Talk:NetWare

Not giving MS credit.. again...
The section on Netware's current status goes on about their decline and seemingly only indicated it was because of Microsofts marketing. To me this section reeks of a "poor loser" mentality and excuses of a pro Netware author. It also goes on to say how subjective remarks like the Windows alternative was less reliable and more expensive.

I think an improvement would be to actually give MS some credit as the market really changed when Windows 2000 with its own directory service (Active Directory) with also the option of running the new Exchange 2000 that was integrated with AD (unlike Exchange 5.5) which gave potential clients some very good reasons for considering switching. I was the IT manager at a college and I know switching from Netware/Notes to AD/Exchange was a very refreshing change for us - which was also cheaper (albeit at educational pricing) to purchase and maintain.

djambalawa (talk) 03:17, 13 February 2009 (UTC)

NDS as an strategic mistake?
This part makes no sense. NDS was a great, easy to use tool. I've worked with a lot of NetWare instalations, and people complain about a lot of things, but not NDS. It was a painless upgrade from bindery, and much more useful. Much better than NT Domains. I'm removing this part190.10.22.214 21:30, 14 August 2007 (UTC)

IIRC NDS was the very best thing Novell has ever made. And it still is, now called eDirectory, abbrev. eDir. Very scalabe and reliable. IMHO still today much more robust and scalable product related with AD or the Domain model. —Preceding unsigned comment added by 212.63.70.206 (talk) 09:04, 27 September 2007 (UTC)

I think may be worthy to consider that NDS was being viewed by some (at least when it comes to forced upgrade) as an unnecessary over-complication (as it already was mentioned that small businesses resisted NDS, so this probably should be moved under the strategy section.) Thus regarding to this, someone probably should add information about Netware Migration Agent for including bindery servers under NDS.

Frankly, I think this is more of an important issue than trivial things like "new users preferred GUI," which again doesn't make much sense considering that the console part is dealt by netops anyway and users who "preferred GUI" probably didn't install servers. If anything, there seemed to be a sentiment against the newer Java based ConsoleOne and GUI Nwadmin and a nostalgia for SYSCON and NETADMIN (In the 9th edition of Upgrading and Repairing PCs I believe it was mentioned that it was really a shame Netware moved from a text sufficient server to one that required graphics.) Furthermore, does anyone have any evidence of DOS partition being a real issue? If you CAN install a Netware server then creating a DOS partition probably isn't an issue. I have even seen Netware 5 packaged with bootdisks that do this for you.

technical decline reasons too verbose
the section on the technical reasons for NetWare's decline is too lengthy, and too pedantic. There are numerous technical reasons which could be advanced for the rise and fall of any product. You don't need to mention them all. Maybe some people thought "SYS" was confusing as a default mapped drive - maybe, but who cares? It's too trivial. That section should be shortened, and the point about the difficulty of drives being resized cut out altogether. This article is not the place to go into such details. 210.48.84.234 06:31, 24 April 2007 (UTC)retroguy

Last statement has no corroboration
The closing statement in this article makes a conclusion and has an almost accusatory tone, yet there are no supporting references. I have added a peer review request regarding this.

Peer review tag removed
The offending statement was removed some time ago.

NetWare beyond 6.x
As I recall, shortly after Novell's acquisition of Ximian and SuSE, they announced on their web site that 6.x would be the last version of NetWare. No proof, but can anyone corroborate this? -- Enigmatick 22:50, 23 August 2006 (UTC)

TCP/IP and NetWare v4.x
NetWare v4.x, released in 1995, does not support TCP/IP for NCP communications. It does support TCP/IP for services such as FTP, LPR/LPD and the NetWare HTTPD webserver. Novell created NetWare/IP, called NWIP, which could encapsulate IPX into TCP/IP packets and provided a bridge from IPX to TCP/IP for clients. NetWare v4.x is an obsolete and elderly version of NetWare, and was EOLed in 2000. The current product is OES-NetWare.

Novell and beyond NetWare v6.x
Novell has stated that NetWare v6.5 is end-of-the-line for "traditional" NetWare. The NetWare product line has been succeeded by Open Enterprise Server, which offers the same services atop either a NetWare or a Linux kernel.

Novell has stated that only the Linux kernel would be enhanced and developed and that NetWare "upgrades" are a move to Linux, not newer versions of NetWare. For example, the following will only run on the Linux kernel of OES and will never run on the NetWare kernel:

- Multiple instances ("virtual directories") of eDirectory v8.8 - iFolder 3.x - JVM 1.5x and above - 64-bit support - Future "versions" of NetWare will run as a "virtual machine" on XEN on Linux

OES is not NetWare
In line with the above section referring to the Linux kernel, while "OES NetWare" really refers to "NetWare as included with Open Enterprise Server," it is not yet another NetWare version, nor does it represent further development of the NetWare platform. This section - particularly referring to OES 2.0 - implies that NetWare was moved to a 64-bit kernel (or an entire 64-bit archtecture) which is wholly inaccurate. NetWare as included with OES is 32-bit NetWare 6.5, no more, no less. OES has the ability to run it virtualized - as a 32-bit guest on a 32-bit or 64-bit hypervisor - but still as 32-bit. Likewise, iFolder 3.x never shipped as a NetWare component; iFolder 3.x was only ever available under the OES Linux kernel, again, as referenced above.

In short, the entire OES section should come out except as it pertains to the virtualization of NetWare 6.5 and the included licensing for NetWare with OES. — Preceding unsigned comment added by Beckysdad (talk • contribs) 03:45, 12 February 2019 (UTC)

Rearrange the page?
How about describing the present product first, and the history later?
 * I agree. I don't think it's good to start the history section with the decline of NetWare, either. It should be rearranged chronologically. Rhobite 00:40, 10 December 2006 (UTC)
 * Have a look back to before 23 Nov. Up to then it was setup fine. All this decline stuff went in then - it's probably valid enough in some ways - but most certainly should be down the bottom - ie in chrono order (imho)Snori 06:18, 10 December 2006 (UTC)
 * Didn't realize it was a new addition.. that explains the lack of talk. I think it is valid, but there are some speculative parts of it. I toned it down a little, but it still suggests that NetWare was technically superior to LAN Manager / NT and the only reason for NetWare's decline was Microsoft's marketing. Rhobite 22:46, 10 December 2006 (UTC)
 * I plan to go forward with the re-org suggestions, as well as some material expansion regarding strategic decisions that led to NetWare's decline (not just chalking it up to Microsoft's "superior" marketing) -- unless someone objects within the next few days. EJSawyer 05:46, 7 March 2007 (UTC)

Error in Early Years section
In the Early Years section, there is this comment: "Prior to the 80286 CPU, servers were based on either the Intel 8086/8088 or Motorola 68000 8-bit processors that were limited to 640k RAM and lacked pre-emptive multi-tasking". The 68000 is neither 8-bit, nor limited to 640kb ram. Don't want to edit myself, as unsure if this error is due to confusion between the 68000 and the 6800 or if someone's assumed the attributes of the 8086/8088 apply to the 68k too.

Also, being pedantic, the 8086/8088 aren't limited to 640kb of RAM, rather the original IBM PC architecture imposed this limit.

Davidprior 21:28, 7 January 2007 (UTC)


 * The 6800 had 64KB max addressing capability, like 6502. The 68000 and 68008 CPUs had 6800 signals to allow seamless use of 6800 peripherals, could have been plugged into their original 6800 design. The 8086 had internal and external 16-bit buses (the 8088 had an external 8 bit bus), 16-bit data ALU, offsetting 2 16bit registers in a 20bit address ALU to give a 20 bit address, some instructions operated on 32bit data. The 68000 had internal and external 16-bit buses (the 68008 had an external 8 bit bus and external 20-bit address bus), 2 16bit data ALUs, one 16bit address ALU (two operations for 32bit address generation) and physical address truncated to 24 external pins, some instructions operated on 32bit data. Original Apple Mac was 128KB/512KB/1MB due to machine design choices. (A customized 68000 with modified microprogramming was used in the XT/370. Shjacks45 (talk) 11:35, 27 June 2013 (UTC)

SuperNOS/Gemini
Does anyone have access/links to articles regarding the origins of SuperNOS/Gemini? Specifically, during the period around 1992-1993, even prior to the Unix deal, there was a lot of noise from Novell about moving NetWare's services to a stable app server kernel. I remember this clearly because the kernel that was being discussed was the first time I'd ever heard the name "Linux" (it's truly ironic that that failed decision has now come full circle...) I want to add some info about this, but without citations it's going get serious scrutiny/criticism. EJSawyer 05:53, 7 March 2007 (UTC)
 * Not sure if it helps in regard to SuperNOS (a Novell project to merge NetWare 4.1 with UnixWare 2.0 around 1995), but the first time I heard about Linux as the possible core for a potential future Novell NOS has been in conjunction with Novell Corsair (1993-1995) and Novell Exposé (1994), which led its way to Caldera Network Desktop in 1995 and OpenLinux since 1996/1997. Caldera NetWare for Linux 1.0 became available in 1998. The second time was much later in conjunction with Novell SUSE Linux (since 2003) or Novell Open Enterprise Server (since 2005).
 * SCO Gemini, the OpenServer/UnixWare merger was rumored since 1996, released in 1998 as UnixWare 7.
 * Thinking about the 1992/1993 timeframe, I remember Univel Destiny (what became UnixWare 1.0) and Novell Star Trek, but there was no Linux involved in these projects, and the second was a project to replace the desktop, not the server.
 * --Matthiaspaul (talk) 14:15, 19 May 2014 (UTC)

Difficulties In Developing for Netware
Nothing is said about the lack of compilers, materials and support for developing for the Netware platform. It has improved greatly but was an extremely closed group. Administrators were prisoner to the few apps available from 3rd parties.

Explanation of Acronyms
"Novell servers could be assembled using any brand system with an Intel 80286 or higher CPU, any MFM, RLL, ESDI, or SCSI hard drive and any 8-bit or 16-bit network adapter, subject to availability of suitable drivers."

Great, but what is MFM, RLL, and ESDI??? Now I googled them all and found out, but us younger folks have no idea what these mean without further explanation. Perhaps they should be explained or at least be given a link to follow. I would even say you could change this to avoid using these terms at all. I won't change it due to my lack of knowledge of 80's computer equipment and Netware, but maybe someone should look at it. —The preceding unsigned comment was added by 68.93.155.35 (talk) 14:42, 9 May 2007 (UTC).

MFM and RLL are different  VERY  old ways to "talk" to the precedors of the later IDE disks. IIRC the encoding of data on the physical disk was done in totally different way. The controllers for RLL and MFM disks were different an 100% incompatible, there I'm sure.

Basically this might be a CITE from a NetWare 2.2 whitepaper. This should be moved to some "museum / history section", as in these days it WAS remarkeble that an OS was supporting any available disk type on the market. —Preceding unsigned comment added by 212.63.70.206 (talk) 09:01, 27 September 2007 (UTC)

Strategic Mistake Wording
It lists the strategic mistakes as starting around 1995, and then in the next sentence lists FreeBSD and Linux as viable alternatives. FreeBSD 4 didn't rear it's ugly head until 2000 (according to wiki), and I am not entirely convinced linux was all the rage in 1995 either. I don't know that the article intends to say that these OSes were popular in 1995, but the wording is unclear. 69.143.88.44 (talk) 06:13, 20 February 2008 (UTC)

DOS partition
Even if you decide to run NetWare 6.5 without a bootable DOS partition, the DOS partition will still be created and the OS loads files from that partition during boot. BTW, is there a list over all released versions of NetWare?

Ximalas (talk) 08:11, 27 October 2008 (UTC)

I'm not so sure that I remember anybody complaining about the DOS partition at all. I honestly don't believe it was an overbearing factor. --210.185.71.34 (talk) 05:20, 9 June 2009 (UTC)

Strategic Mistakes ??
Late adoption of a GUI and a more publicly accepted programing environment were probably the two biggest failures. Java capability came along in version 5 and by then the Windows environment was embeded on every school kids desktop. If you were learning how to program it was windows that you were programming, not NLMs (Netware Loadable Modules). Writing for Windows95 was not too much different to writting for Windown NT. Subsequently Microsoft, like Linux later, had itself armies of programmers developing meaningful applications for their server platform to run. On the other hand, with only a few exceptions, Novell was left to develop its own applications for Netware servers. Jave popularity came eventually but Windows and Linux could run that too, so as I see it, Java provided no strategic advantage to Novell. --210.185.71.34 (talk) 05:20, 9 June 2009 (UTC)

NOS or Not NOS - then What?
The artice suggests that Netware is a Network Operating System. However the Network Operating System article itself says that it is not. So what is it? --210.185.71.34 (talk) 05:41, 9 June 2009 (UTC)
 * It is a NOS. I correct the other article which seemed to contradict itself. 64.102.254.33 (talk) 13:57, 10 September 2009 (UTC)

netware
what is netware —Preceding unsigned comment added by 59.97.96.37 (talk) 14:42, 17 July 2009 (UTC)
 * Lots of information here: http://en.wikipedia.org/wiki/Novell_NetWare Aloysius-Fibonacci (talk) 17:40, 23 October 2009 (UTC)

Long file name support
Please see the bold or in the following passage:

".. long name support (standard filenames were limited to 8 characters plus a three letter extension, matching MS-DOS) or Macintosh style files."

or / for? --Vssun (talk) 11:50, 17 August 2012 (UTC)

Novell begins to recover
See the section NetWare 4.1x and NetWare for Small Business: Novell begins to recover. I think the section title is inappropriate as the section following the above one (NetWare 5.x) says that NetWare 5 was released during a time when NetWare market share dropped precipitously. --Vssun (talk) 08:09, 20 August 2012 (UTC)

Some qualms
1.) It would be nice to have a table of versions like wikipedia has for Linux and MS Software. There were differences from Netware/286 and Netware 2.15; Netware 3.0, 3.1, 3.11 (support for 4.0), 3.12 (supported 4.11), 3.2; Netware 4.0, 4.02, 4.1, 4.11, 4.12, 4.2; etc. 2.) Note no native IP support in release 4.0; IPX encapsulated in IP until 4.2 3.) Java console in 5 was very slow   4.) Artisoft Lantastic but also ("LAN Manager client" or "MS Client" then) Workgroup Add-on for MSDOS and Banion Vines 5.) I added Novell's sale, "Novell, Inc." implies a robust independent company. Attachmate split Netware and SUSE Linux divisions (strongly supported by IBM; IBM devices like cash registers use SUSE images). Fate of Netware and support of Netware users is not guaranteed.  6.) No note of Novell Small Business Server; the redirect comes here with no information but the Microsoft Small Business Server has its own article. Like MS SBS had lower cost of entry. Shjacks45 (talk) 13:32, 27 June 2013 (UTC)

Purpose: R to anchor
Note: This is about a recent reverted edit to a redirect that targets NetWare

I was the one who added to the rcat text: "or it might be an old section header anchored within the new header to prevent link breaks," with this edit less than a month ago. Before that, it just read, "to an anchored part of a page on the subject, other than a section." And I was wrong to change that sentence. The category's caveat is clear – it is only for anchors that are not in or near section headers. Before I just tweaked that caveat, its previous wording was, "Caveat: This category and its template are not meant to tag Anchors located within or at (just before or after) a section title of an article. Use R to section for such links to Anchors." So again, I was wrong to change the text at R to anchor that allowed it to include section-header anchors. Those types of anchor redirects should be tagged with R to section, but not with R to anchor. That is why I reverted my previous edit at the R to anchor template, and that is why I made the edit to this redirect, ".vlm", that targets a section of this NetWare article. So please revert your revert of my self-revert and let's make things right again. –  Paine Ellsworth   C LIMAX ! 00:03, 12 March 2014 (UTC)

PS. Thank you for restoring the Redr template on the redirect. Now we just need to use R to section rather than R to anchor. This redirect targets a section header, so R to section is the correct rcat to use. PS added by –  Paine Ellsworth   C LIMAX !
 * Hi guys, Mr Ellsworth asked me to comment on the importance of using the one template over the other, and I'm saddened to inform you it probably doesn't matter a hill of beans to anyone. I have a vague recollection of discussing the &lt;span tags&gt; scope versus the workings of Anchor, but those grey cells recall nothing now that would technically matter, especially since mid-section spans versus a end one are all formed the same, after wikimarkup is done.


 * Hence,... These are merely tracking categories that are a convenience to some, mainly bots I think (IIRC), but the "real impetuous" if I recall that correctly, was the excited push to put out a DVD version of the Wikipedia. Such pages were to be processed to establish criteria for which articles were going to be honored by being included, and so forth. Made for some crazy efforts too. This was around the time we'd become 'respectable' again after installing the early measures to better vett sources and got kind of militant about booting articles without multiple sources. Sad times, many topics (i.e. contemporary culture) stuff got chopped. Hope this helps, keep tagging but don't put any special effort into it. best regards // Fra nkB 05:34, 12 March 2014 (UTC)


 * Thank you, Fra nkB ; as the creator of the anchor category you've pretty much fixed everything about all this. We very much appreciate your input!  Joys! –   Paine Ellsworth   C LIMAX ! 14:42, 12 March 2014 (UTC)
 * I found this discussion from Template talk:R to anchor, hoping to get clarification on the wording "Also, the anchor might be an old section header that has been changed..." in the template's documentation. Unfortunately, it doesn't look like a definite conclusion was reached one way or the other. Would it help to create a new rcat like ?
 * The specific case I'm struggling with is Concurrent PC DOS ([//en.wikipedia.org/w/index.php?title=Concurrent_PC_DOS&redirect=no redirect=no]). I [//en.wikipedia.org/w/index.php?title=Concurrent_PC_DOS&diff=602021179&oldid=557154885 changed an rcat] and [//en.wikipedia.org/w/index.php?title=Concurrent_PC_DOS&diff=prev&oldid=604406357 self-reverted]; now I'm unsure whether the reversion was justified. (Note that that redirect is tagged as "with possibilities" and hence printworthy. As I mentioned in the first of those edit summaries, I wanted to avoid a conflicting categorization of printworthiness.) --SoledadKabocha (talk) 05:35, 16 April 2014 (UTC)
 * See also 1976 Olympic boycott ([//en.wikipedia.org/w/index.php?title=1976_Olympic_boycott&redirect=no redirect=no]) for another case of conflicting printworthiness where I wasn't confident enough to be bold. --SoledadKabocha (talk) 07:20, 10 May 2014 (UTC)
 * I think the problem is down to R to anchor making a redirect unprintworthy at the same time. Either the templates should auto-detect, if such implicitly added unprintworthiness gets overridden by another Rcat (like f.e. R printworthy) and consequently suppress the unprintworthiness, or we should remove the unprintworthiness from R to anchor. In my judgement, it's a different property altogether, and it shouldn't have been mixed up in the first place. I know many examples of printworthy redirects to anchors, as well as unprintworthy redirects to section headers etc.
 * --Matthiaspaul (talk) 13:31, 18 May 2014 (UTC)
 * I think that this is a great idea. Auto-detect would be great, but just stopping the default would be helpful. It's my impression that an anchor redirect is more likely to be printworthy if a user took the time to create the anchor. Moreoever, I don't see them as analogues to WP:SHORTCUTs, which are made up abbreviations, not existing terminology as are many anchor redirects. —Ost (talk) 20:00, 25 March 2015 (UTC)
 * Regarding the original discussion about the exact purpose of R to anchor and R to section, I obviously have a different understanding of their original purpose than Paine (and this is why I reverted him when he changed the description a while ago). In my judgement, R to section is to be used for redirects to actual section headings (that is, those created using the ==section title== syntax), not to alternative anchors placed inside or nearby section headings via anchor or similar means. Likewise, R to anchor is to be used for any kind of anchors (including those nearby or inside of section headings), which are not actual section headings themselves. This way, bots could check and possibly even resync redirects with section headings specifically by searching for ==section title==-like constucts in articles and R to section in redirects. This doesn't work, if nearby anchors are also allowed to be targets for R to section, because "nearby" is a vague criterium, making it difficult for a bot to distinguish between other nearby, but independent anchors and those meant as alternative section headers.
 * It might be useful to distinguish "alternative section anchors" (that is, those not created by ==section title==) from independent anchors as well. Therefore, we seem to need another type of Rcat like R to alternative section anchor.


 * On a different note, there are also redirects targetted at currently non-existing (independent) anchors in an article. This may indicate an error that needs to be fixed. Consequently, the German Wikipedia classifies them as error and the #target gets removed from such links over there. However, given that they don't cause any harm, they could also be used deliberately to help establish a future logical organization of contents even before the corresponding contents (and thereby the corresponding anchors) has been added to an article - in this case, they should not be counted as errors and removed, but left intact. Therefore, in order to allow bots to check and verify anchors as well, it might be useful to add another Rcat like R to potential anchor or R to anchor with possibilities to our repertoire.
 * --Matthiaspaul (talk) 13:31, 18 May 2014 (UTC)
 * I agree that R to anchor should go to specifically defined anchors and R to section specifically to sections, especially if there is now no technical reason to not. Full disclosure that that's how I've been using them. Incidentally, I've noticed this conflation of uses as a problem with other redirect templates, which seem to have generic or self-evident names until one reads the note or documentation.
 * I could see the usefulness of the other templates, but not convinced that they are needed if we're not removing bad sections from redirects. Personally, I'm glad that we don't do that, as the section artifact often needs to be repaired rather than removed. —Ost (talk) 20:00, 25 March 2015 (UTC)

Motorola 6800 or 68000?
Which one was early NetWare running on? The article says 68000 in one place and 6800 in another. They're not at all the same thing. -- 89.182.220.207 (talk) 11:12, 22 April 2014 (UTC)
 * Definitely not, thanks for spotting this. I have changed it to 68000 now, as this is more plausible, given that there was a CP/M-68K, but no CP/M for 6800 processors. Mark Dixon's blog mentions a 68000 as well. Nevertheless, we need a more reliable source for that. --Matthiaspaul (talk) 13:47, 19 May 2014 (UTC)

Tidy, get some refs
I've just done some edits, adding some references, and moving some material about to keep the story a bit clearer. More needs to be done though... Snori (talk) 07:42, 3 September 2014 (UTC)

External links modified
Hello fellow Wikipedians,

I have just added archive links to 1 one external link on NetWare. Please take a moment to review my edit. If necessary, add after the link to keep me from modifying it. Alternatively, you can add to keep me off the page altogether. I made the following changes:
 * Added archive https://web.archive.org/20001205045800/http://www.informationweek.com:80/673/73iunet.htm to http://www.informationweek.com/673/73iunet.htm

When you have finished reviewing my changes, please set the checked parameter below to true to let others know.

Cheers. —<sup style="color:green;font-family:Courier">cyberbot II <sub style="margin-left:-14.9ex;color:green;font-family:Comic Sans MS"> Talk to my owner :Online 14:42, 18 October 2015 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 3 external links on NetWare. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20160728221417/http://www.textmodegames.com/download/snipes.html to http://www.textmodegames.com/download/snipes.html
 * Added archive https://web.archive.org/web/20061125020459/http://www.novell.com/documentation/ncs6p/pdfdoc/orionenu.pdf to http://www.novell.com/documentation/ncs6p/pdfdoc/orionenu.pdf
 * Added archive https://web.archive.org/web/20050219225024/http://krakau-inc.com/199406.htm to http://www.krakau-inc.com/199406.htm

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 15:06, 12 January 2018 (UTC)

Client footprint
One thing this article fails to mention is the size of the client footprint. In the days of the DOS operating system (which includes Windows until 3.0), the available memory was a paltry 640K. Novell had a massive advantage here with a client footprint of about 60K, Microsoft LanManager was 120K with minimum features or 200K with full features. 120/200K meant that there wasn't enough room for other applications to run rendering LanManager as useless. This was the primary reason why Novell established such a dominant position in the early days. This dominance made them arrogant and they didn't innovate sufficiently to stay competitive in the post DOS memory era. the one innovation they did create was NDS which was a great concept and product and they could have done more with it.
 * And if you had a 386, you could run Novell Client32 for DOS, which only occupied 4K of conventional memory, compared to MS's stack which occupied 200K+ as you've noted. It accomplished this with a mechanism similar to how Helix Cloaking or Digital Research DPMS worked.
 * Lemme guess? That's an unbelievable claim and citation needed.

Incorrect statement Removed
In the section about NetWare v4, there was a statement:

The upgrade was not without its flaws – initially NetWare 4 could not coexist with earlier versions on the same network because of incompatibilities.

That is not correct. I personally ran a mixed NetWare v3/v4 environment.

139.138.6.121 (talk) 07:38, 4 July 2021 (UTC)


 * Netware 4.0 was very seriously flawed, not only in that respect. 4.01 followed very closely, fixing most issues. But without proper WP:RS, claims like that may be removed (can't find another source either). --Zac67 (talk) 20:55, 24 September 2021 (UTC)