Hello. I recently acquired a NetWinder...

Daniel Gimpelevich daniel at gimpelevich.san-francisco.ca.us
Tue Mar 7 11:05:27 EST 2006


On Mar 7, 2006, at 3:19 AM, Ralph Siemsen wrote:

> Daniel Gimpelevich wrote:
>
>> Regarding the website, perhaps, but you would know more about who 
>> ended  up with the licenses and copyrights, which is what I 
>> originally asked.
>
> NetWinder Inc. is (was?) a subsidiary of Thor Solutions based in Nova 
> Scotia, and as far as I know, they are the owners of the technology. 
> Admittedly I have not heard from anyone there in at least a year, 
> perhaps longer.  The current WHOIS for netwinder.net does not match up 
> at all, I have no idea who is running that blog-style site, or if they 
> are actually related at all.
>
>> Evidently, any such discussion took place during an era of the 
>> mailing  list archives that I have not yet found. This synopsis seems 
>> slightly  paradoxical, since a lack of interest in that property is 
>> hardly  consistent with actively turning down requests for permission 
>> to use it  rather than taking advantage of opportunities to get rid 
>> of it.
>
> We got so far as to have an email saying it was okay.  I decided to 
> cover my butt and asked for it in writing signed by an authoritative 
> person, but around the same time, netwinder inc closed their office 
> and moved all the equipment (presumably to Nova Scotia), and became 
> unresponsive at that time.  I tried for a year to "complete" the deal 
> but eventually gave up.

They seem to still be going through some flux. In the Google-cache copy 
of their Management page, retrieved a week ago Sunday, there is this 
text under V.P. of Technology:
> Winston is President, CEO of NetWinder Inc. NetWinder Inc. is a 
> privately held corporation focused on leveraging "innovation at the 
> edge" of the Internet. NetWinder Inc., focus is to develop leading 
> embedded application platforms, distributed application 
> infrastructures, and maintains a market leading service and support 
> team.
In the current version of the page, the V.P. of Technology has moved to 
President while the President has moved to V.P. of Sales. There is no 
longer a mention of the "V.P. of Technology" title, and the text I 
quoted has been replaced with a blurb about "National Privacy 
Services." The PDF version of the page contains the old text to this 
day.

>> The list archives in the old FTP site stop in early 2000, but the  
>> current pipermail begins in late 2004. I have not yet seen the  
>> pipermail archives for the intervening time period.
>
> I'll try to complete the collection from my own archive.

Thanks.

>> So the board was fully populated at one point? Any idea how that  
>> interfaced with software? Also, I don't see a Winbond video codec 
>> chip  anywhere in these pictures. Where would it have been? And what 
>> kind of  signals would be accessible through that little card-edge 
>> connector  under the fan?
>
> Early cards were populated, but there was no software support, due to 
> licensing costs for the 56k modem technology.  As far as I know the 
> modem never worked.

I guess I misstated my question about the interface. I have not yet 
found any evidence that any of the pins connecting the daughterboard to 
the mainboard carried signals that could communicate with a modem, 
other than the ones connected to ttyS0. If they didn't, it would 
explain why it never worked but lead to questions about how that could 
possibly have come to be. If they did, any info identifying them might 
be useful.

> See below for the video codec story.
>
> The card-edge connector under the fan is a JTAG port.
>
>> Ouch! You even deleted the changelogs! I guess now I'd better hope 
>> the  firmware in the machine never gets corrupted...
>
> Ah but now OSU matches all the other mirrors :)
>
>> Seems slightly wasteful if libfloat can't use hard-float support even 
>>  in the few instances when it's there...
>
> I would think that writing a libfloat with this behaviour would be 
> possible, but its not the same as having the opcodes of floating point 
> instructions embedded in your executable.  And that is really the 
> issue, you really can't have it both ways.

It's not the same, but it would still be an improvement over soft-only 
float. I remember ancient benchmarks for m68k comparing them.

> For netwinder, all the software is readily re-compiled from source, so 
> I don't see any real issue with going to soft-float, and many other 
> ARM architectures have long ago reached the same conclusion.

Going to soft-float is a positive step, as long as NWFPE and FastFPE 
remain loadable modules for freak cases of binary collaging. But I 
would compare the former need for such modules to a current need for 
libfloat to behave as I mentioned. If you have a rule with a common 
special-case exception that you find more common than instances 
conforming to the general rule, and you therefore declare the exception 
to be the new rule, the instances conforming to the former general rule 
then become the new special-case exception.

>> http://lkml.org/lkml/2006/1/20/427
>> It's not all that discouraging, yet...
>
> At least there seems to be interest... that's the biggest challenge!

Perhaps it would be beneficial to let them know authoritatively that 
we're still here?

>> That's the most disappointing thing I've heard so far. I think  
>> framebuffer reliance is a rather steep price for a software upgrade.
>
> We're been this way since early 2.4.  If you want VGA that badly, I 
> will dig through old mail to see what PatB told us had to be done 
> (actually you seem pretty good at googling, so you can probably chase 
> down the last discussion on this)...

Couldn't find that. I expect it would be in the missing period of 
archives.

>> I would think its presence would be coupled with the presence of the  
>> Philips chip. Where on the board was this Winbond chip supposed to 
>> be?
>
> Just below the cyberpro (where the philips chip is just above).  Its 
> footprint is present on rev4 boards; on rev5 there is a glue logic 
> chip there instead.

Does this mean that the vast majority of machines did not include it? 
That's a crying shame. I suppose lspci would show it if it were there. 
No wonder video capture performance is so abysmal. There go most of my 
ideas about uses for that...

>> Maybe the first step, but everything Corel-specific and 
>> Rebel-specific  in the DM images should ideally be available.
>
> Lofty goal.  We haven't had any volunteers step forward...  Just 
> building NW-9 as it stands today took a good few months of my time and 
> effort, and while I enjoy this sort of thing, I do have other 
> interests besides NetWinder :)

Perhaps one of the reasons nobody's volunteered is that there's no 
indication of what one would be volunteering for. I have seen no 
definitive list of DM stuff that never made it beyond the DM images. If 
such a list were available, it may be more palatable to divide & 
conquer the tasks on it.

>>> You can also do some clever cable-folding to make use of "free" 
>>> space  in the case.  Using a two-headed IDE cable, run it on top of 
>>> the HDD,  then do a 90 degree diagonal bend, and you'll have room on 
>>> the side.   The only issue is supporting it mechanically.
>> Room on the side? You mean put the second drive in sideways? I don't  
>> quite see how that's possible...
>
> I didn't say it was easy :)  One thing to note, the HDD is usually on 
> rails, to adapt the spacing of the support screws.  This can be 
> improved upon if one is mechanically inclined.  Heat is an issue 
> though, with the CPU right under the hard drive - a second drive would 
> only make it worse.

Indeed, the rails take up quite a bit of space, but I believed this to 
be necessary, perhaps even insufficient, to keep the HD away from the 
CPU. Still, a flash drive would generate far less heat than a real HD. 
I would be interested in how anybody else may have managed to fit two 
drives in the case.

>> As for the power supply, I have two with the Corel label, and their  
>> plugs barely fit the power connector, becoming entirely dislodged at  
>> the slightest touch. One of them I have not tried with the unit 
>> because  I was told that its operational status was unknown, and 
>> indeed, its  voltage without load is slightly lower than that of the 
>> one that  currently works.
>
> Hmm I seem to recall for a while they shipped psu's where the 
> connector was too short, and prone to falling out.  In fact I think I 
> have a few of such supplies myself.  Even the "right" ones are pretty 
> prone to falling out.  It was a common complaint, but the replacement 
> (automotive) power connector used on the rackmounts drew its own 
> complaints.  Get out your soldering iron ;)

The shortness of the connector only initially appears to be at fault. I 
took a blade to some of the insulation at the base of it, thereby 
making it long enough to have a small portion of its length exposed 
even when pushed all the way in, but that only improved the situation 
very slightly. The real culprit is that the diameter of the plug's 
positive terminal is too big and the negative too small. I brought up 
the PSU issue to ask about its suitability to drive two IDE devices.




More information about the Netwinder mailing list