Matthew Ayres | 9 Aug 2010 12:05
Picon

Quick distro question

Hi all

My current distro's minimum KDE version is KDE 4.4.4.  I suppose I could add in some KDE3 libraries, but really would rather not.  So I'm planning to install a small KDE3 distro under Virtual Box.  Does anyone have a suggestion to which distro I use, particularly so that I can aide in porting KTechLab?

------------------------------------------------------------------------------
This SF.net email is sponsored by 

Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev 
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel
Julian Bäume | 9 Aug 2010 18:40
Picon
Gravatar

Re: Quick distro question

moin,

On Monday 09 August 2010 12:05:07 Matthew Ayres wrote:
> My current distro's minimum KDE version is KDE 4.4.4.  I suppose I could add
> in some KDE3 libraries, but really would rather not.  So I'm planning to
> install a small KDE3 distro under Virtual Box.  Does anyone have a
> suggestion to which distro I use, particularly so that I can aide in porting
> KTechLab?
There is a live-distro that ships the old KTechLab. It's the
openSUSE-Education live-dvd. I also played around with the
openSUSE build-studio to create small images that work within
vmware/virtualbox. I think I have to re-create them, if you want
to use it. At least it worked some time ago, so I guess it shouldn't be
much effort to create them again. (The config files should still be there.)

Is that an option for you? Then I will have a look at it ;)

bye then
julian
------------------------------------------------------------------------------
This SF.net email is sponsored by 

Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev 
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel
Matthew Ayres | 9 Aug 2010 21:29
Picon

Re: Quick distro question

> My current distro's minimum KDE version is KDE 4.4.4.  I suppose I could add
> in some KDE3 libraries, but really would rather not.  So I'm planning to
> install a small KDE3 distro under Virtual Box.  Does anyone have a
> suggestion to which distro I use, particularly so that I can aide in porting
> KTechLab?
There is a live-distro that ships the old KTechLab. It's the
openSUSE-Education live-dvd. I also played around with the
openSUSE build-studio to create small images that work within
vmware/virtualbox. I think I have to re-create them, if you want
to use it. At least it worked some time ago, so I guess it shouldn't be
 much effort to create them again. (The config files should still be there.) 
 
Is that an option for you? Then I will have a look at it ;)


That's an excellent option for me, thank you!
------------------------------------------------------------------------------
This SF.net email is sponsored by 

Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev 
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel
Zoltan Padrah | 9 Aug 2010 22:54
Picon

Re: Quick distro question



2010/8/9 Matthew Ayres <solar.granulation <at> gmail.com> 
Does anyone have a suggestion to which distro I use, particularly so that I can aide in porting KTechLab?


For porting, I'm using kde4, kdevelop4, git and and the new code base. The kde3 version is useful only to know how should various features work.


------------------------------------------------------------------------------
This SF.net email is sponsored by 

Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev 
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel
Matthew Ayres | 10 Aug 2010 00:07
Picon

Re: Quick distro question

On Mon, Aug 9, 2010 at 9:54 PM, Zoltan Padrah <zoltan.padrah <at> gmail.com> wrote:


2010/8/9 Matthew Ayres <solar.granulation-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> 
Does anyone have a suggestion to which distro I use, particularly so that I can aide in porting KTechLab?


For porting, I'm using kde4, kdevelop4, git and and the new code base. The kde3 version is useful only to know how should various features work.


Ah, excellent.  I've found the wiki article for developing ktechlab with kdevelop4.  Having some problems right now but hopefully when I'm more awake (tomorrow) I'll be able to sort that out.

Meanwhile, I downloaded the openSUSE-Education live DVD.  It doesn't seem to have KTechLab though :(
------------------------------------------------------------------------------
This SF.net email is sponsored by 

Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev 
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel
Alan Grimes | 20 Aug 2010 07:39
Favicon

Elements inside a 2A3

Lets review pins and elements using my favorite part, the 2A3

The 2A3 is a filamentary triode with 3 electrodes and 4 pins.

The pins are:

Anode (a)
Grid (g)
and the filament (f1, f2)

Note:
g means Conductance

A first approximation of the filament has a 1-ohm resistor between f1
and f2. (powered by a 2.5v supply) Several problems immediately present
themselves. First, this model ignores the temperature coefficient of the
filament. Second, there's the problem of computing the critical working
voltage of the tube. ( V_a - avg(V_f1, V_f2) ) And third, there is the
parasitic capacitance. All voltages in the analysis of a vacuum tube are
referenced to the cathode.

There are two ways to approach the second problem. The first is to split
the resistor in half and add a node to the circuit, then compute all the
triode properties off of that node. The second is to write a generalized
triode equation that can handle a 2-node cathode (filament in this case)
and perform the necessary averaging...

So a model for f1 and f2 would have one resistor for each filament node,
connected to an internal hidden pin which would also have two capacitors
(the parasitic capacitance between the filament and the grid, and the
filament and the anode respectively). The next set of internal elements
are a pair of thermionic diodes (I don't know how these work just yet),
The obvious one between the filament and the anode, and the second
between the filament and the grid. Both of these diodes have a transfer
function based on the voltage between said electrode and the filament.
Finally, these diodes interact with each other, the grid will affect the
transfer function to the anode at any practical voltage, and the anode
will affect the transfer function to the grid when it is conducting.
This is called transconductance. The 2A3 has a relatively small
transconductance (about 4), so it can require a full-scale swing of as
much as 95 volts between V_p (pinch-off voltage), and positive grid
voltage (where the tube enters class A2 operation and the grid begins to
conduct).

My point is to show an interesting example where the number of elements
dwarfs the number of pins.

Attached is an attempt to simulate a 2A3 based on an elemental analysis.

-- 
DO NOT USE OBAMACARE.
DO NOT BUY OBAMACARE.
Powers are not rights.

Attachment (2A3equiv.circuit): application/x-circuit, 20 KiB
------------------------------------------------------------------------------
This SF.net email is sponsored by 

Make an app they can't live without
Enter the BlackBerry Developer Challenge
http://p.sf.net/sfu/RIM-dev2dev 
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel
Zoltan Padrah | 24 Aug 2010 11:42
Picon

Adding nodes() to IDocumentModel ?


 Hello,

As implementing the simulator, I've found a problem with the IDocumentModel interface: it doesn't expose the set of nodes, that don't belong to any component/element. This is a problem, because those nodes are needed by the connectors is some cases. When two connectors are connected, a new node is created at the connection point, and one connector is split in two, finally resulting in three connectors and one node. This way, in the model, a connector can always connect two nodes.

As I've seen in the CircuitModel class, the nodes had to be inserted in that class, too. 

Theoretically it's possible to deduce all nodes from the end-node-id-s of the connectors, but I'm not sure if that is the best solution.

So should be the nodes exposed in IDocumentModel, or not? What is your opinion?

 Zoltan



------------------------------------------------------------------------------
Sell apps to millions through the Intel(R) Atom(Tm) Developer Program
Be part of this innovative community and reach millions of netbook users 
worldwide. Take advantage of special opportunities to increase revenue and 
speed time-to-market. Join now, and jumpstart your future.
http://p.sf.net/sfu/intel-atom-d2d
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel
Julian Bäume | 24 Aug 2010 14:09
Picon
Gravatar

Re: Adding nodes() to IDocumentModel ?

Moin,

On Tuesday 24 August 2010 11:42:19 Zoltan Padrah wrote:
> As implementing the simulator, I've found a problem with the IDocumentModel
> interface: it doesn't expose the set of nodes, that don't belong to any
> component/element. This is a problem, because those nodes are needed by the
> connectors is some cases. When two connectors are connected, a new node is
> created at the connection point, and one connector is split in two, finally
> resulting in three connectors and one node. This way, in the model, a
> connector can always connect two nodes.
Yes, I’m facing this situation, too. So I would be fine with exposing this 
information from the IDocumentModel and sync to it, everytime something 
changes.

> As I've seen in the CircuitModel class, the nodes had to be inserted in
> that class, too.
It’s no problem to move this information one level higher into IDocumentModel.

> Theoretically it's possible to deduce all nodes from the end-node-id-s of
> the connectors, but I'm not sure if that is the best solution.
agreed!

> So should be the nodes exposed in IDocumentModel, or not? What is your
> opinion?
+1 from me

bye then
julian
------------------------------------------------------------------------------
Sell apps to millions through the Intel(R) Atom(Tm) Developer Program
Be part of this innovative community and reach millions of netbook users 
worldwide. Take advantage of special opportunities to increase revenue and 
speed time-to-market. Join now, and jumpstart your future.
http://p.sf.net/sfu/intel-atom-d2d
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel
Zoltan Padrah | 24 Aug 2010 17:57
Picon

Re: Adding nodes() to IDocumentModel ?

Change committed to GIT:


Note that also I've merged my repository's master branch with yours. This caused some conflicts, so please review the previous commit, too.



------------------------------------------------------------------------------
Sell apps to millions through the Intel(R) Atom(Tm) Developer Program
Be part of this innovative community and reach millions of netbook users 
worldwide. Take advantage of special opportunities to increase revenue and 
speed time-to-market. Join now, and jumpstart your future.
http://p.sf.net/sfu/intel-atom-d2d
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel
Julian Bäume | 24 Aug 2010 22:18
Picon
Gravatar

Re: Adding nodes() to IDocumentModel ?

hey,

On Tuesday 24 August 2010 17:57:32 Zoltan Padrah wrote:
> Change committed to GIT:
> 
> http://ktechlab.git.sourceforge.net/git/gitweb.cgi?p=ktechlab/ktl-zoltan_p;
> a=commit;h=e7bfb95c62b747f49b510ec8f215f31d6f166283
> 
> Note that also I've merged my repository's master branch with yours. This
> caused some conflicts, so please review the previous commit, too.

I merged your master and tested it. It crashes, when starting with a session 
containing opened files. This is, because the simulatormanager isn’t 
initialized. So in main.cpp the manager needs to be initialized before the 
core. (i. e. 3 lines above) This could cause trouble in the future, though, 
when the simulatormanager needs the core to be running. But I don’t see this 
use-case, yet.

Before merging your branch, I’d like another thing sorted out. There’s a lot 
of trailing-whitespace in your code. It’s not a big deal, but it would be 
nice, if master is clean. KDevelop is capable of removing it, when saving and 
kate can visualize it. You should enable these options. If you mention 
something like that before you push your changes, you can commit a fix and use 
git rebase -i to mark the fix as “fix-up”. This will merge these patches into 
one. However, this is not easily possible after changes have been pushed, 
because you can’t merge “fast-forward”. I’d like to do this before merging the 
KDE4 version into the official repo (and break our master branches with this…) 
but to limit this work, master should be as clean as possible.

Once we got this official master branch, we should do all this before merging 
to master. Since the branch will be deleted afterwards or won’t be used widely 
anyway, the “fast-forward” problem has little to none impact.

All in all, nice work, thanks! :) I hope to finish my routing branch, soon 
(still have some crashes, that need to be sorted out ;)) and after that, I 
will start implementing items that will be attached to ConnectorItems in a 
circuit and visualize the data from the simulator.

bye then
julian
------------------------------------------------------------------------------
Sell apps to millions through the Intel(R) Atom(Tm) Developer Program
Be part of this innovative community and reach millions of netbook users 
worldwide. Take advantage of special opportunities to increase revenue and 
speed time-to-market. Join now, and jumpstart your future.
http://p.sf.net/sfu/intel-atom-d2d
_______________________________________________
Ktechlab-devel mailing list
Ktechlab-devel@...
https://lists.sourceforge.net/lists/listinfo/ktechlab-devel

Gmane