Slashdot Mirror
Spolsky Stands Firm on Linux on the Desktop
erlando writes: "SoftwareMarketSolution is running an interview with Joel Spolsky (from JoelOnSoftware) in which he responds to this earlier thread here on Slashdot. In short: He defends his position and makes some interesting remarks on Linux and the desktop."
You have what I would call a superiority complex. I would say that is completely normal for a technically inclined or computer competant person. One thing you dont have, which is quite evident in your last statement, is a simple respect for your peers.
Now, I may not directly consider you a peer, for the simple fact that we differ on a number of simple comparitive issues. You are stuck in the past. Since you will most definitely ask for some type of proof of this, allow me to explain.
In your responce to the question regarding the recode of a fortran program, you responded:
Do you hold on to an old 286, even though it could easily crash purely because of its old-hardware, just because it works? Do you hold on to a 1982 Toyota Corolla, just because it works? No. For the safety of you data, and for the safty of your family, you go and get new stuff.
Of course, your responce will be well, programs dont degrade with time. I say they do. Allow me to elaborate once again.
I've worked on a number of programming projects. One of which works perfectly, but we are still planning a complete rewrite of. For a matter of context, I'll involve the name. The Bahamut IRCd happens to be presently supporting over 130,000 simultanious connections between about 30 servers. It will probably be quite good until well over 200,000 connections. But we're rewriting it anyway. Because we've learned one thing the hard way. Dont wait for problems to come to you. Just because your software works fine with 10,000 users, doesnt mean its going to scale to 11,000 users.
Programming in *any* arena is not cut and dry. Programs are *never* "perfect", as you as a software developer should know. If you've worked on any project of any reasonable size, then you realize that there is a point where picking through old code to improve software isnt worth your time, and you can simply rewrite the whole damn thing to do what you want and accually have spent less time doing it.
.
"Sounds like he could save quite a bit of time and maintanence nightmare if he just invested the effort to write it well the first time."
You people seem to be going out of your way to miss Joel's point on this.
Routine code maintenance (ie, refactoring code, adding small improvements, etc.) is a standard part of the development cycle that every good software team follows. Doing so does not mean that the product wasn't written well the first time. (You Linux people should know that better than any of us, but I digress...)
Code maintenance is not the same thing as a complete code rewrite, which is what Joel is arguing against. Incidentally, lack of proper code maintenance over time will likely result in the unwieldy mess that Stephen Eick describes.
So does code rust? Of course not. Does poorly-developed, unmaintained code rust? Maybe, but that still doesn't mean that everything should be rewritten from scratch.
DirectX
IE5
Windows Media Player
Windows Installer
Active Directory
All of these modify, or extend the Win32 API under Win95 when installed.
Right, because executing untrusted scripts is the core functionality of Outlook. And here I was thinking it was composing emails, dealing with contacts, and scheduling appointments. Joel is right, and you're not understanding his point.
Outlook has a problem with attachments to email, which is probably actually less than 1% of the codebase of the product. Therefore he is right, and saying that rewriting everything in the program is the only way to fix it is wrong. Sure, let's redo the spellchecker, the datetime picker for the calendar, the user interface components - ONLY by rewriting all of those things will Outlook be secure!
It only takes one bad line of code to crash everything.
Yes, thanks. So should the solution be to rewrite that one line or the whole app? Hmmmmmm
Please subscribe to see the more insightful version of th
No one uses the out-of-the-box version of Win95 anymore, do they?
Two things about Win95:
1. Microsoft no longer supports Windows 95.
2. USB is not natively supported in the early versions of Win95 (and later versions are spotty, they get it right by Win98).
So on that level, Joel's arguments are flawed. Less and less people will be testing on Win95 because MS doesn't support it any more. The WIN baseline now starts at Win98.
Another thing: if you DO use Win95, do you test the off-the-CD version or the highly patched (practically Windows 98) version? Grey grey grey area, Joel.
----- rL