I'm not bashing MCSE-qualified people at all. I'm just saying that these are a known quantity and they are far more plentiful than a truly qualified Linux admin. I wouldn't even say that they tend to be less capable, only that they are playing a different game than I think most Linux admins are. An MCSE person (a generic one) will have a rough day trying to administer a Linux box, and a Linux admin might have an almost equally tough time administering a M$ box.
Linux folks missed the starting gun on standard qualifications for technicians. In part, this is because the folks Linux was made for were technical people themselves, trying to solve individual problems. It's interesting that the group that pushes open standards for computer applications comes in much later when discussing standard qualifications for IT staff.
My own standards for hiring are maybe a bit bizarre. First, they have to be able to write a resume. If it's longer than 2 pages, that counts heavily against them. A commercially prepared "standard form" resume counts heavily against them. Second, they have to pass muster with all the senior staff at my place of work, which amounts to 4 of us. That covers personality and technical knowledge. Third, they have to pass a test that demonstrates their ability to perform the task required. For an IT staff, we would break a server and let them figure out how to fix it. For a programmer, we have them build a program. The score is based on completion time, or on how far they got before the time is up.
Back to the pimped penguin, I don't have a problem with anyone who wants to do it. I just don't think every feature should be turned on with the default install. I think a lot of us would be perfectly happy with a nice lightweight window manager and a pimp-my-penguin-HOWTO.
Sick of Gentoo
Message
- humbletech99
- Veteran
- Posts: 1229
- Joined: Thu May 26, 2005 9:30 am
- Location: London
No, I have to disagree there, actually. MCSEs are not as skilled as unix dudes in general, I know from experience.
For window managers, I recommend Fluxbox, it's minimal, comes with plain themes, but is ultra configurable and there are more snazzy themes out there than you can shake a stick at and you even write your own without complexity. (voted best wm by the way 2 years in a row at linuxquestions.org)
I'm intrigued about your hiring, how do you break the servers for them to fix?
For window managers, I recommend Fluxbox, it's minimal, comes with plain themes, but is ultra configurable and there are more snazzy themes out there than you can shake a stick at and you even write your own without complexity. (voted best wm by the way 2 years in a row at linuxquestions.org)
I'm intrigued about your hiring, how do you break the servers for them to fix?
The Human Equation:
value(geeks) > value(mundanes)
value(geeks) > value(mundanes)
I don't think I would want one that is "pimped out" either. I have tried some others but I always come back to KDE. I couldn't get Gnome to do some things that I wanted to do, because I didn't know what to do really. The Ice thingy wasn't for me either but it was REALLY fast.1clue wrote: Back to the pimped penguin, I don't have a problem with anyone who wants to do it. I just don't think every feature should be turned on with the default install. I think a lot of us would be perfectly happy with a nice lightweight window manager and a pimp-my-penguin-HOWTO.
What I like about Linux is the choices. Look at automounted CDs and such. If supermount doesn't work for you, try submount or the dbus, ivman hal combo. If you have trouble with something in windoze, your basically stuck waiting M$ to fix it.
Oh, someone asked about how much I know about windoze, I was working on computers when 3.1 came out. I'm 38 and have been messing with computers and electronics in general since I was in school. I recently threw away my old Vic-20.
My thoughts for the day.
My rig: Gigabyte GA-970A-UD3P mobo, AMD FX-8350 Eight-Core CPU, ZALMAN CNPS10X Performa CPU cooler,
G.SKILL 32GB DDR3 PC3 12800 Memory Nvidia GTX-650 video card LG W2253 Monitor
60TBs of hard drive space using LVM
Cooler Master HAF-932 Case
G.SKILL 32GB DDR3 PC3 12800 Memory Nvidia GTX-650 video card LG W2253 Monitor
60TBs of hard drive space using LVM
Cooler Master HAF-932 Case
Consider that most linux hacks won't apply for that sort of job if they can't hold their own. Either they already have a job managing a network and they start with an old PC with Linux on it, and only put it in production when they made it work, or they get the expertise before they get the job.
A lot of people assume they're smart enough to get the job and THEN learn how to do it. Maybe true for an entry level position with a small company, but on Linux you don't have the configuration tools set up for idiots, you have a text file or a tool you need to know about.
As a result you have an MCSE, which at the entry level doesn't really say much. So yes, the least common denominator of MCSE is set up for people who change keyboards and clean mice for a living. But it's a certification that somebody can get and use right away. However, that doesn't mean that every MCSE certification can be easily acquired by somebody in 3 nights of studying either.
Trying to boil this whole discussion down, when you go to Barnes and Noble (or whatever book store) and go to the computer section, you WILL see MCSE books. If the place is tiny, you might not see much else other than * for dummies. The MCSE books always have their own section, they all sort of look like they belong on the same shelf, they always have prominent shelf space (called slotting) and there's always a bunch of them compared to anything else. Somebody who doesn't know much to start with but wants to get organized comes in, and they see a big organized display of all the things they can get certified for. They also might see a whole lot of other books which DO NOT look like they belong on the same shelf, which tends to look disorganized.
And, for what it's worth, most big companies specify minimum job requirements for IT staff as MCSE standard certifications. It's like not hiring somebody who has gone no further than a high school diploma, or buying a car without a warranty. BTW, I'm a guy who took 6 years of college but never actually came out of it with a degree. At the time I went, they didn't have "design-your-own" degrees, and since then I've always interviewed with somebody who understood the job I was interviewing with.
It's easy to misconfigure a server so it doesn't work or in some cases give it bad data or even stick in bad hardware on a non-production box that will be thrown out in a few weeks anyway. You could also give them a REAL broken box that you know the diagnosis of, and tell them to fix it or diagnose it. Another option is to give them a new bit of hardware and have them install it, or if they're a person skilled in, say, Business Objects, give them a server and a database and a universe and tell them to make some reports. In other words, we design a situation where they have to DO their job for 1 hour, then give them about 2 or 3 hours of work. You grade them on how much they got done, and how much of that works, and on how elegantly they did it. The test needs to be designed to test the things you want them to do, and on the skill level they profess to have.
This sort of test does wonders. Somebody you thought to be really solid in the interview turns out to be unable to actually do the job. It takes the test out of the realm of answering questions from A, B, C or D and into real problem solving ability.
A lot of people assume they're smart enough to get the job and THEN learn how to do it. Maybe true for an entry level position with a small company, but on Linux you don't have the configuration tools set up for idiots, you have a text file or a tool you need to know about.
As a result you have an MCSE, which at the entry level doesn't really say much. So yes, the least common denominator of MCSE is set up for people who change keyboards and clean mice for a living. But it's a certification that somebody can get and use right away. However, that doesn't mean that every MCSE certification can be easily acquired by somebody in 3 nights of studying either.
Trying to boil this whole discussion down, when you go to Barnes and Noble (or whatever book store) and go to the computer section, you WILL see MCSE books. If the place is tiny, you might not see much else other than * for dummies. The MCSE books always have their own section, they all sort of look like they belong on the same shelf, they always have prominent shelf space (called slotting) and there's always a bunch of them compared to anything else. Somebody who doesn't know much to start with but wants to get organized comes in, and they see a big organized display of all the things they can get certified for. They also might see a whole lot of other books which DO NOT look like they belong on the same shelf, which tends to look disorganized.
And, for what it's worth, most big companies specify minimum job requirements for IT staff as MCSE standard certifications. It's like not hiring somebody who has gone no further than a high school diploma, or buying a car without a warranty. BTW, I'm a guy who took 6 years of college but never actually came out of it with a degree. At the time I went, they didn't have "design-your-own" degrees, and since then I've always interviewed with somebody who understood the job I was interviewing with.
It's easy to misconfigure a server so it doesn't work or in some cases give it bad data or even stick in bad hardware on a non-production box that will be thrown out in a few weeks anyway. You could also give them a REAL broken box that you know the diagnosis of, and tell them to fix it or diagnose it. Another option is to give them a new bit of hardware and have them install it, or if they're a person skilled in, say, Business Objects, give them a server and a database and a universe and tell them to make some reports. In other words, we design a situation where they have to DO their job for 1 hour, then give them about 2 or 3 hours of work. You grade them on how much they got done, and how much of that works, and on how elegantly they did it. The test needs to be designed to test the things you want them to do, and on the skill level they profess to have.
This sort of test does wonders. Somebody you thought to be really solid in the interview turns out to be unable to actually do the job. It takes the test out of the realm of answering questions from A, B, C or D and into real problem solving ability.
- humbletech99
- Veteran
- Posts: 1229
- Joined: Thu May 26, 2005 9:30 am
- Location: London
my current employer took the approach of chatting and a bit of grilling in technical information and then gave me what they called "impossible questions" to see how I reacted, they were less interested in the exact figures I gave as they were in my thought process and how I came to conclusions. Whenever I gave an answer, they changed the problem to break it and made me recalculate. I guess that's why they called it "impossible" since they didn't expect you to win because they were shifting the goal posts. Really they were testing intelligence and analytical skills rather than pre-programmed memorised book work, since you can always gleam missing bits of information (unless you're uninitiated in the field).
In fact they never once asked me about all the quals I had, Smart people though, the company has no dead wood unlike everywhere else I've ever worked. I didn't have a degree though (waste of time in retrospect - when I compare myself to people of similar age, I'm years ahead of them, so having it tough and missing uni kinda paid off in a way...)
In fact they never once asked me about all the quals I had, Smart people though, the company has no dead wood unlike everywhere else I've ever worked. I didn't have a degree though (waste of time in retrospect - when I compare myself to people of similar age, I'm years ahead of them, so having it tough and missing uni kinda paid off in a way...)
The Human Equation:
value(geeks) > value(mundanes)
value(geeks) > value(mundanes)
That sort of stuff happens in the interview as far as I'm concerned. Asking a question and getting a response is only a true test if your job involves answering questions that somebody asks you. Our goal is to be sure they can actually do the job. No matter who it is, we try to make a test that sees if what they do actually works in a real situation. We make the situation as real as we know how, and see how they fix it.
I am as interested in the process as in the results, though. That says a lot about the applicant.
We also don't have dead wood. We're a highly team-oriented company, and if you're not on the team you don't work here. We don't all have to agree on everything though. In fact, we rarely agree on everything. However everyone will listen to everyone else, and we combine into a very effective group.
I am as interested in the process as in the results, though. That says a lot about the applicant.
We also don't have dead wood. We're a highly team-oriented company, and if you're not on the team you don't work here. We don't all have to agree on everything though. In fact, we rarely agree on everything. However everyone will listen to everyone else, and we combine into a very effective group.
Q&A doesn't really display ability. There's nothing like a live hands on test to examine someone's mantle. Theory is all well and good, but what may apply in theory may not always apply in practice. For e.g. Quamtum theory dictates that the whole earth could just instantly (more or less) pop into the other side of the universe, however in practice that isn't gonna happen in our life time or indeed the lifetime of the universe itself. An extreame example granted, but it shows my point.1clue wrote:That sort of stuff happens in the interview as far as I'm concerned. Asking a question and getting a response is only a true test if your job involves answering questions that somebody asks you. Our goal is to be sure they can actually do the job. No matter who it is, we try to make a test that sees if what they do actually works in a real situation. We make the situation as real as we know how, and see how they fix it.
I am as interested in the process as in the results, though. That says a lot about the applicant.
We also don't have dead wood. We're a highly team-oriented company, and if you're not on the team you don't work here. We don't all have to agree on everything though. In fact, we rarely agree on everything. However everyone will listen to everyone else, and we combine into a very effective group.
Sometimes in practice you need to try things outside currently known theory. Team building is easy to work on when you have a bunch of people who know what they're doing, can apply themselves and not just talk about it. The other problem with Q&A is if the questions are vaugely defeined (which they usually are) it's gonna leave someone stuck for an answer since you could spit out hundreds of different solutions and none of which might actually be correct for the problem queried since the problem was poorly presented and defined. And then you might have someone who can work problems out in their head but not be able to articulate solutions very well, and thus you can miss out on an extreamly tallented individual for that reason alone.
You're right in that Q&A does not display ability. You can tell how much somebody knows about a topic, but not anything about their reasoning ability.
Two of the more notable examples of my approach:
The first trial of this testing thing was for a programmer. Being a programmer shop, we made a test we figured few people if any would be able to finish in an hour in an unfamiliar environment. We got a new guy in, still wet behind the ears. He finished the test in 40 minutes. At that point we figured that the only problem we would really have with him was his ego.
Another one was for a Business Objects person. We had a broken BO server, none of us had the expertise to fix it and the BO support didn't know what was wrong either. The girl came in, that was her test. We didn't expect her to make it go, but wanted to see what she did. She fixed the server without using support. Whether it was a fair test or not, that's the sort of thing that gets a job.
Two of the more notable examples of my approach:
The first trial of this testing thing was for a programmer. Being a programmer shop, we made a test we figured few people if any would be able to finish in an hour in an unfamiliar environment. We got a new guy in, still wet behind the ears. He finished the test in 40 minutes. At that point we figured that the only problem we would really have with him was his ego.
Another one was for a Business Objects person. We had a broken BO server, none of us had the expertise to fix it and the BO support didn't know what was wrong either. The girl came in, that was her test. We didn't expect her to make it go, but wanted to see what she did. She fixed the server without using support. Whether it was a fair test or not, that's the sort of thing that gets a job.
- humbletech99
- Veteran
- Posts: 1229
- Joined: Thu May 26, 2005 9:30 am
- Location: London
Chicago.
We write extensions to accounting apps for big businesses.
We have a REASONABLY talented team that works hard. No really big guns, they tend to be a bit too arrogant to deal with and so they don't make really good team players. It's a small company, every voice counts and every voice is heard. We have everything from Q tips to kids, and everybody does both maintenance and new development. We get the new ones in on the design process as soon as we can as well. The pay isn't as good as the interesting nature of the job and the environment is what keeps us here. I could easily make more somewhere else, but then you don't know who you'll be with, and you don't know what sort of work you'll do either.
We write extensions to accounting apps for big businesses.
We have a REASONABLY talented team that works hard. No really big guns, they tend to be a bit too arrogant to deal with and so they don't make really good team players. It's a small company, every voice counts and every voice is heard. We have everything from Q tips to kids, and everybody does both maintenance and new development. We get the new ones in on the design process as soon as we can as well. The pay isn't as good as the interesting nature of the job and the environment is what keeps us here. I could easily make more somewhere else, but then you don't know who you'll be with, and you don't know what sort of work you'll do either.
- humbletech99
- Veteran
- Posts: 1229
- Joined: Thu May 26, 2005 9:30 am
- Location: London
fair enough, interesting...
Remember, it's not arrogance if they are as good as they think, it's only arrogance if they think they are better than they really are... this is by definition.
what OSs do you run mainly, is it heterogenous or only Linux? Is linux what you develop on or are the accounting apps windows that you develop for (I think there are more accounting packages for M$ systems - but I could be wrong as this isn't my thing...)
Remember, it's not arrogance if they are as good as they think, it's only arrogance if they think they are better than they really are... this is by definition.
what OSs do you run mainly, is it heterogenous or only Linux? Is linux what you develop on or are the accounting apps windows that you develop for (I think there are more accounting packages for M$ systems - but I could be wrong as this isn't my thing...)
The Human Equation:
value(geeks) > value(mundanes)
value(geeks) > value(mundanes)
Arrogance: Even if they are as good as they think they are, if they can't work with the rest of us each in their own capacity, then they are not on our team even if they do work here. Those people disrupt a team, and no matter how good they are they aren't as good as a good team.
I tend to train every coder that comes in, and often the QA and support people as well. I have more patience for that than most here, and I think it's crucially important. I'm very helpful at first, and then get dumber and dumber as they get to learn the ropes. I usually start out to be a lot more capable than the younger ones at least, but I work them in so they learn to know their own abilities and the abilities of the team members they have. Mostly, I train them to think about what it is they're doing, and to evaluate the methods being used. I also try to tell them that a good idea that is not brought to the team does us no good at all.
In the process, I've discovered newbies that far surpassed me in one area or the other, and are green as fresh-mown grass on the others. I've also found people who have more experience than me who were caught short in some area. This isn't cause for dismissal, it's cause for training. Also in the process, I use the things I see to train myself.
The goal of a new employee is to learn how to work in the new environment. The additional goal of a green programmer is to figure out the difference between college and the real world. That's about as far as a lot of places go in their reasoning.
On the other hand, if an employer wants a good return on their investment they don't just drop somebody in the lake and see if they learn how to swim. In every case I've seen, a proactive involvement with a good trainer benefits the company enormously in the long run -- to get rid of a bad apple, to teach somebody a critical bit of info or, at least as often, to gain insight into something we didn't know or find a new approach to something we need to improve.
It's mostly a Windows environment. The owner is Linux-savvy but runs Windows. Otherwise I'm the only one who knows much about it. I use Linux only, except for a VMware session for a Windows OS. I'm sort of phasing that out and using Terminal Services more for the Windows testing requirements. We're a Java shop, and some of our clients use Windows and Linux both. The accounting packages we interface with are such as Oracle, JD Edwards, etc. and they need not be on the same system as our software.
I tend to train every coder that comes in, and often the QA and support people as well. I have more patience for that than most here, and I think it's crucially important. I'm very helpful at first, and then get dumber and dumber as they get to learn the ropes. I usually start out to be a lot more capable than the younger ones at least, but I work them in so they learn to know their own abilities and the abilities of the team members they have. Mostly, I train them to think about what it is they're doing, and to evaluate the methods being used. I also try to tell them that a good idea that is not brought to the team does us no good at all.
In the process, I've discovered newbies that far surpassed me in one area or the other, and are green as fresh-mown grass on the others. I've also found people who have more experience than me who were caught short in some area. This isn't cause for dismissal, it's cause for training. Also in the process, I use the things I see to train myself.
The goal of a new employee is to learn how to work in the new environment. The additional goal of a green programmer is to figure out the difference between college and the real world. That's about as far as a lot of places go in their reasoning.
On the other hand, if an employer wants a good return on their investment they don't just drop somebody in the lake and see if they learn how to swim. In every case I've seen, a proactive involvement with a good trainer benefits the company enormously in the long run -- to get rid of a bad apple, to teach somebody a critical bit of info or, at least as often, to gain insight into something we didn't know or find a new approach to something we need to improve.
It's mostly a Windows environment. The owner is Linux-savvy but runs Windows. Otherwise I'm the only one who knows much about it. I use Linux only, except for a VMware session for a Windows OS. I'm sort of phasing that out and using Terminal Services more for the Windows testing requirements. We're a Java shop, and some of our clients use Windows and Linux both. The accounting packages we interface with are such as Oracle, JD Edwards, etc. and they need not be on the same system as our software.
There is an interesting article made by Eric Raymond regarding the subject:
http://www.catb.org/~esr/writings/cups-horror.html
http://www.catb.org/~esr/writings/cups-horror.html
Quidquid latine dictum sit, altum viditur.
[Whatever is said in Latin sounds profound.]
[Whatever is said in Latin sounds profound.]
- Shadow Skill
- Veteran
- Posts: 1023
- Joined: Sat Dec 04, 2004 8:37 pm
That article has expressed what I feel every time I have to go hunting on google or chasing down stuff in the forums in order to get very basic things to work. I basically gave up even bothering with CUPS because it was an utter nightmare to deal with and even if I got it to print it was basically impossible for me to get the alignment to be correct. This is the kind of thing I do not want to spend my waking hours on, I would much rather spend them trying to get a custom audio converter to read files correctly, in fact I spent most of the evening after school and most of the day today doing just that and I am happy to say that the base logic is now functioning properly. I even managed to understand how to use a hash and I had to introduce myself to regular expressions so overall all of that time was worth the effort since my tool is one step closer to being ready for me to use and I actually learned something I deem important.
In short even as someone who spends his free time trying to hack together a batch audio converter and a basic screen capture utitlity hacking things like the cups config file just to get printing to work is not something I have the slightest wish to do, and I know I am not the only one who shares such a sentiment.
In short even as someone who spends his free time trying to hack together a batch audio converter and a basic screen capture utitlity hacking things like the cups config file just to get printing to work is not something I have the slightest wish to do, and I know I am not the only one who shares such a sentiment.
Ware wa mutekinari.
Wa ga kage waza ni kanau mono nashi.
Wa ga ichigeki wa mutekinari.
"First there was nothing, so the lord gave us light. There was still nothing, but at least you could see it."
Wa ga kage waza ni kanau mono nashi.
Wa ga ichigeki wa mutekinari.
"First there was nothing, so the lord gave us light. There was still nothing, but at least you could see it."
- linuxinit
- Tux's lil' helper
- Posts: 84
- Joined: Mon Mar 27, 2006 4:26 am
- Location: Bumkcuf, Egypt
- Contact:
Slackware User
I've been with Slackware for most of my time with Linux. I've dabbled around in other distros, but I always end up going back to Slackware. Gentoo is still installing as I type this, so I won't say much about Gentoo. But I never had a single problem with Slackware. I started out using XFCE, and slowly moved to E17, and now I use E17 100% of the time. I'm a Web Developer and Graphic Designer, and I've never had to use Windows or OSX to get a project done. Between Inkscape, Gimp, SciTE, Firefox, Wine w/ IE6, Safari, and a few self-wrote tools, Linux gets the job done. I'd love a PPC Mac, but just for the hardware aspect. I've used OSX at school and can safely say, I prefer Slackware/Linux. I'm moving to Gentoo mainly because I got drunk last night and decided I would install Gentoo to try it out. Well... almost 24 hours later (thank god it's a weekend), I'm still waiting on parts of KDE to compile. I don't even need KDE, just the KDE libraries for a few apps I use. The same goes for Gnome. I keep thinking KDE is completely done, then some crap like KDEedu will start emerging. I'm getting off-topic now...
Anyways... I've been using Slackware and Microsoft-free since around 97, and have been dabbling in Linux since around 96. Linux has grown a LOT in the last few years, especially since 2000. I don't think it's about to slow down either. Sure Apple's software LOOKS good, and sure most if it works great. But that's not why I use Linux. I use Linux because I can make it do most anything I want, and make it look/feel like I want. I can't stand to use other people's computers, especially Windows. It's just so NOT what I'm used to. I'm sure anyone that's used Windows/OSX most of their life would say the same about Linux/OSX/Windows.
Sure... you may have to fix something occasionally, but that's the way it is with ANY OS. Once you get everything the way you want it, leave it! Stop updating stuff that works, unless there is a vulnerability or feature that you need. What's the point in upgrading your kernel to the 2.6 series if nothing else than the fact that you are using a laptop? I have a server sitting beside me that is still running a 2.2 Kernel/Slackware 4 without a single problem. If it's not broke, don't fix it. I can understand updating vulnerable packages (maybe ~6/year max for a workstation and ~8/year with Slackware, I'm not sure about Gentoo yet), but why on Earth would you have something break every month? Stop chaning stuff! It doesn't break on it's own. If you start fucking around in the Windows Registry eventually... it's gonna break. The same goes for config files, libraries, etc... I've had maybe... 3 problems in the last 5 years with Linux. Once with a wireless card, again with an ATI card, again with a RAID setup, and again with a bandwith shaping router. I've had no problems at all on my main workstation. And I tinker too. I love getting the latest and greatest, things like Compositing, etc... I never had a problem though.
I just thought I'd share my opinion. :) No hard feelings against anyone... Take it with a grain of salt, as I haven't fully experienced Gentoo yet. And we are all adults here. I'm not trying to pull people to Slackware from Gentoo. So please don't flame for that.
I guess while I'm posting, I'll ask:
I've skimmed Doc, FAQs, etc till my eyes hurt...
I'm using the 2006.0 Live Install CD. Is there a file some file that I can copy to my non-readonly filesystem that has a list of what has and has not been installed yet? Or maybe a file that has a list of stuff that is to be installed? I'm really sick of waiting for KDE crap to compile, when I'm not even using it. (I know... I know... I shouldn't have selected it... But I was drunk last night, and yea... You know how that goes.) I'm past all of the basic install, I'm on the Extras right now. I know how to configure partitions, etc... So I don't have a problem just stopping this install right here to stop KDE in it's tracks. Is there some way for me to tell how much is left? I know there is the number of packages, but I'd like to know what they are so I can guestimate how much longer this is going to take since I'll be needing my workstation tomorrow. I was expeting this to take a while, but not over 24 hours on a 1.6GHZ with 1GB ram. KDE and Gnome are the cause of course... ;)
Thanks in advance, and if required, I'll split this into another thread in the Installation forum.
Also... Is there a mailing list I can sign up to so I get notifications of vulnerabilities? Or maybe a forum topic that I've missed? I'm used to getting an email with vulnerabilitiy warnings and updated package locations from Slackware.
Anyways... I've been using Slackware and Microsoft-free since around 97, and have been dabbling in Linux since around 96. Linux has grown a LOT in the last few years, especially since 2000. I don't think it's about to slow down either. Sure Apple's software LOOKS good, and sure most if it works great. But that's not why I use Linux. I use Linux because I can make it do most anything I want, and make it look/feel like I want. I can't stand to use other people's computers, especially Windows. It's just so NOT what I'm used to. I'm sure anyone that's used Windows/OSX most of their life would say the same about Linux/OSX/Windows.
Sure... you may have to fix something occasionally, but that's the way it is with ANY OS. Once you get everything the way you want it, leave it! Stop updating stuff that works, unless there is a vulnerability or feature that you need. What's the point in upgrading your kernel to the 2.6 series if nothing else than the fact that you are using a laptop? I have a server sitting beside me that is still running a 2.2 Kernel/Slackware 4 without a single problem. If it's not broke, don't fix it. I can understand updating vulnerable packages (maybe ~6/year max for a workstation and ~8/year with Slackware, I'm not sure about Gentoo yet), but why on Earth would you have something break every month? Stop chaning stuff! It doesn't break on it's own. If you start fucking around in the Windows Registry eventually... it's gonna break. The same goes for config files, libraries, etc... I've had maybe... 3 problems in the last 5 years with Linux. Once with a wireless card, again with an ATI card, again with a RAID setup, and again with a bandwith shaping router. I've had no problems at all on my main workstation. And I tinker too. I love getting the latest and greatest, things like Compositing, etc... I never had a problem though.
I just thought I'd share my opinion. :) No hard feelings against anyone... Take it with a grain of salt, as I haven't fully experienced Gentoo yet. And we are all adults here. I'm not trying to pull people to Slackware from Gentoo. So please don't flame for that.
I guess while I'm posting, I'll ask:
I've skimmed Doc, FAQs, etc till my eyes hurt...
I'm using the 2006.0 Live Install CD. Is there a file some file that I can copy to my non-readonly filesystem that has a list of what has and has not been installed yet? Or maybe a file that has a list of stuff that is to be installed? I'm really sick of waiting for KDE crap to compile, when I'm not even using it. (I know... I know... I shouldn't have selected it... But I was drunk last night, and yea... You know how that goes.) I'm past all of the basic install, I'm on the Extras right now. I know how to configure partitions, etc... So I don't have a problem just stopping this install right here to stop KDE in it's tracks. Is there some way for me to tell how much is left? I know there is the number of packages, but I'd like to know what they are so I can guestimate how much longer this is going to take since I'll be needing my workstation tomorrow. I was expeting this to take a while, but not over 24 hours on a 1.6GHZ with 1GB ram. KDE and Gnome are the cause of course... ;)
Thanks in advance, and if required, I'll split this into another thread in the Installation forum.
Also... Is there a mailing list I can sign up to so I get notifications of vulnerabilities? Or maybe a forum topic that I've missed? I'm used to getting an email with vulnerabilitiy warnings and updated package locations from Slackware.
For mailing lists:
http://www.gentoo.org/main/en/lists.xml
I think the Gentoo announce is the one for security updates.
It takes me about 26 hours to recompile all the packages on my system so you may have a little while yet. I think my rig is in my sig. AMD 2500+ and 1GB of ram.
I bought the Slackware and Gentoo CDs at the same time. I booted the Slack CD and then installed Gentoo. Gentoo may have been harder to install but it sure is sweet after the install. It REALLY beats Mandrake. Mandrake was a nightmare to update and everything was slow as syrup, at the North Pole.
Later
http://www.gentoo.org/main/en/lists.xml
I think the Gentoo announce is the one for security updates.
It takes me about 26 hours to recompile all the packages on my system so you may have a little while yet. I think my rig is in my sig. AMD 2500+ and 1GB of ram.
I bought the Slackware and Gentoo CDs at the same time. I booted the Slack CD and then installed Gentoo. Gentoo may have been harder to install but it sure is sweet after the install. It REALLY beats Mandrake. Mandrake was a nightmare to update and everything was slow as syrup, at the North Pole.
Later
My rig: Gigabyte GA-970A-UD3P mobo, AMD FX-8350 Eight-Core CPU, ZALMAN CNPS10X Performa CPU cooler,
G.SKILL 32GB DDR3 PC3 12800 Memory Nvidia GTX-650 video card LG W2253 Monitor
60TBs of hard drive space using LVM
Cooler Master HAF-932 Case
G.SKILL 32GB DDR3 PC3 12800 Memory Nvidia GTX-650 video card LG W2253 Monitor
60TBs of hard drive space using LVM
Cooler Master HAF-932 Case
- linuxinit
- Tux's lil' helper
- Posts: 84
- Joined: Mon Mar 27, 2006 4:26 am
- Location: Bumkcuf, Egypt
- Contact:
Coolios! Thanks a lot.
But yea. I'm really looking forward to getting this done installing so I can experience Gentoo. :) I see so many people with problems... It sort of freaks me out... But I'm pretty sure it'll be fine. I'm not a Linux noob like most. If I can handle the simpleness that is Slackware, I think I can handle Gentoo. :) But yea... I think KDE is all done and Gnome-base is going now (don't need it either :S but I guess it doesn't hurt) Gnome seems go compile MUCH faster than KDE. ;)
Thanks for the link, and I agree with you on Mand(rake/riva). ;)
But yea. I'm really looking forward to getting this done installing so I can experience Gentoo. :) I see so many people with problems... It sort of freaks me out... But I'm pretty sure it'll be fine. I'm not a Linux noob like most. If I can handle the simpleness that is Slackware, I think I can handle Gentoo. :) But yea... I think KDE is all done and Gnome-base is going now (don't need it either :S but I guess it doesn't hurt) Gnome seems go compile MUCH faster than KDE. ;)
Thanks for the link, and I agree with you on Mand(rake/riva). ;)
- Shadow Skill
- Veteran
- Posts: 1023
- Joined: Sat Dec 04, 2004 8:37 pm
Avoding using the system doesn't help anything at all and claiming something is effective because you never update it or otherwise use it is just ridiculous. Most of the time it becomes apparent that applications are successes or failures when people actually try to use them. I can't count the number of times I have tried to do something with an application only to find that it actually fails miserably when I try to do something even slightly interesting. Of course every application has this problem since non are actually perfect so I then ask myself just how annoying is the problem and is a suitable solution currently available. Simply never using systems or just not upgrading them in no way shows the effectiveness of the update tool or the programs one is using.
Ware wa mutekinari.
Wa ga kage waza ni kanau mono nashi.
Wa ga ichigeki wa mutekinari.
"First there was nothing, so the lord gave us light. There was still nothing, but at least you could see it."
Wa ga kage waza ni kanau mono nashi.
Wa ga ichigeki wa mutekinari.
"First there was nothing, so the lord gave us light. There was still nothing, but at least you could see it."
- linuxinit
- Tux's lil' helper
- Posts: 84
- Joined: Mon Mar 27, 2006 4:26 am
- Location: Bumkcuf, Egypt
- Contact:
Sorry but you are missing my point. I'm saying most of the apps that I use are already stable and I've never had a problem with not updating. I may update something like my webbrowser or mail client, but not something like gaim, scite, or inkscape, because they have already reached a point of stability that suits me. It's all a matter of opinion. I'm not saying don't update, and I'm not saying updating screws things up, I'm saying I've never had a problem with not updating something that already works fine.
Chill would ya? :)
Chill would ya? :)
- Shadow Skill
- Veteran
- Posts: 1023
- Joined: Sat Dec 04, 2004 8:37 pm
Exactly what do you call that then? Updating/configuring IS a part of using software so you are in fact saying that things don't break if you do not actually try to use them in the first place. Its like saying unplug your ethernet cable if you want security and I have never had a security problem with a box that cannot go out to the internet. Then asking directly or indirectly in your case why people have problems with using various nics etc. What I am saying is simply this the success or failure of a program is based on two things:Sure... you may have to fix something occasionally, but that's the way it is with ANY OS. Once you get everything the way you want it, leave it! Stop updating stuff that works, unless there is a vulnerability or feature that you need. What's the point in upgrading your kernel to the 2.6 series if nothing else than the fact that you are using a laptop? I have a server sitting beside me that is still running a 2.2 Kernel/Slackware 4 without a single problem. If it's not broke, don't fix it. I can understand updating vulnerable packages (maybe ~6/year max for a workstation and ~8/year with Slackware, I'm not sure about Gentoo yet), but why on Earth would you have something break every month? Stop chaning stuff! It doesn't break on it's own. If you start fucking around in the Windows Registry eventually... it's gonna break. The same goes for config files, libraries, etc... I've had maybe... 3 problems in the last 5 years with Linux. Once with a wireless card, again with an ATI card, again with a RAID setup, and again with a bandwith shaping router. I've had no problems at all on my main workstation. And I tinker too. I love getting the latest and greatest, things like Compositing, etc... I never had a problem though.
1. Doing what you wanted of it in the first place.
2. Not coming apart if you decide to utilize a feature you were not aware of at first or dying on you the moment you decide to upgrade it for any reason.
Since number one is almost a given if number two proves to be false there is a real problem.
I can point to multiple oh so wonderful experiences with ALSA deciding to stop working properly after an upgrade requiring me to dance around for as much as sixteen hours on one occassion trying to get sound to behave again. This last time it died on me I was trying to get sound with flash movies and although I did succeed it took me about a half hour of screwing around with alsaconf and recompiling my kernel at least five times before realizing alsaconf had decided to write junk data to its config file. Once I took out the junk data everything including the site I had wanted the sound for in the first place worked perfectly. The problem is that the problem started when I dared to enable oss support in ALSA which I needed for sound to actually work for the site I was trying to view.
This is the kind of thing I see with relative frequency on Linux updates that just nuke config files or otherwise send things that were once working straight to hell [Multimedia applications are especially guilty of this.] believe it or not there was a time when acc/mp4 audio streams actually worked with xine-lib on 64bit systems apparently that is never going to happen again. Reverting to the last version that supports those type of streams is not an option either since that would break all the other nice features/fixes that have come down the pipe since then. I just updated Xine-lib because I noticed someone at least tried to fix the aac issues but I can safely say that it didn't end up working. [on my box anyway.]
This whole upgrade=breakage thing is something that has happened fairly often to me in Linux even when working with the more stable distros out there far more than it has ever happened with any windows programs I use with any regularity which really tells me there is something in general about the second point I brought up that people releasing software for Linux need to look at.
Ware wa mutekinari.
Wa ga kage waza ni kanau mono nashi.
Wa ga ichigeki wa mutekinari.
"First there was nothing, so the lord gave us light. There was still nothing, but at least you could see it."
Wa ga kage waza ni kanau mono nashi.
Wa ga ichigeki wa mutekinari.
"First there was nothing, so the lord gave us light. There was still nothing, but at least you could see it."
Upgrading software SHOULD improve software stability, not decrease it. New versions are released because they fix old bugs. Code is cleaned up and some applications become quicker and more efficient with new versions.
Not to mention added features!
Today I updated compiz and Xgl. Sure, nothing was "broken" (well....), but the update added a ton of new plugins that are simply amazing. Not only that, but I'm sure it fixed bugs I didn't even know existed.
I've had systems where I haven't updated anything in over 6 months and had it get to the point where I couldn't properly update the box without spending hours and hours fixing things that had to be changed. If I'd kept maintaining the box on a monthly basis it wouldn't get to the point where it's almost a better idea if I completely reinstall everything.
Plus my machine is a desktop machine. I run a lot of bleeding edge software because lets face it.. desktop use is fast and ever changing. There are some very clever minds coming up with great, new ideas on a constant basis, and keeping your software up-to-date lets you appreciate that.
In conclusion.... run Xgl+compiz ! Holy shit the things that sucker can do now... simply amazing.
Not to mention added features!
Today I updated compiz and Xgl. Sure, nothing was "broken" (well....), but the update added a ton of new plugins that are simply amazing. Not only that, but I'm sure it fixed bugs I didn't even know existed.
I've had systems where I haven't updated anything in over 6 months and had it get to the point where I couldn't properly update the box without spending hours and hours fixing things that had to be changed. If I'd kept maintaining the box on a monthly basis it wouldn't get to the point where it's almost a better idea if I completely reinstall everything.
Plus my machine is a desktop machine. I run a lot of bleeding edge software because lets face it.. desktop use is fast and ever changing. There are some very clever minds coming up with great, new ideas on a constant basis, and keeping your software up-to-date lets you appreciate that.
In conclusion.... run Xgl+compiz ! Holy shit the things that sucker can do now... simply amazing.
- humbletech99
- Veteran
- Posts: 1229
- Joined: Thu May 26, 2005 9:30 am
- Location: London
I agree that it's better to stay up to date, and incremental is easier than all in one go, but you can't deny that although some apps gain a little efficiency here and there, generally newer programs require more power to run, an OS of 5 years ago takes less the run that the most modern OSs. I tend to stay up to date...
The Human Equation:
value(geeks) > value(mundanes)
value(geeks) > value(mundanes)
- linuxinit
- Tux's lil' helper
- Posts: 84
- Joined: Mon Mar 27, 2006 4:26 am
- Location: Bumkcuf, Egypt
- Contact:
I see what you guys are saying. :) But I just generally never updated a package unless it was known to have a vulnerability. Each person has his own way of doing things. I was just sharing mine. Gentoo finished installing last night, and I'm starting to dig in just now. I'm sure Gentoo is way different than Slackware, but I never had problems with Slackware, and I hope to never have problems with Gentoo. The more advanced your software gets, the more chances it will fail or have some problem. I just prefer to keep it simple. As I said in another post... Each to his own. :)
You guys are on opposite ends of the same discussion.
New releases of software don't always fix bugs. New features introduce new bugs and, unfortunately, even bug fixes introduce new bugs.
How effective the QA process is depends on the ability of the QA department and the application engineers to design tests for the application. If you have poor testing facilities, you can alternate back and forth between two different bugs over and over. Fixing the first causes the second, and fixing the second causes the first.
I hate to say it, but if you truly need stability and don't have any extra hardware, then the only way to get that stability is to find a setup that works for you and then just leave it alone.
Most consumers in the past have pretty much adopted that stance just because they don't know any different. A computer is an appliance, you buy one and get Windows 95 on it, and when you get sick of that you buy a new one that has Windows 98. People don't go buy new operating systems just because they can, the new OS often requires a lot more resources (can you say bloat?) and might not run very well anyway. You're just trading old bugs for new ones.
On the other hand, a computer application (use, not software) often dictates that the application (software, not use) be continuously competitive. In that case, the current approach of updating software incrementally becomes important. Most of us who keep pounding the Gentoo mirrors do it just because an incremental upgrade is easier than a large upgrade though, and some of us have not thought out the ramifications either way.
New releases of software don't always fix bugs. New features introduce new bugs and, unfortunately, even bug fixes introduce new bugs.
How effective the QA process is depends on the ability of the QA department and the application engineers to design tests for the application. If you have poor testing facilities, you can alternate back and forth between two different bugs over and over. Fixing the first causes the second, and fixing the second causes the first.
I hate to say it, but if you truly need stability and don't have any extra hardware, then the only way to get that stability is to find a setup that works for you and then just leave it alone.
Most consumers in the past have pretty much adopted that stance just because they don't know any different. A computer is an appliance, you buy one and get Windows 95 on it, and when you get sick of that you buy a new one that has Windows 98. People don't go buy new operating systems just because they can, the new OS often requires a lot more resources (can you say bloat?) and might not run very well anyway. You're just trading old bugs for new ones.
On the other hand, a computer application (use, not software) often dictates that the application (software, not use) be continuously competitive. In that case, the current approach of updating software incrementally becomes important. Most of us who keep pounding the Gentoo mirrors do it just because an incremental upgrade is easier than a large upgrade though, and some of us have not thought out the ramifications either way.