What debugger should I download and install for debugging the Windows VM? (I have a bug in my VI4 version). I downloaded the latest source and compile tools from squeak.sourceforge.net (Thank you Andreas Raab!). I downloaded and tried Cygwin Insight gdb 5.0 from http://www.objectcentral.com/downloads.htm, but it complains that "this executable has no debugging information" even though my VM compiled successfully with the -g option included (the default in Andreas makefile). I also downloaded the gdb source from http://www.gnu.org/software/gdb but to compile it on Windows the Readme said I needed the DJGPP development environment. I this is what I will have to do? Is there an easier way? Why doesn't Cygwin Insight gdb 5.0 work?
Thanks, Anthony
P.S. I'm not a Windows programmer.
Hello, Anthony,
While I've never done the Windows VM debugging under gdb, the "strip" command invocation in Makefile.mingw32 looks suspicious to me.
-- Yoshiki
Anthony Hannan wrote:
What debugger should I download and install for debugging the Windows VM? (I have a bug in my VI4 version). I downloaded the latest source and compile tools from squeak.sourceforge.net (Thank you Andreas Raab!). I downloaded and tried Cygwin Insight gdb 5.0 from http://www.objectcentral.com/downloads.htm, but it complains that "this executable has no debugging information" even though my VM compiled successfully with the -g option included (the default in Andreas makefile). I also downloaded the gdb source from http://www.gnu.org/software/gdb but to compile it on Windows the Readme said I needed the DJGPP development environment. I this is what I will have to do? Is there an easier way? Why doesn't Cygwin Insight gdb 5.0 work?
Thanks, Anthony
P.S. I'm not a Windows programmer.
You should be able to use gdb and the tk frontend that comes with cygwin (is this Insight?). I was able to debug the mpegPlugin with it with few problems. Make sure not to have make strip your exes/dlls
A break from the ideological wars -- here's what may be a real (at least less controversial) bug.
I seem to be unable to upgrade the most current published version of Squeak 3.2 to 3.3. Am I missing something?
1. Take a raw 3.2 image 2. DoIt: "SqueakVersion setImage" 3. Help -> updates
This loads a few updates, but then stops on or about update 4681.
Please advise where I am missing the boat, or what I should do to effect the update to current alpha using the updaters.
Apologize -- step 2, I changed "Squeak3.2" to "Squeak3.3alpha"
On Wednesday, August 28, 2002, at 09:41 PM, Andrew C. Greenberg wrote:
A break from the ideological wars -- here's what may be a real (at least less controversial) bug.
I seem to be unable to upgrade the most current published version of Squeak 3.2 to 3.3. Am I missing something?
- Take a raw 3.2 image
- DoIt: "SqueakVersion setImage"
- Help -> updates
This loads a few updates, but then stops on or about update 4681.
Please advise where I am missing the boat, or what I should do to effect the update to current alpha using the updaters.
"Andrew C. Greenberg" werdna@mucow.com asked...
I seem to be unable to upgrade the most current published version of Squeak 3.2 to 3.3. Am I missing something?
Yes.
- Take a raw 3.2 image
- DoIt: "SqueakVersion setImage"
- Help -> updates
This loads a few updates, but then stops on or about update 4681.
Please advise where I am missing the boat, or what I should do to effect the update to current alpha using the updaters.
A choice was offered, at update #4653, to advance to 3.3 and the maintainer of the 3.2 fork chose NO. There are no provisions for advancing to 3.3 after that point.
It's true that this differs from most earlier releases, because 3.2 continued to evolve after the fork, rather than merely receiving a few fixes that might have allowed for later advancing to 3.3.
It is conceivable that you could fool your system into thinking that it was at update 4653 (remove or rename all changeSets after this #), *then* change your version and go forward from there, but I have no certainty that it would be successful.
Hope this helps.
- Dan
Hi Dan,
I thought that Squeak Central was in control of the update process. Who exactly is the maintainer of the 3.2 fork and why did this fork take place? Is Squeak Central now out of the loop?
Cheers, Roger.....
On Friday, August 30, 2002, at 07:57 PM, Dan Ingalls wrote:
"Andrew C. Greenberg" werdna@mucow.com asked...
I seem to be unable to upgrade the most current published version of Squeak 3.2 to 3.3. Am I missing something?
Yes.
- Take a raw 3.2 image
- DoIt: "SqueakVersion setImage"
- Help -> updates
This loads a few updates, but then stops on or about update 4681.
Please advise where I am missing the boat, or what I should do to effect the update to current alpha using the updaters.
A choice was offered, at update #4653, to advance to 3.3 and the maintainer of the 3.2 fork chose NO. There are no provisions for advancing to 3.3 after that point.
It's true that this differs from most earlier releases, because 3.2 continued to evolve after the fork, rather than merely receiving a few fixes that might have allowed for later advancing to 3.3.
It is conceivable that you could fool your system into thinking that it was at update 4653 (remove or rename all changeSets after this #), *then* change your version and go forward from there, but I have no certainty that it would be successful.
Hope this helps.
- Dan
"Roger Vossler" rvossler@qwest.net asked...
I thought that Squeak Central was in control of the update process. Who exactly is the maintainer of the 3.2 fork and why did this fork take place? Is Squeak Central now out of the loop?
Hi, Roger -
Sorry about the misleading use of words. That maintainer was me. I had originally worded my reply as though Andrew had been maintaining his own image, and it said something like "you chose NO". Then I realized this is the case for any 3.2, so I changed it to this impersonal reference to the election of the fork to 3.2 in update 4653.
And to answer the bigger question, no, SQC is not out of the loop (as far as we know ;-). On the contrary I expect various of us to be more active again as we get into the fall. I'll try to write down my thoughts about some of the current activities and how they might play together in a coherent manner as we go forward.
- Dan --------------------------
On Friday, August 30, 2002, at 07:57 PM, Dan Ingalls wrote:
"Andrew C. Greenberg" werdna@mucow.com asked...
I seem to be unable to upgrade the most current published version of Squeak 3.2 to 3.3. Am I missing something?
Yes.
- Take a raw 3.2 image
- DoIt: "SqueakVersion setImage"
- Help -> updates
This loads a few updates, but then stops on or about update 4681.
Please advise where I am missing the boat, or what I should do to effect the update to current alpha using the updaters.
A choice was offered, at update #4653, to advance to 3.3 and the maintainer of the 3.2 fork chose NO. There are no provisions for advancing to 3.3 after that point.
It's true that this differs from most earlier releases, because 3.2 continued to evolve after the fork, rather than merely receiving a few fixes that might have allowed for later advancing to 3.3.
It is conceivable that you could fool your system into thinking that it was at update 4653 (remove or rename all changeSets after this #), *then* change your version and go forward from there, but I have no certainty that it would be successful.
Hope this helps.
- Dan
Hi Dan,
I'd be particularly interested in your thoughts about how block closures, and a number of related issues, are going to be integrated into Squeak and when. Also, there was another surge of interest in modules in June (305 messages) which seemed to have decayed off rather rapidly (49 messages in July and 6 messages in August). While everyone may be either on vacation or furiously writing code, I'd like to hear your thoughts on where the modules effort is headed, when it is likely to happen, and what the final (or interim) results are likely to look like. Comments from others would also be welcome.
Cheers, Roger.....
On Saturday, August 31, 2002, at 05:01 AM, Dan Ingalls wrote:
"Roger Vossler" rvossler@qwest.net asked...
I thought that Squeak Central was in control of the update process. Who exactly is the maintainer of the 3.2 fork and why did this fork take place? Is Squeak Central now out of the loop?
Hi, Roger -
Sorry about the misleading use of words. That maintainer was me. I had originally worded my reply as though Andrew had been maintaining his own image, and it said something like "you chose NO". Then I realized this is the case for any 3.2, so I changed it to this impersonal reference to the election of the fork to 3.2 in update > 4653.
And to answer the bigger question, no, SQC is not out of the loop (as far as we know ;-). On the contrary I expect various of us to be more active again as we get into the fall. I'll try to write down my thoughts about some of the current activities and how they might play together in a coherent manner as we go forward.
- Dan
[snip]
Roger:
There has been an extensive amount of work on modules integrated into the image. You should review the hundreds of changesets that comprise the present module system in the 3.3alpha testpilot drive -- there is extensive documentation on the Swiki as well.
On Sunday, September 1, 2002, at 06:16 PM, Roger Vossler wrote:
Hi Dan,
I'd be particularly interested in your thoughts about how block closures, and a number of related issues, are going to be integrated into Squeak and when. Also, there was another surge of interest in modules in June (305 messages) which seemed to have decayed off rather rapidly (49 messages in July and 6 messages in August). While everyone may be either on vacation or furiously writing code, I'd like to hear your thoughts on where the modules effort is headed, when it is likely to happen, and what the final (or interim) results are likely to look like. Comments from others would also be welcome.
Cheers, Roger.....
On Saturday, August 31, 2002, at 05:01 AM, Dan Ingalls wrote:
"Roger Vossler" rvossler@qwest.net asked...
I thought that Squeak Central was in control of the update process. Who exactly is the maintainer of the 3.2 fork and why did this fork take place? Is Squeak Central now out of the loop?
Hi, Roger -
Sorry about the misleading use of words. That maintainer was me. I had originally worded my reply as though Andrew had been maintaining his own image, and it said something like "you chose NO". Then I realized this is the case for any 3.2, so I changed it to this impersonal reference to the election of the fork to 3.2 in update > 4653.
And to answer the bigger question, no, SQC is not out of the loop (as far as we know ;-). On the contrary I expect various of us to be more active again as we get into the fall. I'll try to write down my thoughts about some of the current activities and how they might play together in a coherent manner as we go forward.
- Dan
[snip]
There has been an extensive amount of work on modules integrated into the image. You should review the hundreds of changesets that comprise the present module system in the 3.3alpha testpilot drive -- there is extensive documentation on the Swiki as well.
I think Andreas is on something really substantial.
His SqueakScript image is only the tip of an ice-berge.
We can expect a greater SqueakScript image real soon now. And by greater I mean smaller and bootstrap-able. And extensible to a minimal morphic Squeak image. Where MVC gets its death sentence. As well as something like 99% of the morphs.
In his own words, 'I am busy like hell'.
Let's hope that he's been busy chopping up the morphs. And when he has a random cycle to spare. Then maybe we will hear all these from the horse's mouth.
Cheers,
PhiHo.
----- Original Message ----- From: "Andrew C. Greenberg" werdna@mucow.com To: squeak-dev@lists.squeakfoundation.org Sent: Sunday, September 01, 2002 9:29 PM Subject: Re: [BUG?]Upgrading 3.2 to 3.3alpha
Roger:
There has been an extensive amount of work on modules integrated into the image. You should review the hundreds of changesets that comprise the present module system in the 3.3alpha testpilot drive -- there is extensive documentation on the Swiki as well.
On Sunday, September 1, 2002, at 06:16 PM, Roger Vossler wrote:
Hi Dan,
I'd be particularly interested in your thoughts about how block closures, and a number of related issues, are going to be integrated into Squeak and when. Also, there was another surge of interest in modules in June (305 messages) which seemed to have decayed off rather rapidly (49 messages in July and 6 messages in August). While everyone may be either on vacation or furiously writing code, I'd like to hear your thoughts on where the modules effort is headed, when it is likely to happen, and what the final (or interim) results are likely to look like. Comments from others would also be welcome.
Cheers, Roger.....
On Saturday, August 31, 2002, at 05:01 AM, Dan Ingalls wrote:
"Roger Vossler" rvossler@qwest.net asked...
I thought that Squeak Central was in control of the update process. Who exactly is the maintainer of the 3.2 fork and why did this fork take place? Is Squeak Central now out of the loop?
Hi, Roger -
Sorry about the misleading use of words. That maintainer was me. I had originally worded my reply as though Andrew had been maintaining his own image, and it said something like "you chose NO". Then I realized this is the case for any 3.2, so I changed it to this impersonal reference to the election of the fork to 3.2 in update > 4653.
And to answer the bigger question, no, SQC is not out of the loop (as far as we know ;-). On the contrary I expect various of us to be more active again as we get into the fall. I'll try to write down my thoughts about some of the current activities and how they might play together in a coherent manner as we go forward.
- Dan
[snip]
Roger wrote,
Hi Dan,
I'd be particularly interested in your thoughts about how block closures, and a number of related issues, are going to be integrated into Squeak .
Apparently, Anthony's work ran into some troubles in debugging C codes with a C debugger.
It's a pity that the Interpreter Simulator is not working with Squeak 3.2, otherwise the interpreter could have been debugged in Slang codes with the Squeak debugger. Isn't that the whole idea of writing the interpreter in C-ized Smalltalk ?
Besides, if a Squeak VM can not be debugged inside a Squeak image then it's not Squeak anymore, right ?
Also, there was another surge of interest in modules in June (305 messages) which seemed to have decayed off rather rapidly (49 messages in July and 6 messages in August). While everyone may be either on vacation or furiously writing code, I'd like to hear your thoughts on where the modules effort is headed, when it is likely to happen, and what the final (or interim) results are likely to look like.
If by some miracles, the modularizing efforts fail, we can always fall back to the latest and greatest Majorshrink script that Dan promised (?). I believe that this latest and greatest script will give us the smallest shrinked image we ever had.
Cheers,
PhiHo.
----- Original Message ----- From: "Roger Vossler" rvossler@qwest.net To: squeak-dev@lists.squeakfoundation.org Sent: Sunday, September 01, 2002 6:16 PM Subject: Re: [BUG?]Upgrading 3.2 to 3.3alpha
Hi Dan,
I'd be particularly interested in your thoughts about how block closures, and a number of related issues, are going to be integrated into Squeak and when. Also, there was another surge of interest in modules in June (305 messages) which seemed to have decayed off rather rapidly (49 messages in July and 6 messages in August). While everyone may be either on vacation or furiously writing code, I'd like to hear your thoughts on where the modules effort is headed, when it is likely to happen, and what the final (or interim) results are likely to look like. Comments from others would also be welcome.
Cheers, Roger.....
On Saturday, August 31, 2002, at 05:01 AM, Dan Ingalls wrote:
"Roger Vossler" rvossler@qwest.net asked...
I thought that Squeak Central was in control of the update process. Who exactly is the maintainer of the 3.2 fork and why did this fork take place? Is Squeak Central now out of the loop?
Hi, Roger -
Sorry about the misleading use of words. That maintainer was me. I had originally worded my reply as though Andrew had been maintaining his own image, and it said something like "you chose NO". Then I realized this is the case for any 3.2, so I changed it to this impersonal reference to the election of the fork to 3.2 in update >
4653.
And to answer the bigger question, no, SQC is not out of the loop (as far as we know ;-). On the contrary I expect various of us to be more active again as we get into the fall. I'll try to write down my thoughts about some of the current activities and how they might play together in a coherent manner as we go forward.
- Dan
[snip]
Apparently, Anthony's work ran into some troubles indebugging C codes with a C debugger.
We fixed this problem today. It turned out that there was a slight mismatch in boundary conditions with #to:do: loops - Squeak evaluates the upper bound only once whereas the equivalent C code (using for(;;)) may do this multiple times. In situations like these you just have to resort to a C level debugger since the simulator ran the same code perfectly nice.
Cheers, - Andreas
On Friday, August 30, 2002, at 10:57 PM, Dan Ingalls wrote:
A choice was offered, at update #4653, to advance to 3.3 and the maintainer of the 3.2 fork chose NO. There are no provisions for advancing to 3.3 after that point.
Perhaps a #4652-level Squeak should be available, with a README describing the situation and a HOWTO for building a v.3 from a v.2?
squeak-dev@lists.squeakfoundation.org
-
Andreas Raab -
Andrew C. Greenberg -
Anthony Hannan -
Dan Ingalls -
Jason Dufair -
Ohshima, Yoshiki -
PhiHo Hoang -
Roger Vossler