Siig / initio inic1620 sata controller driver?
I've purchased a pair of SATA (SATA 3Gb/s, actually) hard drives, with plans to build a SUSE 10 server around them. Last week I bought a Siig SATA II-150 PCI controller card to use with them. I clearly recall verifying Linux-compatibility before buying the card, but that now seems to be a complete hallucination. As of last night, I find no reference anywhere online that indicates I can use this controller card with Linux. I can find no linux driver for the card. I've emailed Siig, as well as Initio (the chipset manufacturer), and neither has plans for a linux driver. At this point I'm assuming that's $40 down the drain, but thought I'd see if anybody here could offer any possibility for this card I haven't considered. A) Maybe I've missed an online resource somewhere, and there is in fact a driver that will work with this card? While Suse is my first choice, I'm not opposed to using a different distro if necessary. B) I have absolutely no idea how easy or difficult it is to write a device driver, especially if the hardware/API isn't publically documented. Is there a particular group of coders that take up 'wish-list' requests like this? (I assume not.) If I'm somewhat bright, and good at figuring out logic puzzles, but have practically no programming experience, is it ridiculous to consider the possibility of writing a device driver myself? I assume so; I have no idea where to even begin with such a thing. Thanks, Steve
Steve Jacobs wrote:
I've purchased a pair of SATA (SATA 3Gb/s, actually) hard drives, with plans to build a SUSE 10 server around them.
Last week I bought a Siig SATA II-150 PCI controller card to use with them. I clearly recall verifying Linux-compatibility before buying the card, but that now seems to be a complete hallucination. As of last night, I find no reference anywhere online that indicates I can use this controller card with Linux.
I can find no linux driver for the card. I've emailed Siig, as well as Initio (the chipset manufacturer), and neither has plans for a linux driver.
At this point I'm assuming that's $40 down the drain, but thought I'd see if anybody here could offer any possibility for this card I haven't considered.
A) Maybe I've missed an online resource somewhere, and there is in fact a driver that will work with this card? While Suse is my first choice, I'm not opposed to using a different distro if necessary.
B) I have absolutely no idea how easy or difficult it is to write a device driver, especially if the hardware/API isn't publically documented. Is there a particular group of coders that take up 'wish-list' requests like this? (I assume not.)
If I'm somewhat bright, and good at figuring out logic puzzles, but have practically no programming experience, is it ridiculous to consider the possibility of writing a device driver myself? I assume so; I have no idea where to even begin with such a thing.
Wow, that's some serious illusion! If you are willing to throw probably hundreds of hours of work into such a project in order to rescue $40 of bad investment without any realistic prospect of a return I advise you not to work for anyone that might ask you how effective you organize your resources!! In order to write a driver you need a lot of knowledge which you yourself profess not to have. My advice is to get your priorities straight. $40 might annoy you but that is a cheap lesson compared to many frustrating hours working on a project that will probably never see the light of success. Much more effective just to ask the listmembers what kind of controller they are using with what capabilities and what problems they encountered. I advise you to first write down the list of requirements the server has to fulfill. If you don't know all the requirements you might try to find out by asking the list what is necessary for a server with x users and y services, backup, reliability etc. Sandy -- List replies only please! Please address PMs to: news-reply2 (@) japantest (.) homelinux (.) com
On 4/4/06, Sandy Drobic <suse-linux-e@japantest.homelinux.com> wrote:
Steve Jacobs wrote:
B) I have absolutely no idea how easy or difficult it is to write a device driver,
If I'm somewhat bright, and good at figuring out logic puzzles, but have practically no programming experience, is it ridiculous to consider the possibility of writing a device driver myself? I assume so; I have no idea where to even begin with such a thing.
Wow, that's some serious illusion! If you are willing to throw probably hundreds of hours of work into such a project in order to rescue $40 of bad investment without any realistic prospect of a return I advise you not to work for anyone that might ask you how effective you organize your resources!! In order to write a driver you need a lot of knowledge which you yourself profess not to have.
My advice is to get your priorities straight. $40 might annoy you but that is a cheap lesson compared to many frustrating hours working on a project that will probably never see the light of success.
Sandy
Maybe I'm taking your response the wrong way, but it seems to me you're almost attacking me, and accusing me of having my priorities way out of whack. i originally stated that I have no idea how easy or difficult writing a driver is - therefore, for all I know/knew, it's pretty simple, and I may have been mistaken by *assuming* it was beyond me. Now, having said that, I also stated that I was pretty sure it was a ridiculous notion; I was just trying to confirm that. Seems to me I've got things pretty straight - I just wanted to verify what I believed to be the case before acting on it. And if the job was realistically 'doable', even if it wasn't trivial, I'd be happy to challenge myself and learn something new in the process. Obviously there are folks in the community that *DO* write drivers for unsupported hardware, so it's clearly possible. Most likely I myself haven't got the skills, admittedly.
Steve Jacobs wrote:
On 4/4/06, Sandy Drobic <suse-linux-e@japantest.homelinux.com> wrote:
Steve Jacobs wrote:
B) I have absolutely no idea how easy or difficult it is to write a device driver,
If I'm somewhat bright, and good at figuring out logic puzzles, but have practically no programming experience, is it ridiculous to consider the possibility of writing a device driver myself? I assume so; I have no idea where to even begin with such a thing.
Wow, that's some serious illusion! If you are willing to throw probably hundreds of hours of work into such a project in order to rescue $40 of bad investment without any realistic prospect of a return I advise you not to work for anyone that might ask you how effective you organize your resources!! In order to write a driver you need a lot of knowledge which you yourself profess not to have.
My advice is to get your priorities straight. $40 might annoy you but that is a cheap lesson compared to many frustrating hours working on a project that will probably never see the light of success.
Sandy
Maybe I'm taking your response the wrong way, but it seems to me you're almost attacking me, and accusing me of having my priorities way out of whack.
No reason for me to attack you. Though I definitely think that your priorities are out of whack. (^-^) The answer is pretty simple when you compare how expensive one hour of your paid time is for your employer to the price of another controller. When our telephone system went berserk and had to be restored, the technician I called asked if I really wanted him to backup and restore the mailboxes of all the employees. He would need about one and a half hour to do that which would cost me about $100 for each hour, and the employees could do that themselves. I had no trouble with that. If the employees all had to do it again, they would each spend about 10-15 minutes, and a dozen would ask me to help them, so I would end up using my time AND the employees time to restore the mailbox. Even taking into account that the time of an employee doesn't cost half as much as the technicians time, in the end our company would have spend a LOT more money on the restoration if all the employees had to restore their own mailbox.
i originally stated that I have no idea how easy or difficult writing a driver is - therefore, for all I know/knew, it's pretty simple, and I may have been mistaken by *assuming* it was beyond me.
Now, having said that, I also stated that I was pretty sure it was a ridiculous notion; I was just trying to confirm that.
Seems to me I've got things pretty straight - I just wanted to verify what I believed to be the case before acting on it.
And if the job was realistically 'doable', even if it wasn't trivial, I'd be happy to challenge myself and learn something new in the process. Obviously there are folks in the community that *DO* write drivers for unsupported hardware, so it's clearly possible. Most likely I myself haven't got the skills, admittedly.
Writing a driver is slow and frustrating work even if you do have solid programming knowledge ond additionally expert knowledge of the protocols involved. Writing a driver without having all the specs of the hardware is extremely difficult. I really do admire your willingness to contribute to the community but I also think that this is a task that's not worth the effort. Sandy -- List replies only please! Please address PMs to: news-reply2 (@) japantest (.) homelinux (.) com
On Tue, 2006-04-04 at 22:29 +0200, Sandy Drobic wrote:
i originally stated that I have no idea how easy or difficult writing a driver is - therefore, for all I know/knew, it's pretty simple, and I may have been mistaken by *assuming* it was beyond me.
Now, having said that, I also stated that I was pretty sure it was a ridiculous notion; I was just trying to confirm that.
Seems to me I've got things pretty straight - I just wanted to verify what I believed to be the case before acting on it.
And if the job was realistically 'doable', even if it wasn't trivial, I'd be happy to challenge myself and learn something new in the process. Obviously there are folks in the community that *DO* write drivers for unsupported hardware, so it's clearly possible. Most likely I myself haven't got the skills, admittedly.
Writing a driver is slow and frustrating work even if you do have solid programming knowledge ond additionally expert knowledge of the protocols involved. Writing a driver without having all the specs of the hardware is extremely difficult.
Also note that if the driver was that easy the hardware company probably would have one. Writing a driver would probably take you minimally 20 man hours. At even $15/hour that is as much as you could buy a nice 3ware or some other supported SATA controller and even have hardware raid. Just my 2 cents. Brad Dameron SeaTab Software www.seatab.com
participants (3)
-
Brad Dameron -
Sandy Drobic -
Steve Jacobs