Kurobox NAS-Central Forums

Welcome to the Kuro Community
It is currently Fri Jun 22, 2018 6:48 am

All times are UTC-06:00




Forum locked  This topic is locked, you cannot edit posts or make further replies.  [ 36 posts ]  Go to page 1 2 3 Next
Author Message
 Post subject: next version
PostPosted: Mon Dec 13, 2004 6:10 pm 
Offline

Joined: Mon Dec 13, 2004 5:32 pm
Posts: 9
i was wondering if there are plans for what the next version of the Kuro Box will be? new hardware and software for example?


Top
   
 Post subject:
PostPosted: Wed Dec 15, 2004 6:25 pm 
Offline

Joined: Tue Sep 28, 2004 8:44 am
Posts: 55
Location: Austin, TX
Well, I just started shipping a limited number of HGs. Hopefully by mid-Spring we'll have a processor at 400+.

I'm a big proponent of the additional LAN interface, RAM and expandable memory among other things. Hopefully, we can get a new board to offer these and some of the other great suggestions as well.

If you stop by to visit at CES, I"ll have a TeraStation to check out as well. I'm not optimistic that I'll be able to offer it as a Kuro style product though, but we'll see...


Top
   
 Post subject:
PostPosted: Wed Dec 15, 2004 6:55 pm 
Offline

Joined: Mon Dec 13, 2004 5:32 pm
Posts: 9
sounds good. I like the extra power :) why though would you not be able to offer it as a kuro style product? well i really hope it all works out.

-chris


Last edited by poundsmack on Wed Dec 15, 2004 7:18 pm, edited 1 time in total.

Top
   
 Post subject:
PostPosted: Wed Dec 15, 2004 7:08 pm 
Offline

Joined: Tue Sep 28, 2004 8:44 am
Posts: 55
Location: Austin, TX
I'm pushing, but we'll see. I definitely won't have a better feel for whether there's a chance until January. There's alot of different factors that will have an effect on the decision, so we'll just have to overcome each one individually. The biggest factor of all is the existence of more IP code than GPL. I think everyone understands that one...

Again, though, I'm going to give it a shot.


Top
   
 Post subject:
PostPosted: Wed Dec 15, 2004 7:15 pm 
Sven wrote:
If you stop by to visit at CES, I"ll have a TeraStation to check out as well. I'm not optimistic that I'll be able to offer it as a Kuro style product though, but we'll see...


WOW, at that price it will develop its own hacker community in no time.


Top
   
 Post subject:
PostPosted: Wed Dec 15, 2004 7:25 pm 
Offline

Joined: Sun Dec 12, 2004 2:06 pm
Posts: 60
Location: UK
If the community can become fairly self-sufficient in any particular way or can provide some value to Buffalo, that may help Svens' case. The thing to remember is, it doesn't have to be fixing bugs. For instance, Sun Microsystems sees real value in the OpenOffice community because they have translated the manuals into a huge number of languages and assisted in the localization of the product itself.

_________________
Michael


Top
   
 Post subject:
PostPosted: Thu Dec 16, 2004 4:46 am 
Sven wrote:
I'm a big proponent of the additional LAN interface


Additional *two* LAN interfaces, please - even if only as bare pads!

A second socket would enable a simple firewall or router. A thrid socket means that a proper firewall product could be created with a DMZ network for mail/web servers and the like, or two independant private networks.


Top
   
 Post subject:
PostPosted: Fri Jan 28, 2005 7:26 am 
The key thing for Kuro is _to keep it cheap_ : the mistake would be adding too many bells and whistles and upping the unit price.

I'd suggest: make the power supply an external brick, cheaper and easier to deal with future psu problems; make it easier to upgrade RAM (perhaps, SMT a socket, and allow the RAM to be unplugged and replaced); would be nice ATA speed were increased (not ATA66, but ATA100);

Did I mention: keep the price down? Don't add multiple more NIC's (let people buy a soekris, etc instead); Don't add more than one more USB port; if people want more than 2, then let them use a hub; Reduce power consumption, this should be possible (17W is a high future, I assume it's not nomimal but worst case); Did I mention: keep the price down ?

Would it be possible to update the casing so that they could be joined together for those with multiple boxes ?

If you organised the source/binary distributions to make it easier for hackers to start "off the shelf" with developing the product (notice all the comments on the forums about how to do this/that/etc just to get a basic development environment/etc going).


Top
   
PostPosted: Fri Jan 28, 2005 10:59 am 
I agree with the Guest post above... who cares about more CPU/memory/network adapters... The cool thing about the Kuro is that its small efficent and fairly cheap. If it could be made even cheaper I bet it'd sell faster. Let the whole micro-ATX / MacMini crowd work on trying to build a desktop replacement, keep the Kuro small and efficent.

The other main thing lacking right now is a safe and supported route to rebuilding the kernel. Not only can you not download a buildable kernel or .config off the website, but if you do manage to get one built flashing it might end up bricking your Kuro.

Revolution/Buffalo want us to hack this thing to pieces and make it do everything, but how easy is it to that without a rebuildable kernel? Until we get there the Kuro will be stuck in the mud...


Top
   
 Post subject:
PostPosted: Fri Jan 28, 2005 7:25 pm 
Offline

Joined: Thu Nov 11, 2004 5:57 pm
Posts: 1
I really like the idea of the expandable case.. Add a thin "shell" to add a second hard drive to the Kuro.. Is it possible, or would that over juice the power supply? Imagine a pair of 400GB Seagates in a KuroHG.. That would make the Kuro a life saver for me..


Top
   
 Post subject:
PostPosted: Sat Jan 29, 2005 10:51 pm 
I want it to stay cheap too but you guys don't know what you are talking about. :roll:
I want more USB ports and I want them to work.
The Kuro and Linkstation already have 5 USB2.0 ports in them.
Buffalo only put connectors on one or two. Add a few more connectors or places to put them. Cost about $0.50. If they brought out more connectors no hub would be needed. That's cheaper than a hub!
The internal PCI to ATA controller can be replaced by an PCI to SATA controller for about the same price. The connector is cheaper for SATA. Maybe bring an SATA connector for a second drive out the back. Leave the power supply alone. You would have to make it bigger and cost a lot more to connect to more drives, so don't.
Now for nothing heres what I get:
Let me plug an external SATA drive and enable RAID. No Cost :wink: (Firmware available)
Let me plug in multiple USB2.0 drives externally and enable RAID No Cost :wink: (Firmware available)
Let me connect up two USB2.0 printers. No Cost :wink: (firmware already there)
Let me plug in a Wireless card into the USB for wireless router. No cost :wink: (Firmware available) Just let it work! :shock:
Let me plug in a second 10/100 Ethernet USB card. No cost :wink: (Firmware available) Just let it work!
Put the internal memory into a socket so I can upgrade it later.
We can pretty much build anything we want with what you got plus firmware. If you go to the KURO WIKI you can see how to add the other USB ports to your current KURO for a few bucks if you can solder. :roll: Then you can add the drivers to build what you need already.


Top
   
PostPosted: Wed Feb 02, 2005 12:04 am 
Offline

Joined: Thu Dec 30, 2004 2:50 pm
Posts: 44
Location: seattle
HG User wrote:
I agree with the Guest post above... who cares about more CPU/memory/network adapters...


Actually, I (and more importantly Sven too it seems) care a lot about more CPU/memory and almost consider it an imperative to widespread dev appeal. For those of us that don't have a Mac (clear majority of dev community) it's pretty brutal having to deal with recompiling the kernel using the 266mhz (i have hg) stock cpu and 128 meg memory. Maybe I'm spoiled with modern day arch since it was only five years ago that overnight compiles were the modus operandi, but these days, just slogging through the coma-like sludge of emerging simple utilities is enough to raise my blood pressure a lot. :oops: Not to mention the thought of recompiling a 2.6 kernel a bunch of times as a thorougly unappealing and miserable experience imho. Cross compiling might be an answer if i had extra cycles and resources to spare, but it would seem to make much more sense to improve onboard cpu and ram, or at least provide the leads on the board that makes it an option.

Quote:
Let the whole micro-ATX / MacMini crowd work on trying to build a desktop replacement, keep the Kuro small and efficent.
[\quote]

It sounds like the product you're looking for is the linksys NSLU2, since it provides all the features and affordability that you seek at an attractive $70 price. The kuro is priced in an entirely different market that's looking for far better baseline performance. [/kernel]

Quote:
The other main thing lacking right now is a safe and supported route to rebuilding the kernel. Not only can you not download a buildable kernel or .config off the website, but if you do manage to get one built flashing it might end up bricking your Kuro.


I've been shocked that the kuro isn't better supported in these areas. With 3rd party firmware dev becoming such a force with mad potential, a well documented method for flash upgrade should be mandatory.


Quote:
Revolution/Buffalo want us to hack this thing to pieces and make it do everything, but how easy is it to that without a rebuildable kernel? Until we get there the Kuro will be stuck in the mud...


Gentoo total conversion is the solution for most here, but again, it also underscores my first point on caring about increased cpu/memory.

And even just one additional NIC would provide a plethora of additional utility to the kurobox. That was my first instinct when discovering this box.


Top
   
 Post subject: Crosscompile
PostPosted: Thu Feb 10, 2005 1:23 pm 
Offline

Joined: Thu Dec 23, 2004 8:33 am
Posts: 1
ydef, you should setup a ppc toolchain on your desktop and crosscompile for the powerpc/kurobox (i see you mention it as an option - the only hard part is setting it up initially. as for spare cycles it could always compile while you sleep)

Joe


Top
   
 Post subject:
PostPosted: Thu Feb 17, 2005 7:41 pm 
I just checked, you're right: there are a 5 USB ports: yes, I think a good idea to just lead these out to the back panel.

CPU and RAM are obvious targets; but kernel compile times are a red herring: the primary aim of the kuro is not as a development machine, so additional CPU is simply not necessary for the 95% of the time that the box is just serving up files, etc. The only reall CPU/RAM burner could be transcoding audio/video say when serving up media. I've loaded my kuro up with http, ftp, svn, mt-daapd, nfs, etc and it's still only taking up 24Mb; and on a heavy loaded NFS write, samba is 40/50% CPU loaded.

No one is going to buy a kuro as the primary machine: they;re going to buy it in additional to something else they have that has higher performance. Most hackers are going have another linux or bsd box anyway. If anything, making the kernel/system distribution cross-compilable would help.

In addition, for all the talk about better and faster hard drive speeds, remember that the bottleneck is the 100Mb interface anyway; so in some respects, you could say that the Kuro HG is actually overengineered for its job: the additional transfer speed performance is localised within the box and not making any difference for external clients. To really do something about performance, you'd want to start thinking about gigabit ethernet.

So upgrading RAM/CPU, okay, nice, but in reality, for a NAS box, isn't going to do a great deal. It will do something if the NAS box is being used as a more substantial multimedia server.

Upgrading the IDE interface: again, same argument: NFS reads are still bottlenecked by the 100Mbps interface, not by the ATA interface.

However, these may help with performance between USB and ATA, say to increase the speed of backups. But is that really an issue? Backups are usually semi-offline activity and aren't horrendeously performance critical.

Upgrading from 100Mb to Gig, that would really open up the speed, but not a lot of people have Gig ethernet yet (though, my Dell GX270 does).

I've made a lot of engineering arguments, but I guess a lot of people buy these kind of stuff and don't think like an engineer, so doing some of those upgrades could be a good marketing/sales tactic.

I agree with the other poster about making the kernel/system easier to hack/work with: safer and easier to rebuild and have the kernel load.

I think supporting an additional IDE drive is really the big issue, because it would allow either (a) bigger NAS box, e.g. 2x400GB drives, (b) another way for backups from one IDE to the other (rather than to an external USB drive), (c) RAID.


Top
   
 Post subject:
PostPosted: Fri Feb 18, 2005 4:44 am 
Offline

Joined: Thu Dec 30, 2004 2:50 pm
Posts: 44
Location: seattle
Anonymous wrote:
In addition, for all the talk about better and faster hard drive speeds, remember that the bottleneck is the 100Mb interface anyway; so in some respects, you could say that the Kuro HG is actually overengineered for its job: the additional transfer speed performance is localised within the box and not making any difference for external clients. To really do something about performance, you'd want to start thinking about gigabit ethernet.


Actually, the HG comes stock with a gigabit ethernet interface, at least mine did.


Top
   
Display posts from previous:  Sort by  
Forum locked  This topic is locked, you cannot edit posts or make further replies.  [ 36 posts ]  Go to page 1 2 3 Next

All times are UTC-06:00


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Powered by phpBB® Forum Software © phpBB Limited