RE: Fitness for a particular purpose Nathan T. Freeman 11.Dec.02 11:13 AM a Web browser Notes Client 6.0Linux - RedHat, Linux - SuSE
Fine, I'll be more precise. Should Notes for Linux be written with GTK+ or
Qt? What configurations for window managers should be supported?
Or are you proposing that Lotus grow all their own directly from XFree86?
Admittedly, using Qt would probably allow the greatest chance for
cross-platform availability. But even so, that's a HUGE QE effort to be
Seriously, our company has had to do many web-only apps in the past and I
know how bad notes it is in this aspect.
I've been building web-only apps for the last 4 years. As much as browsers
suck, there are some enormous advantages they offer over the native client
beyond availability. How about dynamic tables? How about the UI event model?
How about the ability to target any other object? How about media-specific CSS
overlays? How about drag & drop control over any object? How about control
over right-click menus? How about keystroke events?
Browsers offer substantially better UI control to developers. Is it harder to
do? You bet. Is it worth it? Absolutely.
WIth Notes 6 it's getting there and if there's really iNotes and the new
Rich Text control for Linux, that would be a good first step.
Notes 6 is a considerable improvement. There's still a lot more to do.
If the cost of Linux is a 10% lower than the one of Windows, then that's
a 10% more profit for us.
That sounds like programmer accounting. First off, both IDC and Meta agree
that the *total* cost of a Linux system is *higher* than a Windows system.
Second, if you think you can take all the saved cost as profit in a competitive
environment, I'd suggest reexamining your pricing policies.
Since when a 10% profit is unimportant?
As long as it's there, it's important. Is it there? IDC and Meta Group say no.
However, if you think ofering Linux to our clients is not a good way to
make money, then your clients are very different from ours.
I have little doubt of that. And I'm not proposing that it's not a good way
for *you* to make money. It clearly is. What I question is whether your
*clients* make money.
5 years of migration expenses? Why? You don't have to migrate all
computers to linux at once, if you think that's what I was saying.
You do if your objective is to save on Windows licensing costs, which is the
only direct financial benefit to a Linux migration on the desktop.
However, you can dump Exchange and go for Domino+Notes because you know
that you'll be able to slowly migrate to Linux if you have the need.
You can dump *Exchange* and migrate to Domino now, if you want. Exchange is
simply the back-end server. If you want to run Linux Domino servers in a
Windows desktop environment, there's nothing stopping you. This makes a lot of
sense and is widely done.
Or if you already have Domino we can sell you support to slowly migrate
your machines to Linux.
Good news. I'd like to see more Domino servers moving to Linux.
Anyway, I don't really see the point why you don't want IBM to develop a
Because I think it would be an expensive, resource-consuming project that would
ultimately prove fiscally unsuccessful, threatening an already shakey future
for the native client. How much have people talked here about the questionable
future of the Notes/Domino platform? Do you really think that a $100 million
investment in a high-risk platform would *help* that position?
I want to see IBM put those resources into better return projects. This would
1) a GPLed local XML/XSLT transform object for Mozilla, that can be called
similarly to the Microsoft XML Transform ActiveObject.
2) XSLT engines for Domino, that allow for something like
/myView?ReadViewDesign&Format=XSLT to put out a stylesheet for the view design
that can be applied to the /myView?ReadViewEntries result set.
3) componentization of the iNotes features so they can be used in custom Domino
applications. This includes the drop-down menus, the calendar control,
rich-text control, attachment control and all those other cool toys.
4) enabling of all core web functions for Domino on Mozilla, which is
DEFINITELY achievable in the Mozilla DOM (considerably better than prior
In fact, most of us understand that it might be too costly and we are
just asking for a first winelib port to test the market, so that when there's a
real need for it, there's already an stablished market.
I'm a huge advocate of this idea. It's a pain to QE, because of the number of
layers, but it's certainly easier than a direct port. And it would give a good
metric on the market opportunity.
And before you say it might be too costly, transgaming (just to say
someone else apart from codeweavers) are porting Windows games to Linux using
Wine. So, a small software company can port and give support to The Sims and
some other games, but IBM can't do the same with Lotus Notes? Come on!
Well, I'll point out that Notes is almost definitely more complicated than just
about any game on the market. But yeah, I agree that a Wine/Crossover effort
would make a lot of sense, and would encourage IBM to pursue it vigorously.