Did RedHat et. al jump the gun too quickly

[ral]

Just read this article:

 

"Will companies make wholesale moves to other platforms? It is still early to make that judgment, according to Park. "I haven't seen any mass defections yet," he said.

 

But Microsoft's licensing gaffe has caused enterprises to begin looking at the competition and performing cost analyses on whether it would be beneficial to move to alternate products, Park said. "That type of movement, if it occurs, is just now beginning," he noted. "It will be important to watch in the next two years how many [companies] move to alternative-source or open-source platforms."

 

x x x

 

"Ultimately, Microsoft's best friend is corporate inertia. IT departments that want to ditch Microsoft will have to make a strong case to executives that replacement costs will be recouped in a timely manner. If they fail to do so, companies are likely to pay the extra licensing charges and stay the course with Microsoft."

 

http://linuxtoday.com/it_management/200304...042902526NWMSLL

 

Watching our painful Linux migration process at the UP College of Law where the decision has been to deploy RedHat 7.3 on servers/workstations and Mandrake 9 or 9.1 on desktops, I think some Linux developers jumped the gun too quickly.

 

Migration costs money and now Linux is starting to costs more too (i.e RedHat). When looking at Unix options, Linux is clearly cheaper. When looking at Windows options, it would take several years to recoup the cost.

 

With Windows you pay for the software and support. With RedHat Linux you pay for the updates and support. If not for Mandrake, this whole migration might be in jeopardy.

 

Anyway 50% of the machines had Linux installed by February 2003. But all the desktops are now back to Windows with a slower migration process on desktops planned (to Mandrake 9 or 9.1).

 

Looking at things as a whole, I think a Linux, MS Windows and StarOffice combination might be the most cost effective approach.

 

At our Internet Cafe, we just opted to invest in WinXP and use OpenOffice. Our time keeping software does not run on Linux and the cost of porting it was too high. Retraining staff would consume time. Also with the trend appearing that Linux will cost more over time. And frankly :( the clients don't like it (can't blame them... born on MS...).

 

On the MS side, they have intorduced new scheme like Open Value and Open license which make things cheaper (or at least spread otu the costs).

 

Still we are derving great benefits from OpenSource. RedHat 9 still powers my desktop PC, Mandrake 9 resides in our ECS Desknote (kind of like a lap top) and RedHat 7.3 in our one of our servers. Switching to the Mozilla browser made our Windows PC's more resistant to malware and OpenOffice is now used in 70% of our PC's.

 

I than came to realize, the problem was not really Linux, but the consumers. Going to Linux for many of us is a cost reduction process. So we are willing to pay very little for Linux. On the other had Linux companies have employees who do need to earn a living too so how long can they keep giving things for free.

 

It may be ironic, but Linux best contribtuion may not be to replace MS but to keep MS in check. And that might not be a bad thing at all.

[DragonMage]

Well.. in case of your internet cafe.. I don't blame the consumers for wanting XP more than linux. People come to the internet cafe to surf, email and maybe do a little bit of text editing and such. Since their time is somewhat limited and they want to maximize their computer use, they tend to go into the thing that is most familiar. In terms of the crew also, retraining them takes time, even though everything is already preinstalled and preconfigured.

 

Another example I see in your case is the UP College of Law. Well, I hate to be mean but law students are not exactly rocket scientists. They just want things that are familiar to them. I heard that the standard wordprocessor for the lawyers in the US is wordperfect, not ms word. If microsoft cannot even change the US lawyers' mind, what chance do you think linux have? I notice that 50% of the machines that still keep linux are servers/workstations. This is perfect for linux in the beginning.. keep them to the backend. All the people notice is that the uptime is longer and everything that is network related (such as print server, web server, file server, etc) runs smoother.

 

Like it or not.. people must be willing to learn in order to use something different, even changing OS. I guess that's why linux is huge among the Computer Science and Engineering dept in larger universities. The students there are required to learn unix-like OS because like it or not, their future jobs depend on it. I mean Visual Studio is nice and all, but do you want to rework and recompile everything after a crash? Not to mention the code created by Visual Studio are generally huge mess. Even industries that are moving to linux are somewhat wary of linux desktop. They know linux's power as servers, workstations, and specialized apps such as point of sale terminals, but the picture of the general linux desktop is kinda iffy at the moment.

 

The good thing is that linux moves at a really fast rate, even in desktop. Compare today's desktop from say.. 2 years ago.. lots of changes already. Maybe linux on desktop is a hard sale today, but by next year, it should be at least a lot easier as people are more familiar with it.

[ramfree17]

Another example I see in your case is the UP College of Law.

 

for a moment there i thought you were referring to the University of the Philippines College of Law, which puzzled me since the University of the Philippines president is pushing for the adoption of linux.

 

then i guessed that you might have meant University of Pennsylvania or something. :#:

 

ciao!

[aRTee]

Read my successful transition article.

 

In some cases it does not make sense to use linux on the desktop.

 

The standard home desktop can often be run with linux, if all necessary tools can be had for linux.

 

First use the stuff on windows that is available on linux: OOo, mozilla (web and mail). Then it should be little hassle to migrate.

 

I don't see the problem with linux in an internet cafe, for the users (I realise it is costly to convert software and/or retrain staff): put 4 icons on the desktop, nothing else, labeled Web, Write, AudioCD or whatever.

Users should hardly notice what kind of OS it is running, just that it doesn't crash.

 

If you are migrating for the right reasons, you can make linux work.

That it's cheap is not necessarily the right reason.

 

That it's free is (free speech), and many other reasons.

 

Linux is not there to keep MS in check, it is there for those people who care to replace MSWin.

If linux only has 5% of marketshare in 2005 it's still fine with me. I will be among them.

With that kind of marketshare all hardware makers will have to provide drivers or info to make drivers.

That's all I want, to be able not to check before I buy hardware if it's supported under linux.

 

It's going the right way, linux was ahead of windows for bluetooth for instance, and also on dvd+/-rw.

 

If >90% of the computer users want MSwin, fine with me too. Not in my house though.

 

Linux is not for everyone.

Nor is MSWin.

Nor is MacOSX.

Nor is UNIX (whatever flavour).

 

Linux is for me.

At work I have a win pc connecting to the unix domain with reflection.

In 2 years I will have a linux pc at work. 90% sure.

 

RH, Mdk are not too early? Why do you say that maybe devvers jumped the gun too quickly? Maybe it was you who jumped too quickly?

(I'm not trying to lay any blame on you or so, it's an honest question)

 

Linux is perfectly suitable in many cases, server and desktop.

But you have to know what you are doing...

 

In an internet cafe, I would not know what to do, I don't have a clue what people do there. But my bet would be to have a mix with XP and linux, see how it goes...

 

BTW I disagree that you have to pay more and more. You don't. At least, I don't.

[ral]

Very funny RAMFREE17.

 

Well, like Ramfree17 said, we are committed. Its just not been recieved well by the students. I don' know how the other Colleges are doing.

 

But Mandrake won out over RedHat for the desktop.

[ral]

Did we jump the gun too quickly... yes and no.

 

We began looking at Linux only 3rd and 4th quarter of last year. This year the trend seems to be Linux will cost more. Less will be given free. So we were actually a little "late".

 

RedHat is going that route. Decapitate the lower end products by cutting their life and offer higher ticket items ("Linux Enterpirse"). Mandrake, I am not sure yet.

 

So when you are nto sure... well you maintain the status quo. I love Linux guys, but business is business.

 

Some products (relevant to us are raid controllers)... only run on RedHat 7.2 or 7.3. &.2 support is now over. 7.3 end at the end of this year (and so does 9). By the time support for 9 is out... well its near EOL too.

 

RedHat solution: Linux Enterprise Servers

 

Ours: Software Raid or Win2K.

 

But like I said, the fault is the consumer. For my personal box I left Windows for reasons other than cost, I liked it better though I still have a Win98 partition for my games).

 

For the shop, I am only going Linux if it will save money. I will wait and see what RedHat and Mandrake do next.

 

Microsft launched as very friendly Open License scheme here.

 

"BTW I disagree that you have to pay more and more. You don't. At least, I don't."

 

For myself, the shop, well I bought a copy of RH8 personal... I should spend some money on Mandrake too... when 9.1 gets here... I owe them that much.

 

But I keep wondering if the freebies are going to end soon...

[ral]

Cool, Highpoint has RedHat 8 raid drivers (0/1). Wonder who the Highpoint distributor is over here.

 

No Mandrake drivers though... but Highpoint never releases Mandrake drivers :(

[Michel]

I also think that using opensource apps on windows will help the crossover to linux.

However..to my opinion..the network maintainers at universities....are also reponsible....At our school we have RedHat on the desktop pcs, but they are not maintained at all...The version of mozilla is OOOLLLLDDDDD. On the windowspc's we have also Mozilla, but as long as nobody knows that Mozilla is a browser, nobody will click it!! The theme of Mozilla id classic, while the modern one (to my opinion is better)...Just some little details...liek also the pop-upblocking of the new versions.....You have to make it userfriendly....You can't expect of users that have never jheard of linux,will change settings or discover programs....

[aRTee]

ral, do I understand correctly that the highpoint drivers only work for standard (stock) RH versions? What if RH users recompile their kernel for some reason?

 

If you can't recompile the highpoint drivers, I don't see the point to have them at all. Stay clear of HPT.

If you can, you can use them on Mdk too.

 

I think you tried to convert things very fast, understandable it didn't go smoothly, but still a pain that you didn't get what you wanted from linux in the end...

 

Instead of buying copies of Mdk or becoming a club member, there are plenty of other things that you can do. For me a clubmembership is peanuts, but wages on the Philippines are no doubt not what they are here; if I take my girlfriend out for dinner it's (sometimes) more expensive than a membership.

If it comes close to a days salary, things are very much different.

 

Other things (that just cost some effort and time) to repay for using a free distro (and not paying) are for instance: sending bugreports, helping with translations, promotion (for me, I decided to help colleagues convert; they wanted more info, I wrote long emails,... then I thought: might as well stick it on a website...etc), helping out people on forums, etc.

 

It makes sense that for your shop you take what helps you make the most money. Doing anything else is bad management. However, keep an eye on what is really cheaper. If not today, linux may be really cheaper tomorrow, taking into account all aspects (acquisition, maintenance, personnel, whatever). The more you motivate your personnel to play with linux, the less the conversion may have to cost when the time comes....

 

 

BTW the freebies can't end, it's open source.

If mandrake stops being free (which for me is one of its greatest 'selling points'), I'll move to debian or so...

Maybe I'm ready for LFS.... ;) although my girlfriend already finds I spend too much time with my computer...

[ral]

"ral, do I understand correctly that the highpoint drivers only work for standard (stock) RH versions? What if RH users recompile their kernel for some reason?

 

If you can't recompile the highpoint drivers, I don't see the point to have them at all. Stay clear of HPT.

If you can, you can use them on Mdk too. "

 

I am not really sure. We have one motherboard with built in highpoint raid controllers... lemme try it out.

 

"I think you tried to convert things very fast, understandable it didn't go smoothly, but still a pain that you didn't get what you wanted from linux in the end..."

 

In the Univesity, yes, way too fast. But, now deployment is starting again but in a phased manner. The new RedHat support policies are a problem so we may have to consider deploying the Linux Enterprise products on the servers, but this is unlikely. Why, becuase the move was being made to save money. Most likely, the "free" version will continue to be used even past there end of life, which could lead to security issues.

 

In the shop we did tests we too on a few machines. I would deploy a few now and see how it goes if we could get our timekeeping software (locks the machines and keeps track of usage) to work on it.

 

----------------

 

I guess my fear is that we will invest a lot of time and effort in making the move, and in the end using Linux will not be free anymore. Of course we could move on to some other distro... but if have to do all that, well in the ends Windows does not cost that much either.

 

But like I said, the problem is the consumers and matbe the disto makers too. Most consumers move to lower cost, distro makers need to earn a living.

 

I know a lot say costs savings is not a "good" reason to move to Linux. In the Univesity and the Shop it would be the only reason to move to Linux.

 

On my desktop, cost has nothig to do with it. I enjey learning new things, so Windows got boring after awhile.

 

I think another reason why costs is different here and lets say in the US, is the US is that in the US, Microsoft pushes Licesing 6 and Open Value. Here Microsoft is liberal with cheap OEM licenses (can be bought with any hardware nto necesarily a new PC)and now launched Open License (equivalent to OEM pricing).

 

In the end, Microsoft is more sensitive to local conditions (i.e. we earn less) than Linux distros.

 

The cost of 10 RedHat enterprise Workstation is twice the cost of a Windows 2000 server with 10 client access license, here at least. Over three years its seven times the cost.

 

Now MS support costs a lot (and OEM copies dont have any), but after 7 years in business we never really had need for it. It just works. With Linux, some things dont work and we don't have the ability to fix them ourselves, hence we would need support.

 

If all you need are patches and updates, MS does not charge you for that. Pay thel ciene fee and you get patches for 5 years. Linux gives the software for free and is now tryign to charge for the "patches".

[zero0w]

I think I read somewhere there is going to be mid tier security update solutions for EOL distros. The Register has an article on this one:

 

http://www.theregister.co.uk/content/61/29330.html

 

Maybe we will have some up2date clone in the making provided by 3rd party or community as low cost solutions. It's more crucial when you have to support server operations which needs constant maintenance, unlike on desktop where you simply downgrade or find another app when the one you are using is not working.

[ral]

"Maybe we will have some up2date clone in the making provided by 3rd party or community as low cost solutions."

 

It exists, AptGet+Synaptic, well almost. The problem here is more of accpetance. No one get fired for buying Microsoft. No get fired when you put in an MS patch and it messes things up (altough you should test any patch pefore deploying it). Testing could take a few days to a month.

 

Now getting stuff from somewhere that has no accoutability to you. Me I know it works, the Linux community is made up of reponsible group of persons. The hard part is convincing a CEO that its safe.

 

Deployment thorugh a network is also an issue, but that I am pretty sure can be fixed.

 

"unlike on desktop where you simply downgrade or find another app when the one you are using is not working"

 

You will take several months of testing before deploying a new desktop OS. Lets say 2-6 months. By the time that you do, you got 6-10 months left of RH support and 10-16 months of Mandrake support. Than you go through the cycle again.

 

Hhhhmmm... who want to do that. If you deployed Windows 2000 back in 2000 (or even late in 1999), you would not have ot worry about replacing it till 2007 or 2008. After all, XP does not really work better.

 

Also Deploying it to the 150 or so machines at the University is time consuming.

 

"It's more crucial when you have to support server operations which needs constant maintenance"

 

Its problematic when the OS your using is about to be EOL'ed and the hardware companies have not release drivers for the newer distro yet.

 

Come on Promise, release some RH8 drivers (actually found a third party fix for this... but Raid 0/1 only), but its hard to convince even IT people that it works when you go to the official site and they say it does not. Besides, RAID 5 is what people want these days.

 

Maybe Mandrake is reading in on this, I am not asking for a freebie:

 

Price: $40 to $100.

 

EOL should be 3 to 5 years.

 

No need to provide email and telephone support. This who want that can pay extra.

 

The great thing about Linux is that you probably never have to reinstall.... so you can deploy it once and have it running for 3-5 years till it EOL's, but if you EOL it in 12-18 months thats a problem... than it becomes a bit like Windows (which properly used should last more than a year... altough in a public environment, reisntall need ot be done more often).

[zero0w]

It exists, AptGet+Synaptic, well almost. The problem here is more of accpetance. No one get fired for buying Microsoft. No get fired when you put in an MS patch and it messes things up (altough you should test any patch pefore deploying it). Testing could take a few days to a month.  

 

Now getting stuff from somewhere that has no accoutability to you. Me I know it works, the Linux community is made up of reponsible group of persons. The hard part is convincing a CEO that its safe.

 

That's the job of Linux advocacy group and community based organizations. While the business world trusted credibility, it is the reality that things changed and big company who could afford will afford, and while other non-profit organizations will have to organize around cost-effective community based solutions. You can't have all the goodies or otherwise people will deploy Linux a lot faster than they have already. There isn't really a 'free' (or very cheap) solution to critical task that people had to depend on a daily basis - but I wonder even for commercial software (namely Windows) how long can they be up and running and not subject to this maintenance schedule.

 

You will take several months of testing before deploying a new desktop OS. Lets say 2-6 months. By the time that you do, you got 6-10 months left of RH support and 10-16 months of Mandrake support. Than you go through the cycle again.

 

I am not sure about that, but then again it's about the task involved that requires the amount of testing time.

 

Hhhhmmm... who want to do that. If you deployed Windows 2000 back in 2000 (or even late in 1999), you would not have ot worry about replacing it till 2007 or 2008. After all, XP does not really work better.

 

Also Deploying it to the 150 or so machines at the University is time consuming.

 

Well, I think massive deployment of Linux is more difficult than small scale (< 50 seats) deployment. But Linux for now is more geared to custom jobs such as cluster computing and properly configured environment. For a general purpose desktop environment Windows certainly has its own merits. You are right if you recommend Windows for those who are less inclined to learn and adjust then. As for 2007/08, I think DRM solution and Palladium will force you upgrade already, don't bet on that long.

 

Its problematic when the OS your using is about to be EOL'ed and the hardware companies have not release drivers for the newer distro yet.

 

No argument on that. Make sure the hardware you purchase/use already has Linux support, if not, no one can tell when or how well it will be supported. However, keep in mind Linux driver support mostly would not have hardware obsolesence - even the hardware company is gone Linux and open source programmer can still make it work in the kernel driver.

 

Come on Promise, release some RH8 drivers (actually found a third party fix for this... but Raid 0/1 only), but its hard to convince even IT people that it works when you go to the official site and they say it does not. Besides, RAID 5 is what people want these days.

 

I need to study more about that. But RAID support seems crucial to commercial and critical task environment. However, I believe the increase of Linux programmers in the future will help a bit on that.

 

The great thing about Linux is that you probably never have to reinstall.... so you can deploy it once and have it running for 3-5 years till it EOL's, but if you EOL it in 12-18 months thats a problem... than it becomes a bit like Windows (which properly used should last more than a year... altough in a public environment, reisntall need ot be done more often).

 

The EOL for Linux will become longer in the future, I suppose, because many packages will become more mature and requires less maintenance. I could be wrong, perhaps someone will come along and make much more improvement to the kernel and user space apps and then more maintenance is needed. I think if you want commercial quality or stability of upgrade cycle, the only way to do that is to convince and request commercial software vendors to make Linux port using GUI widget such as Qt or wxWindows. Of course it won't be easy, but I think it's what most commercial shop will face in the future: cross-platform app support.

[Guest JaseP]

Another example I see in your case is the UP College of Law. Well, I hate to be mean but law students are not exactly rocket scientists. They just want things that are familiar to them. I heard that the standard wordprocessor for the lawyers in the US is wordperfect, not ms word. If microsoft cannot even change the US lawyers' mind, what chance do you think linux have? I notice that 50% of the machines that still keep linux are servers/workstations. This is perfect for linux in the beginning.. keep them to the backend. All the people notice is that the uptime is longer and everything that is network related (such as print server, web server, file server, etc) runs smoother.  

 

Hey, watch what you say about lawyers...

 

I AM one,... (don't practice anymore, though, switched to financial consulting).

 

Many of the OLDER lawyers are very set in their ways, but the younger set, those in their 20's and 30's and some in their early 40's are very technically savy. There was an attorney who wrote for Linux Journal magazine until just recently. I know of several small lawyer offices that have adopted Linux because of cost.

 

The biggest obstical in adopting Linux in a law office is that the time management and case management software is all written for Windoze (they had a seven year head-start). The secretaries are the hardest to convert. That is where the problem lies. They are familiar with one thing and have no desire to switch. If they are going to switch, it makes no difference whether they switch to Linux or M$ Word. The same learning curve exists. Linux just makes certain things easier. The lawyers themselves rarely use the technology, most of them anyway. The ones who do have enough technical background to switch to Linux with a few months learning curve. What Linux needs is time-billing software and robust contact management/case management software in order to woo Lawyers. If someone were to develop that,... or make their exisiting software multi-platform, then more lawyers would use Linux.

 

It's interesting to note that there are a small minority of Lawyers who use MAC. Many of them started with MAC in college and fell in love with the platform. So lawyers as a whole are not married to any particular platform (we tend to be cost conscious), but are married to a particular way of doing things. If you can have a Linux word processor that accepted WordPerfect keystrokes, files and macros, then you would have an easy time converting them to Linux.

[DragonMage]

Sorry about that lawyer crack JaseP. I just want to make a point about inertia :)

 

Anyway, are you sure there is no case management or time management software created for linux? Or even in emulation mode such as win4lin (I don't think those kind of thing works with wine so let's move it a step up).