I suppose Linux will have to do for the meantime. The trouble is which linux distro? Looks like I'd have to try CentOS.
If you can afford ESX, getting RHEL just for the support and/or the approval of the big boss is the way to go.
That said, the host OS will be stripped down to the bare minimum for both performance and security issues by anyone who is capable of managing ESX properly.
I agree on a technical level, one of the *BSDs would be better, though not by much. It's not like deciding between Linux and Windows as the host.
Skype charges about $0.02/USD per minute long distance to north america and much of Europe. You pre-pay 10 or 25/EUR for the time and it expires in 6 months. I've used it to cut my cell phone costs and drop my land line entirely.
True, when it doesn't work it doesn't work at all, though I've only had 1 call in my first week of usage that was not usable...about the same as my cell phone.
While I agree with you on the price @ $30/month, the termination fees for calling a real phone do cost money. It's a monopoly in most cases, true, and it's silly as the wires involved can be the same, though they have charged for 100+ years. Expecting them to charge nothing all the sudden isn't reasonable.
If you want to 'call' someone using the Internet only, use one of the free programs out there or Skype and just call them.
... At most other places, I'd agree with you and be all for locking the users down and not giving them any admin privledges.
I've been there; mainly social/'business requirements' issues. (Horror story: One database for a 50K/day, $1M/day check processing system had the default admin account with no password -- and accessed through about a dozen non-integrated apps that way; everyone used the admin account. 6 months of pointing out carefully to key people that this was a really bad idea made no difference. Some got it. None that got it could change anything. It's probably still that way. The rest of the systems weren't much better. I wouldn't touch them as I was not the admin on that project or one of the development staff; not my place to make that unilateral decision. I keep within my role even if it makes me jump up and down in frustration. [def: Frustration; the repression of the need to choke the living $#!7 out of some poor bastard that desperately needs it.])
I don't consider the servers special because they have better hardware, but because that's where the data lives. No data lives on the workstations here....
Good. Servers no doubt have a few weeks of backups, right?:}
Also because configuration problems in the server affect everyone, not just the user of the single workstation. Yes, a workstation can cause problems for the whole network, but a compromised server immediately puts all the data/configs for the whole network in question.
True. With that environment, you probably do system isolation at the router and elsewhere. (Ex: DB server A is used by web app server B and web app server B is used by everyone: Expose B on port 80 to the intranet, and set a route between A & B but not A & the intranet. DB A is exposed by tunnel/login, other DBs and apps are exposed on a 'need to know' basis.)
Not too restrictive, though, as you don't want folks to have a reason to sneak around and undo/workaround what you have spent time setting up. (Ex: Getting root and setting everything to 777 on your *nix box because of frequent premissions errors.)
If you are one of those rare programmers with sysadmin skills, get a job as a sysadmin and you will quickly learn that most users should not be let anywhere near a computer, let alone given admin.
BSEG! Yep, I assume every person using a system and every system connected to the network is about to cause problems with every other system it can see. It's so much easier to be nice to people when things can't be destroyed or -- at a minimum -- can be recovered.
Sorry to burst your bubble there buddy but basic sys admin stuff is well...trivial...Note I said most...and basic...not all. Programming is a much harder skill to master for most people.
I've worked with some top-notch programmers, and others that didn't know that there were things such as diff or that installing malware was a bad thing. I'd be a poor programmer...and there are plenty of poor programmers out there, just as there are plenty of admins that shouldn't be allowd on the systems they 'manage'.
Our users are the admins of their machines. They can load whatever software they want...
That's the only way to run a network of computer-savvy users. Imagine a metalworking shop that wouldn't let the machinists adjust their own wrenches. You'd have to put a call-ticket in to "Tool Technology Support" and after a few hours (if you are lucky) or days (if you aren't) some kid comes over who doesn't know anything and tries to adjust your hammer.
Not really. There are different levels of dangerous;
Total novice: Harmless and they admit ignorance. (Assumes you've configured the machine moderately first!)
Moderately experienced: Typically cautious, though not always. Usually will not go out and do anything dumb.
Begining experts:Tend to think they are experts in all software. Cocky and will assume they know what they don't and will argue that something they don't know is broken (sometimes it is...not usually!).
Experts: Have gotten burned and will be extra cautious. This group assumes they are about to make a big mistake and will plot ways of undoing the dammage before they create it. Humble but knowledgable; this group knows what they don't know.
If you put in controls and make people responsible when they screw up, you'll reduce the dammage caused by the beginning expert group. Some things should be handled by the IT department, even if it's a Jr. IT monkey.
Apparently you can't read. He didn't say they were Administrator if their DOMAIN. He said they were the admin of their own machine. HUGE difference. Apparently you have no clue how MS domain/security works.
Not to raise your ire any more...an honest observeration and question from another network admin.
Most of the networks that I've managed had desktop systems that were about as powerful -- and sometimes more powerful -- than the servers in the back room. Except for the development and customer deployment systems (mostly for custom software), the servers are not heavy duity speed machines. Most are old and taxed to the limits we can push them. To me, this is as it should be as long as we can expand the backend as necessary.
The main differences between desktop and server are the quality of the hardware, sometimes the disk array and memory, and the networking gear that ties it all together. Everything else is configuration and network design (router, physical cable, login authentication/scripts,...).
Having seen and delt with how most admins manage networks, that last step seems to be what suffers the most. There's this idea that a 'server' is something unique and special...thus, what the client systems do is not nearly as important. Yet, the client systems can and do cause most of the problems -- and are plenty capable of replacing the backend servers if folks were sneaky enough and the admins aren't doing a good job.
In the case of malware and just dumb mistakes on the user level, why allow most people to admin the desktop machines?
(Got bitten with allowing admin access to a server over the weekend, btw. Someone created and then deleted MySQL accounts on a *nix server...causing multiple problems. The person is knowledgeable enough to be dangerous, though I'm not going to yank any admin access from him yet. I think he learned not to be sloppy. [crosses fingers] I'm attempting to lead by example by documenting everything as I go along and asking for an OK before acting on larger changes. Even a minor thing such as shutting down/restarting a service for a few moments gets reported and logged before it happens.)
Horsepuckey. If you (and slashdot) found out MS was running anything else BUT MS products for internal use you'd throw a fit and condemn 'em. And personally, I'd rather use a product that the developer themselves would trust end to end.
In general, I agree -- though I can't speak for the person you're replying to. This article is on Microsoft's IT department and they should 'eat thier own dog food'. If nobody, anywhere, in Microsoft has Linux, Solaris, HPUX, and *BSD networks and an IBM mainframe or two that would be very short sighted of them. I'm betting that they have these and others as well.
That said, they do make Mac products, so the Mac group in MS should be running on nearly all Apple gear and OS up to the routers. The IT department for that part of MS would support a non-MS network...and they should.
I agree. When people rant against TCO it means that they don't get what it means or how important it is. TCO is a building block for more important parts of a business. It can not be ignored -- and should not be ignored.
Yet, people who complain about TCO tend to know that there's something missing. They're right too.
The link is to Michael Tiemann's talk on TCO, process, and product improvement. While he lays out an open source view of the world, there's nothing in what he says that requires open source at all. The same things could be done with or without software at all -- open or closed -- if you spend a little time thinking about it.
Go watch it and if you're curious, read on. If not...that's good too as I'm only going to ramble a bit;
What I take from it is that the developer should reject the impulse to build everything from scratch and build just the core tool kit for others to use. After all, you can't know what other people are thinking or what they want...even if they tell you.
Along those lines, I look for projects like Plone that build on the work that preceeded it (Python to Zope to Plone) and make it easy to design extentions (Plone Products) that interoperate with the lower levels. I avoid monolythic projects that don't seem to be flexable enough to incorporate other toolkits. This is not pre-made integration, though. Quite the opposite.
Having the lower levels available and modifiable (Python source of Zope and Plone) means that you're not locked into one and only one way of doing things if you need to make changes. The vendor or core developer(s) don't dictate what you do or how you do it. Yet, along the chain each part works well with the levels above and below it.
If your open office file is put on a disk and the disk portion with your data on it gets even the slightest bit corropted then doesn't doom any chance of recovering that file? Maybe I just spend too much time recovering files from old floppy disks gone bad that people send me and this isn't much of a problem anymore.
Nope Zip files can be recovered either entirely or in part...depending on the dammage. A minor amount of corruption may not lead to any data loss -- something that isn't true if the original uncompressed data is dammaged by the same amount.
Since the contents of the zip are text files, at worst they could be edited by hand to correct them. I can't think of a more stable document format that doesn't involve having multiple copies of the document.
Please help me.. Tabbed browsing is huge? Yes it's a huge ANNOYANCE for me. I hate programs that have this. That's what the taskbar is for. I want to see a tab down there for every single page/app/etc I have open. Do not group them together, do not only have them within each app, put one down there for each and every instance.
You sound as if you haven't used Firefox as tabbed browsing isn't enabled by default.
Tabbed browsing is so useful that it's hard to descibe to someone who hasn't tried it.
Do this: From Firefox, browse to a couple web pages you tend to use frequently, opening each up in a tab. Once you have a set of say 3-6 tabs open, bookmark all tabs and set that group of as your home page.
Now, use Firefox for a week with your multi-tab 'home page'.
Occasionally, open up new pages in tabs instead of other windows. See if it grows on you.
Will you change your mind? Come back in a week or two and tell us.
Slashdot Mirror
If you give me some idea what you really need it would be more possible to answer your question.
If you can afford ESX, getting RHEL just for the support and/or the approval of the big boss is the way to go.
That said, the host OS will be stripped down to the bare minimum for both performance and security issues by anyone who is capable of managing ESX properly.
I agree on a technical level, one of the *BSDs would be better, though not by much. It's not like deciding between Linux and Windows as the host.
True, when it doesn't work it doesn't work at all, though I've only had 1 call in my first week of usage that was not usable...about the same as my cell phone.
If you want to 'call' someone using the Internet only, use one of the free programs out there or Skype and just call them.
I've been there; mainly social/'business requirements' issues. (Horror story: One database for a 50K/day, $1M/day check processing system had the default admin account with no password -- and accessed through about a dozen non-integrated apps that way; everyone used the admin account. 6 months of pointing out carefully to key people that this was a really bad idea made no difference. Some got it. None that got it could change anything. It's probably still that way. The rest of the systems weren't much better. I wouldn't touch them as I was not the admin on that project or one of the development staff; not my place to make that unilateral decision. I keep within my role even if it makes me jump up and down in frustration. [def: Frustration; the repression of the need to choke the living $#!7 out of some poor bastard that desperately needs it.])
Good. Servers no doubt have a few weeks of backups, right? :}
True. With that environment, you probably do system isolation at the router and elsewhere. (Ex: DB server A is used by web app server B and web app server B is used by everyone: Expose B on port 80 to the intranet, and set a route between A & B but not A & the intranet. DB A is exposed by tunnel/login, other DBs and apps are exposed on a 'need to know' basis.)
Not too restrictive, though, as you don't want folks to have a reason to sneak around and undo/workaround what you have spent time setting up. (Ex: Getting root and setting everything to 777 on your *nix box because of frequent premissions errors.)
BSEG! Yep, I assume every person using a system and every system connected to the network is about to cause problems with every other system it can see. It's so much easier to be nice to people when things can't be destroyed or -- at a minimum -- can be recovered.
Being an expert in one field doesn't make you knowledgeable in another. If you really believe what you say, guess where you'd end up on my list?
I've worked with some top-notch programmers, and others that didn't know that there were things such as diff or that installing malware was a bad thing. I'd be a poor programmer...and there are plenty of poor programmers out there, just as there are plenty of admins that shouldn't be allowd on the systems they 'manage'.
Great reply. Agreed with all parts of it.
That's the only way to run a network of computer-savvy users. Imagine a metalworking shop that wouldn't let the machinists adjust their own wrenches. You'd have to put a call-ticket in to "Tool Technology Support" and after a few hours (if you are lucky) or days (if you aren't) some kid comes over who doesn't know anything and tries to adjust your hammer.
Not really. There are different levels of dangerous;
If you put in controls and make people responsible when they screw up, you'll reduce the dammage caused by the beginning expert group. Some things should be handled by the IT department, even if it's a Jr. IT monkey.
Not to raise your ire any more...an honest observeration and question from another network admin.
Most of the networks that I've managed had desktop systems that were about as powerful -- and sometimes more powerful -- than the servers in the back room. Except for the development and customer deployment systems (mostly for custom software), the servers are not heavy duity speed machines. Most are old and taxed to the limits we can push them. To me, this is as it should be as long as we can expand the backend as necessary.
The main differences between desktop and server are the quality of the hardware, sometimes the disk array and memory, and the networking gear that ties it all together. Everything else is configuration and network design (router, physical cable, login authentication/scripts, ...).
Having seen and delt with how most admins manage networks, that last step seems to be what suffers the most. There's this idea that a 'server' is something unique and special...thus, what the client systems do is not nearly as important. Yet, the client systems can and do cause most of the problems -- and are plenty capable of replacing the backend servers if folks were sneaky enough and the admins aren't doing a good job.
In the case of malware and just dumb mistakes on the user level, why allow most people to admin the desktop machines?
(Got bitten with allowing admin access to a server over the weekend, btw. Someone created and then deleted MySQL accounts on a *nix server...causing multiple problems. The person is knowledgeable enough to be dangerous, though I'm not going to yank any admin access from him yet. I think he learned not to be sloppy. [crosses fingers] I'm attempting to lead by example by documenting everything as I go along and asking for an OK before acting on larger changes. Even a minor thing such as shutting down/restarting a service for a few moments gets reported and logged before it happens.)
In general, I agree -- though I can't speak for the person you're replying to. This article is on Microsoft's IT department and they should 'eat thier own dog food'. If nobody, anywhere, in Microsoft has Linux, Solaris, HPUX, and *BSD networks and an IBM mainframe or two that would be very short sighted of them. I'm betting that they have these and others as well.
That said, they do make Mac products, so the Mac group in MS should be running on nearly all Apple gear and OS up to the routers. The IT department for that part of MS would support a non-MS network...and they should.
Was the file on the network? If so, it's Windows that is caching the unencrypted file, not Google Desktop Search.
...to cancel the rest of the magazines, I'd be happy. (OK...happier!)
It was fun reading it when I was a kid and dreaming about flying cars...but come on now!
Preload it then.
That isn't an operating system issue, it's a desktop issue.
This is for you.
Yet, people who complain about TCO tend to know that there's something missing. They're right too.
This is one amazing example of how to go beyond basic TCO.
The link is to Michael Tiemann's talk on TCO, process, and product improvement. While he lays out an open source view of the world, there's nothing in what he says that requires open source at all. The same things could be done with or without software at all -- open or closed -- if you spend a little time thinking about it.
Part of my previous /. comment;
Go watch it and if you're curious, read on. If not...that's good too as I'm only going to ramble a bit;
What I take from it is that the developer should reject the impulse to build everything from scratch and build just the core tool kit for others to use. After all, you can't know what other people are thinking or what they want...even if they tell you.
Along those lines, I look for projects like Plone that build on the work that preceeded it (Python to Zope to Plone) and make it easy to design extentions (Plone Products) that interoperate with the lower levels. I avoid monolythic projects that don't seem to be flexable enough to incorporate other toolkits. This is not pre-made integration, though. Quite the opposite.
Having the lower levels available and modifiable (Python source of Zope and Plone) means that you're not locked into one and only one way of doing things if you need to make changes. The vendor or core developer(s) don't dictate what you do or how you do it. Yet, along the chain each part works well with the levels above and below it.
If only I could run Evolution on Windows...I'd probably have even more people switching to Linux eventually.
Do you actually do any work on Ruby or do you just spend your time trolling message boards and mailing lists?"
The sad part is that I know you are serious as well as telling the truth.
I think the same thing virtually everytime I see him post.
I call BS. I checked his post history...and he mentions Ruby infrequently.
Are there actually any programs other than OpenOffice.org that can read/write in OOo formats?
Right now? Yes. Others are in the works.
The cry of an AC...how noble.
So, you agree with me? :/
Nope Zip files can be recovered either entirely or in part...depending on the dammage. A minor amount of corruption may not lead to any data loss -- something that isn't true if the original uncompressed data is dammaged by the same amount.
Since the contents of the zip are text files, at worst they could be edited by hand to correct them. I can't think of a more stable document format that doesn't involve having multiple copies of the document.
Run one of the videos. If you have Plone give it a try. Slick.
You sound as if you haven't used Firefox as tabbed browsing isn't enabled by default.
Tabbed browsing is so useful that it's hard to descibe to someone who hasn't tried it.
Do this: From Firefox, browse to a couple web pages you tend to use frequently, opening each up in a tab. Once you have a set of say 3-6 tabs open, bookmark all tabs and set that group of as your home page.
Now, use Firefox for a week with your multi-tab 'home page'.
Occasionally, open up new pages in tabs instead of other windows. See if it grows on you.
Will you change your mind? Come back in a week or two and tell us.