Bryan,
Feel free to get as annoyed as you want, but I beg to differ on some of your points of view. Still, I respect the fact that you have a right to them, and I'm not saying that mine is better than yours but consider this,...
<snip>
Bryan Phinney wrote:
Given that someone wrote it and it appears to work for them, you might consider it a starting point and go through it line by line to see what they are doing in the script and then try to apply it to your installation. I know that is a lengthy process, however, there probably is a quicker way and that is to hire someone who has done it before and have them do it. If you go the cheap route, you should be prepared to expend some time and effort on it.
I find that many times, scripts are written from the personal point of view. That is, from the perspective of a person who has done several installs of a program, and has prepared the system in advance. Many things which would normally be included in a script file, are omitted on purpose - since the person writing the script didn't need it for their own install (assuming that their database or apache config had already been completed), or by accident simply due to the fact that the script writer felt that anyone using the script would have already done these preparations. I can think of a multitude of other reasons, all of which would point to the human-factor.
I don't mind trying to help out where I can but if I need to tell someone exactly what to do, step by step, I am probably going to be sending them an invoice after the fact. Others here might not feel the same way. If I were already using the package and you were asking for assistance with specific problems that I thought I could help with, I would try to help. If you waltz in and say, I can't get something to work, can someone give me exact instructions to make it work in my environment, with my installed software, telling me exactly what to do, my first impulse is to turn on the timer for my hourly bill rate.
You can count me into the "Others" category here. Helping others is one of the way that my company contributes back to the Linux community. I make it the responsibility of every employee to contribute something, even though I'm the one paying their salaries while they're out helping on a volunteer basis. They select the person or persons or groups they will assist and we allot a salary incentive to those staff members who can track and vouch for that time.
When I "waltz" in and ask for help, I'm not asking for a lecture or perspective on whether or not the help is "billable" or not. I've taken hours and days out of my time to help others on this list (past and present), and will continue to do so in the future. For me, it's not always a question of money. If you're curious, I'd be willing to send you a short list of some of the most recent times I've helped others.
Fortunately for both of us, I wasn't saying "I can't get something to work, can someone give me exact instructions to make it work in my environment, with my installed software, telling me exactly what to do", I was asking if anyone had managed to get it working in Mandrake and whether they could help or not. If you need a reminder of that, I'll be happy to re-post my original message.
My earlier comments about ease of installation refer to the fact that the installation instructions on the pages your mentioned are not written in a logical order, and the scripts don't fully complete a workable installation.
I am not running OGO, however, my understanding is that it is an enterprise level package that provides a large amount of functionality with many components and tying to a large variety of background archictectures. My expectation for such a package, especially one as new as OGO would be that it would require some effort to integrate it into any particular environment. If your needs are not so complex to warrant the use of such a package, there are other options available. If your needs are such that you do need such a complex package, you should be prepared to spend the time necessary to get it to work. Another option would be to buy a package, pay for support and get someone to do it for you.
FYI, I consider 4 days sufficient time. However, like any smart consumer, I am not about to buy a package or pay for support for something that I haven't seen, and I would hope that you wouldn't either. All I've seen are a few screen shots which don't tell me whether or not the product is stable or flexible or how customizable it may be.
As an example, they don't check to see if PostGreSQl is installed, don't create the database or add the user to PostGreSQL or for that matter don't detect and finsh configuration of Apache.
They mention that Postgresql can be used as well as other types of SQL databases. So, they don't assume postgresql on the off chance that you want to use something else. The small part of installation instructions that I read indicate that you must install postgresql and create the database separately, or provide some other database. Detecting which version of Apache, 2 or 1.3 and then properly configuring it including creating virtual directory entries, setting access and getting past whatever security (within Apache) you have or would desire would on the surface, appear to be complex too. Add in whether you are going to do purely SSL access or allow open http access. Throw in the diverse number of environments, flavors of Linux, security, etc. and I can totally understand why a script that someone wrote for themselves might not be totally working for you out of the box.
I can relate to your points here, but again it brings me back to the point that if the script file can't find these answers or ask the person installing it for the answers, then they should make a point of locking down the dependencies so that it's easier to install. I've seen many scripts do exactly that, and they have worked flawlessly for me and probably for many others.
That may sound like a lot, but I've seen other install scripts work for
effectively, such as the ones for webmin, macromedia flash, and adobe's
acrobat reader. Even the installers for ATI and NVidia drivers are a bit more friendly than this one.
Funny you mention that, since the most recent release of Nvidia drivers installation script doesn't work with Mandrake.
This makes perfect sense, since they are written for XFree and not necessarily for Xorg, even though the two are relatively interchangeable. However, I've been getting consistent results from Xorg including 3D acceleration, so I won't be bothering with NVidia's drivers anymore, unless Xorg finds a way to screw it up badly or something.
And install scripts for Microsoft Exchange might be easier than ogo as well. That is very much totally beside the point. You are not working with those packages or those developers, you are working with ogo.
Well, since I'm not now, nor never have used MS Exchange, despite acquiring and maintaining my MCSE status for the last 6 years, but my point was simply that programs that install on MS-Based PC's and Servers are specifically written so that installation and configuration are as easy as possible. With all the additional functionality and power than Linux and Open-Source provide, putting together a clearly defined list of dependencies (Ie; MySQL and PostgreSQL, Apache 1.3.X and Apache2) would narrow down the variations in the script file, thereby alloowing the writer to put a single, comprehensive install script together. After all, it's not like we're unfamiliar with dependencies, now are we?
I totally understand someone getting frustrated when something doesn't work. I also understand someone REQUESTING help. What I do not accept is the tone that one takes when they EXPECT something to work or that if it doesn't work for them, then there is something wrong with Linux or Open Source.
Feel free to handle that in any which way you like, but tone or no tone, I expect certain things to work. If they don't I'll take some very expensive time and try to make it work. If I can't, I'll then ask for help, and yes, maybe I'll vent a little about what I perceive as the problem.
Whether or not you or I or anyone else in the Linux community likes it or not, Linux will be getting compared to Microsoft-based products for some time to come. That's just the reality of it. They got there first, and now Apple, Unix, Linux and Novell are all being compared to the plague from Redmond. Consumers are more concerned about getting it to work quickly and easily than they are about our perspectives on which is better. Companies and businesses are more concerned about deploying it quickly, and getting it into production.
The trick is to make Linux easier. It's already better than the commercial closed-source alternatives - no question. But it needs to surpass, not to be the same.
Linux requires more effort than something that you buy from someone. That is the nature of it. Comparing a free product that was given to you with a product that earned the authors millions of dollars in revenue is simply not a material comparison. Give me money and I have an incentive to make you happy. Give me nothing, and my only incentive is what makes me happy. If that also makes you happy, then great, if not, oh well.
Your absolutely right about Linux requiring more effort, but my point is that it doesn't necessarily have to require more effort. Linux is/was developed by technical people with technical minds and perspectives. This was/is a necessary part of the whole process. But Linux needs people who are just regular people, who can look at it and say that this or that is too complicated and needs to be made easier. If the technical powers that be are smart enough to pay attention to that, we'll see a faster adoption of Linux throughout the world.
It is a balancing act to be sure, but not impossible. In fact, one of the reasons for the popularity of Mandrake over others is the fact that they've been taking the time to build and develop apps and software that make configuration easier - such as the Mandrake Wizards, MCC, etc. In fact, as a MandrakePartner I recently received a letter from Mandrake about helping them make it easier to install and use. I figure that if Mandrake sees it that way, who am I to argue?
But to be more specific, the fact that OGO is included as a significant or even a contributed application implies that it should be ready and able to install with little help, not that it should be virtually useless. From a common sense standpoint, why bother making Mandrake specific RPM's if it doesn't work?
You implied that it has to do with the fact that I've made a mistake, since the author of the script got it to work, but looking at the script, it seems fairly simple, and yet 4 of my developers who are far smarter than you AND I put together could figure it out either.
So, while I can understand your views, I also see that you haven't tried installing it which means that you don't know what the difficulties are any more than I do.
IMHO, many Microsoft users and admins who consider switching from Windows to Linux in a production environment are going to expect software to install a bit easier than OGO does.
And, IMHO, if they expect it to work the same, they should BUY a copy of a distribution from Mandrake, Red Hat, IBM, Novell, etc. and pay for support to get it properly installed and configured or expect to learn how to do it themselves. Why is the bar for Linux immediately higher than for MS, Unix, Apple? All of those products cost money and have R&D put into them supported by their customers, one way or another. With Linux, you have to put something into it as well. Besides bitching/griping because if that were the same as currency, Linus would own the world by now.
There is no such thing as a free lunch. And that is not a poor reflection on Linux/OS, it is merely the reality of the world that Linux/OS resides in.
Well, here we are at a crossroads again. The world that the Linux OS resides in IS a Microsoft world. That's just a fact that we have to deal with. None of us like it much, but that's the way it is. Until Linux dominates the OS world and Open-Source applications dominate the software world, we'll just have to bear the stress on our collective shoulders.
In order to succeed, it will have to show the world how MUCH better it is. Until it becomes more install and config and user friendly, the battle will be a lot harder than it NEEDS to be. End of story.
I don't have a problem with your "free lunch" point of view, since I pay for virtually everything as it is. But I'm not going to buy OGO or anything else without seeing it perform and without knowing the ins and outs of the install and configuration process. No one should have to.
Considering that the script has to be run as root, and considering that
Linux is a lot more powerful and flexible in what it can do, these
problems shouldn't exist.
Pray tell why not? I can buy a commercial software product in the Windows world and have it fail to do something that I need it to do. Why would Linux be any different? I can buy a commercial product for Windows and have the installation fail because of conflicts with some other software in my environment. Again, why would Linux be any different? My only recourse is to return the software and get my money back. How is Linux any different?
Linux is more powerful, flexible and that is why the problems exist. Because the power and flexibility are partly a product of it not being a monolithic, closed environment where everything is the same and controlled and can be depended upon to be the same. You say flexibility, they say uncertainty. It is, in reality, the same animal. It is both a boon and a curse. It creates problems and can be used to solve problems as well.
Give me a small, simple tool that requires no other components, software and I can write an installation script that I guarantee will work on any Linux platform in the world without any problem at all. Make the project more complex, tie in other dependencies and you increase the potential problems exponentially.
I'll tell you something that may shock you, and anyone who can't handle the truth should plug their ears and close their eyes, cuz it won't be pretty. People don't want that much choice. Of course, I'm not talking about us Linux newbies, experts, geeks or guru's, but about the general consumers. They certainly don't want as much choice as Linux and open-Source provides.
What they want is someone to tell them what they need, and to take care of getting it and setting it up. I am constantly promoting to my clients as well as a long list of Open-Source applications and all i see is those clients sticking with KDE and the more popular applications. They want software that looks and feels like Microsoft so that they can be as productive as possible in as short a time as possible.
Once they feel comfortable with it, THEN they start looking at other desktops and programs, but they always come back to what they feel most comfortable with. The percentage of users who prefer choice can be counted on one or two fingers (on a global scale that is) compared to those that want Linux to look and run like Microsoft. That's what we need to change. That's the biggest conflict that Linux and O-S software faces.
The only way to get past that phobia, is to make Linux easier to install, configure and use than Microsoft. Personally, I don't think that this is impossible, but there are going to be some choices that need to be made for this to succeed. OGO for instance will have to decide to limit the database systems and possibly even the distros that it can be installed on, if they can't adapt their install scripts to work properly on those other distros. And they need to make the O-S version installable by almost anyone who's into Linux. That way they can sell support for the O-S version a lot more. It would even make more sense to charge a higher rate to support the O-S version. If I was looking for support, I'd be content with paying the higher fees since the original suite was free.
Also, the comercial version apparently has properly built scripts that set up a fully functional install so it
seems strange that a modified version of that script isn't included with
the GPL'ed version.
GRRRRRRRR. Again. Gee, a company spent some time and effort to do something, invested their energy,effort, resources and want compensation for it. So what? GPL says that source is available when you use other GPL software. They have made source available. Their install scripts do NOT include GPL software so they are not REQUIRED to release it to the public. Do you have some other copy of the GPL that says that an OS developer becomes an indentured servant to you by virtue of you wanting his product? Perhaps you know something that I don't.
Calm down Bryan. You're gonna wear out the G & R keys on your keyboard! First, I used the word "apparently", which means that it's not necessarily the case. Secondly, I return to my statement about being able to test and evaluate the O-S version before buying the commercial version. I have NO problem with buying a good product, but unfortunately I can't determine if it's good or not and you already know why. Thirdly, it would also be a good idea if they considered selling the install script for a small fee instead of only including it in the commercial version, if my previous statement holds water.
Are you being paid to spread FUD for proprietary software companies? Because from my POV, it sounds an awful lot like the same old story.
You need to work on your POV dude! My company buys 50 powerpacks each month (it's a standing order) and another 40 OEM copies every two months. We only do Microsoft-related work if a customer refuses to even consider Linux, and in fact we've been turning clients away if they want a Microsoft solution. I have Windows XP on one system here and it's only used for comparison purposes. I stay informed about the Microsoft stuff so I can educate my clients as to the pitfalls of it. I'm as Microsoft compatible as Linus Torvalds is!
But even I have to face reality and deal with the real world, and that world is a Microsoft one. I'm very good at showing clients why Linux is better. Part of my job is to evaluate the possible solutions out there, and to define an easy installation and procedural process for any new products we take on. If I can't install it for testing purposes, I'm not going to consider it as a viable option. If my developers can't get it to work, then it's not going to be easy for my technical or sales staff to sell or maintain. That is what will hurt Linux and O-S software in the long run. The inability for the general consumers to try it for themselves before buying it. Plain and simple. The fact that OGO is either missing a clear and precise installation document or simply doesn't work without some significant step that seems to be missing, means it's not quite there yet. End of story.
Just my 2 cents.
And mine.
Looks like your two cents turned out to be wooden nickels Bryan. There was NO reason to be insulting or demeaning in your reply. I asked for help not for judgments. Perhaps you should reconsider your comments. I deal in the real world, and it would be nice if you did too. We're in for a hell of a fight against the Big Bad in Redmond, and it would be nice if we could stop fighting amongst ourselves before we try to take them on. No matter how good linux is or will become, it won't make a bit of difference if the community as a whole is constantly at odds with each other. Like the saying goes,... if you're not part of the solutions, you're part of the problem.
Lanman Registered Linux User #190712
____________________________________________________ Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com Join the Club : http://www.mandrakeclub.com ____________________________________________________
