From: Dom Lachowicz (doml@appligent.com)
Date: Mon Mar 25 2002 - 12:33:44 EST
> Still here? :) OK. I must admit that I've had doubts about my continued
> use of time on AbiWord. Not the project itself or the gang - which I
> really love - but I'm feeling we're not getting anywhere (FAOD I'm fully
> aware of my personal contribution the last months which is very close to
> zero, AWN aside). Bugzilla keeps fighting back every attempt at reducing
> the Bug counts, and doing QA, I see the same old bugs reappear after a
> while.
>
> Maybe it's just because writing a WP is tricky. Maybe it's because the
> backend is stretched beyond its limits.
>
> Whatever it is, I know that Miguels rant about Ximian's reasoning for
> Mono (the massive time spent on Evolution development - I forget the
> number, but it was a factor >10 over what they'd expected) is starting
> to get to me.
FWIW, C# or Java isn't going to be much better. In fact, no programming
language is a godsend and will make all of our work easy and make all of
our bugs go away. Mono will be better for Evolution, probably, but you
have to consider that these guys are using a self-created object
heirarchy based on C and structs-as-vtables, without many of the
language features that BOTH C++ and C# offer.
If Miguel is reasoning that C# will be a magic bullet, then he is on
crack. We've all seen magic bullets come and go. Each one had something
to offer, and something to not like about it. It's highly unlikely that
any programming language change will drop *most* product's development
time by a factor of 10, or even a substantial fraction thereof. I do not
believe Abi or Evolution to be exceptions to this general statement. I
suggest that for further insight, you go read Herb Brook's "Mythical Man
Month." There is no escaping the fact that systems are inherently
complex. AbiWord and Evolution are such systems.
> Bottom line: when I'm now thinking about what to spend my precious spare
> time on, I'm not sure C/C++ programming comes into the picture anymore.
> I think, for me, the performance of computers have come to the point
> where I no longer mind the thought of running my apps on a VM.
Which is fine. I'll also go suggest that everyone who complains that Abi
is slow on their PII 233 with 64MB RAM go talk to you about the merits
of this argument :) (hrm.. didn't we get this email today???) VMs and
interpreted languages are great and certainly have their places. So are
other compiled languages. The important part is to use the correct tool
for the job, and I've seen NO argument (viable and substantiated,
anyway) for either keeping C++ or moving to another language,
interpreted or otherwise.
> So here's the question: Is there room for liberating thoughts such as
> building an additional AbiSource framework, maybe using a spreadsheet as
> test app, that is both programming language and platform neutral (the
> parts of .NET/Mono that I've understood - and really like. I don't know
> if there's more hype to it than that, but I wouldn't be surprised).
The point and main advantage of using C#, in my opinion, would be that
we wouldn't have to build Yet Another Framework.
Quite frankly, this argument sounds an awful lot like the original "why
won't abi just do gtk+ and then port gtk+ elsewhere" arguments. Or the
wxWindows arguments. Or the QT arguments. The difference here is that
the argument is coming from someone whom I deeply respect and who has
actually contributed to the project in a meaningful way.
The fact is that Mono is immature and slow, if it even works at all.
Last time I checked, it only works on ia32 linux systems, and poorly at
that. C# has no runtime or virtual machine to speak of on Solaris, BSD,
MacOS, QNX, BeOS, and arguably Linux. Its gui classes are malformed and
misadjusted, especially with regard to its container classes (or total
lack thereof).
Mono is also being built with many of the same tools and programming
languages as Evolution, which took 10x the manpower that was originally
expected. Should we expect a pattern here? I honestly don't know and I
do hope for the best. But I'm a youthful optimist.
Should we change? Quite frankly, I don't know. But I'd rather spend the
rest of eternity in this bugfixing feature-freeze hell than rewriting
yet another framework and starting everything from scratch. There are
proven and tested frameworks out there. Let's use them.
Dom
This archive was generated by hypermail 2.1.4 : Mon Mar 25 2002 - 12:36:10 EST