Re: gEDA-user: Problems with subnets containing short-circuits
Steve, Thanks for confirming my suspicion. Should I report this somewhere, or is this list enough? I'm not familiar with how the gEDA development is handled. /Arvid Steve Meier skrev: Arvid, Your are correct the produced netlist is incorrect. The last two nets unnamed_net2 and unnamed_net1 should have been combined. As previously demonstrated the net list for test.sch is correct and so I suspect the bug is in the code where the the hierarchical boundary is crossed to combine the nets. unnamed_net9Rload-1 Rtop4-1 Rtop2-2 GND Rload-2 Vin-2 unnamed_net8Rtop3-2 Rtop1-1 Vin-1 unnamed_net4Utest/R34-1 Rtop3-1 unnamed_net3Utest/R34-2 Rtop4-2 unnamed_net2Rtop1-2 unnamed_net1Rtop2-1 Steve Meier ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
gEDA-user: Starting a SM board
I want to convert a small thru-hole design to a surface mount. Started with a netlist and placed ic, 603 & 1206 parts from geda lib on board. It appears that the parts are all on the component side for viewing but solder side for connectivity. Auto route works. Using pcb 20050609. Time for a beer or two because this can't be real. George ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
Arvid, Your are correct the produced netlist is incorrect. The last two nets unnamed_net2 and unnamed_net1 should have been combined. As previously demonstrated the net list for test.sch is correct and so I suspect the bug is in the code where the the hierarchical boundary is crossed to combine the nets. unnamed_net9Rload-1 Rtop4-1 Rtop2-2 GND Rload-2 Vin-2 unnamed_net8Rtop3-2 Rtop1-1 Vin-1 unnamed_net4Utest/R34-1 Rtop3-1 unnamed_net3Utest/R34-2 Rtop4-2 unnamed_net2Rtop1-2 unnamed_net1Rtop2-1 Steve Meier ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
Steve Meier skrev:
send me your test1.sym and I will run it for you.
Steve Meier
Oh, I forgot to include it. Sorry.
Here it is!
/Arvid
On Thu, 2006-07-20 at 18:43 +0200, Arvid Rosén wrote:
Ok. I am not sure what that means as I am not familiar with the PCB backend.
What happens if you run in on the top-level file (top.sch), which
includes test.sch as a subnet?
/Arvid
Steve Meier skrev:
When I run gnetlist -gPCB test.sch -o test.net I get
unnamed_net3R34-2 4-1
unnamed_net21-1 2-1
unnamed_net1R34-1 3-1
On Thu, 2006-07-20 at 17:28 +0200, Arvid Rosén wrote:
Stuart Brorson wrote:
I guess I don't see how shorting around a circuit involves connecting
two nets which don't connect to anything else. Can you post a
simplified version of what you are trying to do? We can look at it
during the code sprint.
Ok, I attached two sch-files which I think illustrates the problem quite
well.
I tried this with the gnet-geda backend, and it fails discover that
Rtop1 and Rtop2 should be connected.
Meanwhile, why not stick a zero ohm resistor into the shorting net,
similar to what John suggests? That way gnetlist has a component
which will preserve the net.
Stuart
Yeah, sure. This is no problem as long as you are aware of it. But I
just spent one day looking for errors in my netlister-backend.
I still think this is a strange behavior.
/Arvid
Stuart Brorson wrote:
Just out of curiosity, why would you want to have two nets which
connect to nothing other than each other?
Before we label it a bug, what is the goal of this construct? =20
Stuart
=20
Hi Stuart!
That's a good question. The construct itself is quite pointless.=20
However, I was working on a quite complex design with two=20
filter-subcircuits. I wanted to test some aspects of my circuit, so I=20
shorted the first filter just to bypass it, which resulted in an open=20
circuit instead of a bypass! That is quite annoying, don't you think?
The possibility to bypassing stuff is always important when debugging.
Cheers
Arvid
=20
John Luciani wrote:
=20
On 7/20/06, Arvid Ros=3DE9n <[EMAIL PROTECTED]> wrote:
=20
However, I also tried to make a 2-pin subnet containing only a
short-circuit. This connection was completely ignored by gnetlist!!
Could someone confirm this?
=20
You are correct. If you have two nets that only connect to each other
the connection does not show up in the netlist.
I get around this by connecting one of the nets to a component.
(* jcl *)
=20
Ok, thanks!
Shouldn't that be considered as a pretty severe bug? I am new here,=3D=
20
maybe it's already reported.
/Arvid
--=3D20
Arvid Ros=3DE9n
Softube AB
Teknikringen 7
S-583 30 Link=3DF6ping
Sweden
Phone: (+46) 13 21 16 23
Mobile: (+46) 73 25 12 34 7
Mail: [EMAIL PROTECTED]
URL: www.softube.se
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
=20
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
=20
--=20
Arvid Ros=E9n
Softube AB
Teknikringen 7
S-583 30 Link=F6ping
Sweden
Phone: (+46) 13 21 16 23
Mobile: (+46) 73 25 12 34 7
Mail: [EMAIL PROTECTED]
URL: www.softube.se
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
plain text document attachment (test.sch)
v 20060123 1
C 1000 4500 1 0 0 in-1.sym
{
T 1000 4800 5 10 1 1 0 0 1
refdes=3
}
C 4600 6100 1 180 0 in-1.sym
{
T 4600 5800 5 10 1 1 180 0 1
refdes=2
}
C 1000 5900 1 0 0 in-1.sym
{
T 1000 6200 5 10 1 1 0 0 1
refdes=1
}
N 1600 6000 4000 6000 4
C 4600 4700 1 180 0 in-1.sym
{
T 4600 4400 5 10 1 1 180 0 1
refdes=4
}
C 2300 4500 1 0 0 resistor-1.sym
{
T 2500 5000 5 10 1 1 0 0 1
refdes=R34
T 2500 4800 5 10 1 1 0 0 1
value=10k
}
N 2300 4600 1600 4600 4
N 3200 4600 4000 4600 4
plain text document attachment (top.sch)
v 20060123 1
N 5000 9500 4300 9500 4
C 5000 9200 1 0 0 test.sym
{
T 5400 10300 5 10 1 1 0 0 1
refdes=Utest
T 5400 9200 5 10 1 1 0 0 1
device=test
T 5000 9200 5 10 1 0 0 0 1
source=test.sch
}
C 1700 8200 1 0 0 va
Re: gEDA-user: Twister2 made with gEDA is working
Congrats on that Karel. John G Karel Kulhavy wrote: I used gEDA and PCB to degign Ronja Twister2. I have tested it and it's working, with 0% packetloss ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: OT: UV exposure?
On Jun 27, 2006, at 10:34 PM, Steve Meier wrote: It takes a fair amount of time for the UV light to break down the nylon of a tent on a high mountain. I suspect, that UV light won't effect the performance of your board. I would be more worried about moisture and corrosion. Any possibility you could put a lot of epoxy all over your board except for the photo diodes? That's a very good possibility. I can mount the other components on the other side of the board, put the photodiode array on the sunlight-facing side, and mask off everything except the photodiodes. That might be a good idea anyway, since many semiconductors (especially diodes in glass housings) are light sensitive, and behave differently in full sunlight than they do on the bench in a darkish lab. JDB [cue story about leaving 2716s out in the sun to erase them] -- LART. 250 MIPS under one Watt. Free hardware design files. http://www.lartmaker.nl/ ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
send me your test1.sym and I will run it for you. Steve Meier On Thu, 2006-07-20 at 18:43 +0200, Arvid Rosén wrote: > Ok. I am not sure what that means as I am not familiar with the PCB backend. > What happens if you run in on the top-level file (top.sch), which > includes test.sch as a subnet? > > /Arvid > > Steve Meier skrev: > > When I run gnetlist -gPCB test.sch -o test.net I get > > > > unnamed_net3R34-2 4-1 > > unnamed_net21-1 2-1 > > unnamed_net1R34-1 3-1 > > > > > > > > On Thu, 2006-07-20 at 17:28 +0200, Arvid Rosén wrote: > > > >> Stuart Brorson wrote: > >> > >>> I guess I don't see how shorting around a circuit involves connecting > >>> two nets which don't connect to anything else. Can you post a > >>> simplified version of what you are trying to do? We can look at it > >>> during the code sprint. > >>> > >>> > >> Ok, I attached two sch-files which I think illustrates the problem quite > >> well. > >> I tried this with the gnet-geda backend, and it fails discover that > >> Rtop1 and Rtop2 should be connected. > >> > >>> Meanwhile, why not stick a zero ohm resistor into the shorting net, > >>> similar to what John suggests? That way gnetlist has a component > >>> which will preserve the net. > >>> > >>> Stuart > >>> > >>> > >> Yeah, sure. This is no problem as long as you are aware of it. But I > >> just spent one day looking for errors in my netlister-backend. > >> I still think this is a strange behavior. > >> > >> /Arvid > >> > >> > >> > >> > >> > >>> > >>> > Stuart Brorson wrote: > > > > Just out of curiosity, why would you want to have two nets which > > connect to nothing other than each other? > > > > Before we label it a bug, what is the goal of this construct? =20 > > > > Stuart > > =20 > > > > > Hi Stuart! > That's a good question. The construct itself is quite pointless.=20 > However, I was working on a quite complex design with two=20 > filter-subcircuits. I wanted to test some aspects of my circuit, so I=20 > shorted the first filter just to bypass it, which resulted in an open=20 > circuit instead of a bypass! That is quite annoying, don't you think? > The possibility to bypassing stuff is always important when debugging. > > Cheers > Arvid > > > > =20 > > > > > >> John Luciani wrote: > >>=20 > >> > >> > >>> On 7/20/06, Arvid Ros=3DE9n <[EMAIL PROTECTED]> wrote: > >>> =20 > >>> > >>> > However, I also tried to make a 2-pin subnet containing only a > short-circuit. This connection was completely ignored by gnetlist!! > > Could someone confirm this? > =20 > > > >>> You are correct. If you have two nets that only connect to each other > >>> the connection does not show up in the netlist. > >>> > >>> I get around this by connecting one of the nets to a component. > >>> > >>> (* jcl *) > >>> > >>> =20 > >>> > >>> > >> Ok, thanks! > >> Shouldn't that be considered as a pretty severe bug? I am new here,=3D= > >> > >> > 20 > > > >> maybe it's already reported. > >> > >> /Arvid > >> > >> --=3D20 > >> Arvid Ros=3DE9n > >> Softube AB > >> Teknikringen 7 > >> S-583 30 Link=3DF6ping > >> Sweden > >> Phone: (+46) 13 21 16 23 > >> Mobile: (+46) 73 25 12 34 7 > >> Mail: [EMAIL PROTECTED] > >> URL: www.softube.se > >> > >> > >> > >> ___ > >> geda-user mailing list > >> geda-user@moria.seul.org > >> http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > >> > >>=20 > >> > >> > > ___ > > geda-user mailing list > > geda-user@moria.seul.org > > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > > =20 > > > > > --=20 > Arvid Ros=E9n > Softube AB > Teknikringen 7 > S-583 30 Link=F6ping > Sweden > Phone: (+46) 13 21 16 23 > Mobile: (+46) 73 25 12 34 7 > Mail: [EMAIL PROTECTED] > URL: www.softube.se > > > > ___ > geda-user mailing list > geda-user@moria.seul.org > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > > > > >>> > >>> ___ > >>> geda-user mailing list > >>> geda-user@moria.seul.org > >>> http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > >>> >
Re: gEDA-user: Problems with subnets containing short-circuits
Ok. I am not sure what that means as I am not familiar with the PCB backend.
What happens if you run in on the top-level file (top.sch), which
includes test.sch as a subnet?
/Arvid
Steve Meier skrev:
When I run gnetlist -gPCB test.sch -o test.net I get
unnamed_net3R34-2 4-1
unnamed_net21-1 2-1
unnamed_net1R34-1 3-1
On Thu, 2006-07-20 at 17:28 +0200, Arvid Rosén wrote:
Stuart Brorson wrote:
I guess I don't see how shorting around a circuit involves connecting
two nets which don't connect to anything else. Can you post a
simplified version of what you are trying to do? We can look at it
during the code sprint.
Ok, I attached two sch-files which I think illustrates the problem quite
well.
I tried this with the gnet-geda backend, and it fails discover that
Rtop1 and Rtop2 should be connected.
Meanwhile, why not stick a zero ohm resistor into the shorting net,
similar to what John suggests? That way gnetlist has a component
which will preserve the net.
Stuart
Yeah, sure. This is no problem as long as you are aware of it. But I
just spent one day looking for errors in my netlister-backend.
I still think this is a strange behavior.
/Arvid
Stuart Brorson wrote:
Just out of curiosity, why would you want to have two nets which
connect to nothing other than each other?
Before we label it a bug, what is the goal of this construct? =20
Stuart
=20
Hi Stuart!
That's a good question. The construct itself is quite pointless.=20
However, I was working on a quite complex design with two=20
filter-subcircuits. I wanted to test some aspects of my circuit, so I=20
shorted the first filter just to bypass it, which resulted in an open=20
circuit instead of a bypass! That is quite annoying, don't you think?
The possibility to bypassing stuff is always important when debugging.
Cheers
Arvid
=20
John Luciani wrote:
=20
On 7/20/06, Arvid Ros=3DE9n <[EMAIL PROTECTED]> wrote:
=20
However, I also tried to make a 2-pin subnet containing only a
short-circuit. This connection was completely ignored by gnetlist!!
Could someone confirm this?
=20
You are correct. If you have two nets that only connect to each other
the connection does not show up in the netlist.
I get around this by connecting one of the nets to a component.
(* jcl *)
=20
Ok, thanks!
Shouldn't that be considered as a pretty severe bug? I am new here,=3D=
20
maybe it's already reported.
/Arvid
--=3D20
Arvid Ros=3DE9n
Softube AB
Teknikringen 7
S-583 30 Link=3DF6ping
Sweden
Phone: (+46) 13 21 16 23
Mobile: (+46) 73 25 12 34 7
Mail: [EMAIL PROTECTED]
URL: www.softube.se
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
=20
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
=20
--=20
Arvid Ros=E9n
Softube AB
Teknikringen 7
S-583 30 Link=F6ping
Sweden
Phone: (+46) 13 21 16 23
Mobile: (+46) 73 25 12 34 7
Mail: [EMAIL PROTECTED]
URL: www.softube.se
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
plain text document attachment (test.sch)
v 20060123 1
C 1000 4500 1 0 0 in-1.sym
{
T 1000 4800 5 10 1 1 0 0 1
refdes=3
}
C 4600 6100 1 180 0 in-1.sym
{
T 4600 5800 5 10 1 1 180 0 1
refdes=2
}
C 1000 5900 1 0 0 in-1.sym
{
T 1000 6200 5 10 1 1 0 0 1
refdes=1
}
N 1600 6000 4000 6000 4
C 4600 4700 1 180 0 in-1.sym
{
T 4600 4400 5 10 1 1 180 0 1
refdes=4
}
C 2300 4500 1 0 0 resistor-1.sym
{
T 2500 5000 5 10 1 1 0 0 1
refdes=R34
T 2500 4800 5 10 1 1 0 0 1
value=10k
}
N 2300 4600 1600 4600 4
N 3200 4600 4000 4600 4
plain text document attachment (top.sch)
v 20060123 1
N 5000 9500 4300 9500 4
C 5000 9200 1 0 0 test.sym
{
T 5400 10300 5 10 1 1 0 0 1
refdes=Utest
T 5400 9200 5 10 1 1 0 0 1
device=test
T 5000 9200 5 10 1 0 0 0 1
source=test.sch
}
C 1700 8200 1 0 0 vac-1.sym
{
T 2400 8850 5 10 1 1 0 0 1
refdes=Vin
T 2400 8650 5 10 1 1 0 0 1
value=dc 0 ac 1
}
C 3400 9900 1 0 0 resistor-1.sym
{
T 3600 10400 5 10 1 1 0 0 1
refdes=Rtop1
T 3600 10200 5 10 1 1 0 0 1
value=10k
}
C 1900 7300 1 0 0 gnd-1.sym
N 2000 8200 2000 7600 4
N 2000 9400 2000 1 4
N 6800 9500 7600 9500 4
N 7600 1 6800 1 4
N 2000 8000 9000 8000 4
N 9000 9100 9000 1 4
N 3400 1 2000 1 4
N 4300 1 5000 1 4
C 4300 9600 1 180 0 resistor-1.sym
{
T 4100 9100 5 10
gEDA-user: Is anyone going to DAC
Is anyone here going to the "Design Automation Conference"? http://www.dac.com ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Is anyone going to DAC
Looks like fun and I am over in Fremont so I can get over there at least to see the Exhibits. Steve Meier On Thu, 2006-07-20 at 12:09 -0400, al davis wrote: > Is anyone here going to the "Design Automation Conference"? > > http://www.dac.com > > > ___ > geda-user mailing list > geda-user@moria.seul.org > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
When I run gnetlist -gPCB test.sch -o test.net I get
unnamed_net3R34-2 4-1
unnamed_net21-1 2-1
unnamed_net1R34-1 3-1
On Thu, 2006-07-20 at 17:28 +0200, Arvid Rosén wrote:
> Stuart Brorson wrote:
> > I guess I don't see how shorting around a circuit involves connecting
> > two nets which don't connect to anything else. Can you post a
> > simplified version of what you are trying to do? We can look at it
> > during the code sprint.
> >
> Ok, I attached two sch-files which I think illustrates the problem quite
> well.
> I tried this with the gnet-geda backend, and it fails discover that
> Rtop1 and Rtop2 should be connected.
> > Meanwhile, why not stick a zero ohm resistor into the shorting net,
> > similar to what John suggests? That way gnetlist has a component
> > which will preserve the net.
> >
> > Stuart
> >
> Yeah, sure. This is no problem as long as you are aware of it. But I
> just spent one day looking for errors in my netlister-backend.
> I still think this is a strange behavior.
>
> /Arvid
>
>
>
>
> >
> >
> >> Stuart Brorson wrote:
> >>
> >>> Just out of curiosity, why would you want to have two nets which
> >>> connect to nothing other than each other?
> >>>
> >>> Before we label it a bug, what is the goal of this construct? =20
> >>>
> >>> Stuart
> >>> =20
> >>>
> >> Hi Stuart!
> >> That's a good question. The construct itself is quite pointless.=20
> >> However, I was working on a quite complex design with two=20
> >> filter-subcircuits. I wanted to test some aspects of my circuit, so I=20
> >> shorted the first filter just to bypass it, which resulted in an open=20
> >> circuit instead of a bypass! That is quite annoying, don't you think?
> >> The possibility to bypassing stuff is always important when debugging.
> >>
> >> Cheers
> >> Arvid
> >>
> >>> =20
> >>>
> John Luciani wrote:
> =20
>
> > On 7/20/06, Arvid Ros=3DE9n <[EMAIL PROTECTED]> wrote:
> > =20
> >
> >> However, I also tried to make a 2-pin subnet containing only a
> >> short-circuit. This connection was completely ignored by gnetlist!!
> >>
> >> Could someone confirm this?
> >>=20
> >>
> > You are correct. If you have two nets that only connect to each other
> > the connection does not show up in the netlist.
> >
> > I get around this by connecting one of the nets to a component.
> >
> > (* jcl *)
> >
> > =20
> >
> Ok, thanks!
> Shouldn't that be considered as a pretty severe bug? I am new here,=3D=
>
> >> 20
> >>
> maybe it's already reported.
>
> /Arvid
>
> --=3D20
> Arvid Ros=3DE9n
> Softube AB
> Teknikringen 7
> S-583 30 Link=3DF6ping
> Sweden
> Phone: (+46) 13 21 16 23
> Mobile: (+46) 73 25 12 34 7
> Mail: [EMAIL PROTECTED]
> URL: www.softube.se
>
>
>
> ___
> geda-user mailing list
> geda-user@moria.seul.org
> http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
>
> =20
>
> >>>
> >>> ___
> >>> geda-user mailing list
> >>> geda-user@moria.seul.org
> >>> http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
> >>> =20
> >>>
> >> --=20
> >> Arvid Ros=E9n
> >> Softube AB
> >> Teknikringen 7
> >> S-583 30 Link=F6ping
> >> Sweden
> >> Phone: (+46) 13 21 16 23
> >> Mobile: (+46) 73 25 12 34 7
> >> Mail: [EMAIL PROTECTED]
> >> URL: www.softube.se
> >>
> >>
> >>
> >> ___
> >> geda-user mailing list
> >> geda-user@moria.seul.org
> >> http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
> >>
> >>
> >
> >
> >
> > ___
> > geda-user mailing list
> > geda-user@moria.seul.org
> > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
> >
>
>
> plain text document attachment (test.sch)
> v 20060123 1
> C 1000 4500 1 0 0 in-1.sym
> {
> T 1000 4800 5 10 1 1 0 0 1
> refdes=3
> }
> C 4600 6100 1 180 0 in-1.sym
> {
> T 4600 5800 5 10 1 1 180 0 1
> refdes=2
> }
> C 1000 5900 1 0 0 in-1.sym
> {
> T 1000 6200 5 10 1 1 0 0 1
> refdes=1
> }
> N 1600 6000 4000 6000 4
> C 4600 4700 1 180 0 in-1.sym
> {
> T 4600 4400 5 10 1 1 180 0 1
> refdes=4
> }
> C 2300 4500 1 0 0 resistor-1.sym
> {
> T 2500 5000 5 10 1 1 0 0 1
> refdes=R34
> T 2500 4800 5 10 1 1 0 0 1
> value=10k
> }
> N 2300 4600 1600 4600 4
> N 3200 4600 4000 4600 4
> plain text document attachment (top.sch)
> v 20060123 1
> N 5000 9500 4300 9500 4
> C 5000 9200 1 0 0 test.sym
> {
> T 5400 10300 5 10 1 1 0 0 1
> refdes=Utest
> T 5400 9200 5 10 1 1 0 0 1
> device=test
> T 5000 9200 5 10 1 0 0 0 1
> source=test.sch
> }
> C 1700 8200 1 0 0 vac-1.sym
> {
> T 2400 8850 5 10 1 1 0 0 1
> refdes=Vin
> T 2400 8650 5 10 1 1 0 0 1
> value=dc 0 a
Re: gEDA-user: Problems with subnets containing short-circuits
Stuart Brorson wrote:
I guess I don't see how shorting around a circuit involves connecting
two nets which don't connect to anything else. Can you post a
simplified version of what you are trying to do? We can look at it
during the code sprint.
Ok, I attached two sch-files which I think illustrates the problem quite
well.
I tried this with the gnet-geda backend, and it fails discover that
Rtop1 and Rtop2 should be connected.
Meanwhile, why not stick a zero ohm resistor into the shorting net,
similar to what John suggests? That way gnetlist has a component
which will preserve the net.
Stuart
Yeah, sure. This is no problem as long as you are aware of it. But I
just spent one day looking for errors in my netlister-backend.
I still think this is a strange behavior.
/Arvid
Stuart Brorson wrote:
Just out of curiosity, why would you want to have two nets which
connect to nothing other than each other?
Before we label it a bug, what is the goal of this construct? =20
Stuart
=20
Hi Stuart!
That's a good question. The construct itself is quite pointless.=20
However, I was working on a quite complex design with two=20
filter-subcircuits. I wanted to test some aspects of my circuit, so I=20
shorted the first filter just to bypass it, which resulted in an open=20
circuit instead of a bypass! That is quite annoying, don't you think?
The possibility to bypassing stuff is always important when debugging.
Cheers
Arvid
=20
John Luciani wrote:
=20
On 7/20/06, Arvid Ros=3DE9n <[EMAIL PROTECTED]> wrote:
=20
However, I also tried to make a 2-pin subnet containing only a
short-circuit. This connection was completely ignored by gnetlist!!
Could someone confirm this?
=20
You are correct. If you have two nets that only connect to each other
the connection does not show up in the netlist.
I get around this by connecting one of the nets to a component.
(* jcl *)
=20
Ok, thanks!
Shouldn't that be considered as a pretty severe bug? I am new here,=3D=
20
maybe it's already reported.
/Arvid
--=3D20
Arvid Ros=3DE9n
Softube AB
Teknikringen 7
S-583 30 Link=3DF6ping
Sweden
Phone: (+46) 13 21 16 23
Mobile: (+46) 73 25 12 34 7
Mail: [EMAIL PROTECTED]
URL: www.softube.se
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
=20
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
=20
--=20
Arvid Ros=E9n
Softube AB
Teknikringen 7
S-583 30 Link=F6ping
Sweden
Phone: (+46) 13 21 16 23
Mobile: (+46) 73 25 12 34 7
Mail: [EMAIL PROTECTED]
URL: www.softube.se
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
___
geda-user mailing list
geda-user@moria.seul.org
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
--
Arvid Rosén
Softube AB
Teknikringen 7
S-583 30 Linköping
Sweden
Phone: (+46) 13 21 16 23
Mobile: (+46) 73 25 12 34 7
Mail: [EMAIL PROTECTED]
URL: www.softube.se
v 20060123 1
C 1000 4500 1 0 0 in-1.sym
{
T 1000 4800 5 10 1 1 0 0 1
refdes=3
}
C 4600 6100 1 180 0 in-1.sym
{
T 4600 5800 5 10 1 1 180 0 1
refdes=2
}
C 1000 5900 1 0 0 in-1.sym
{
T 1000 6200 5 10 1 1 0 0 1
refdes=1
}
N 1600 6000 4000 6000 4
C 4600 4700 1 180 0 in-1.sym
{
T 4600 4400 5 10 1 1 180 0 1
refdes=4
}
C 2300 4500 1 0 0 resistor-1.sym
{
T 2500 5000 5 10 1 1 0 0 1
refdes=R34
T 2500 4800 5 10 1 1 0 0 1
value=10k
}
N 2300 4600 1600 4600 4
N 3200 4600 4000 4600 4
v 20060123 1
N 5000 9500 4300 9500 4
C 5000 9200 1 0 0 test.sym
{
T 5400 10300 5 10 1 1 0 0 1
refdes=Utest
T 5400 9200 5 10 1 1 0 0 1
device=test
T 5000 9200 5 10 1 0 0 0 1
source=test.sch
}
C 1700 8200 1 0 0 vac-1.sym
{
T 2400 8850 5 10 1 1 0 0 1
refdes=Vin
T 2400 8650 5 10 1 1 0 0 1
value=dc 0 ac 1
}
C 3400 9900 1 0 0 resistor-1.sym
{
T 3600 10400 5 10 1 1 0 0 1
refdes=Rtop1
T 3600 10200 5 10 1 1 0 0 1
value=10k
}
C 1900 7300 1 0 0 gnd-1.sym
N 2000 8200 2000 7600 4
N 2000 9400 2000 1 4
N 6800 9500 7600 9500 4
N 7600 1 6800 1 4
N 2000 8000 9000 8000 4
N 9000 9100 9000 1 4
N 3400 1 2000 1 4
N 4300 1 5000 1 4
C 4300 9600 1 180 0 resistor-1.sym
{
T 4100 9100 5 10 1 1 180 0 1
refdes=Rtop3
T 4100 9300 5 10 1 1 180 0 1
value=10k
}
N 3400 9500 3000 9500 4
N 3000 9500 3000 1 4
C 7600 9900 1 0 0 resistor-1.sym
{
T 7800 10400 5 10 1 1 0 0 1
refdes=Rtop2
T 7800 10200 5 10 1 1 0 0 1
value=10k
}
C 8500 9600 1 180 0 resistor-1.sym
{
T 8300 9100 5 10 1 1 180 0 1
refdes=Rtop4
T 8300 9300 5 10 1 1 180 0 1
value=10k
}
N 9000 1 8500 1 4
N 8500 9500 9000 9500 4
C 8900 9100 1 270 0 resistor-1.sym
{
T 9400 8900 5 10 1 1 270 0 1
refdes=Rload
T 9200 8900 5 10 1 1 270
Re: gEDA-user: Problems with subnets containing short-circuits
My understanding of the gnetlist algorithm used to traverse a page is to find a component... find a pin then follow the nets around to other pins... A net by itself would never be seen. Steve Meier On Thu, 2006-07-20 at 16:48 +0200, Arvid Rosén wrote: > Stuart Brorson wrote: > > Just out of curiosity, why would you want to have two nets which > > connect to nothing other than each other? > > > > Before we label it a bug, what is the goal of this construct? > > > > Stuart > > > Hi Stuart! > That's a good question. The construct itself is quite pointless. > However, I was working on a quite complex design with two > filter-subcircuits. I wanted to test some aspects of my circuit, so I > shorted the first filter just to bypass it, which resulted in an open > circuit instead of a bypass! That is quite annoying, don't you think? > The possibility to bypassing stuff is always important when debugging. > > Cheers > Arvid > > > > > >> John Luciani wrote: > >> > >>> On 7/20/06, Arvid Ros=E9n <[EMAIL PROTECTED]> wrote: > >>> > However, I also tried to make a 2-pin subnet containing only a > short-circuit. This connection was completely ignored by gnetlist!! > > Could someone confirm this? > > >>> You are correct. If you have two nets that only connect to each other > >>> the connection does not show up in the netlist. > >>> > >>> I get around this by connecting one of the nets to a component. > >>> > >>> (* jcl *) > >>> > >>> > >> Ok, thanks! > >> Shouldn't that be considered as a pretty severe bug? I am new here,=20 > >> maybe it's already reported. > >> > >> /Arvid > >> > >> --=20 > >> Arvid Ros=E9n > >> Softube AB > >> Teknikringen 7 > >> S-583 30 Link=F6ping > >> Sweden > >> Phone: (+46) 13 21 16 23 > >> Mobile: (+46) 73 25 12 34 7 > >> Mail: [EMAIL PROTECTED] > >> URL: www.softube.se > >> > >> > >> > >> ___ > >> geda-user mailing list > >> geda-user@moria.seul.org > >> http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > >> > >> > > > > > > > > ___ > > geda-user mailing list > > geda-user@moria.seul.org > > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > > > > ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
On 7/20/06, Stuart Brorson <[EMAIL PROTECTED]> wrote: Just out of curiosity, why would you want to have two nets which connect to nothing other than each other? Before we label it a bug, what is the goal of this construct? My goal was to avoid snaking wires to connect subcircuits. Here are two simple examples --- 1. I have a microcontroller subcircuit, that I reuse many times, that has an analog input port with pins P6.0-P6.7 that I want to connect to an analog front end (which I also reuse many times) that has two outputs AOUT0 and AOUT1. I include these schematics in hierachical blocks. To show connections on the same sheet as the blocks I could do P6.0 -- AOUT0 P6.1 -- AOUT1 I do not have to alter either block or run a rat's nest of wires to show connectivity. 2. If separate grounds are used you could do a DGND --- AGND to show connectivity. (* jcl *) -- http://www.luciani.org ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: OT: UV exposure?
> http://www.tecaeromex.com/ingles/destilai.html Hmmm, $16,000 for a machine that makes stuff that goes BOOM if you're not careful. No thanks ;-) ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: OT: UV exposure?
DJ, Have you seen this page about making your own 99% H2O2? John G http://www.tecaeromex.com/ingles/destilai.html DJ Delorie wrote: When the gov't comes to get you, just open your eyes wide and say "I want to be an astronaut when I grow up!" ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
gEDA-user: Reminder: gEDA-dev: gEDA code sprint, freeedaug, and cookout - Sunday July 23rd in NH
[My wife says she can bring the innocent bystanders to the beach down the road if the weather is good, so bring bathing suits] The next gEDA code sprint, Freeedaug meeting, and general gEDA cookout will all happen on Sunday July 23rd at my house in Deerfield NH! That's this Sunday, and about an hour north of Boston. The official start will be 10:00 AM EST, it ends when the last guest goes home :-) If you wish to attend in person, bring your family (kids too!) (optional of course ;) and we'll make a day of it. I don't want directions to my house to end up in the archives, so send me a private email ([EMAIL PROTECTED]) and I'll send you directions. The adventurous could probably figure it own with a little googling, but at least let me know if you plan on coming so we can plan food. If you want to participate in the code sprint, bring your laptop. Both wired and wireless connections will be available. If you want to show off your latest project, bring it along! Lawn games and lawn chairs would be welcome too. If you wish to participate via IRC, you may find us on the #geda channel. Irc server: irc.seul.org, channel: #geda We'll provide generic soda, chips, hot dogs, and burgers, but please bring any other food or drink you may prefer. Note: This event replaces this month's regular Freeedaug meeting in Cambridge. The vision behind the code sprint is to spend a day working on whatever pet gEDA coding project you've been wishing to attack. You can sit down with the rest of the gEDA developer community, and crank out that new feature and/or bug fix [1]! Since we'll all be working together (either in person or over the net), the power of community may be brought to bear on any problems we encounter while hacking. Free Dog is an association of like-minded hackers and engineers interested in free and open EDA tools. We hold monthly meetings near MIT (and other locations) featuring informal networking, speakers, and camaraderie. Our goals are to learn more about CAD, engineering and scientific software, share ideas about our current projects, and -- most importantly -- have fun with like-minded people. We welcome new members and participants of all ages. Students are particularly welcome! DJ Delorie [EMAIL PROTECTED] PS: If it rains, we'll have a cook-in instead :-) ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
I guess I don't see how shorting around a circuit involves connecting two nets which don't connect to anything else. Can you post a simplified version of what you are trying to do? We can look at it during the code sprint. Meanwhile, why not stick a zero ohm resistor into the shorting net, similar to what John suggests? That way gnetlist has a component which will preserve the net. Stuart > > Stuart Brorson wrote: > > Just out of curiosity, why would you want to have two nets which > > connect to nothing other than each other? > > > > Before we label it a bug, what is the goal of this construct? =20 > > > > Stuart > > =20 > Hi Stuart! > That's a good question. The construct itself is quite pointless.=20 > However, I was working on a quite complex design with two=20 > filter-subcircuits. I wanted to test some aspects of my circuit, so I=20 > shorted the first filter just to bypass it, which resulted in an open=20 > circuit instead of a bypass! That is quite annoying, don't you think? > The possibility to bypassing stuff is always important when debugging. > > Cheers > Arvid > > > > =20 > >> John Luciani wrote: > >>=20 > >>> On 7/20/06, Arvid Ros=3DE9n <[EMAIL PROTECTED]> wrote: > >>> =20 > However, I also tried to make a 2-pin subnet containing only a > short-circuit. This connection was completely ignored by gnetlist!! > > Could someone confirm this? > =20 > >>> You are correct. If you have two nets that only connect to each other > >>> the connection does not show up in the netlist. > >>> > >>> I get around this by connecting one of the nets to a component. > >>> > >>> (* jcl *) > >>> > >>> =20 > >> Ok, thanks! > >> Shouldn't that be considered as a pretty severe bug? I am new here,=3D= > 20 > >> maybe it's already reported. > >> > >> /Arvid > >> > >> --=3D20 > >> Arvid Ros=3DE9n > >> Softube AB > >> Teknikringen 7 > >> S-583 30 Link=3DF6ping > >> Sweden > >> Phone: (+46) 13 21 16 23 > >> Mobile: (+46) 73 25 12 34 7 > >> Mail: [EMAIL PROTECTED] > >> URL: www.softube.se > >> > >> > >> > >> ___ > >> geda-user mailing list > >> geda-user@moria.seul.org > >> http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > >> > >>=20 > > > > > > > > ___ > > geda-user mailing list > > geda-user@moria.seul.org > > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > > =20 > > > --=20 > Arvid Ros=E9n > Softube AB > Teknikringen 7 > S-583 30 Link=F6ping > Sweden > Phone: (+46) 13 21 16 23 > Mobile: (+46) 73 25 12 34 7 > Mail: [EMAIL PROTECTED] > URL: www.softube.se > > > > ___ > geda-user mailing list > geda-user@moria.seul.org > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
So, It's a short circuit between two pins of a sub-block that is not netlisted correctly? Can you craft a small example that does not work and post it? Does this only happen in the spice netlister? Mike Arvid Rosén wrote: Stuart Brorson wrote: Just out of curiosity, why would you want to have two nets which connect to nothing other than each other? Before we label it a bug, what is the goal of this construct? Stuart Hi Stuart! That's a good question. The construct itself is quite pointless. However, I was working on a quite complex design with two filter-subcircuits. I wanted to test some aspects of my circuit, so I shorted the first filter just to bypass it, which resulted in an open circuit instead of a bypass! That is quite annoying, don't you think? The possibility to bypassing stuff is always important when debugging. Cheers Arvid John Luciani wrote: On 7/20/06, Arvid Ros=E9n <[EMAIL PROTECTED]> wrote: However, I also tried to make a 2-pin subnet containing only a short-circuit. This connection was completely ignored by gnetlist!! Could someone confirm this? You are correct. If you have two nets that only connect to each other the connection does not show up in the netlist. I get around this by connecting one of the nets to a component. (* jcl *) Ok, thanks! Shouldn't that be considered as a pretty severe bug? I am new here,=20 maybe it's already reported. /Arvid --=20 Arvid Ros=E9n Softube AB Teknikringen 7 S-583 30 Link=F6ping Sweden Phone: (+46) 13 21 16 23 Mobile: (+46) 73 25 12 34 7 Mail: [EMAIL PROTECTED] URL: www.softube.se ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user -- -- Mike Jarabek FPGA/ASIC Designer http://www.istop.com/~mjarabek -- ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
Stuart Brorson wrote: Just out of curiosity, why would you want to have two nets which connect to nothing other than each other? Before we label it a bug, what is the goal of this construct? Stuart Hi Stuart! That's a good question. The construct itself is quite pointless. However, I was working on a quite complex design with two filter-subcircuits. I wanted to test some aspects of my circuit, so I shorted the first filter just to bypass it, which resulted in an open circuit instead of a bypass! That is quite annoying, don't you think? The possibility to bypassing stuff is always important when debugging. Cheers Arvid John Luciani wrote: On 7/20/06, Arvid Ros=E9n <[EMAIL PROTECTED]> wrote: However, I also tried to make a 2-pin subnet containing only a short-circuit. This connection was completely ignored by gnetlist!! Could someone confirm this? You are correct. If you have two nets that only connect to each other the connection does not show up in the netlist. I get around this by connecting one of the nets to a component. (* jcl *) Ok, thanks! Shouldn't that be considered as a pretty severe bug? I am new here,=20 maybe it's already reported. /Arvid --=20 Arvid Ros=E9n Softube AB Teknikringen 7 S-583 30 Link=F6ping Sweden Phone: (+46) 13 21 16 23 Mobile: (+46) 73 25 12 34 7 Mail: [EMAIL PROTECTED] URL: www.softube.se ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user -- Arvid Rosén Softube AB Teknikringen 7 S-583 30 Linköping Sweden Phone: (+46) 13 21 16 23 Mobile: (+46) 73 25 12 34 7 Mail: [EMAIL PROTECTED] URL: www.softube.se ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
Just out of curiosity, why would you want to have two nets which connect to nothing other than each other? Before we label it a bug, what is the goal of this construct? Stuart > > John Luciani wrote: > > On 7/20/06, Arvid Ros=E9n <[EMAIL PROTECTED]> wrote: > >> However, I also tried to make a 2-pin subnet containing only a > >> short-circuit. This connection was completely ignored by gnetlist!! > >> > >> Could someone confirm this? > > > > You are correct. If you have two nets that only connect to each other > > the connection does not show up in the netlist. > > > > I get around this by connecting one of the nets to a component. > > > > (* jcl *) > > > Ok, thanks! > Shouldn't that be considered as a pretty severe bug? I am new here,=20 > maybe it's already reported. > > /Arvid > > --=20 > Arvid Ros=E9n > Softube AB > Teknikringen 7 > S-583 30 Link=F6ping > Sweden > Phone: (+46) 13 21 16 23 > Mobile: (+46) 73 25 12 34 7 > Mail: [EMAIL PROTECTED] > URL: www.softube.se > > > > ___ > geda-user mailing list > geda-user@moria.seul.org > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
John Luciani wrote: On 7/20/06, Arvid Rosén <[EMAIL PROTECTED]> wrote: However, I also tried to make a 2-pin subnet containing only a short-circuit. This connection was completely ignored by gnetlist!! Could someone confirm this? You are correct. If you have two nets that only connect to each other the connection does not show up in the netlist. I get around this by connecting one of the nets to a component. (* jcl *) Ok, thanks! Shouldn't that be considered as a pretty severe bug? I am new here, maybe it's already reported. /Arvid -- Arvid Rosén Softube AB Teknikringen 7 S-583 30 Linköping Sweden Phone: (+46) 13 21 16 23 Mobile: (+46) 73 25 12 34 7 Mail: [EMAIL PROTECTED] URL: www.softube.se ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Problems with subnets containing short-circuits
On 7/20/06, Arvid Rosén <[EMAIL PROTECTED]> wrote: However, I also tried to make a 2-pin subnet containing only a short-circuit. This connection was completely ignored by gnetlist!! Could someone confirm this? You are correct. If you have two nets that only connect to each other the connection does not show up in the netlist. I get around this by connecting one of the nets to a component. (* jcl *) -- http://www.luciani.org ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
gEDA-user: Problems with subnets containing short-circuits
Hi! I have experimented a bit with the hierarchy functions in gschem/gnetlist and I think I found a problem. If I make a subnet, with a bunch of components in it, it seems to work alright. However, I also tried to make a 2-pin subnet containing only a short-circuit. This connection was completely ignored by gnetlist!! Could someone confirm this? Cheers Arvid -- Arvid Rosén Softube AB Teknikringen 7 S-583 30 Linköping Sweden Phone: (+46) 13 21 16 23 Mobile: (+46) 73 25 12 34 7 Mail: [EMAIL PROTECTED] URL: www.softube.se ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Specifications
[snip] >How large a schematic can the schematic capture handle?, That is, how >many pages, what sheet size. There are no limits that I'm aware of on schematic size. > >Does that schematic handle hierarchial schematics? Yes. Just keep in mind that 1) bus hierarchy is not supported and 2) output netlists are typically flattened. -Ales ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Installation problem
[snip] >Ran into these errors on the installation. Can someone tell me what the >problem is? This is on Suse 10.1 32-bit. > [snip] >make[3]: *** [o_picture.lo] Error 1 >make[3]: Leaving directory >`/home/afore/geda-sources/gedagaf/libgeda-20060123/src' >`/home/afore/geda-sources/gedagaf/libgeda-20060123/src' >make[2]: *** [install] Error 2 >make[2]: Leaving directory >`/home/afore/geda-sources/gedagaf/libgeda-20060123/src' >make[1]: *** [install-recursive] Error 1 >make[1]: Leaving directory >`/home/afore/geda-sources/gedagaf/libgeda-20060123' >make: *** [libgeda_install] Error 2 Unfortunately, I don't see why the compiler is failing here (no failure details in the above output). The first thing to do, if you haven't done so already is to make sure you have all the packages listed at: http://geda.seul.org/wiki/geda:suse_10.0 installed. Another possibility is to download gEDA/gaf directly from: http://geda.seul.org/devel/20060123/ and follow the build instructions given in either: http://geda.seul.org/devel/20060123/README or http://geda.seul.org/devel/20060123/gaf-20060123-relnotes.html -Ales ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
Re: gEDA-user: Installation Problem
On Thursday 20 July 2006 02:13, Art Fore wrote: > Still having problem with gEDA installation on Suse 10.1. I get the > following. Noting shows up under libtermcap in yast, libreadline-java > exists though. Any suggestions on what the problem is? I am not that > knowledgeable on these things. > > Art > > > /home/afore/geda-sources/gnucap/gnucap-2006-01-17 > ./configure --prefix=/home/afore/geda-install > Configuring gnucap using --prefix=/home/afore/geda-install > usr/lib/gcc/i586-suse-linux/4.1.0/../../../../i586-suse-linux/bin/ld: cannot > find -ltermcap > collect2: ld returned 1 exit status > either libtermcap or libreadline is missing - not using readline The configure script did not find the library, so it disabled the features that require it. It will compile ok, but missing some features. The missing feature is command line editing. === g++ -DHAS_READLINE -DUNIX -O2 -DNDEBUG -I.. -I. -W -c ../globals.cc ../e_card.h: In function 'CARD_LIST::fat_iterator findbranch(CS&, CARD_LIST*)': ../e_card.h:152: error: cannot convert 'CARD_LIST::fat_iterator' to 'CARD_LIST*' for argument '2' to 'CARD_LIST::fat_iterator findbranch(CS&, CARD_LIST*)' ../e_compon.h: In member function 'void COMPONENT::attach_common(COMMON_COMPONENT*)': ../e_compon.h:139: error: '::attach_common' has not been declared ../e_compon.h: In member function 'void COMPONENT::detach_common()': ../e_compon.h:140: error: '::detach_common' has not been declared make[2]: *** [globals.o] Error 1 = (I replaced the one you sent with one that doesn't have color changes, because it is more readable in email that way.) This relates to a change in the C++ language spec, regarding friends. The compiler was changed with gcc 4.1, breaking code done to the old standard. From the gcc documentation: `-ffriend-injection' Inject friend functions into the enclosing namespace, so that they are visible outside the scope of the class in which they are declared. Friend functions were documented to work this way in the old Annotated C++ Reference Manual, and versions of G++ before 4.1 always worked that way. However, in ISO C++ a friend function which is not declared in an enclosing scope can only be found using argument dependent lookup. This option causes friends to be injected as they were in earlier releases. = Gnucap was done according to the "old Annotated C++ Reference Manual". The best way to handle the problem is to get the latest development version of gnucap. Official site: http://www.gnucap.org/devel/gnucap-2006-07-08.tar.gz Mirrors: http://www.geda.seul.org/dist/gnucap-2006-07-08.tar.gz Alternatively, you can patch the one you have: Change the file src/configure, line 22: from: echo "-DUNIX -O2 -DNDEBUG -I.. -I. -W" >>Make.ccflags to: echo "-DUNIX -O2 -DNDEBUG -I.. -I. -W -ffriend-injection" >>Make.ccflags This adds the compiler option "-ffriend-injection", which reverts to the old behavior. A third way to deal with this is to use an older compiler. It works on 4.0. ___ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
