Open Source: Where's the proof?

1999-06-03

The Open Source advocates always like to claim that open source produces better software.  My question is where's the proof?  It seems as though everybody takes it for granted that it's true without actually checking into it.  The arguements are compelling.  More eyes on the code should mean that more features get added and more bugs get fixed and so we should get better software but I think if we actually look that's not the case.

The Open Source people usually point Emacs, GCC, Linux, Perl and Apache. That's about as far as they get. Maybe they add sendmail or fetchmail in there. A few of the new desktops might make it too like KDE or GNOME.

Lets take GCC. It seems like the commercial products of both VC++ by Microsoft and the high−end development tools for the SGI are both far superior to GCC. I can't say I know all of what's available for GCC but for example all the browsing stuff in VC++ and the ability to change code while debugging and execute the changes without recompiling. The SGI ones have a awesome debugger too.  Of course if GCC is better just show me the proof.  Maybe it generates better code.  But if it does is it so much better as to outway the benefits the other compilers might offer interms of getting your job done quicker.  For example is it easier to write an end user application with GUI using GCC or VC++?  I suppose it depends on your definition of a UI.  Lets say you want to duplicate WordPad from Windows (not: for the GCC case I'm assuming you are under Linux with some X desktop.  I think VC++ is going to win pretty easily.

How about Emacs?   That's probably a religious arguement. Some people like their editors. Some people even prefer "vi" and "edlin/ed" but I think it can be argued that for example in several areas Visual SlickEdit are currently better than emacs.  If Open Source is so great and Emacs is one it's greatest achivements why can just a few programmers (probably less than 10) create a better product so quickly while Open Source Emacs with its thousands of eyes lags behind?

Sendmail doesn't in anyway compete with MS Exchange in terms of features (and in my opinion very very useful features) I think a simple demo would convince nearly anybody of that. With Sendmail you get basic e−mail, administrator only managed mailing lists and alias.  With MS Exchange you get e−mail, non−administrator managed mailing lists and alias AND you get Web access to your e−mail, IMAP support, a newsgroup server which can run as a full NNTP partner or just a polling type (ie, with no more access than a NNTP client) You get shared public folders that can track updates it individual entries.   The list goes on and on.

Apache I don't know at all so I can't comment except to say again it appears that "out of the box" it is much easier to setup the Microsoft product and out of the box many many more features work without major configuration.

Perl? Well it's just a language of which there are hundreds.  The one getting the most press is Java which is the product of a commercial endevor.  Not Open Source.   Note: I'm not going to try to debate whether Perl is better or worst than other languages.

The leaves Linux and the desktops available for it.  As for the desktops it's pretty arguable none of them have caught up with Windows or the Mac yet but of all the programs above they do seem to be making most progress.  It will be interestng to see if they can get enough applications to support those desktops to make them approchable by the average user but clearly Windows is still better at that particular feature.

Linux is also sorely behind NT in terms of features.

Of course I'm open to the possibility that I just don't have enough experience with all of these products but my impression is that it's really the Open Source people missing the experience with the non Open Source world and just accepting on faith that it's better.

Comments
Race
dating_before 1999-10-22