On Tue, 8 Jul 1997, Greg Vence wrote: > Most of the time, certification is a great way for the vendor to make an > extra buck. I've found in the M$ world that scores on a multiple-guess > test that has been heavly coached is a _poor_ indicator real-world > results. It seems that the portfolio of recient projects usually > out-weighs the certification. Its the same reason that fresh grads have > fewing fun options. Most advertisments are looking for experiance and > really want results.
Based on my current paradigm (17 year old student willing to became an enginner, working as a sysadmin and coding as a hobby), i figured a certification, by an (free) organization, _really_ puts a lot of imparciality into the test's reputation. Remember the NIST POSIX's Compliance Testing? The Plum Hall Validation suite? (i'm not home, i really haven't opened any books before writing this. :]) there's another point: free software can be made to compete against commercial software, not just via performance and cost, but using all the available solutions. I'm almost sure a few of us can build a sofisticated (uucp+smtp+byhands) sendmail setup with little or no trouble, at an awesome price. But face this against a Win95 user? he could easily choose WinNT, put a reasonable mail server, configure with a few clicks. A lot faster, if have lots of money. A certified professional could easily help convince the unix option, by (weakly, i admit) proving he really played with it before. Given the 'hacker' nature of free software, a debian/gnu/fsoftware certified professional could add a new behaviour to the environment, by tweaking by itself all the free code available, saving a lot of money too. You know, if you have a hammer, all the troubles you see are nails. <- Apply this to a CNE, like i saw two months ago. He was heavily interested in building an ISP by using Novell Netware. > However, maybe the commercial software world should consider a POSIX > certification instead. The reasoning I'd provide for this instead of > more specific is that those specifics are fairly easly found by > individuals with enough initiative to RTFM (my def: Read The _Fine_ > Manual). Even if there is only enough initiative to ask the group on a > Linux or Debian list, there is usually sufficient quick turn-around to > support a 'flavor novice.' POSIX seems a good option, at least for me. :] > Just my view from Atlanta. replied from brazil. :] done. Aldrin Leal <[EMAIL PROTECTED]> -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .