They don't make money on Pros. The pro stuff is there as an advertisement for Joe Consumer.
This "technology" is semi-common in skis. Companies like, hmmm, HEAD.
The folks who put on LEDs to "show the dampening" stopped cause it was just a stupid gimmick that cost too much per ski.
Me? I think it makes no real different in skis. My ski's are attached to my feet and legs. They don't need help with the minor vibrations (eg. not the bumps). The dampening is minor. This isn't a power generator, it's supposed to just reduce the little "buzz" vibrations.
But there are perhaps wrist/elbow injuries associated with the "buzz" of wacking those balls around.
Gimmick still?
Oh yeah, but people will buy it. People like my uncle (who I can still beat with my 1978 racket).
1) study a little kerberos. Then ad public/private key tech.
Once upon a, I worked on an early e-commerce project. This MIGHT have been S.E.T., or maybe just a variation. Anyhow:
You select an item or items to buy, lets say $70.00 worth
You go to "check out" and offer your address, CC number, name, etc
I (or my program) takes that info, format it "right" and digitally sign it using MY key.
I encrypt that chunk with VISA's cert and pass it to VISA for validation. Getting validation, I ack (to charge) and then ship the goods.
I keep no CC info that I can read (or crackers, presumably).
In detail:
I now have a (signed) chunk of data with MY company info, how much you spent, your CC number and personal info (and NOT what you actually bought). I've encrypted that with VISAs (whoever) public key and wipe the unencrypted info. I send that encrypted data to VISA for approval.
VISA will decrypt it and check the signature. They will validate that yes, your address matches your CC number and that you can take the $70 charge. They send back a "YES" and I return an ACK that causes the charge to actually occur.
At this point, I have a record of what you bought, your address, how much. That's for my records for shipping. I have an encrypted blob that VISA has accepted and the ACK code that they returned. I DO NOT have your credit card number stored in anyway that even I can read. If there are questions, I can resend the encryption blob to VISA and they can validate it, but I need never have that info.
Note also that Visa needs not know what you bought.
We had this running in 1995 or so, on a test system. It seemed obviously "right." Is this really hard to get?
Bad guy gets my computer, they get a lot of encrypted info that even I cannot decrypt.
Best answer is a "smart CC" with challenge response. "here is your valid CC number for this transaction, never to be repeated" but we don't have that yet.
Had a friend who got an MCSE as an exercise. He's been working with Unix for a decade or more but hadn't touched Windows since 3.0 (not even 3.1). He did all his studying by book - never touched a Windows machine. All theory, zero practice.
Took the test and passed it.
He's terribly amused that he had an MSCE and would struggle to set the clock on a windows box. Achieved his goal of showing that the MSCE meant close to nothing.
OTOH, there are certificate programs like Cisco's where you do actually learn and gain skills. Those certifications are respected.
The real bottom line, however, is experience. College is helpful, especially when it comes to more advanced stuff later - you hopefully "learned how to think" during college and gained critical thinking and debugging skills, but right now people are (1) seeing a glut of somewhat experienced people from the.com blowout and (2) seeing that you've never really run machines in real world scenarios.
The classic ways into the market are to get experience by cheating:
Take a job that's beneath that and start to pick up some of the system admin slack where you can.
Do volunteer project oriented work. Having "coordinated internet connectivity and firewalls at $myTown's school system - implementing IMAP, mail and web access using LDAP for authentication and perl to report on web access" would be killer. And good, real experience.
It's something that can be done in a finite time and takes up some useful resume space. You also often work with interested parents who, often, work at or own companies.
Thinking outside the box and all will show that you have initiative as well.
Good luck
Once upon a time, the ArpaNet needed better protocols than NCP to run. The DoD wanted something that was Open. They gave a grant to Berkeley to develop network protocols that were not proprietary.
Further back
The DoD needed a way to keep an infrastructure going during and after a war. Blah Blah Blah, arpanet was born.
They needed Operating Systems for it to run on - prefererably something not closed and owned by IBM, DEC, Prime, etc.
They ended up funding a lot of the development of young Unix
Many many tools that are "just there" are there because they were developed under public grants - often not the point of the research, just tools developed in the processes of doing the work. If I develop CoolTool(TM) while being paid by taxes to do something, I can't just sell CoolTool. Frankly, often it was put out because "here was something handy and I want it to keep going but am moving on."
(GnuPlot came from Dartmouth after being written to plot data under a weather grant or something)
My point is that Open Source and the gub'mint (esp the DoD and military) have a long history together. The fact that free software is auditable and readable is often mandatory - especially for systems that will never get third party support.
No licenses I've glanced at have ever said "If you make changes for your own use, you must give them back." If this ends up being sold one day (and many military technologies (besides Tang) HAVE made it back into the civilian world), then you may have issues.
Re:I must be missing something
on
Review: SliMP3
·
· Score: 1
It's a lovely thought, but it's wrong.
Most of the time, that audio hum is from really long cables making like an antenna and picking up 60Hz Hum (50Hz in Europe - that threw me once).
How to avoid: Used BALANCED signals. One is positive, one negative (the 3rd pin is to the shield). I've successfully run a linelevel signal (1V) over 1/2 mile (1km) with this.
You can get op-aps, or just hit most pro-sound stores and pick up a transformer. It's got a two wire plug in on end (usually 1/4" plug) and a 3 pin XLR (or Canon) plug on the other. They go for around US$7.
No buzz, no hum.
OTOH, sound cards are inside an electrically noisy computer and generally suck. I'd pay for solid, good amplifiers in an isolated area away from the computer. If that's 4" from my receiver, then great, I'll just use RCA cables.
Cute, but it's using that Old IP
on
Review: SliMP3
·
· Score: 1
DHCP is fine and all, but IPv4 is so, well, so 90s.
Every modern kernel (and windows) supports
IPv6. I expect it of all my net connected boxes.
And with 64 times the Internet addressed to my house,
I'll not need that NAT crap either.
Get with the 00's (pronounced UhOh's) and get an IPv6 MP3 player.
Remember back when Unix was 12 vendors all yelling about each other that their competitors' Unix sucked and that theirs was best? Who won?
NT
It came up behind while the big boys of Unix were standing in their circle peeing at each other.
In corporate-land, the ones that have mainframes already and are facing huge IT costs and a recession, the ones who are winning the mailboxes are Exchange and Notes. They had virtually no share 10 years ago, now they have lots of network share. They also cost a lot to run (Gartner says $25+ per mailbox per month).
Now here's a company that runs on Unix, that has an IMAP server that can scale HUGELY on one (or many) boxes. That can give Secretary Joe the ability to do the admin on his group's 100 users and do that for 200 groups so that the system admin can do more important things than deal with adding a mailbox for this month's temp receptionist.
QMail? Postifix? Who? Go talk to the CEO's, the stockholders. Given Dan's support group a call at 4AM when your TLS mail isn't working right or general stability of the organization, this isn't a choice for those who don't really want to spend all their money running their computers.
Recall that when you're trying to run mail for 500+ people, there just aren't a lot of options out there. Notes and Exchange tack on the IMAP letters on their product and claim it supports standards.
For those in the Real World, take a look around at how many actual standards based tools there are with solid commercial support.
So Sendmail's MTA, IMAP server and Webmail client run on the Mainframe!? Bitchin', now I have something to counter those MSCE's who claim that we must run Exchange to survive.
but on my bike, if I'm going 80, (not that I would;), I can stop my 450 pound (200kg) bike in the same distance as that 4000lb (1800kg) SUV/boat when it's doing 50.
And STOPPING distance is really the number that we care about when we set speed limits.
Unix IS an IDE.
Help is at your fingertips, jumping from compile to debug to profile - it's all there.
If you don't have XWindows running, use EMacs for a closer integration.
Yeah, it takes a bit of effort to get really good at it, but how long did it take you to learn the details of VBasic?
Once you've learned Unix well, those skills are useful on the next language you need
It's all just a big festival of knowledge that builds on itself.
Um, guaranteed Quality of Service
Lots of host addresses (wait until every cell phone gets an IP address)
Faster stacks (no really. 6 was not designed for a PDP-11 with no RAM).
IPSEC. Well, IPSEC for IPv4 was back ported
Auto Addressing on ethernet - dhcp becomes moot
Routers that don't SMOKE with the number of routes being run through them (you haven't run a multihomed router with BGP, have you?)
And much much more.
IPv4 met our needs for a while. NAT let us get around some of the address shortage problems (and introduced its own problems).
Now the REST OF THE WORLD wants in. China could use the whole v4 address space itself (ever wonder why so much of the work is coming from Asia? See KAME.net).
Just like TCP replaced NCP. Time moves on. We went from 256 hosts to an unlimited 32 bit address space. Next stop, 128bit.
Golly! Why waste 10 minutes on Google and Yahoo when I can Ask Slashdot (tm)?
Okay, background, SMTPE, described before, tosses timecode onto the tape (in video blanking interval?). Critical to professional editing.
There are consumer versions of time code. Not as locked down to the frame, but for most purposes, enough.
Professional editing:
2 or more VCRs playing to one recording VCR. (actually in modern times, VCRs are replaced by these computer things - loaded from 1" digital tape for holding the volumes of data that no computer can).
These VCR's (or VTR - the C just came in after reel tapes disappeared)
all talk to each other over something that looks serial like. Modern ones may have RS-422 or 232.
Let's now visit consumer land:
Control-S (sony) is simply a wire send of their IR protocol.
This is actually really useful. Same signals as IR means that if you can demodulate the IR. A device called the (slink-e does this, albeit the software is on Windows only last I checked. (cool toy)).
Control-L, otherwise known as LANC, stands for Local Application Control. This protocol is used as a pro-sumer edit control protocol on a variety of Sony equipment. Other manufacturers such as Nikon and Canon also use LANC in some of their equipment.
This 2-way communication happens at 9600 baud, and is organised as 8 byte data packets sent every field.
For a good reference on LANC see, with details on the data packet format and various commands - see:
http://home.t-online.de/home/mb.koenig/lanc.htm
There are serial adapters. Or you can do your own and save the money and spend the time.
The only reason doc is a defacto 'standard', which all it's blemishes, is because Word was a good tool for general users.
Note the use of was. In the days of 1 MB Macs, Windows 3.0 and DOS, Word allowed USERS (not programmers) to get their stuff done. USERS don't want to care if they are using a computer, or a typewriter or a pen, then want to get their words out and be able to manage those words.
WordPerfect offered a fine interface - if you were a programmer from Utah. SHIFT-F7, Y Y to exit - nice. Invisible formatting commands ++ick. Who thought this was still a good idea. Beats WordStar, but so does Emacs and VI.
Word won the mind share.
In order to win back any of that, you need
tools that USERS can handle well
tools that save to and read from these formats
Given a choice, I'd be fine to start with a DOC -> XML and XML->DOC cli converters. Similiar to the graphics conversion tools that abound on Unix (pnmtoFOO and FOOto(pnm|png)), these could leave us time to actually make XML (or whatever) a USABLE standard. Get "save as" into the open tools (Star Office, Koffice, etc). Pressure the "friendly" vendors to get it into WordPerfect, FrameMaker, etc. Perhaps MS will come around.
If they don't - well, we make sure there are conversion tools that do it for them.
I'm sick of being on the road with a non-GUI ssh interface to my mail and having someone send me info I need in Word. WordPerfect won't read Word 2k doc's yet. I've taken to replying to salesmen in Applix, Makers and LaTeX just to annoy them.
So a Standard that nobody uses is quite useless (see recent IETF forays into printing standards for examples). We need a standard, but we need tools that can actually use them. Tools that USERS can use.
The internet will fail in 1997(?) from too much growth.
The stock market will crash last fall (and 11k dow and 3k Nasdaq don't count).
He's not often right. But yeah, hear him as representing OSS backlash. He makes good points. He's just wrong a bunch.
Open Source as "everyone will work for free for the good of society and make money based on charitable feelings of corporations" will fail.
Open Source as "groups of people can do good project work and will do it for personal gain and better products" DOES work (apache comes to mind - people wanted a web server that worked and worked on bits that made their own lives better - personal gain - PLUS they got other people's hacks on top of that.)
Other models about (I want the freaking source for the product I just paid a lot for because I need to support, debug, improve it).
I've been paid by my bosses to do work which has made it back into OSS projects. Not out of charity but out of "I need to fix this; I'll also share those changes" and there's little problem recognizing that.
Ya know, I used to believe that. Then the vendors who had to do with X DID get together and dictated policy.
And we got CDE. (The committee Designed Environment).
No, better to let groups like GNOME, or KDE dictate policy now.
In retrospect, yeah, we needed style dictates, but in 1989, not 2000. We NEEDED a heavy hand to say "All applications shall have a menu, and EXIT shall be under the FIRST menu item." That randomness still haunts us (xv, WordPerfect, FrameMaker, xterm, etc EACH have different ways to exit). The big change is that the Vendors are now on the sidelines. Sun has Pissed DEC out of existance, SGI is wallowing in a ditch, IBM has begun to figure out what happened. HP, well, they
still have HP-UX.
In the meantime, the Open Sores guys have taken the reins from them and started to DO something to counter Windows. Frankly, I trust them more than [vendor of choice here].
gnome/kde efforts have dealt with a lot of this quite nicely, without the brutal overhead of CDE type things. Fine. And about time. Let the best interface win.
The G4 is a Apple computer model name. Like the
Ford Mustang. The G4 is
not a chip. The G3 uses a PowerPC 775 (770?) as the CPU; combines with ROM, RAM, a backplane, case and power supply, you have a G4.
So if they want to put 2 G4's into a box, then you
have two motherboards, two set of RAM, and so forth. If you want two PPC CPU's in a box, that's a different story.
I used to do LOTS of stuff with what became OpenVision's HA and Sun's HA.
To add new services, you pretty much need a start, stop and test script. The data reside on shared disks that get transfered between the machines on failover.
On failover, the new primary will run the start script and serve PostGresql. On failure, it will try to run the STOP script before it dies and the other machine become primary (see beginning of paragraph until bored).
Of course, dedicating two machines to an HA setup and running (good fast hardware) RAID with SCSI joined to the two machines means then buying/hacking your HA software means that you're dropping some cash on this. And you're running PostGresql.
If the service is worth dropping that sort of cash on, do you really want to use a product with nobody to call when some stored procedure causes a crash? Or might it be worth putting a little cash into support as well? (and count the time when you spend 4 days debugging a problem. if it takes 4 people 4 days @ (say) US$600/day, then you've just spent around $10000. And that's cheap labor.
So, that said, it's not hard to add jo-blow services to the HA systems I've worked with.
I've seen this when people use one large partition for everything but he shows a severe lack of experience here. What was he on?
Root on my OpenBSD and FreeBSD machines are around 30 MB. This is big to allow for a couple core files. /var as needed (usually 60 meg is more than enough, large mail sites benefit from separating mail from logs anyhow. The rest (/usr,/usr/local,/tmp,/usr/X11R6,/home, whatever) depend on what you install.
I thought Solaris was bad in that you needed 40MB root partitions, but this bozo uses 300MB.
For the record, I have firewalls with 20MB root, 50MB/usr and 60MB/var where/var is the only thing writable (okay swap too). (Mounting / and/usr readonly limits bad things and helps avoid mistakes.) I've gotten SunOS down to 65MB, with a bunch of stuff on disks pinned readonly (break that!)
So wouldn't it be neat if some of these OS reviews were by professionals rather than hobbyists? Someone who is fairly adept in 3 or more different Unixes who has some background to critisize in context?
How do they get these jobs? Hey, I've had almost 12 months experience using a single operating system and I'd like to review some others.
And they get hired! How about someone who has experience with some sysV derivitives (HP-UX, Solaris) some BSD things (bsd 4.2 stuff like SunOS, NeXTstep, perhaps another BSD4.4 take), and some of the odd stepchildren like Linux and AIX. I'd expect at least 3 of the above to be understood before a person was considered to be in a position of critic of the FreeBSD flavor.
Well, let's see. the M4 front end for generating configs came with 8.6(?). I'm not saying it's pretty, but I have a 12 line m4 file that generated configs used on 2000 machines (and three more.mc files for the rest of them).
Oh yeah, change OSTYPE and the 4 config files were used on 12 different architectures. Easier than figuring out each vendor's little hacks.
Open Source is cool!
So m4 isn't really pretty, but frankly, once you get the sendmail config file done, unless you redesign your network or move the machine, the cf file just doesn't get changed. Do I want some GUI that shows me some of the stuff? How good will it be as a side project from someone who, once they've written the GUI, doesn't really need it?
Frankly, i'm MUCH more efficient with vi and the sendmail operations guide next to me than I am trying to find something in a GUI. Do I really want "what would you like Timeout.Urgent set to [1d]?" queries (times 1000 options)?
If you really WANT a gui and the other stuff pro offers, get it. It costs about what I charge in a day. It's worth while to let jimmy the operator have at it.
If you want a happy fluffy system that holds your hand each step of the way, well, unix just ain't it. Sorry. It's here for us geek mechanics who like the silicon grease under our fingernails. Oh yeah, it doesn't drive as slow and pokey as those prefabs, it can do things that perhaps nobody thought of 2 years ago (the biggest Windows innovations I've seen have been streaming media and that's been around, better, for YEARS with multicast).
So grab a wrench, dig in and see that by putting some effort into learning the m4 tool that you can get some really powerful things done. sendmail, I use it to pre-process ipfilterd and syslog.conf files.
8.10 beta is out now and, given a long beta cycle, would probably be released near year end or early next year (who's gonna change key software a month before y2k?). 8.10 has LOTS of changes, partly cause sendmail.com is paying people to work fulltime on the OSS version (kool). Reading the notice they sent, as a beta release, they can still make feature changes, so there may be more (if (client==exchange) {sleep 10;continue;}). So...
based on that, I'd guess that 8.11 would come out sometime in 2002 or so.
How long before 8.10.0beta7 is out? No guess. It's beta software. Real beta, not like Netscape/AOL or microsoft beta, but beta like "help us find bugs and work on this test version of the software" beta. It's running on the home machine as of noon, handling 10's of messages a month.
I think the security issue has been answered. Don't run sendmail, or DNS, or libc's from 1985.
As far as configuration goes, I don't think any of these posters have LOOKED at Sendmail Pro or formerly MetaInfo sendmail (the big deal about this release is that it's more Sendmail Pro than MetaInfo).
The GUI interface makes much of the m4 stuff go away. If you're still creating config files with vi and editting.CFs, then you really needed to move on 5 year ago into m4 based files (or you've been working with a vendor's mediocre sendmail.)
In 1993, at a large financial institute, I compiled up OSS sendmail and built 4 config files which ran on 12 OS's and upwards of 1500 machines. If an admin wanted to use, say, SGI's sendmail he could, I just wouldn't help him. Use mine, get support.
Only one of the 4 config files used more than 20 lines of M4.
NT Sendmail? Well, OS under it is still going to be a limit. I don't know how much mail you can get through NT's TCP stack before THAT maxes out. I'm pretty confident that it's less than Unix derivatives on the same (or lesser) hardware).
Slashdot Mirror
This "technology" is semi-common in skis. Companies like, hmmm, HEAD.
The folks who put on LEDs to "show the dampening" stopped cause it was just a stupid gimmick that cost too much per ski.
Me? I think it makes no real different in skis. My ski's are attached to my feet and legs. They don't need help with the minor vibrations (eg. not the bumps). The dampening is minor. This isn't a power generator, it's supposed to just reduce the little "buzz" vibrations.
But there are perhaps wrist/elbow injuries associated with the "buzz" of wacking those balls around.
Gimmick still?
Oh yeah, but people will buy it. People like my uncle (who I can still beat with my 1978 racket).
Once upon a, I worked on an early e-commerce project. This MIGHT have been S.E.T., or maybe just a variation. Anyhow:
- You select an item or items to buy, lets say $70.00 worth
- You go to "check out" and offer your address, CC number, name, etc
- I (or my program) takes that info, format it "right" and digitally sign it using MY key.
- I encrypt that chunk with VISA's cert and pass it to VISA for validation. Getting validation, I ack (to charge) and then ship the goods.
- I keep no CC info that I can read (or crackers, presumably).
In detail:I now have a (signed) chunk of data with MY company info, how much you spent, your CC number and personal info (and NOT what you actually bought). I've encrypted that with VISAs (whoever) public key and wipe the unencrypted info. I send that encrypted data to VISA for approval.
VISA will decrypt it and check the signature. They will validate that yes, your address matches your CC number and that you can take the $70 charge. They send back a "YES" and I return an ACK that causes the charge to actually occur.
At this point, I have a record of what you bought, your address, how much. That's for my records for shipping. I have an encrypted blob that VISA has accepted and the ACK code that they returned. I DO NOT have your credit card number stored in anyway that even I can read. If there are questions, I can resend the encryption blob to VISA and they can validate it, but I need never have that info.
Note also that Visa needs not know what you bought.
We had this running in 1995 or so, on a test system. It seemed obviously "right." Is this really hard to get?
Bad guy gets my computer, they get a lot of encrypted info that even I cannot decrypt.
Best answer is a "smart CC" with challenge response. "here is your valid CC number for this transaction, never to be repeated" but we don't have that yet.
Took the test and passed it.
He's terribly amused that he had an MSCE and would struggle to set the clock on a windows box. Achieved his goal of showing that the MSCE meant close to nothing.
OTOH, there are certificate programs like Cisco's where you do actually learn and gain skills. Those certifications are respected.
The real bottom line, however, is experience. College is helpful, especially when it comes to more advanced stuff later - you hopefully "learned how to think" during college and gained critical thinking and debugging skills, but right now people are (1) seeing a glut of somewhat experienced people from the .com blowout and (2) seeing that you've never really run machines in real world scenarios.
The classic ways into the market are to get experience by cheating:
- Take a job that's beneath that and start to pick up some of the system admin slack where you can.
- Do volunteer project oriented work. Having "coordinated internet connectivity and firewalls at $myTown's school system - implementing IMAP, mail and web access using LDAP for authentication and perl to report on web access" would be killer. And good, real experience.
Thinking outside the box and all will show that you have initiative as well.It's something that can be done in a finite time and takes up some useful resume space. You also often work with interested parents who, often, work at or own companies.
Good luck
Further back
- The DoD needed a way to keep an infrastructure going during and after a war. Blah Blah Blah, arpanet was born.
-
They needed Operating Systems for it to run on - prefererably something not closed and owned by IBM, DEC, Prime, etc.
Many many tools that are "just there" are there because they were developed under public grants - often not the point of the research, just tools developed in the processes of doing the work. If I develop CoolTool(TM) while being paid by taxes to do something, I can't just sell CoolTool. Frankly, often it was put out because "here was something handy and I want it to keep going but am moving on."They ended up funding a lot of the development of young Unix
(GnuPlot came from Dartmouth after being written to plot data under a weather grant or something)
My point is that Open Source and the gub'mint (esp the DoD and military) have a long history together. The fact that free software is auditable and readable is often mandatory - especially for systems that will never get third party support.
No licenses I've glanced at have ever said "If you make changes for your own use, you must give them back." If this ends up being sold one day (and many military technologies (besides Tang) HAVE made it back into the civilian world), then you may have issues.
Most of the time, that audio hum is from really long cables making like an antenna and picking up 60Hz Hum (50Hz in Europe - that threw me once).
How to avoid: Used BALANCED signals. One is positive, one negative (the 3rd pin is to the shield). I've successfully run a linelevel signal (1V) over 1/2 mile (1km) with this.
You can get op-aps, or just hit most pro-sound stores and pick up a transformer. It's got a two wire plug in on end (usually 1/4" plug) and a 3 pin XLR (or Canon) plug on the other. They go for around US$7.
No buzz, no hum.
OTOH, sound cards are inside an electrically noisy computer and generally suck. I'd pay for solid, good amplifiers in an isolated area away from the computer. If that's 4" from my receiver, then great, I'll just use RCA cables.
Every modern kernel (and windows) supports IPv6. I expect it of all my net connected boxes.
And with 64 times the Internet addressed to my house, I'll not need that NAT crap either.
Get with the 00's (pronounced UhOh's) and get an IPv6 MP3 player.
NT
It came up behind while the big boys of Unix were standing in their circle peeing at each other.
In corporate-land, the ones that have mainframes already and are facing huge IT costs and a recession, the ones who are winning the mailboxes are Exchange and Notes. They had virtually no share 10 years ago, now they have lots of network share. They also cost a lot to run (Gartner says $25+ per mailbox per month).
Now here's a company that runs on Unix, that has an IMAP server that can scale HUGELY on one (or many) boxes. That can give Secretary Joe the ability to do the admin on his group's 100 users and do that for 200 groups so that the system admin can do more important things than deal with adding a mailbox for this month's temp receptionist.
QMail? Postifix? Who? Go talk to the CEO's, the stockholders. Given Dan's support group a call at 4AM when your TLS mail isn't working right or general stability of the organization, this isn't a choice for those who don't really want to spend all their money running their computers.
Recall that when you're trying to run mail for 500+ people, there just aren't a lot of options out there. Notes and Exchange tack on the IMAP letters on their product and claim it supports standards.
For those in the Real World, take a look around at how many actual standards based tools there are with solid commercial support.
So Sendmail's MTA, IMAP server and Webmail client run on the Mainframe!? Bitchin', now I have something to counter those MSCE's who claim that we must run Exchange to survive.
but on my bike, if I'm going 80, (not that I would ;), I can stop my 450 pound (200kg) bike in the same distance as that 4000lb (1800kg) SUV/boat when it's doing 50.
And STOPPING distance is really the number that we care about when we set speed limits.
Unix IS an IDE.
Help is at your fingertips, jumping from compile to debug to profile - it's all there.
If you don't have XWindows running, use EMacs for a closer integration.
Yeah, it takes a bit of effort to get really good at it, but how long did it take you to learn the details of VBasic?
Once you've learned Unix well, those skills are useful on the next language you need
It's all just a big festival of knowledge that builds on itself.
Lots of host addresses (wait until every cell phone gets an IP address)
Faster stacks (no really. 6 was not designed for a PDP-11 with no RAM).
IPSEC. Well, IPSEC for IPv4 was back ported
Auto Addressing on ethernet - dhcp becomes moot
Routers that don't SMOKE with the number of routes being run through them (you haven't run a multihomed router with BGP, have you?)
And much much more.
IPv4 met our needs for a while. NAT let us get around some of the address shortage problems (and introduced its own problems).
Now the REST OF THE WORLD wants in. China could use the whole v4 address space itself (ever wonder why so much of the work is coming from Asia? See KAME.net).
Just like TCP replaced NCP. Time moves on. We went from 256 hosts to an unlimited 32 bit address space. Next stop, 128bit.
Gonna collapse from all this
Gonna run out of IP addresses by 1996 or so
Is IPv6 going to help this all?
Only if we give out prefixes in a sensible manner (for some definition of sensible - country? Provider? Height of user? You decide).
This is the Good Times Virus (tm) of Routing.
Perhaps Sun offers a site with lots and lots of connectivity for free.
It took 1/2 a day.
No init levels (I've used those about 4 times in the last 11 years of Unix admin), but start/stop scripts are simple.
Okay, background, SMTPE, described before, tosses timecode onto the tape (in video blanking interval?). Critical to professional editing.
There are consumer versions of time code. Not as locked down to the frame, but for most purposes, enough.
Professional editing:
2 or more VCRs playing to one recording VCR. (actually in modern times, VCRs are replaced by these computer things - loaded from 1" digital tape for holding the volumes of data that no computer can).
These VCR's (or VTR - the C just came in after reel tapes disappeared) all talk to each other over something that looks serial like. Modern ones may have RS-422 or 232.
Let's now visit consumer land:
Control-S (sony) is simply a wire send of their IR protocol.
This is actually really useful. Same signals as IR means that if you can demodulate the IR. A device called the (slink-e does this, albeit the software is on Windows only last I checked. (cool toy)).
LAN-C/Control-L. Described HERE, and I quote:
Note the use of was. In the days of 1 MB Macs, Windows 3.0 and DOS, Word allowed USERS (not programmers) to get their stuff done. USERS don't want to care if they are using a computer, or a typewriter or a pen, then want to get their words out and be able to manage those words.
WordPerfect offered a fine interface - if you were a programmer from Utah. SHIFT-F7, Y Y to exit - nice. Invisible formatting commands ++ick. Who thought this was still a good idea. Beats WordStar, but so does Emacs and VI.
Word won the mind share.
In order to win back any of that, you need
- tools that USERS can handle well
- tools that save to and read from these formats
Given a choice, I'd be fine to start with a DOC -> XML and XML->DOC cli converters. Similiar to the graphics conversion tools that abound on Unix (pnmtoFOO and FOOto(pnm|png)), these could leave us time to actually make XML (or whatever) a USABLE standard. Get "save as" into the open tools (Star Office, Koffice, etc). Pressure the "friendly" vendors to get it into WordPerfect, FrameMaker, etc. Perhaps MS will come around.If they don't - well, we make sure there are conversion tools that do it for them.
I'm sick of being on the road with a non-GUI ssh interface to my mail and having someone send me info I need in Word. WordPerfect won't read Word 2k doc's yet. I've taken to replying to salesmen in Applix, Makers and LaTeX just to annoy them.
So a Standard that nobody uses is quite useless (see recent IETF forays into printing standards for examples). We need a standard, but we need tools that can actually use them. Tools that USERS can use.
The internet will fail in 1997(?) from too much growth.
The stock market will crash last fall (and 11k dow and 3k Nasdaq don't count).
He's not often right. But yeah, hear him as representing OSS backlash. He makes good points. He's just wrong a bunch.
Open Source as "everyone will work for free for the good of society and make money based on charitable feelings of corporations" will fail.
Open Source as "groups of people can do good project work and will do it for personal gain and better products" DOES work (apache comes to mind - people wanted a web server that worked and worked on bits that made their own lives better - personal gain - PLUS they got other people's hacks on top of that.)
Other models about (I want the freaking source for the product I just paid a lot for because I need to support, debug, improve it).
I've been paid by my bosses to do work which has made it back into OSS projects. Not out of charity but out of "I need to fix this; I'll also share those changes" and there's little problem recognizing that.
Instant Karma.
And we got CDE. (The committee Designed Environment).
No, better to let groups like GNOME, or KDE dictate policy now.
In retrospect, yeah, we needed style dictates, but in 1989, not 2000. We NEEDED a heavy hand to say "All applications shall have a menu, and EXIT shall be under the FIRST menu item." That randomness still haunts us (xv, WordPerfect, FrameMaker, xterm, etc EACH have different ways to exit). The big change is that the Vendors are now on the sidelines. Sun has Pissed DEC out of existance, SGI is wallowing in a ditch, IBM has begun to figure out what happened. HP, well, they still have HP-UX.
In the meantime, the Open Sores guys have taken the reins from them and started to DO something to counter Windows. Frankly, I trust them more than [vendor of choice here].
gnome/kde efforts have dealt with a lot of this quite nicely, without the brutal overhead of CDE type things. Fine. And about time. Let the best interface win.
So if they want to put 2 G4's into a box, then you have two motherboards, two set of RAM, and so forth. If you want two PPC CPU's in a box, that's a different story.
To add new services, you pretty much need a start, stop and test script. The data reside on shared disks that get transfered between the machines on failover.
On failover, the new primary will run the start script and serve PostGresql. On failure, it will try to run the STOP script before it dies and the other machine become primary (see beginning of paragraph until bored).
Of course, dedicating two machines to an HA setup and running (good fast hardware) RAID with SCSI joined to the two machines means then buying/hacking your HA software means that you're dropping some cash on this. And you're running PostGresql.
If the service is worth dropping that sort of cash on, do you really want to use a product with nobody to call when some stored procedure causes a crash? Or might it be worth putting a little cash into support as well? (and count the time when you spend 4 days debugging a problem. if it takes 4 people 4 days @ (say) US$600/day, then you've just spent around $10000. And that's cheap labor.
So, that said, it's not hard to add jo-blow services to the HA systems I've worked with.
I've seen this when people use one large partition for everything but he shows a severe lack of experience here. What was he on?
Root on my OpenBSD and FreeBSD machines are around 30 MB. This is big to allow for a couple core files.
/var as needed (usually 60 meg is more than enough, large mail sites benefit from separating mail from logs anyhow. /usr/local, /tmp, /usr/X11R6, /home, whatever) depend on what you install.
The rest (/usr,
I thought Solaris was bad in that you needed 40MB root partitions, but this bozo uses 300MB.
For the record, I have firewalls with 20MB root, 50MB /usr and 60MB /var where /var is the only thing writable (okay swap too). (Mounting / and /usr readonly limits bad things and helps avoid mistakes.) I've gotten SunOS down to 65MB, with a bunch of stuff on disks pinned readonly (break that!)
So wouldn't it be neat if some of these OS reviews were by professionals rather than hobbyists? Someone who is fairly adept in 3 or more different Unixes who has some background to critisize in context?
How do they get these jobs?
Hey, I've had almost 12 months experience using a single operating system and I'd like to review some others.
And they get hired!
How about someone who has experience with some sysV derivitives (HP-UX, Solaris) some BSD things (bsd 4.2 stuff like SunOS, NeXTstep, perhaps another BSD4.4 take), and some of the odd stepchildren like Linux and AIX. I'd expect at least 3 of the above to be understood before a person was considered to be in a position of critic of the FreeBSD flavor.
(plus you could keep some really cool chips).
Not cheap. Last year's often is, if you know folks at a place with an old one.
Oh yeah, change OSTYPE and the 4 config files were used on 12 different architectures. Easier than figuring out each vendor's little hacks.
Open Source is cool!
So m4 isn't really pretty, but frankly, once you get the sendmail config file done, unless you redesign your network or move the machine, the cf file just doesn't get changed. Do I want some GUI that shows me some of the stuff? How good will it be as a side project from someone who, once they've written the GUI, doesn't really need it?
Frankly, i'm MUCH more efficient with vi and the sendmail operations guide next to me than I am trying to find something in a GUI. Do I really want "what would you like Timeout.Urgent set to [1d]?" queries (times 1000 options)?
If you really WANT a gui and the other stuff pro offers, get it. It costs about what I charge in a day. It's worth while to let jimmy the operator have at it.
If you want a happy fluffy system that holds your hand each step of the way, well, unix just ain't it. Sorry. It's here for us geek mechanics who like the silicon grease under our fingernails. Oh yeah, it doesn't drive as slow and pokey as those prefabs, it can do things that perhaps nobody thought of 2 years ago (the biggest Windows innovations I've seen have been streaming media and that's been around, better, for YEARS with multicast).
So grab a wrench, dig in and see that by putting some effort into learning the m4 tool that you can get some really powerful things done. sendmail, I use it to pre-process ipfilterd and syslog.conf files.
Sendmail 8.9 came out about 2 years ago.
8.10 beta is out now and, given a long beta cycle, would probably be released near year end or early next year (who's gonna change key software a month before y2k?). 8.10 has LOTS of changes, partly cause sendmail.com is paying people to work fulltime on the OSS version (kool). Reading the notice they sent, as a beta release, they can still make feature changes, so there may be more (if (client==exchange) {sleep 10;continue;}). So...
based on that, I'd guess that 8.11 would come out sometime in 2002 or so.
How long before 8.10.0beta7 is out? No guess. It's beta software. Real beta, not like Netscape/AOL or microsoft beta, but beta like "help us find bugs and work on this test version of the software" beta. It's running on the home machine as of noon, handling 10's of messages a month.
Don't run sendmail, or DNS, or libc's from 1985.
As far as configuration goes, I don't think any of these posters have LOOKED at Sendmail Pro or formerly MetaInfo sendmail (the big deal about this release is that it's more Sendmail Pro than MetaInfo).
The GUI interface makes much of the m4 stuff go away. If you're still creating config files with vi and editting .CFs, then you really needed to move on 5 year ago into m4 based files (or you've been working with a vendor's mediocre sendmail.)
In 1993, at a large financial institute, I compiled up OSS sendmail and built 4 config files which ran on 12 OS's and upwards of 1500 machines. If an admin wanted to use, say, SGI's sendmail he could, I just wouldn't help him. Use mine, get support.
Only one of the 4 config files used more than 20 lines of M4.
NT Sendmail? Well, OS under it is still going to be a limit. I don't know how much mail you can get through NT's TCP stack before THAT maxes out. I'm pretty confident that it's less than Unix derivatives on the same (or lesser) hardware).