[WinMac] Re: MacOS 9: Upgrade now or wait? ALSO: NT4/SP6 is out
Daniel L. Schwartz(expresso[at]snip.net)
Fri, 29 Oct 1999 17:50:38 -0400
Ahhh... I love a spirited debate! We're not making this personal amongst
list readers - Just among those worth >$1 billion. Besides, they can afford
to be insulated from the rants of us pee-ons! :)
First off, I'm running Win2k RC1 on one of my 7 Alpha's - A DIGITAL
AlphaPC 164SX @ 533 mHz. And it runs quite nicely. There are copies of RC2
floating around, too!
Now, on to the really spirited debate!
At 04:21 PM 10/29/99 -0400, you wrote:
>My turn
>
>> From: "Daniel L. Schwartz" <expresso@snip.net>
>> Date: Fri, 29 Oct 1999 15:47:20 -0400
>> To: winmac@lists.best.com
>> Subject: Re: Re: MacOS 9: Upgrade now or wait? ALSO: NT4/SP6 is out
>>
>>
>> OK, I'll respond inline...
>>
>> At 01:02 PM 10/29/99 -0400, you wrote:
>>> Okay, Microsoft hasn't *Charged* for the bugfixes...but what new
>>> features have they added as compared to what YOU are calling *bugfixes*
>>>
>>> hmm....well, sp3 made you do registry hacks to keep clear text
>>> passwords...that's a fun feature...not much else
I use .REG files instead of Regedit hacking to roll out Registry
modifications...
>>> now, let's just see what Apple's *bugfixes* did, starting with 7.6.1 as
>>> the base
>>>
>>> 8.0
>>> multithreaded finder
>>> PPC native finder
>>> sticky menus
>>
>> -> Shareware "borrowing"
>
>Um...name me the shareware native finder replacement. NOT add-on, but a
>piece of shareware that duplicated the Finder, and let you *delete* the
>Finder, and was PPC native
CopyDoubler, part of the Connectix SpeedDoubler troika...
>>
>>> popup windows
>>
>> -> Shareware "borrowing"
>>
>>> threaded file copies and trash empties
>>
>> -> Long a staple of SpeedDoubler
>
>Which breaks with every OS update...getting rid of that made my life SO nice
I sorta agree; but then like with any OS upgrade (NT .AND. Mac) you need
to look for updated Ring zero drivers before proceeding...
>>> cut out the '030 macs, long time coming, unlike the intel bolt-on
>>> theory
>>
>> Wrong: They cut out non-32 bit clean machines.
>
>To quote you...WRONG...7.6 cut out the non-32 bit clean Macs...8.0 was 040
>and better ONLY....
I'll have to re-check this... Even *I* don't have any more `030 Macs!
>>>
>>> 8.1 (free, but you *forgot* this one)
...And I left it off my $99 per bugfix roster, too
>>> New file system, HFS+
>>
>> -> With *no* repair tools until NUM 3.5 many months later...
>
>Again, not true...you forget TechTool Pro, which came out within a month of
>HFS+, and the ever present Disk First Aid...not as good, but not to be
>discounted either
TechTool Pro wasn't really proven. And Alsoft "fouling the footpath" with
their crap didn't help, either...
>>> VM system improved, using actual paged VM, nice performance increase
>>
>> -> Baloney: VM still requires - To this day - a contiguous block
>> equal to the size of the desired VM size is required, unlike
>> OS/2, NT, *nix... Or even windoze 3.1!
>
><sigh> it reserves that space on the HD, however, the actual VM ops are 4k
>pages. Deal, read the technote, get the facts straight. In truth, by
>reserving HD space for the VM swap file, you ensure you always have it
>available, UNLIKE NT/9X where you can happily fill the hard drive(s) to the
>point of not having any swap left....result...KEERASH!
NT uses fixed VM ranges, by default with the minimum equal to
(installed RAM) + (maximum Registry size). But NT will run nicely with just
a 2 MB paging file; and unlike MacOS that 2 MB gets added to the heap. [I
dare not comment about other windoze crap-ola... 3.1/95/98 is for the
unwashed masses and kiddie gamers!]
>>> Open Transport upgrade, allows for single link multihoming
>>> ppc native filemanager
>>
>> -> Yeah, it took over FOUR years for this.
>
>And how many years will it have taken NT to be able to dynamically change
>network configs...MUCH more than 4...
Sorry, it's included in Win2k. Besides, you don't *really* need to reboot
NT4 as often as prompted: Just stop and restart the affected services, and
*most* of the time it'll work. Now, I need to see how well it works with
hot-swap PCI cards - I have a brand new IBM Netfinity 5500 with this
feature sitting in my kitchen that I'm setting up as a 1/7 terabyte
NT4/SP6-AppleShare IP server...
[Don't get me wrong: Changing IP & AppleTalk configs on the fly with Open
Transport is indeed nice...]
>>> 8.5
>>> no more 68K macs!!(yay!)
>>
>> -> WRONG! 68k Macs (including the just-previously released
>> PB190) were Steveified with 8.1. Besides, you call this
>> an "Upgrade?!"
>
>Gee Dan, then that Quadra I was using really WASN'T running 8.1 with a 2GB
>HFS boot disk and a 10 GB HFS+ application disk? (No PPC card) Again, get
>your facts straight....this faux pas was just plain dumb, you must be busy.
Yeah, after I measured ~35% speed hit on a Quadra 840AV between 7.6.1 &
8.0, I killed off 8.0 on all `040 Macs I support...
>>> PPC native applescript
>>
>> -> Yeah, it took over FOUR years for this.
>
>Still beats the heck out of anything winders has
MSVCRT42.DLL
>>> actual appearance controls, allowing for better customization
>>
>> -> Shareware rolled in...
>
>your point is?
>
>>
>>> Desktop printing and Network setups scriptable
>>> Radically faster Finder copies and Network I/O
>>
>> -> Again, "borrowed" from SpeedDoubler
>
>You really need to slow down...SpeedDoubler had scripting additions for
>Desktop printing and Network setups?
No, I was commenting on the remark "Radically faster Finder copies and
Network I/O"
>I think not. Also, they didn't *borrow*
>speed doubler code, but rewrote the entire Finder algorithms...(That from
>the guy who actually wrote it. ) Speed double was the inspiration...but not
>the code source.
>
>>
>>> Sherlock
>>
>> -> "Borrowed" from Copernic...
>
>Again...so?
It was a freebie rolled into System 8.1.
BTW, if you're an NT user check out Copernic at <http://www.copernic.com>
- You'll really like it, especially if you have a decent Internet
connection...
>
>>
>>> SNMP
>>
>> -> Aw, shucks, MacSNMP has been around since Apple Internet
>> Router 3.0
>
>Not working under Open Transport Sparky
Apple Internet Router 3.0.1 works nicely under Open Transport - Been
there, done her. In fact, AppleShare Server 4.2.2 works *extremely* fast on
a Yikes G4, too...
>>> 8.5.1
>>> FREE bug fixes
>>> Fixed random, and uncommon disk corruption
>>> Fixed some applescript memory leaks
And it was *strongly* recommended to apply 8.5.1 ASAP, too...
>>> 8.6
>>> Free
>>> Fixed a lot of DHCP errors
>>
>> -> Including Apple's incorrect implementation, per the chairman of
>> the IETF's DHCP subcommittee...
>
>Oh, you're sitting in MS land whining about poor standards
>implementation....Doesn't that make you feel a LITTLE creepy?
Not really, because Internet Engineering Task Force is unix-centric... And
those guys sure have a hard-on for Micros~1, too...
>>> Fixed a *long* standing ethernet driver bugs
>>
>> -> Why was it so long standing to begin with?!
>
>Because until the rest of the system could handle the fix, it would have
>broken more than it fixed...(driver death under EXTREMELY heavy loads, i.e.
>max loading a 100mbit line)...so it was decided to wait until the pain level
>was acceptable. MS could learn from this
Or, you have a choice of NIC vendors and their driver code. Personally, I
prefer i82557 - based NIC's for NT/x86, and DIGITAL's 2114x - series
("Tulip") cards for PCI Macs and Alpha's...
>>> more code transitioned to PPC Native
>>> Cron - type scheduler included on CD
>>> FTP/HTTP up and downloads now scriptable
>>>
>>> 8.6.1
>>> Actually, a fix for a single problem, font corruption, not an OS
>>> update. 8.6.1 is what folks called it.
>>
>> -> This is why you have regression testing. Even PageMaker 6.52 - One
>> of the most popular MacOS apps - failed the FontFixer test.
>
>This is why ADOBE has regression testing...Quark worked jes' peachy
About the only thing Quack did right with their cruddy XPress...
>>>
>>> 9.0
>>
>> [OS9 stuff cut, because I haven't deployed this ".0" release yet...]
>
>Then why do you comment on it
*Because* it's a ".0" release!
>>> Gee, some actual new features...worth paying for, and a more than a few
free
>>> bugfixes. So in 2 years, we went from 8.0 to 9.0, shed a boatload of 68k
>>> code, fixed a LOT of longstanding bugs.
>>
>> Ever hear of "Fat Binary?! Of course, the 68k code was long overdue
>> in being shed... Because the 68k OS support was also shed.
>
>Yeah, it was a transition concept, the transition is done, 68k is dead, long
>live PPC
>
>>
>> .AND.
>>
>> Paying for bugfixes?! Gimme a break!
>>
>>> Did some stuff break, yes. Is that a
>>> surprise, no. Has NT done any better? Not even close, and heck, Win2k
is not
>>> even half the OS that NT 5 was promised to be.
>>
>> So tell me, have you actually *used* NT5 (W2k) yet? In a production
>> environment? I have (b3RC1), and it works nicely... Especially the enhanced
>> SFM implementing AFP/IP directly?!
>
>(Not on one of your vaunted Alphas you don't)
OH YES I DO!! [SEE ABOVE!]
>No, because
>a) I don't EVER do betas in production REGARDLESS of who makes them. I have
>lived by that for over 10 years. Works real well too. And
OK, what about the 8500 & 9500 fiasco with Open Transport 1.0.8? I was the
support engineer for a digital imaging company in early 1996 when I
inherited a bunch of 9500/132-based $50,000 digital imaging systems. I
*had* to deploy 7.5.3b7 and then 7.5.3b9 JUST TO GET THE DAMN SYSTEMS TO
WORK. Even Spindler admitted that they jumped the gun by releasing hardware
(PCI bus) before the software (Open Transport) was ready to be shipped...
>b) face facts, any rewrite from MS is ALWAYS a beta until the first service
>pack.
Well, you better get ready for production-ready Win2k. Even RC1 - From
back in May - worked flawlessly on both x86 and Alpha.
>Besides, RC2 won't install onto the Adaptec SCSI card drives I'm
>using. But I have gotten it to work in VPC.
Why do you think I call it "AdapCrap" or "InepTech!?" Switch over to
Symbios 53c875 - series SCSI controllers and your problems will vanish...
>and Win2k is NOT NT v 5.0, it's Win2K 1.0
>
>>
>> Yeah, some of the enterprise stuff in W2k will be missing in the original
>> release, such as failover/clustering... But they will be free add-ons
>> included in a Service Pack. After all, with 65 million lines of code
>> *something* has to give!
>
>Exactly, it was over promised, will under perform, (classic Microsoft) and
>if you think I am going to trust MS to write 65 million lines of quality
>code, you are REALLY fooling yourself. MS has NOT shown any indication of
>improved quality, as the continuing flurry of security and bug fixes for IE
>and Office show me. If they can't get a browser right, I'm supposed to buy
>into the OS? Oh please, get real. (NOTE: that's IE for WINDOWS...the Mac
>versions of IE/OE are light years past the winders versions, in both
>quality, and functionality).
Well, IE5 really flies under NT. Micros~1 yanked a lot of garbage out of
the rendering engine from IE4, and the difference is really nice...
>I am done bending over in front of Bill and screaming..."Thank you sir, may
>I have another"
Well then, ask him to put vaseline on it and turn the paddle lengthwise!
>So how many Alphas ARE you going to have to replace with Pentiums....damn,
>that must really suck...bet your Alpha clients are freaking thrilled...
I could fill volumes on Compaq's gaffe on AlphaNT... But fortunately there
are (essentially) three replacement OS' to run on these machines: OpenVMS,
Tru64 Unix (formerly DIGITAL Unix), and Linux. Plus, you can use Tru64
libraries with Linux... [For more info, go to:
<http://www.AlphaNT.com>.
As to the "Enrico-ifying" of NT on Alpha, basically it was a palace coup
foisted by the Tru64 group. Compaq was paying Micros~1 $100 million per
year, plus maintaining DEQwest with over 100 AlphaNT engineers, just to
keep AlphaNT going. It was a bad deal left over from MICA; and like IBM &
MIPS before it (with NT/PPC & MIPS/NT) Q saw the light and shoved it back
into Bill's court...
Cheers!
Dan
-----------------------------------------------------------------
While you're spamming me, don't forget to include these guys:
Chairman Reed Hundt: rhundt@fcc.gov Comm. James Quello: jquello@fcc.gov
Comm. Susan Ness: sness@fcc.gov Comm. Rachelle Chong: rchong@fcc.gov
US Postal Service: customer@email.usps.gov Fraud Watch: fraudinfo@psinet.com
Federal Trade Commission: consumerline@ftc.gov
Oh, and while you're at it, pound some sand too!
admin@loopback $LOGIN@localhost $LOGNAME@localhost $USER@localhost
$USER@$HOST -h1024@localhost root@mailloop.com
-----------------------------------------------------------------
*** Windows-MacintoshOS Cooperation List ***
FAQ: http://www.darryl.com/winmacfaq/
Archive: http://www.darryl.com/winmac/
To unsubscribe, send mail to winmac-request@lists.best.com
with just the word "unsubscribe" in the body of the message.
This archive was generated by hypermail 2.0b2
on Fri Oct 29 1999 - 16:32:43 PDT
|