4 ms·
I get that they have to make money somehow, but I can't help but feel that they should be focusing on polish, bug fixes, and regression management instead of pi
by acabal 14y ago
I get that they have to make money somehow, but I can't help but feel that they should be focusing on polish, bug fixes, and regression management instead of piling on more features like this. I'd be a lot happier if they were adding these features after fixing things like broken Compiz quirks or Unity commandeering the super key and breaking your personalized keyboard shortcuts.
Hell, I'd even pay cash money for Ubuntu if they could guarantee regression-free updates and a truly solid desktop (i.e., no "oh yeah it works just fine, except for this long list of annoying problems"). But their priorities seem to be elsewhere. Blame can certainly be assigned outside Canonical but the buck has to stop somewhere.
Personally I think this cheapens the desktop quite a bit. I'm already advertised at for 90% of my day, now I get to be advertised at in my one sanctuary, my desktop.
- mneumegen 14y agoAgreed that it cheapens it. Surely there's a lot of money to be made in server service/support now that Ubuntu is becoming a major player in the server market.
- jamesjporter 14y agoI would be willing to shell out top dollar for a UNIX OS that's as stable as OSX and "just works" the way OSX does but is also as customizable and hacker-friendly as Linux, even if it's closed source. I've been wondering if there's any room for a company in this space—its obviously a very niche market, so you'd have to charge a lot, but I (and maybe other nerds?) would certainly be willing to pay.
- habosa 14y agoThere is definitely room for such a company, and the market is not as niche as you'd imagine. If you could provide an environment as user friendly as OS X with top-notch security and all of the customization features of Linux, most software development companies would pursue this option for their teams. I know I'd use it every day, and I'd happily pay for a non-commercial license in the $50 range.
- devcpp 14y agoIt's niche enough to not justify the hundreds of millions of dollars such a project would require and the high risks. OSX "just works" because it has a full suite of software that comes loaded on it and "just works" very well and makes a lot of Windows programs jealous (except for some Office products). If people want customization and general freedom, they already have Linux for free. I think it is safe to say that the market for people unhappy enough with both OSX and Linux, and willing to pay for that, enough to justify years and years of expensive and risky development, is _relatively_ small.
- habosa 14y agoI don't think it would require hundreds of millions if it combined pieces from existing Linux distros. This could definitely get some venture funding.
- m0wfo 14y agoMaybe there would have been 5 years ago prior to the tablet+mobile (iOS/Android) revolution. I agree though; personally I love Debian + Solaris but neither them nor any other *NIX variant has the slick feel of OS X. Not to mention a nice development environment like ObjC+Cocoa.
- gizmo686 14y agoI do not think you can have a "hacker friendly" OS that is closed source. I suspect that if a hacker friendly OS that "just works" does come out, it will be from a computer company trying to offer a non Windows option (either for price and/or because people start not liking Windows). Given the volume of computers such large companies make, the unit cost associated with a polished OS will be very low. Although this OS would be designed for the average user, the company would likely choice to use open source programs for cost.
- acabal 14y agoDell might be edging in to that space with Project Sputnik. It's too bad that the specs on their machine are underwhelming for their target market. If it fails because of that, Dell will be hesitant to enter the market again--but for precisely the wrong reasons. I'd pay $2k+ for a 14 inch laptop, 1600x1200 IPS screen, SSD, i7, dedicated graphics, and a good keyboard (with home, end, pgup, pgdown) that could run Linux. Hell, I'd be happy with the dedicated graphics just working in Windows, and I'd dual-boot to game. But no such machine exists right now.
- agildehaus 14y agoHaiku is the closest existing OS to that I can think of (one "true" distribution that "just works", but nothing stopping you from making your own). It's not Linux though.
- drcube 14y agohttp://www.slackware.com/ http://www.slackware.com/ Slackware is simple, easy to use, and "it just works". It's free software so of course it's hackable. It's not as GUI-y as an OSX fan might like, but otherwise, it fits your requirements. It even has the "top dollar" price tag ($30-$50 depending on if you want the 6 CD set or a single DVD, and whether you sign up for the subscription or not), although you can download it for free if you want.
- ggordan 14y agoI completely agree. I know this whole story is tired, and it has been said before but I'd really like to be able to update to a x.04 version without fear that it will break everything.
- Aardwolf 14y agoThere are other distros than Ubuntu you know... There's a whole world of modern, active, distros open to explore, that don't include shopping options in them!
- subsystem 14y agoI personally think they are missing a great opportunity to reinvent the workstation. If we imagine there are a number of different user patterns in computing: entertainment users (consuming), business users (data processing), power users (creation and research), application developers, system/server developers and infrastructure (servers). (A bit simplified of course). To me it seems like they are targeting entertainment users, when they really should be targeting power users and application developers. The first two (entertainment & business users) already being dominated by the competition and the last two (system/server development and infrastructure) already having a good track record. The nice thing about researchers and creators are that they mostly want the same things like low-latency processing of multimedia, stability, flexibility, power and value. Features that will trickle down to entertainment users that do care about it, but are unable to evaluate it exclusively. Paradoxically enough this used to be Apples territory before they switched to Unix. With the exception of value and therefor also power, leaving the an opening for hobbyists (creators) and large scale deployments (research) to Microsoft and Linux respectively. To do this you'll need application developers to bridge the gap between system & server developers and power users, which there seem to be a surprisingly small overlap between, at least seen to activity. These are unfortunately the people you piss off when you value sponsored searches and flashy menus over consistency.