Compiles and works. Archive includes compiler and library for building.
v0.004:
All build paths are relative now, you can extract the archive to any
desired location.
Patched source files to work around Borland compiler bug tickled by
NTVDM. Now you can build on WINXP, without DOSBox.
ed2k://|file|PCBoard%2015.3%20source%20code%20v0.004.zip|25893726|3F682E866794414A012F74295B335A78|h=AITR6VVLDS52J4FE7UED44VY3K5UYM3T|/
this include the pplc bits I added?
You've confirmed it works [builds] for you?
I was sitting here fixing the relative path thing in my diffs as you typed >this... :-)
You've confirmed it works [builds] for you?Yes.
Using \PCBS\ as you did, I don't think you ever built .\main, did you?
It was pretty messed up for a relative path build, as it was. Fixed now though.
If you can apply your pplc work to v0.004 and diff it, I will bundle it
into v0.005.
On Tuesday, February 26, 2013 9:04:47 PM UTC-5, Trifle Menot wrote:
You've confirmed it works [builds] for you?Yes.
At this point, I'm assuming you parsed 'it' as v0.004 vs. my PPLC bits.
Using \PCBS\ as you did, I don't think you ever built .\main, did you?
It was pretty messed up for a relative path build, as it was. Fixed now
though.
Both main and pplc build in my tree... I fixed paths to match my needs as
I went when I first built main close to a year ago.
If you can apply your pplc work to v0.004 and diff it, I will bundle it
into v0.005.
ACKd. Had to reinstall ed2k... waiting for it to download now.
If you can apply your pplc work to v0.004 and diff it, I will bundle it
into v0.005.
The build is easier now. I added a batch file to build all the library pieces. Please see src153.txt for new instructions. It also mentions pcbxdot. I planned to put that in the archive too, but that crashed dos pkzip (out of memory), and pcbxdot is still not in the archive. So now
I have a documentation bug in src153.txt. :-\
4) I got rid of the *.lib->*.386 stuff you were doing. This was just
and edit of your new batch script and a modification to the pcboard
makefile, compile.bat and renaming a file in MAIN/153. I'm going to
say we don't care about TC30 at this point and removed the conditional
the forced that rename. .LIB is more natural to linker/etc anyway.
At some point, I'll do further cleanup and likely just get rid of all
of the TC30 conditionals/directories/etc.
5) I had to patch LIB/SOURCE/COUNTRY to include an additional component for >PPLC.
5) I had to patch MAIN/SOURCE/PPL/SCRMISC.CPP to add another conditional for >PPLC.
LIB, MAIN and PPLC all build now. It's late, so I'll do a unified diff >tomorrow and validate it against a fresh copy of 0.004 then post it.
As an aside, in looking at PCBXDOT that you provided elsewhere, the contents >of pcbkit_l that I made for PPLC is very different than what they were >shipping. So, I probably should rename the pcbkit_l I make in the PPLC >makefile just for future posterity.
Their PCBKIT_L is simply most of LIB/SOURCE/TOOLKIT/ as a single library. >Their kit does include 2 things we don't appear to have source for: COMMDRV.OBJ
and FOSSIL.OBJ.
On Tue, 26 Feb 2013 21:54:23 -0800 (PST), Efudd <jasonbrent@gmail.com> wrote: >4) I got rid of the *.lib->*.386 stuff you were doing. This was just >and edit of your new batch script and a modification to the pcboard >makefile, compile.bat and renaming a file in MAIN/153. I'm going to >say we don't care about TC30 at this point and removed the conditional >the forced that rename. .LIB is more natural to linker/etc anyway. When I compiled with TC30 it produced an executable that crashed. So I abandoned that idea and focused on BC31. >At some point, I'll do further cleanup and likely just get rid of all >of the TC30 conditionals/directories/etc. If you want to. I wanted to have the whole thing building so I could hack on the source and see the result. I don't care how messy the build system is. >5) I had to patch LIB/SOURCE/COUNTRY to include an additional component for >PPLC. >5) I had to patch MAIN/SOURCE/PPL/SCRMISC.CPP to add another conditional for >PPLC. > >LIB, MAIN and PPLC all build now. It's late, so I'll do a unified diff >tomorrow and validate it against a fresh copy of 0.004 then post it. > >As an aside, in looking at PCBXDOT that you provided elsewhere, the contents >of pcbkit_l that I made for PPLC is very different than what they were >shipping. So, I probably should rename the pcbkit_l I make in the PPLC >makefile just for future posterity. > >Their PCBKIT_L is simply most of LIB/SOURCE/TOOLKIT/ as a single library. >Their kit does include 2 things we don't appear to have source for: COMMDRV.OBJ >and FOSSIL.OBJ. COMMDRV is the modem multiport board support. CDC omitted that from the source, we can never build it. Doesn't matter anyway, nobody's going to use PCBoard in a high density modem setup. There is code that provides fossil support for standard COM ports, I use it. Perhaps FOSSIL.OBJ was the fossil interface to COMMDRV, so that won't matter either.nice. I think commdrv was just thier custom digiboard driver. and how many are still using the 8 bit digigboards now. I can find the edonkey to download this.
I can find the edonkey to download this.
where is it?
1) I've patched in my PPLC changes.
2) I moved all of my changes to relative pathing.
3) I've done a teeny bit of cleanup in the pcboard mak file.
diff -Nru --ignore-file-name-case 004\lib 005\lib >lib.005
diff -Nru --ignore-file-name-case 004\main 005\main >main.005
pkzip -rP src153.zip @src153.lst
di.bat
dobc31.bat
src153.txt
b\c31\*.*
lib\codebase\*.*
pcbsrc\*.*
pcbsrcv\*.*
f) Delete lib and main from 004. We only need to keep 000, 001, and the
most recent version, in this case, 005.
di.bat
dobc31.bat
src153.txt
b\c31\*.*
lib\codebase\*.*
pcbsrc\cleanit.bat
pcbsrcv\*.*
*snip*
On Wednesday, February 27, 2013 2:14:41 PM UTC-5, Trifle Menot wrote:
*snip*
ed2k://|file|PCBoard%2015.3%20source%20code%20v0.005.zip|33549818|5EA0F9023E15B70C4CBAB8E073D7DD91|/
Let's see how this one goes. I've not shared via ed2k before, so hopefully that's a proper link....
I modified DOBC31.BAT to add a "root" variable. It defaults to empty.
VERSION = 153
ROOT = ..
PROJ = $(ROOT)\MAIN
LIBDIR = $(ROOT)\LIB\BCDOS\$(BCCOMPILER)
I tested it at root, under subdirectories and through directory renames. I >think all the scripts clean up properly in that case now.
del obj\pplc\\pcbkit_l tlib obj\pplc\\pcbkit_l + obj\pplc\\h2name TLIB 3.02 Copyright (c) 1992 Borland International
Warning: 'H2NAME' already in LIB, not changed!
tlib obj\pplc\\pcbkit_l + obj\pplc\\dosfopen TLIB 3.02 Copyright (c) 1992 Borland International
Warning: 'DOSFOPEN' already in LIB, not changed!
tlib obj\pplc\\pcbkit_l + obj\pplc\\dosfgets TLIB 3.02 Copyright (c) 1992 Borland International
Warning: 'DOSFGETS' already in LIB, not changed!
tlib obj\pplc\\pcbkit_l + obj\pplc\\crypt TLIB 3.02 Copyright (c) 1992 Borland International
I'm probably done with this for a few days now unless there is something >busted.
Having an environment variable of the same name, "root," suggests it can override the makefile value, but I know that is not the intent.
I tried a pplc build, got a path not found error, and some other
problems:
del obj\pplc\\pcbkit_l tlib obj\pplc\\pcbkit_l + obj\pplc\\h2name TLIB 3.02 Copyright (c) 1992 Borland International
Warning: 'H2NAME' already in LIB, not changed!
tlib obj\pplc\\pcbkit_l + obj\pplc\\dosfopen TLIB 3.02 Copyright (c) 1992 Borland International
Warning: 'DOSFOPEN' already in LIB, not changed!
tlib obj\pplc\\pcbkit_l + obj\pplc\\dosfgets TLIB 3.02 Copyright (c) 1992 Borland International
Take a break. I'll look at it closer tomorrow, and see what I can do.
Thanks for your work.
trifle - you opposed to git?
I loathe this 001/002/003...
At least for myself, I've moved this all to a unix base, I'm going to put the >revisions under git branches, i've lowercased everything and then just use samba
to mount on windows XP for builds (until I decide to try to do that via dosemu >or something later).
If you are opposed to git, then I'll just write a script on my side to >properly generate the current format upon release....
diff -Nru --ignore-file-name-case 004/MAIN/153/PCBOARD.MAK 005/MAIN/153/PCBOARD.MAK
--- 004/MAIN/153/PCBOARD.MAK 2013-02-26 23:46:00.000000000 -0500
+++ 005/MAIN/153/PCBOARD.MAK 2013-02-27 18:21:30.802593024 -0500
@@ -136,19 +136,13 @@
GEN = $(GEN) -DPCB_DEMO
!endif
-!if $d(386) && ! $d(TC30) # TC 3.0 does not support 386 compilation
COPT = $(COPT) -3 /DCPU386
ASMOPT = $(ASMOPT) /DCPU386
-EXT = 386
-LIBEXT = 386
-RES = PCB_$(BCCOMPILER).386
-!else
-EXT = RES
+EXT = LIB
LIBEXT = LIB
RES = PCB_$(BCCOMPILER).RES
-!endif
-LIBS = $(VERSION)\LIBS$(BCCOMPILER).$(EXT)
+LIBS = $(VERSION)\LIBS_$(BCCOMPILER).RES
#source code generation options
CODEOPT = -D$(GEN) -D$(COM) -D$(STATS) -D$(NUMNODES) -D$(DRV) -D$(DEFFIDO) -DKEY=KEY_$(NODES)
Seems like a good time to discuss our respective goals. We perhaps have
some in common, but not all.
On this project I'm like an archaeologist. I want to sift through the
ruins without disturbing them. Trying to clean up the build system is futile, it's a big mess that cannot be fixed. I don't want to obscure
its history with aesthetic changes that don't fix bugs. Such as:
..snip..
Don't get me wrong, I appreciate your work on pplc. But once I study it
and understand what you did, I may revise it and release v0.006.
di.bat and dobc31.bat worked well for me, but not when distributed to
other environments. So now I've merged them into one file, pcbsrc.baz,
which is copied to pcbsrc.bat where local changes are kept.
How to proceed, I'm not sure. One step at a time I guess ...
Fair enough. Not sure my goals. For the 15+ years I was seeking the source, >my mental model was "pull it current!". Now my goal is just to hack on it.
As an aside, I read earlier threads re: copyright. I ended up finding
some mention that fred had tried to sell the source rights for 40k under >another company name post-bankrupcy. Then other indications that Zion bank >owned the rights... did you ever find out anything any different?
Zero intention here other freely sharing code
just trying to evaluate personal liability when sharing ~18 year old
dead code from a dead company with an effectively dead user base.
Personally I'd really love to see a drop-in data file compatible version
with native tcp/ip and no comm support for modern OSes exist. The one person >who cares probably would enjoy it anyway. :-)
On Fri, 1 Mar 2013 07:49:34 -0800 (PST), Efudd <jasonbrent@gmail.com> wrote: >Fair enough. Not sure my goals. For the 15+ years I was seeking the source, >my mental model was "pull it current!". Now my goal is just to hack on it. Trying to do more than that would be overly ambitious. It required a dedicated team some years to create it. It would require much more than a few isolated enthusiasts to renovate it. >As an aside, I read earlier threads re: copyright. I ended up finding >some mention that fred had tried to sell the source rights for 40k under >another company name post-bankrupcy. Then other indications that Zion bank >owned the rights... did you ever find out anything any different? No. >Zero intention here other freely sharing code "Freely sharing" is not possible, considering the license of PCBoard, bcc, and codebase. That's one reason I'm not interested in maintaining a git repo that might start a party. Better to stay under the radar. >just trying to evaluate personal liability when sharing ~18 year old >dead code from a dead company with an effectively dead user base. I doubt CDC/Fred will ever reappear to assert copyright. The license to codebase is a question I have no desire to probe. >Personally I'd really love to see a drop-in data file compatible version >with native tcp/ip and no comm support for modern OSes exist. The one person >who cares probably would enjoy it anyway. :-) An interesting goal, if only in our dreams. :-)I might like git. the jackass just sits for hours and never gets my files. geez, I hope no one ever buys the rights to pcboard. I think I am the only one that ever bought a source license. so everything points to me as letting the source out. good thing it is just a hobby system now. I dought any companys are still using it anymore. clark used to make special versions for people, I heard. even the warez people dont want it anymore. can you send me the updated zip to my email?
Trying to do more than that would be overly ambitious. It required a dedicated team some years to create it. It would require much more thanYea, likely true...
a few isolated enthusiasts to renovate it.
"Freely sharing" is not possible, considering the license of PCBoard,Oh. I meant a .git repository in the released source.. those who know how
bcc, and codebase. That's one reason I'm not interested in maintaining
a git repo that might start a party. Better to stay under the radar.
I doubt CDC/Fred will ever reappear to assert copyright. The license to codebase is a question I have no desire to probe.Agreed. codebase isn't a huge concern to me. I haven't looked, but I'm betting it wouldn't be too difficult to replace the codebase calls with something free-er like sqlite.
Indeed. I'm sure my ADD will kick in soon and I'll switch to something else again for a few months...Personally I'd really love to see a drop-in data file compatible version >with native tcp/ip and no comm support for modern OSes exist. The one person >who cares probably would enjoy it anyway. :-)
An interesting goal, if only in our dreams. :-)
the jackass just sits for hours and never gets my files
can you send me the updated zip to my email?
I might like git. the jackass just sits for hours and never gets my files.
geez, I hope no one ever buys the rights to pcboard.
I think I am the only one that ever bought a source license.
can you send me the updated zip to my email?
I meant a .git repository in the released source.. those who know how
to use it can.. those who don't, just stick with the current tree. :-)
codebase isn't a huge concern to me. I haven't looked, but I'm betting
it wouldn't be too difficult to replace the codebase calls with something >free-er like sqlite.
ed2k is weird...
??[efudd@wizard] - [~/Projects/pcboard] - [Fri Mar 01, 01:13]
??[$] <git:(feature/build_cleanup*)> git tag
v0.000
v0.001
v0.002
v0.003
v0.004
v0.005
v0.005a
TBH, I screwed up directory structures before v0.005a... :-)
If you have a new release before I do, go ahead and use v.006. If you
beat me to it, I'll just bump to v0.007 ...
But let's not consume numbers for every little change, rather, queue up
a batch of changes before releasing. Say, at least a week's worth.
On Fri, 1 Mar 2013 10:06:17 -0800 (PST), VinylDoctor <vinyldoctorlv@gmail.com> wrote: >the jackass just sits for hours and never gets my files The jackass? LOL. You must have something misconfigured. I got v0.005 within an hour from efudd. Test your port setup to be sure you can talk to the outside world. >can you send me the updated zip to my email? I've already got updates queued for v0.006. When that's ready, I will put it on 4shared if you still can't ride the donkey. :-Dok thanks, I will wait for the 4shared
.. long thread is long ... pcbdiag now builds and runs. It made me realize there is the potential for hard to track run time failures if you link against the wrong thing. For example, there is a function to read the pcboard data file... one in library toolkit and one in library pcb. If you link against the one in toolkit... it doesn't error.. and fills in SOME of the appropriate data structures. If you link against the one in pcb, everything works as one would hope. It looks like the LIB/SOURCE/TOOLKIT were basically stub versions of many functions that were implemented elsewhere. Those stub versions were what got shipped to developers writing software to add on to PCBoard... at least, that is what it appears. Also, it seems that the version of code this is does NOT match the shipping release of 15.3. For example, there is a simple conditional bug in the diag utility code that prevents it from ever returning how the shipping diag utility ran.... -eyes. I think the toolkit was not needed for pcboard. it was like you said, for developers so they could see how to write addons for pcboard. it's just in my stuff because thats what they sent me on disc when I bought my license.
Sysop: | DaiTengu |
---|---|
Location: | Appleton, WI |
Users: | 991 |
Nodes: | 10 (0 / 10) |
Uptime: | 120:08:59 |
Calls: | 12,958 |
Files: | 186,574 |
Messages: | 3,265,642 |