t change. Sun and IBM were rejected to participate in the committee which Microsoft presided and so OOXML passed with a 13-7 vote. (If you’re able to find the number of Microsoft supporters that were present in the committee just by the voting results I’ll give you a candy!)

I guess we all knew this was going to happen but as someone said out there, it’s the journey that counts and not the destiny which is known from the start. For me this is just a step in a long journey! People all over the world now know what happened in Portugal like those hilarious e-mails and letters supporting OOXML that senders have never written and the lack of chairs in the meeting room, and they, these people, these costumers are getting concerned about Microsoft’s business practices.

One Gandhi’s sentence has been crossing my mind in what concerns to this Microsoft-dome-shaped world:

First they ignore you.

Then they laugh at you.

Then they fight you.

Then you win!

If anyone from Microsoft is reading me, can you please stop before freedom wins? It’s for your own good, I say..