Re: The Netscape / Microsoft / Future Quagmire

Scott E. Preece (preece@predator.urbana.mcd.mot.com)
Fri, 18 Oct 1996 11:36:44 -0500


Date: Fri, 18 Oct 1996 11:36:44 -0500
Message-Id: <199610181636.LAA12135@predator.urbana.mcd.mot.com>
From: "Scott E. Preece" <preece@predator.urbana.mcd.mot.com>
To: marnellm@portia.portia.com
CC: smishra@cc.gatech.edu, snowhare@netimages.com, jaobrien@fttnet.com,
In-reply-to: Matthew James Marnell's message of Fri, 18 Oct 1996 11:48:50
Subject: Re: The Netscape / Microsoft / Future Quagmire

 From: Matthew James Marnell <marnellm@portia.portia.com>
| 
| What BS is this?  Clearly, for pages that support CSS and are hence
| MSIE 3.0 for W(95|NT) enhanced, for people that aren't using W(95|NT)
| and have a recourse of Netscape or Other, beating on Netscape to
| add CSS hasn't done s*** and Other generally tends to be a free
| unsupported (for the most part) browser where the persons OS !=
| W(95|NT).  So, what's the recourse then?  Scrap everything and switch
| to W(95|NT)?  Beat on a bunch of students over at NCSA?  Whatever.
---

Well, exactly.  If you discover you're missing something significant
because you're using Netscape, you should at least be thinking about
whether to switch to MSIE or whatever.  That's the point.  If your
vendor isn't getting the job done, you need to think about changing
vendors.  In the CSS case, I'm sure that if enough providers use CSS,
and it really does look better on MSIE, then Netscape will be forced to
support it, too.  Your claim that they haven't responded to pressure so
far is based on the fallacy that the readers of this mailing list are
either enough people or the right people to produce an amount of
pressure that Netscape will notice.  We're not.

---
| :>I don't think browsers are or should be mundane - I'm inclined to think
| :>they should become our primary user interfaces, with a little more
| :>evolution.
| 
| It'll go the other way, before it goes that way.  By the other way,
| I mean that the OS will build the WWW into the OS (it's already
| almost there for both the Mac and Windows).  It'll be all the way there
| for NDs and PDAs based on Java.  But this doesn't get past the fact
| that browsers would have to evolve an awful lot in order for it to
| replace anything as the Primary UI.
---

What you say is what I meant.  It's irrelevant (except maybe to
Netscape) whether you view it as the browser beccoming the shell or the
shell becoming the browser - th point is that the architecture of the
Web will be the architecture of the user interface.

---
| :>The Web is a major step forward in platform-independent integration of
| :>data and process...
| 
| No, according to the opinions that you've espoused in your mail message,
| the WWW isn't a major step forward in platform independant integration,
| but a step sideways if we're constantly downloading new and different
| browsers every time we stumble on a page that our browser doesn't do.
---

I'd like to think we're moving towards browser capabilities that will be
downloadable on demand and will be largely browser-independent.

---
| Which still fools a lot of the people that are more than happy with
| an earlier version of MSIE or Netscape, or a user of MSIE on the
| Mac and Win3.1 which isn't current with the 95|NT versions.  When you
| stumble accross a page that has "This page best viewed with [Netscape
| Now]" and it looks like crap because you're still using 1.1 because
| it does what you need and doesn't eat up 14MB of RAM, but the web page
| doesn't say "This page best viewed with Netscape 4.0b2" you think that
| the webmaster is an idiot, and rightly so, becuase you're using Netscape
| and his pages still look like s***.
---

I agree the bugs should indicate browser capabilities or versions,
rather than brands.  On the other hand, when you say "it does what you
need" is you are missing the point that if the page looks like shit the
browser arguably *isn't* doing what you need.

---
| :>Something else I expect, by the way, is browsers that maintain their
| :>currency automatically by fetching new components from their vendor's
| :>server as needed; this requires a shift to a much more modular
| :>architecture for browsers (to keep upgrade downloading cheap and fast),
| :>but I think it's sure to turn up in a year or so.
| 
| It's sure to turn up in Java based browsers in the immediate future,
| it's going to be much more of a hassle for platform specific browsers,
---

The key is to modularize the browsers so that only the pieces you need
get loaded and so that upgrades can be done piecemeal.  Similarly,
platform-indpendent implementations of functionality have a big
advantage, once the technology gets to the point that the performance
hit is manageable.  Java's not there yet, but it's getting better fast.

---
| it's also likely to be a feature that pisses more people off than
| expected, especailly with apps like netscape who are so resource hungry,
| that Joe Average User with the base Win95 machine with 16MB of RAM
| is going to have to kill some of his other apps in memory in order
| for the Netscape Navigator to run after it's updated itself and added
| another 4 MB to itself in the process.
---

Part of my response is "real OSs have virtual memory..."

But, seriously, the point is to be able to upgrade small, independent
pieces of functionality rather than the whole browser.  This is still
really hard, because the architecture for the pieces isn't clear yet
(e.g., until you've thought about frames you may not have thought about
your browser needing to have a layout module).  My hypothesis is that
large areas of the browser will become relatively static and modular and
that the locus of innovation will be progressively less global.  It's an
unproven hypothesis.

On the other hand, look at the UNIX experience - the architecture
evolved so that it became possible to innovate in independent,
modularized pieces of the system that *could* be released independent of
the whole.  For instance, filesystems can be replaced without changing
the rest of the OS.  So there's at least some hope for modularity.

---
| :>The Web is much to young for anyone to be saying "I've found my browser,
| :>I'm going to stick to it, and I don't want to know what I'm missing."
| 
| While I happen to agree that it's far too early in the game for this
| to happen, it doesn't mean it isn't or won't continue.  People are more
| likely to switch ISPs than browsers.
---

They should - ISPs should be transparent and trivially swapped.

scott

--
scott preece
motorola/mcg urbana design center	1101 e. university, urbana, il   61801
phone:	217-384-8589			  fax:	217-384-8550
internet mail:	preece@urbana.mcd.mot.com