In my book, Sun's officially dead now | Hacker News Reader

In my book, Sun's officially dead now

5 pointslallysingh8 comments

They've been dying for quite some time, but now they've EOL'd all their workstations, save 1 single socket Intel box. Somehow I suspect it's only there b/c it was the only one released in 2008.

Note that this is without ever coming through on their promises to let their AMD boxes handle quad-core CPUs (requiring that they just release a firmware upgrade).

How's anyone supposed to develop for their stuff? How can anyone in the scientific/research community even try now? Why release SPARCs when you don't provide a way to dev for them? Why even maintain Solaris?

It's like they just gave up. Well, I guess a body will atrophy to death when it doesn't have a brain to keep it going.

View on HN

Just out of curiosity...when was the last time you bought a Sun workstation? If you did, was it SPARC or AMD-based?

Heck, when was the last time you bought any desktop system instead of a laptop for development or science work? Sun's hardware is very cool, but desktop machines that neither dirt-cheap nor monster gaming systems are a hard sell these days.

On the other hand, if you want to develop software that targets the SPARC ISA -- and there are lots of reasons to do so -- you can easily pick up an entry-level config of a 1U SPARC server, and rack it up wherever you like.

AMD, the only sparcs available were obscenely slow IIIis.

Without workstations, how's anyone going to dev for solaris? It's not known for its hardware support of non-sun hardware.

I've been in an office with a rackmount in it. The rack is huge and it's loud. They used to just shove the guts of their servers into a desktop box and call them workstations. But I guess that's too much for them now.

Until they return to profitability, it makes some sense to streamline their product line. Although they've narrowed it down to a single workstation, it still comes in different configurations and options.

http://www.sun.com/desktop/workstation/ultra24/index.xml

If they're going to sell x86 workstations, at least sell some more powerful than the macintosh I can buy down the street.

They've been shoving COTS motherboards into x86 boxes for a few years now, how about keeping up with the idea of some computational ooomph?

Seriously, what's a reasonable computer scientist supposed to use these days? I'd rather not spend a day a week fixing/maintaining a linux box for its own problems, or for tracking down the new compatible binaries, or begging for help on undocumented software on mailing lists.

Just use Ubuntu, things pretty much work. For any reasonably sized project there is someone else who has already made it work.

What does that mean for Java if you are right?

It dies. Thank god.

Face it: Java played its part in the Theatre. It dragged people onto GC and provided a simpler C++. It was the logical halfway-path to Lisp. Now, it looks aged, bloated and bad. Note how few fruitful/succesful Java projects there are out there compared to other languages...

I'm not sure how you're defining "fruitful/successful projects", but the impression I get is that there are mountains of Java business logic in the wild... It's practically the new Cobol, and it will be just as hard to kill.

Reply on news.ycombinator.com