> First off, unless your desktop machine is running a full SMTP daemon (e.g. sendmail / postfix/ > exchange / etc.) you're not supposed to be talking to other SMTP servers on port 25. The fact that > you've been allowed to do so is laziness on pretty much everyone's part. Client machines should be > talking to their SMTP server in an authenticated manner using one of the ports like tcp/465 and the > like. Which is not a port that ISPs are blocking.
Really? Which port *would* you talk to an smtp server on? Why should anyone be forced to use their ISP's SMTP server? Many of them are very poorly maintained and are bogged down. Isn't the Internet supposed to avoid centralization?
Many ISPs do more than block incoming connections on port 25 (which they also shouldn't do). They also block outgoing connections to SMTP servers other than the ISPs SMTP server.
The reason they shouldn't block incoming port 25 is that nothing Evil can be done be setting up your mail server to *receive* mail unless it is an open relay, if it's an open relay, there are a number of other methods to prevent the system being used for outgoing spam.
Let's compare the amount of effort in two situations:
1) ISP requires all SMTP traffic to be funnelled through their SMTP
servers and someone on their network is found to be sending SPAM
- The ISP has to check the IP address of the originating system to determine which account holder generated the messages
- The ISP takes action against that particular user
2) ISP allows SMTP traffic to travel normally across the network and
someone on their network is found to be sending SPAM
- Solution is *EXACTLY THE SAME* as above
I really don't see that we're buying ourselves that much here, and I am concerned that many people will start relying too heavily on SPF making it blacklist non-SPF users instead of just using as an additional metric in determining SPAM likelyhood.
Clearly, the "average" user doesn't know where his antivirus software is on Windows either:-/
They also don't know the difference between Word® and "Word Processor" especially since document interchange is not typically something that home users do, they are typing in their word processor so they can print the document out.
Nerds greatly overestimate what "average" users do with their computers.
First point: I help newbies all the time. I do everything I can to help on mailing lists and writing documentation myself.
Second point: This is not about me helping a newbie. This is the newbie saying "Samba documentation is bad" Now, since I have personally looked at the samba documentation a number of times and found it to be quite good, I found this interesting. I went looking for documentation and found not only quantity, but quality and that it was very easy to find (there are lots of products that aren't, I am very happy to admit that).
If I'm correct, and it is high quality, then the rest of the statement follows, correct? So the crux of the issue is whether samba docs are good or no.
So far, no one has said, "Mark: take another look at that documentation, see where it says 'do xyz? There's no way a normal human will understand that."
Instead the blanket statement has been made, and it's supposed to be obvious that this is Bad Documentation.
I say the documentation is Good Documentation and have pointed to very specific useful areas, the only line of discussion I am interested in pursuing is one regarding why this is Bad Documentation.
If I am wrong, and this truly is Bad Documentation, I will recant my statement that PEBKAC.
OK, I call shenanigans on all of you. You're putting me on, right?
You did not even read this documentation that you are claiming is no good.
You also did not look for it very hard.
Let's see...
go to samba.org
click on USA
click on documentation
choose the html version of the howto guide (the first link on that page)
skim through the index and see "2. Fast Start: Cure for Impatience" and click it
Lo and behold, the two configurations you just mentioned are configurations #1 and #2 on that page.
Now, have a look at that documentation, then come back and tell me that it sucks and why, or tell me that it doesn't but that you just couldn't find it, or tell me that you were really just not looking that hard.
[data] comment = Data path = <Your_share_path> force user = <your_username> force group = users read only = No guest ok = Yes
Of course, I don't know what you're trying to accomplish, so this may not be right for you.
There are lots of places you can go to ask real people for real help. I answer questions in mailing lists as often as I can. Many others do too. Your distro almost certainly has a mailing list or support line that will hand you the answer if you don't want to read and understand the documentation.
You're absolutely right. I (not a samba developer) should have answered the question he was asking. Oh, wait. He wasn't asking a question. He was saying that Samba was badly documented in general. I have found the exact opposite to be true. There are plenty of badly documented products out there, Samba is not one of them.
If you think that a new site full of people saying "the documentation is bad" is going to help somehow, you're delusional.
Non-specific criticisms do nothing to help. I welcome questions like "I am trying to do X. I have tried Y, but am getting Z problem" In fact, I volunteer my time to help others in various forums and have written howtos for questions that I have seen come up regularly. But when someone says "Samba sucks, its documentation sucks, I'm outta here" what sort of response do you expect?
Here's an extremely naive one. Just don't expect to be able to mix styles, or use {'s or }'s as anything but indentation markers:-)
import os, sys
if not len(sys.argv) == 2:
print "usage:./evilparser evilfile.py"
sys.exit(0)
e = open(sys.argv[1]) indentlevel = 0
for line in e:
line = line.replace(';', '\n')
line = line.replace('{', '{\n')
line = line.replace('}', '\n}')
lines = line.split('\n')
for l in lines:
l = l.strip()
if not l:
continue
if l[-1] == '{':
l = l[:-1].strip()
if not l:
sys.stdout.write(':\n')
else:
sys.stdout.write('\n' + (' ' * indentlevel) + l + ':')
indentlevel += 1
elif l[-1] == '}':
indentlevel = indentlevel - 1
else:
sys.stdout.write( '\n' + (' ' * indentlevel) + l)
There are so many resources on the web for setting up samba that being unable to use a combination of pre-packaged config files, swat, and the man pages (not to mention any of the books available at actual book stores) really does say more about you than about the documentation.
- right-click the folder, choose "make link" and drag the link to where you want it
- middle-drag (strange, I know) the folder to the desired target and choose "link" from the popup
For a "shortcut" (.desktop file)
- right click in the target directory and do "Create Launcher"
- choose a type of "link"
- give it a name
- type the path in the URL box
This one is interesting: drag the folder to your panel, then from the panel to the target, then delete it from the panel.
Limited resource computing device... is that as opposed to all those unlimited computing devices we have?
-Mark
Re:We're talking about Samba and Linux here...
on
Samba 3 By Example
·
· Score: 1
Why would you care where the data is stored? Have you ever actually set up a network card, or joined a domain, or set up a scsi device by manually editing the registry? I highly doubt it.
The location of the data is not what is all-important. Making the tools that modify that data better is.
Well, that's all well and good, but the question in the Gnome camp has been which language to use to write the *platform* not which languages can be used to develop apps.
> When I do the same with this SUSE 9.0 I get zero > feedback. Zilch. I have no idea if my action was > successful or not and worse I have no idea where > this device was mounted. The process with SUSE > isn't any where as intuitive as it could and > should be.
This just isn't true. I gave SuSE 9.0 a try, and specifically tested this out about a week ago with a friend's USB drive, and it does the exact same thing you describe in OSX. Plug in the drive, an icon appears on the desktop. I was actually fairly surprised by it.
Yup, no problems. Install once with -net switch. All users have a roaming profile that gets pre-filled with a default set of configuration settings (set up as part of the default user profile, as are any number of other applications' settings).
When they log on to a different machine, it just works. I don't even know what global/local registry stuff you're talking about.
I call shenanigans. What are you using, Linux From Scratch? Wireless libraries? what wireless libraries? Wireless support is at the kernel level, the only user-space tool you might need is iwconfig which is included in every semi-recent distro. As for the config file, I strongly suspect you are making that up as well.
I tried SuSE a couple days ago (just to see where the "modern" distros stand compared to my usual Debian install) and after popping in my pcmcia wireless nic, a "New Hardware" window popped up walking me through the steps.
At Novell's Brainshare conference this week they demoed a new migration tool which saved all the user's documents and application settings, wiped the drive, installed SuSE, configured all the apps for the user (Outlook settings map to Evolution, IE bookmarks saved in Epiphany etc) all in the space of 5 minutes or so.
At the moment it looks like a pretty custom job, but I can definitely see a generic tool being in the works.
> As it is, I have to use a script and a reg hack to > make StarOffice work for all users in my labs > without foolishness.
Hmm, that sounds like foolishness to me. I have a both nt4 and win2k3 terminal server farms and OOo works just fine for all users, you have to do a -net install (this is how it has worked since the early days of StarOffice).
> I can kind of understand why Sun may not want it > to work with Terminal Services - they require > you to call them if you want to install it on a > term server
I can't even imagine where you are getting this idea...
> This is the sort of thing MS can pick on because > it's totally true.
If you have ever had the pleasure of having using the Office Resource Kit to generate an installation profile for Office on a terminal server, you might understand why they don't pick that as an issue;-)
Who needs feedback? I consider "mv file1 file2" a proclamation of what _will_ occur. No need to tell me, "your will be done." When I say "Let there be light" there will be light, and it will be good!
Slashdot Mirror
> First off, unless your desktop machine is running a full SMTP daemon (e.g. sendmail / postfix /
> exchange / etc.) you're not supposed to be talking to other SMTP servers on port 25. The fact that
> you've been allowed to do so is laziness on pretty much everyone's part. Client machines should be
> talking to their SMTP server in an authenticated manner using one of the ports like tcp/465 and the
> like. Which is not a port that ISPs are blocking.
Really? Which port *would* you talk to an smtp server on? Why should anyone be forced to use their ISP's SMTP server? Many of them are very poorly maintained and are bogged down. Isn't the Internet supposed to avoid centralization?
Many ISPs do more than block incoming connections on port 25 (which they also shouldn't do). They also block outgoing connections to SMTP servers other than the ISPs SMTP server.
The reason they shouldn't block incoming port 25 is that nothing Evil can be done be setting up your mail server to *receive* mail unless it is an open relay, if it's an open relay, there are a number of other methods to prevent the system being used for outgoing spam.
Let's compare the amount of effort in two situations:
1) ISP requires all SMTP traffic to be funnelled through their SMTP
servers and someone on their network is found to be sending SPAM
- The ISP has to check the IP address of the originating system to determine which account holder generated the messages
- The ISP takes action against that particular user
2) ISP allows SMTP traffic to travel normally across the network and
someone on their network is found to be sending SPAM
- Solution is *EXACTLY THE SAME* as above
I really don't see that we're buying ourselves that much here, and I am concerned that many people will start relying too heavily on SPF making it blacklist non-SPF users instead of just using as an additional metric in determining SPAM likelyhood.
-Mark
seppuku? That's the frisbee thing, right?
"If you succeed, everybody will be like "Holy Crap!""
-Mark
The BDFL is working for Microsoft?!? NO!!!!
(You mean there are other Guidos?)
Clearly, the "average" user doesn't know where his antivirus software is on Windows either :-/
They also don't know the difference between Word® and "Word Processor" especially since document interchange is not typically something that home users do, they are typing in their word processor so they can print the document out.
Nerds greatly overestimate what "average" users do with their computers.
-Mark
I don't know why I am even bothering.
First point: I help newbies all the time. I do everything I can to help on mailing lists and writing documentation myself.
Second point: This is not about me helping a newbie. This is the newbie saying "Samba documentation is bad" Now, since I have personally looked at the samba documentation a number of times and found it to be quite good, I found this interesting. I went looking for documentation and found not only quantity, but quality and that it was very easy to find (there are lots of products that aren't, I am very happy to admit that).
If I'm correct, and it is high quality, then the rest of the statement follows, correct? So the crux of the issue is whether samba docs are good or no.
So far, no one has said, "Mark: take another look at that documentation, see where it says 'do xyz? There's no way a normal human will understand that."
Instead the blanket statement has been made, and it's supposed to be obvious that this is Bad Documentation.
I say the documentation is Good Documentation and have pointed to very specific useful areas, the only line of discussion I am interested in pursuing is one regarding why this is Bad Documentation.
If I am wrong, and this truly is Bad Documentation, I will recant my statement that PEBKAC.
Good grief
-Mark
You did not even read this documentation that you are claiming is no good.
You also did not look for it very hard.
Let's see...
Now, have a look at that documentation, then come back and tell me that it sucks and why, or tell me that it doesn't but that you just couldn't find it, or tell me that you were really just not looking that hard.
-Mark
Very well then. You win.
-Mark
Mine works just fine for win95/98/2000/XP/2003 systems. The only caveat for nt based systems is the "encrypt passwords = yes" directive.
Here's one stolen from the Samba Howto Collection (FastStart section) Of course, I don't know what you're trying to accomplish, so this may not be right for you.
There are lots of places you can go to ask real people for real help. I answer questions in mailing lists as often as I can. Many others do too. Your distro almost certainly has a mailing list or support line that will hand you the answer if you don't want to read and understand the documentation.
-Mark
You're absolutely right. I (not a samba developer) should have answered the question he was asking. Oh, wait. He wasn't asking a question. He was saying that Samba was badly documented in general. I have found the exact opposite to be true. There are plenty of badly documented products out there, Samba is not one of them.
If you think that a new site full of people saying "the documentation is bad" is going to help somehow, you're delusional.
Non-specific criticisms do nothing to help. I welcome questions like "I am trying to do X. I have tried Y, but am getting Z problem" In fact, I volunteer my time to help others in various forums and have written howtos for questions that I have seen come up regularly. But when someone says "Samba sucks, its documentation sucks, I'm outta here" what sort of response do you expect?
-Mark
There are so many resources on the web for setting up samba that being unable to use a combination of pre-packaged config files, swat, and the man pages (not to mention any of the books available at actual book stores) really does say more about you than about the documentation.
-Mark
You can make a symlink by either
- right-click the folder, choose "make link" and drag the link to where you want it
- middle-drag (strange, I know) the folder to the desired target and choose "link" from the popup
For a "shortcut" (.desktop file)
- right click in the target directory and do "Create Launcher"
- choose a type of "link"
- give it a name
- type the path in the URL box
This one is interesting: drag the folder to your panel, then from the panel to the target, then delete it from the panel.
HTH
-Mark
That's the exact same method used to install fonts for over a year now in Gnome. I'm guessing that KDE has the same thing.
-Mark
With windows XP, gtk-wimp can use the system's theming system for native look.
-Mark
Limited resource computing device... is that as opposed to all those unlimited computing devices we have?
-Mark
Why would you care where the data is stored? Have you ever actually set up a network card, or joined a domain, or set up a scsi device by manually editing the registry? I highly doubt it.
The location of the data is not what is all-important. Making the tools that modify that data better is.
-Mark
Well, that's all well and good, but the question in the Gnome camp has been which language to use to write the *platform* not which languages can be used to develop apps.
-Mark
> When I do the same with this SUSE 9.0 I get zero
> feedback. Zilch. I have no idea if my action was
> successful or not and worse I have no idea where
> this device was mounted. The process with SUSE
> isn't any where as intuitive as it could and
> should be.
This just isn't true. I gave SuSE 9.0 a try, and specifically tested this out about a week ago with a friend's USB drive, and it does the exact same thing you describe in OSX. Plug in the drive, an icon appears on the desktop. I was actually fairly surprised by it.
-Mark
Yup, no problems. Install once with -net switch. All users have a roaming profile that gets pre-filled with a default set of configuration settings (set up as part of the default user profile, as are any number of other applications' settings).
When they log on to a different machine, it just works. I don't even know what global/local registry stuff you're talking about.
-Mark
I call shenanigans. What are you using, Linux From Scratch? Wireless libraries? what wireless libraries? Wireless support is at the kernel level, the only user-space tool you might need is iwconfig which is included in every semi-recent distro. As for the config file, I strongly suspect you are making that up as well.
I tried SuSE a couple days ago (just to see where the "modern" distros stand compared to my usual Debian install) and after popping in my pcmcia wireless nic, a "New Hardware" window popped up walking me through the steps.
-Mark
At Novell's Brainshare conference this week they demoed a new migration tool which saved all the user's documents and application settings, wiped the drive, installed SuSE, configured all the apps for the user (Outlook settings map to Evolution, IE bookmarks saved in Epiphany etc) all in the space of 5 minutes or so.
At the moment it looks like a pretty custom job, but I can definitely see a generic tool being in the works.
-Mark
> As it is, I have to use a script and a reg hack to
;-)
> make StarOffice work for all users in my labs
> without foolishness.
Hmm, that sounds like foolishness to me. I have a both nt4 and win2k3 terminal server farms and OOo works just fine for all users, you have to do a -net install (this is how it has worked since the early days of StarOffice).
> I can kind of understand why Sun may not want it
> to work with Terminal Services - they require
> you to call them if you want to install it on a
> term server
I can't even imagine where you are getting this idea...
> This is the sort of thing MS can pick on because
> it's totally true.
If you have ever had the pleasure of having using the Office Resource Kit to generate an installation profile for Office on a terminal server, you might understand why they don't pick that as an issue
-Mark
Who needs feedback? I consider "mv file1 file2" a proclamation of what _will_ occur. No need to tell me, "your will be done." When I say "Let there be light" there will be light, and it will be good!
Hmm, ok, perhaps I've gone too far with it.
-Mark
Sorry, I have to nitpick this one.
> Mod down this rant if you want, but it is an objective opinion of the consensus of this thread.
objective
adj 1: undistorted by emotion or personal bias; based on observable phenomena
opinion
n 1: a personal belief or judgment that is not founded on proof or certainty
> not everyone can program a pix without looking at it.
Uhh... what would looking at a pix tell you about "programming" it?
-Mark