Chad Smith wrote:

MS does it "right" because they "had to". We do it "wrong" because it's inheritied from propriatary software. It's weird, most of the time, I here flossies say that the reason MSO is modualar is because he bought a bunch of different programs and lumped them together. And the same flossies say that OOo is monolithic because it was inheritied from StarOffice.... So MS "inheritied" there correct way, and OOo "inheritied" there wrong way...

I'm not sure you read my email. OOo and MSO are designed quite differently. And I'm not sure what this trade-off you're talking about is. OOo components share most of the code. Putting them together is reasonable because it significantly reduces the ammount of code you (actually, Star Division) have to deal with. MSO components, from what I can tell, seem to share very little. They don't seem to have much to gain from being amalgamated.

I wish you'd understand that these really are different designs. For example, OOo is very strong in vector graphics, and OOo has its own GUI routines (so it can be cross-platform). Right there, you have a lot of code that is in common among at least several applications and should not be duplicated.

Just as modularity is important for good design, so is code reuse. In fact, when done well, modularity can simplify code reuse. For example, if we were making OOo today, we might have chosen to make the graphics routines into a separate library. That increase in modularity would not get in the way of code reuse (each component would use that library) but it would allow other software products to use the same library. Why did Star Division not do this? Probably because they were not interested in other products using the graphics routines they wrote. Personally, I would like UNO to be a separate library. That would reduce the size of OOo (make it more modular) without getting in the way of code reuse.

Another thing you have to realize is that no one here has said that the MS Office model is crap. Based on my understanding of the apps (that they have little to share), it seems just fine to me. Personally I prefer OOo but that's because OOo gives me some things (vector graphics, cross platform, etc) that are important to me.

And it goes further, people dare to call OOo's monolithic design "elegant"! That's BULL!

1) Are you a developer? Do you even follow what we're trying to say here?
2) Who called it "elegant"? I hope it wasn't me. I like the idea of sharing code where it makes sense, the way OOo does. But at the same time, that shouldn't be taken to mean that OOo isn't bloated.

What's "elegant" about downloading 60 MB which expands into 150 MB just so I can type a bookreport and make a budget?

If what you want is a book report, consider using Abiword. OOo provides a lot of very advanced functionality which isn't interesting for simple documents.

Question: How much space does MS Office take? I was sure it was a heck of a lot more than 150 MB. I've heard it's close to 500 MB. Wouldn't this be an example of how sharing code can reduce the total ammount of code?

The fact is, people, if monolithic design is "BAD" then it is always "BAD".

Like Obi-Wan said "only a Sith speaks in absolutes..."   ;-)

I'm sorry my friend, life isn't as simple as a 1-sentence catch phrase. Lumping code together for no reason other than to lump it together is very bad design. Reusing code is good design. Dealing with small self-contained components is good design. These two often go together, some times they don't.

And if these two facts are true, in at least one way, MSO is *better* than OOo.

There are many ways in which MSO is better than OOo. Far be it for me to say otherwise. And I fully expect that any one of its components is more maintainable than OOo, due to a smaller code base.

Just because it beats you in one area doesn't mean you pack it up and go home, but you should at least be honest enough with those who ask the question, and with yourselves, to admit that MSO beats you when and where it beats you. You can't fix a problem that you don't admit *is* a problem.

Why of course.

Stop lying to yourselves and admit that the monolithic design of OOo isn't as good as it could be.

I think you're reading completely different emails from the ones I'm reading. I am not aware of anyone who made that ludicrous suggestion, yet you take it as given.

Cheers,
Daniel.

Reply via email to