ZDNet has an editorial that casts doubt into the whole Open Source thing. The general tone of this piece is that esr is a little too romantic and hopeful for Open Source in general, and that leads to an unfair amount breathless optimism.
True, esr is a heavy with praise for all things Open Source, and quite heavy with the condemnation of all things closed ('specially Redmondware). But, as always, the true answer lies somewhere in the middle.
First, Larry Seltzer asserts that open source runs out of steam when it comes to end-user apps. He cites "[t]he only serious attempts at end-user-oriented open source programs" are things like Mozilla and OpenOffice. He then goes on to describe how neither project would have moved forward without large amounts of corporate investment ("subsidies").
Seltzer then continues to describe how "cathedrals" are better at shipping on schedule than the "bazaar model -- using the 4-year gestation period for Mozilla as proof of that fact. Funny. I seem to remember a long list of projects at Microsoft that never saw the light of day, or were shipped only after many more years of development than originally anticipated.
Seltzer's argument doesn't ring true. KDE is a great example of open source developers working on developing end-user applications. Last I checked, the KDE development team doesn't draw the same hefty "subsidies" that other development teams have had over the years (true, SuSE has funded a good deal of KDE development, but I'm pretty sure the total corporate expenditure on KDE has been significantly smaller than OpenOffice or Mozilla). AFAIK, the gimp is an unfunded open source project to produce an end user app, albeit one for a small audience.
The only point Seltzer makes that holds any water isn't news: that open source favors projects that developers want. That doesn't prove that end-user apps get short shrift, since there is a rather strong argument that end-user apps are a much more difficult problem to solve than, say operating systems, programming languages, web servers or even relational databases.
Like most pundits, Seltzer seems to expect Open Source to compete with or replace proprietary software. What's odd to me is that he even uses the quote from CatB that should have informed him of his erroneous thinking.
Every good work of software starts by scratching a developer's personal itch.
ESR and RMS aside, it has been my experience with Open Source software over the past seven years that the motivations that drive the creation of free (any definition here) software are very different that what drives commercial software. Is it any wonder that developers will create tools to make their lives easier? Of course not. Do developers need office productivity tools? Not very much (by there *is* enough out there [see KDE, Gnome PIM/Palm tools] for SOHO/sole proprietors to get work done).
In the same way I was rebuked for criticizing Docbook's inability to create a nice PDF resume, I would now suggest to Seltzer: What did you expect free software to do? Topple Microsoft? The Open Source Software community is a parallel and separate universe to commercial software. My main desktop has been Linux for a while now. It is to a Linux I turn when I need to get real work done like pay bills (GnuCash), generate invoices (Gnumeric) or do my contract programming (done mostly with Emacs). No doubt this is confusing to most end-users but not all.
Let me put it this way. It is hard to create software. Harder to create working software. It's an order of magnitude more difficult to create software usable by techno-phobes (or the merely techno-uninterested). I know I require money to work on those kinds of projects. Why would anyone endure that kind of pain for free?
