Slashdot Mirror
Linux Growth Doesn't Offset NetWare Decline
steveit_is writes to tell us CommentWire is reporting that the decline in NetWare and Open Enterprise sales is plummeting at a much faster rate than their SUSE Linux sales are growing. It seems that the transition is proving to be every bit as difficult as Novell execs originally suspected. From the article: "When Novell last week announced its financial results for the fiscal first quarter ended January 31, the said that growth in its SUSE Linux and related products was decent, but that sales of its NetWare and Open Enterprise Server--a variant of NetWare that uses Linux as the operating system kernel that was announced last year--declined by 11%."
Novell has declared it's cash cow dead (Netware) long before the new cash cow (Linux) has replaced it. Now Novell has never explicitly said Netware is dead, just that the direction it will be taking is Linux. That's a lesson it learned from WordPerfect. Announce the end of the only product making money long before the new product has replaced the revenue. Ah well, at least Novell was able to use something from one of the many companies it bought. Too bad it was rotten business sense.
You keep using that word, I do not think it means what you think it means
They were originally a mainframe company in the 70's. in fact, they were good size being tied to the IBM mainframe. Afterall, back then nobody got fired for picking IBM mainframes. But by mid-late 80's, they were drying up. IBM was killing off its prodigal children that made it money. Basically, IBM would either buy the companies or would put it out of business in many illegal fashions. But Novell did not move. So they brought in a CEO to take them into the ground and get what they could out of it. And that would be Ray Norda.
....
Of course, Ray found that a small group was working on some interesting items and focused the company on it. Of course, they did lay off a large number of their staff. IIRC, they got down to something like 100 employees. But they came back in flying colors.
Novell will go through some leans times, but they learned to jump ship BEFORE it sank completely. It would have been better had they jumped earlier, but
Novell will be around in 10 years. I doubt that companies like symantic, nortin, intuit, and AOL will.
I prefer the "u" in honour as it seems to be missing these days.
I have to agree.
I got started with networks with Novell Netware 2.12.
From there I moved to 3.12, 4.11.
3.12 was rock solid, the only problem was that it was not an application server, it only did file and print (and did it very well). Having PC's/Solaris/MAC/OS2/Mainframes able to access the same files/print jobs was huge.
During the 2.12 and 3.12 days Microsoft (Win 3.11) was making major inroads with applications which IMO killed Novell. Since Novell did not have a App server to counter it (no one was supporting apps at least on Novell). Novell was left with file and print jobs.
Installing shared apps on a Novell server could be a major nightmare.
When NDS came out it was too little too late.
I see no real reason to migrate to SUSE/Novell at this time
I have bought the 9.2/9.3 version of SUSE (need to buy 10) just to help support Novell.
I had plenty of experience with Netware up to 3.12 and in the 199?-1995 timeframe, it was, for a lot of people, the only place where they could store their stuff, the only option being a floppy. At my university, an IBM PS/2 Model 95 running NW with the Mac storage option (whatever it was called) with TCP/IP as well as IPX serviced a hundred and fifty machines, a mix of PCs and 80s and early 90s Macs. NW also handled all the printers (5 or so) and even a couple of early model plotters (if I recall, Lotus 123 1a would only print the graphs to plotters, but I may be wrong about that).
Good times.
It seems that, more than any other OS, Netware is something whose time has clearly passed; everything Netware provided is now available on the user's desktop, regardless of what it boots to. If I remember correctly, NW has been expanded to also be an application server platform for databases, web servers (I believe Apache can run on it), but it seems that it's a more radical configuration than the most offbeat Unix platform. A friend of mine described programming NLMs as nothing like he'd ever done, and nothing he'd ever like to do again.
tackle the most important Linux problems. The OSDL Linux desktop survey (http://www.osdl.org/dtl/DTL_Survey_Report_Nov2005 .pdf) clearly lists Application support as the first top inhibitor to Linux adoption and Novell's own Cool-Solutions web site (http://www.novell.com/coolsolutions/feature/16798 .html) shows that Quickbooks is the most wanted Linux application. So why doesn't Novell sponsor a real OpenSource alternative?
...
No I don't mean to sponsor GnuCash, I mean to build up a cross-platform solution which is able to compete against Quickbooks on all platform (including Windows). I guess it doesn't need more that just a few developers to create an alternative within halve a year and within a year Quickbooks will notice its business diminish. Well lets see then how all the others Windows-Only vendors will react when they see what happened to Quickbooks.
I'm quite sure these few developers have a much more important impact on the success of Linux that dropping another fifty developers into Suse. It will even be better for Suse if these few developers are taken temporarily away from it.
The way to success is quite easy when you follow a few rules:
- don't have unsolvable obstacles
- don't have killer arguments against you
- don't have inhibitors
- do have something valuable the others don't have
- look at our products with the eyes of your customers or users
-
O. Wyss
See http://wyoguide.sf.net/papers/Cross-platform.html
The Directory itself may be fine but the integration in OES is halfbaked compared to NetWare. Example, Linux User Management - when do you need it, when do you not? Notice that there is no corresponding complexity in NetWare - everything just works out of the box. Example 2 - root password is different on every box. Again on NetWare there is no concept of a box having its own administrator outside of the Directory. The post above is not claiming that these things are wrong, just that to an existing NetWare user they look like a significant step backwards and that Windows looks like a more attractive option.
Don't get me wrong - I would like to see Novell succeed but right now the reality is that previously loyal customers are leaving because Novell have forced them to chose between something they have some familiarity with (Windows) and something that one day might give them the same seamless admin experience as NetWare (OES).
Novell seems to be making deals with SuSE.
Swiss Government
Novell is leading linux in china
I mean come on I don't think the Swiss Gov't is going to pick a company that doesn't know what they are doing.
Redhat is a great example of how a linux company can be successful. Novell is backed by IBM, and has partnerships all over the place like Redhat. I think Novell is going to surprise a lot of people.
Hey even their old CEO is now the CEO of Google. They have too many ties to too many power players for them not to be a success.
If Novell would open up technologies like ZENWorks, they might get some real interest.
This is a great observation. We're migrating off of NetWare because of the hassle of maintaining two directories(We also have ADS), amoung other reasons. We'd love to keep ZENWorks, but guess what? You still have to maintain two directories.
My converstations with supporters of this are almost commical.
ZENWorks Supporter (SUP): ZEN supports ADS!
ZENWorks Skeptic (Skep): Great! I don't have to support multiple directories!
SUP: Well, actually you do, but they sync, so you don't have extra maintenance!
Skep: Ok, so I get to do all my administration in ADS, right?
SUP: Well, no, you have to use Console One, but they're syncing so you don't have to worry about maintaining seperate user accounts!
Skep: So I don't have to worry about eDir logons, right?
SUP: well, you actually have to logon to eDir.
Skep: But it's clientless, right?
SUP: Absolutely! No netware client, you just have to logon to eDir through the ZENWorks client.
Skep: How is that clientless?
SUP: There's no NetWare client, so it's clientless.
Skep: Um.
Skep: I've had problems with maintaining two seperate accounts in the past. For example, password complexity, expiration, etc are all handled differently in the two directories and this has led to inconsistant password policy.
SUP: ZENworks has some great policy options!
Skep: Yes, I understand this, but how do I reconcile eDir's grace logons with ADSs warnings before expiration? This kind of seems like a hassle to me.
SUP: You don't need to worry, the directories are synced! The syncing is so good, you won't have any problems.
Skep: But I am having problems.
SUP: Those aren't really problems. You just don't understand.
Skep: I do understand, and I also don't want a client that replaces my Windows logon GINA.
SUP: The client is not a problem. You'll like it.
Skep: um.
If ZENWorks was fully integrated into ADS and had no seperate logon, we'd use it. ZEN is awesome. But as long as Novell and its supporters fail to see issues with the things that we think are issues, we're going to seek other options.
TW
Novell isnt alone with SuSe.
IBM is buying a lot from Novell, and not from redhat.
For example. Novell has this tiny distributions for Point Of Sale Hardware, called Novell SuSe Linux Point of Sal NSLPOS or NLSpos, it depends how you order the words.
IBM has a HUGE hardware POS, and they build IRES, IBM Retail Enterprise Solution on TOP of Novell`s NLSPOS
Also, Novell has support for the brand new OpenPower, Xseries and so on, also redhat, but the difference, is that, Redhat charges you by instance, and Novell charges you by hardware, and since IBM is preaching the "consolidate on big hardware using LPAR (hardware virtualization) on their machines, Novell offers a much more "from the book to the TCO financial benefits...
So I really think that Novell will survive and will have a huge market, more market than RedHat, they are not so cocky about them self as RedHat, Novell wants money, not fame...
Redhat, seems to seek for fame... and money...
Â_Â
I have been working with Novell for almost four years and have run into many similar arguments. But there are some facts that Slashdoters and others often ignore. And that's a relationship between Novell's products. You can't run ZENworks Desktop management with directories other than eDIrectory. Certain applications work only with other Novell applications, while Novell still gives out as much as they can without geopardizing market shares of their flagship products. Regarding my customer base, it always comes down to sensible advice. Some need MS, others don't. Maybe you needed Novell better than MS? Just an example from one of my clients: three years ago they had some 10 MS servers, 10 Netware, 5 linux servers. They decided to ditch Netware and move to MS. Their current situation: some 40 MS servers, 10 or less linux servers and 7 Netware servers. They still employ eDirectory as their primary directory (they have three), ZENworks. I actually installed a new OES Netware server last week to host their DNS and DHCP services. Will probably be doing 3-node cluster later this year... They took two years of planning the ditch and in the end they have 2 times the servers they did in the beginning. They found no viable option to Novell's services...
*NDS* is too little to late.
THE BEST directory server out there is too little?
/* FUCK - The F-word is here so that you can grep for it */
Novell did have an app server, it was called "Netware". Oracle ran on Netware. Pegasus mail did too. It was a lack of third party support, not a lack of a valid platform.
NDS, if anything, was too much too early.
While I can see advantages to directory services for a 1 person network, DS is very much influenced by the "network effect". The more you use it, the more valuable it is. The more objects it has, the more usefull it is. Its most usefull for large networks. WfW and later NT workstation and 95, with their built in p2p networking, was, in many cases "good enough". Even for networks that used, and liked, NW 3.11, but translated to "good enough" WfW/95, when they grew, grew into other MS products. Netware 4.x, with NDS, was a race to the top, with the bottom falling out compleatly.
I have a co-worker who recently went to a Novell/Suse training class, and from what he tells me they were very good at drinking the Kool-Aid. Lots of talk in terms of not "if" but "when you switch your entire company to Suse Desktop", you should have all your servers upgraded within the year, that kind of stuff. Look, I know Novell has to be behind their stuff, but I doubt there are very many companies out there who can just have all their servers upgraded in that kind of time frame, let alone totally drop Windows on the desktop.
Besides, Novell's immediate problem is not getting Suse out there to it's customers. It's coming, we know it, and even if we don't like it we're going to move there eventually. Novell's big problem is losing current Netware/GW customers, and attracting new ones.
Open source Groupwise. It seems so obvious to me I can't believe Novell isn't doing this, they're pretty much in the process of abandoning GW anyway. Linux is desperate for a full-featured, one-stop Groupware product. How many Suse servers would you sell if open source GW was out there? How many current Netware customers would you save from switching over to Exchange?
Since acquiring Suse, they've seen its share of the Linux market shrink compared to Red Hat's.
You have data to back that up? Everything I've seen in the mid to large corp field defies that, with a few SuSE rollouts I know of, at least in Europe and North America, and no new RH rollouts. The reason seems to be that the Novell brand is an easier sale for IT project managers, it's always Novell SuSE, never SuSE alone, and possibly cheaper support contracts? Don't get me wrong, RH still has more boxes out there, but I'm not getting wind of new RH deployments versus SuSE.
Good people do not need laws to tell them to act responsibly, while bad people will find a way around the laws-Plato
The financial figures are frightening. Novell booked $274.4 m in revenue, of which around $56 m was from open source products, of which around $13 m was pure Linux (the rest was Netware OES), of which only $10 m was from SuSE Linux ( a 22 per cent improvement). The article then quoted an analyst who said that Red Hat's Linux growth was twice as large and their revenue from Linux was five times larger.
Put it another way, a couple of years into their Linux story, Novell is turning over around $1 billion of which pure Linux contributes around $50 m, and much of the rest is declining legacy stuff. This is a drop in the ocean, and all the harder when Red Hat appear to be creaming Novell at the sharp end.
$50 m compared to $1 billion. I don't know how Novell is going to get out of this one, but talking about changes to SuSE or Ximian or yet more sugar-daddy spending on open source projects is like the Titanic and deckchairs. It's very hard to see Novell avoiding a break up.
Las qué passoun
tournoun pas maï
But here's my take on it. Maybe somebody at Novell will read this. The admins seem to be afraid to learn UNIX... So I am assuming that Novell charges for the training, the testing, etc. Now our admins, they get Winders but they really don't get Linux or UNIX. I think it would have been in Novell's best interest to eat the cost for training, especially to those who have already gotten their CNA or CNE in older Netware products. I think FUD and discomfort are major factors in hindering old Netware shops from going to Novell Linux.
Another issue that my company seems to hate about Netware (ver. 6.5) is its poor compatibility with Enterprise software and hardware. Netware clustering doesn't play well with Netbackup (it works, but it's very clunky). It also doesn't do well with SAN stuff... I really don't know what the issues are, but I've heard that expressed on more than one occasion. We do some Heirarchal Storage Management (HSM) (HSM essentially is a user transparent file archiving mechanism) which also is pretty klunky, esp. with Netbackup, etc. I have no idea if the Novell Linux works around these issues, but the FUD and discomfort of going to Linux don't help.
Netware 6.5 isn't all that stable. The servers ABEND fairly often. IMHO, they should have gotten off the DOS base a long time ago. I guess if we were running "workgroup" class Netware systems with less "enterprise" features, they'd be a bit more stable, but Netware needs work fairly well in a mid-level enterprise environment if it wants any real respect in from IT departments.
The old IBM adage, "No one ever got fired for buying IBM," seems to apply now to Microsoft.
But the training is probably the best thing they can do to keep the momentum. I think SUSE, although I haven't used it in a long time, and Linux in general, can and does operate well at the enterprise level, but for non UNIX people, it takes a lot of faith to trust in it.
But that's my two cents.
-Jim
When Novell chose to throw all their eggs into the Linux basket, they took a huge risk. The problem is they didn't really throw all their eggs in it. Here, imho, is what Novell must do to succeed:
1. Give away Zen, or at least parts of it.
Many of the features in Zenworks come part and parcel with active directory. There could be a Zen-lite that does the same things that AD admins can do through group policy. Include the ability to do similar tasks on Linux machines and Novell can go back from "keeping up with Microsoft" to "staying a step ahead of Microsoft". While they're at it, Novell needs to work include support for every aspect of Firefox, including a list of supported plugins and extensions, to amke it manageable through Zen. AD admins can mange the IE settings across their network with GRoup Policy, Linux admins need to be able to do the same thing.
2. Do the same thing with Red Carpet.
Novell either needs to give Red Carpet away or have a limited version that operates the same way SUS does. They could have a professional version that will also use a push architecture in addition to a pull architecture. Personally, I loved Red Carpet when I first heard of it. Patch management for my windows machines and my Linux machines? Score. Here's the problem: I can get patch management on all my windows machines gratis with SUS / WSUS. I've got less than 20 Linux servers in my environment, about 200 windows servers, and around 3500 windows workstations. How could I possibly justify $18 per seat for Red Carpet when I can run SUS for free and just have our admins manually patch the Linux Servers? Yes I know Microsoft is the source of the vulnerabilities in the first place, yes I know Novell shouldn't have to give away a product that cleans up Microsoft's mess for free. Y'know what though, money talks. By having to pay extra cash for Linux patch management, that adds to the TCO of Linux while Windows' TCO stays the same, giving Microsoft marketing more ammo to work with.
3. Improve the Yast firewall interface and add remote management via Zen.
For that matter, everytinhg you can do in Yast needs to be accessible remotely via Zen. In an AD environment I can manage the Windows firewall on all the machines in my domain via Group Policy. I need to be able to do the same thing in a Linux environment. And the Yast firewall interface is the only one I've seen that actually sucks worse than the Windows firewall interface.
4. Ratchet up support for Wine. Partner with Codeweavers, or acquire them.
Novell's Linux support needs to embrace Wine or another emulator to assist with Linux migrations. Their current approach of "Run a Terminal Server that hosts the Windows-only application" isn't going to cut it. Users want icons on desktops that run their applications. Clicking an item on the linux desktop, then logging into a termserver, then clicking an icon on the termserver, then logging into an app, isn't going to fly. If Novell really wants to be successful in migrating companies to Linux, they should partner with or acquire one of the Windows emulation projects, and offer "take your POS custom app that you bought from a vendor or coded in house and make it work on Linux" as a service with a one time fee and optional support.
I think what Novell's trying to do is great, but I see them hanging themselves with it if they don't stay a step ahead of their competition.
Yes, my only tool is a hammer. And you're starting to look like a nail.