Hacker News new | ask | show | jobs
by agentgt 3860 days ago
This is not the first time (Mozilla aka Netscape) and it reminds me back when I first started my programming career and would read good ole Joel:

http://www.joelonsoftware.com/articles/fog0000000069.html

Joel was right and wrong.. Firefox was a huge success (or at least in my mind it was) but some might say the Netscape company was trashed during the process (I don't know if I agree).

Starting over isn't always a bad thing but I agree that is also highly overrated.

1 comments

It was a failure as he said because the goal was to rewrite it for the company to do better. They failed at that. More work was done including by Mozilla. That worked but was too bloated and all-in-one. Eventually, someone trimmed it up to make Firefox and added the customization features. That succeeded.

So, there was a failure, years of struggling changes, several new audiences, and another big change before it made it. Not seeing it as a counterexample as much as good luck for a project that seemed doomed to failure.

I don't disagree it was a bad idea to do the rewrite but rather the company (Netscape) was on a downward spiral regardless (it didn't help but I don't think it was the sole reason).
Definitely wasn't the sole reason, their server software generally sucked, e.g. a lot of people happily switched to Apache once it was perceived as being sufficiently trustworthy. I could see the latter happening in the 1995-6 period.

One thing I've read about the rewrite is that Netscape accuihired? (back before that was a word) a failed company, and put its failed managers in charge, who I guess were good at what really counts in the short term (looking out for themselves). An obvious corollary to the "don't do a rewrite" is "if you're going to do one, employ really good people to do it".

Its funny (and ironic to me given the topic) you mention Netscape and Apache. One of my first jobs out of college was rewriting an old LiveWire application (yes ... the original server side Javascript) to a JSP Tomcat application.
I forgot LiveWire existed. One of few that finally faded from memory. Of old & commercial ones, AOL server + a TCL web framework are still getting updated. Opera is getting redone. The product with the coolest name is still scraping by per .cfm pages I see. Some ancient tech still around but mostly going bye bye.

Only thing left is maybe to redo Mosaic in Ruby or Java to see if it's technically feasible to slow its rendering down any further.