Can we get back to the bugs please?
by Gonzo T. Clown · in Torque Game Engine · 02/18/2004 (12:51 pm) · 12 replies
It's a little discouraging to see that bugs going back to Jul 17, 2002 have not been resolved or haven't been updated in this amount of time. Not to be be rude but, there really is no excuse for it. At the very least there needs to be a review to find which bugs are still issues and which ones have been resolved and get the list cleaned up.
www.garagegames.com/projectmanager/module.php?qc=4415&qs=256&qt=1&qst=15&qmo=28&...
I myself have submitted two repairs to the Demo & engine and would like to see them implemented or rejected before I waste any more time doing more.
I think Torque is a great engine, and would just like to see things get caught up so that all of us can move on with confidence that Torque is not completely full of bugs.
[/endofrant]
www.garagegames.com/projectmanager/module.php?qc=4415&qs=256&qt=1&qst=15&qmo=28&...
I myself have submitted two repairs to the Demo & engine and would like to see them implemented or rejected before I waste any more time doing more.
I think Torque is a great engine, and would just like to see things get caught up so that all of us can move on with confidence that Torque is not completely full of bugs.
[/endofrant]
About the author
#2
www.garagegames.com/projectmanager/issue.php?qis=292
www.garagegames.com/projectmanager/issue.php?qis=291
www.garagegames.com/projectmanager/issue.php?qis=290
www.garagegames.com/projectmanager/issue.php?qis=288
www.garagegames.com/projectmanager/issue.php?qis=295
(Partially fixed, need a VC++ 7.0 user to finish other half)
www.garagegames.com/projectmanager/issue.php?qis=267
02/19/2004 (2:26 am)
Ok, just so nobody thinks I'm a hypocrite, I have addressed and solved these issues myself....www.garagegames.com/projectmanager/issue.php?qis=292
www.garagegames.com/projectmanager/issue.php?qis=291
www.garagegames.com/projectmanager/issue.php?qis=290
www.garagegames.com/projectmanager/issue.php?qis=288
www.garagegames.com/projectmanager/issue.php?qis=295
(Partially fixed, need a VC++ 7.0 user to finish other half)
www.garagegames.com/projectmanager/issue.php?qis=267
#3
But there are a few tasks/bugs which are already solved in the latest head. The only problem is, that they are still marked as open.
But of course it all needs time. GG has to look at the submitted patches and test it. This needs some time.
02/19/2004 (3:22 am)
I'm not going to say that you aren't right.But there are a few tasks/bugs which are already solved in the latest head. The only problem is, that they are still marked as open.
But of course it all needs time. GG has to look at the submitted patches and test it. This needs some time.
#4
We really appreciate any work the community does and submits to us. The biggest problem is that we simply do not have the time or manpower to address everything... even completed patches. Every one of these patches are for bugs that have a stated priority of "Unimportant" or "Minor".
I always think of GarageGames like this: we are emergency medical technicians coming up to a massive multi car wreck on the freeway. Some people are bleeding profusely from the neck, some have broken bones protruding from their thighs, others look broken in half. There are hundreds, maybe even thousands of people wailing and crying from the pain and anguish. We do the only thing we can. We grab our medical bags and start fixing people. Of course, we have to go for the ones bleeding from the neck first to keep them from dying, then we move on to the next level down, etc. After the bleeders are taken care of, we all step back and start building a first aid tent because it will help get everybody fixed in a faster, more efficient manner. However, the people with broken bones or bruises that need attention might not think that is the best thing to do. In the medical field, this practice is called TRIAGE, and it is the only thing we can do.
We absolutely care about all of the issues that people bring up and truly appreciate the contributions from the community. Believe me, when you post came up yesterday, I passed it around the office and we all debated for a while. Your [rant] brought the issue up to the surface, and we all felt really bad.
Believe it or not, we do have plans to solve this issue. However, you have to trust that we are doing the right thing. Sometimes we cannot talk about what is going on behind the scenes, but it will come to light eventually, and hopefully, it will make every one happy.
The part of the answer to this one that I can talk about is that we need to have more maintainers that have the ability to write into CVS. Before we can do that, however, we need a more formal testing procedure. We are open to all of this, and have been talking to some of the community leaders in the background to make it happen. If anybody is interested in becoming a maintainer, let me know at jefft@garagegames.com.
-Jeff Tunnell GG
02/19/2004 (9:28 am)
Kenneth,We really appreciate any work the community does and submits to us. The biggest problem is that we simply do not have the time or manpower to address everything... even completed patches. Every one of these patches are for bugs that have a stated priority of "Unimportant" or "Minor".
I always think of GarageGames like this: we are emergency medical technicians coming up to a massive multi car wreck on the freeway. Some people are bleeding profusely from the neck, some have broken bones protruding from their thighs, others look broken in half. There are hundreds, maybe even thousands of people wailing and crying from the pain and anguish. We do the only thing we can. We grab our medical bags and start fixing people. Of course, we have to go for the ones bleeding from the neck first to keep them from dying, then we move on to the next level down, etc. After the bleeders are taken care of, we all step back and start building a first aid tent because it will help get everybody fixed in a faster, more efficient manner. However, the people with broken bones or bruises that need attention might not think that is the best thing to do. In the medical field, this practice is called TRIAGE, and it is the only thing we can do.
We absolutely care about all of the issues that people bring up and truly appreciate the contributions from the community. Believe me, when you post came up yesterday, I passed it around the office and we all debated for a while. Your [rant] brought the issue up to the surface, and we all felt really bad.
Believe it or not, we do have plans to solve this issue. However, you have to trust that we are doing the right thing. Sometimes we cannot talk about what is going on behind the scenes, but it will come to light eventually, and hopefully, it will make every one happy.
The part of the answer to this one that I can talk about is that we need to have more maintainers that have the ability to write into CVS. Before we can do that, however, we need a more formal testing procedure. We are open to all of this, and have been talking to some of the community leaders in the background to make it happen. If anybody is interested in becoming a maintainer, let me know at jefft@garagegames.com.
-Jeff Tunnell GG
#5
I'd personally rather see the GG guys work on TGE 2.0 than fix minor issues in TGE 1.2. The TGE engine is good, infact its the best for indies that I can see out there. You want to talk bugs? I worked with unreal for about 3 years. Its a great engine, but they had major bugs that required alot of work arounds. The client server stuff worked about half the time. Anything you put in their replication function that had variables set by the ancestor objects was almost guaranteed to give you headaches. They sell that engine for tons of cash. And even with the bugs unreal is still a good engine.
In my opinion TGE is cleaner and has less bugs. Its also cheaper and you get personal attention from the president of GG when you make a post like this. I would be personally very upset if the GG guys went back and took the time to support TGE 1.2. That would push the release date of 2.0 even further out. TGE 1.2 is pretty stable and we have plenty of community members capable of supporting, enhancing and testing it. Leave the GG guys alone. They are busy coming up with a version of TGE that will compete with the modern engines and provide it to the little guys like us. In my opinion this is way more important than supporting an older code base.
Its hard for the indies to commit full time to upgrading patching and testing, but it can be done. Since you fixed many of these bugs in your patch I'd hope and expect you to contact Jeff and offer your services.
02/19/2004 (10:02 am)
Kenneth,I'd personally rather see the GG guys work on TGE 2.0 than fix minor issues in TGE 1.2. The TGE engine is good, infact its the best for indies that I can see out there. You want to talk bugs? I worked with unreal for about 3 years. Its a great engine, but they had major bugs that required alot of work arounds. The client server stuff worked about half the time. Anything you put in their replication function that had variables set by the ancestor objects was almost guaranteed to give you headaches. They sell that engine for tons of cash. And even with the bugs unreal is still a good engine.
In my opinion TGE is cleaner and has less bugs. Its also cheaper and you get personal attention from the president of GG when you make a post like this. I would be personally very upset if the GG guys went back and took the time to support TGE 1.2. That would push the release date of 2.0 even further out. TGE 1.2 is pretty stable and we have plenty of community members capable of supporting, enhancing and testing it. Leave the GG guys alone. They are busy coming up with a version of TGE that will compete with the modern engines and provide it to the little guys like us. In my opinion this is way more important than supporting an older code base.
Its hard for the indies to commit full time to upgrading patching and testing, but it can be done. Since you fixed many of these bugs in your patch I'd hope and expect you to contact Jeff and offer your services.
#6
Kenneth has kindly submitted quite a few patches. Tim has felt guilty for not implementing, testing and releasing those patches.
-Jeff Tunnell GG
EDIT: The biggest problem we have is that testing, integrating, making sure the patch is done correctly, etc. takes a lot of time. If we can come up with a community certificatoin program, then things could progress as a much more rapid pace. Any suggestions?
02/19/2004 (10:43 am)
Don't leave us alone! Let's just work together to come up with a system that works for everybody. Kenneth has kindly submitted quite a few patches. Tim has felt guilty for not implementing, testing and releasing those patches.
-Jeff Tunnell GG
EDIT: The biggest problem we have is that testing, integrating, making sure the patch is done correctly, etc. takes a lot of time. If we can come up with a community certificatoin program, then things could progress as a much more rapid pace. Any suggestions?
#7
I suggest we have a new community version in CVS and some people in the community who are responsible for managing it. The managers have cvs write permission to the repository. When submitting a patch the patch submitter needs to supply some basic test cases. In addition we should have some standard regression tests that can be checked off.
At some fixed interval the managers get together and compile a new version on each platform based on the patches and if they have time new resource submissions and distribute it to the community. The test cases submitted by the patch owner are tested by the community members as well as the regression test cases. Anyone can log into this new forum and submit bugs they find. In addition each test case can be check off with a pass or fail, the torque user perfoming the test case is logged along with remarks if it fails. The release will not be made until each test case has been performed and signed off on by a GG community member. Again any one who is a member here at GG can be a tester.
When bugs are found and test cases fail the managers will look into the issues and find out where the error came from. If neccessary they will roll back the patch and or new submissions. They will contact the submitters and explain the issues. After a second round of tests after the patches are rolled back if their are failures the managers will have to vote go / no go. Because we have limited time I suggest if its a no go we post the bugs and wait until the next interval to see if we can fix the issues.
The interval can be every other month or once a quarter or what ever is deemed reasonable. We would need some regression tests from the GG employees, what ever you guys do on a regular basis. The patch submission's regression tests would be added to the regression tests as well.
I see the regression test list being stored in a DB so the managers should be able to update the test case list as well through some php screen. I'd like to suggest that the more managers the better. Time schedules will force some people away so the more managers we have the more likely it will be done on a long term. There are plenty of people here that would make great managers. Its easy to pick them out they too. They have submitted patches, resources and much more and have consistently been here awhile.
I think if you Jeff, or Jay asked them to join they would be more than willing to accept the responsibility.
02/19/2004 (11:38 am)
Yep,I suggest we have a new community version in CVS and some people in the community who are responsible for managing it. The managers have cvs write permission to the repository. When submitting a patch the patch submitter needs to supply some basic test cases. In addition we should have some standard regression tests that can be checked off.
At some fixed interval the managers get together and compile a new version on each platform based on the patches and if they have time new resource submissions and distribute it to the community. The test cases submitted by the patch owner are tested by the community members as well as the regression test cases. Anyone can log into this new forum and submit bugs they find. In addition each test case can be check off with a pass or fail, the torque user perfoming the test case is logged along with remarks if it fails. The release will not be made until each test case has been performed and signed off on by a GG community member. Again any one who is a member here at GG can be a tester.
When bugs are found and test cases fail the managers will look into the issues and find out where the error came from. If neccessary they will roll back the patch and or new submissions. They will contact the submitters and explain the issues. After a second round of tests after the patches are rolled back if their are failures the managers will have to vote go / no go. Because we have limited time I suggest if its a no go we post the bugs and wait until the next interval to see if we can fix the issues.
The interval can be every other month or once a quarter or what ever is deemed reasonable. We would need some regression tests from the GG employees, what ever you guys do on a regular basis. The patch submission's regression tests would be added to the regression tests as well.
I see the regression test list being stored in a DB so the managers should be able to update the test case list as well through some php screen. I'd like to suggest that the more managers the better. Time schedules will force some people away so the more managers we have the more likely it will be done on a long term. There are plenty of people here that would make great managers. Its easy to pick them out they too. They have submitted patches, resources and much more and have consistently been here awhile.
I think if you Jeff, or Jay asked them to join they would be more than willing to accept the responsibility.
#8
"Every one of these patches are for bugs that have a stated priority of "Unimportant" or "Minor". "
How can you have a bug that is "Priority - Minor (or unimportant)" and yet have "Severity - CRITICAL"? Isn't that a bit of a conflict? I don't see crashing to desktop as unimportant, and neither do quite a few others. I didn't fix them cause they were easy, I fixed them because I went down the line and started grabbing them in the order working from most recent to oldest because I could tell that some of the older ones had been addressed but not removed from the list.
All I'm asking from GG is to have the list updated to reflect the more current status of things as they are today. And I'm challenging the community to come forth and help you guys as I myself am trying to do. I'm not on a crusade or a personal mission, lol. I simply think that this community has the power to address all issues in a timely fashion, although they may not have the drive. I would hate to see v2.0 get to this condition in the future when there is no reason for it to. Can anyone here deny that between the GG staff and the community that purchased the SDK that we have enough programmers to fix everything? I am certainly willing to do any amount of testing it takes to confirm issues both exist and/or have been fixed.
02/19/2004 (12:34 pm)
Jeff, first off let me restate that I love Torque, I love you guys for making Torque, I'd buy it all over again if I had to. Nothing would change that. With that in mind I would like to address this statement..."Every one of these patches are for bugs that have a stated priority of "Unimportant" or "Minor". "
How can you have a bug that is "Priority - Minor (or unimportant)" and yet have "Severity - CRITICAL"? Isn't that a bit of a conflict? I don't see crashing to desktop as unimportant, and neither do quite a few others. I didn't fix them cause they were easy, I fixed them because I went down the line and started grabbing them in the order working from most recent to oldest because I could tell that some of the older ones had been addressed but not removed from the list.
All I'm asking from GG is to have the list updated to reflect the more current status of things as they are today. And I'm challenging the community to come forth and help you guys as I myself am trying to do. I'm not on a crusade or a personal mission, lol. I simply think that this community has the power to address all issues in a timely fashion, although they may not have the drive. I would hate to see v2.0 get to this condition in the future when there is no reason for it to. Can anyone here deny that between the GG staff and the community that purchased the SDK that we have enough programmers to fix everything? I am certainly willing to do any amount of testing it takes to confirm issues both exist and/or have been fixed.
#9
My only suggestions are along the obvious lines....
You need a diverse group that can cover all platforms (Win32, Linux, Mac) and with diverse systems if possible.
I myself have 3 PC's ranging from a Windows 98 machine that would probably choke on Torque, to a well built P3 system with Geforce2 GTS, 512 MB's of ram and Windows XP, to a brand new hand built P4 HyperThreaded with 800 Mhz bus 1 Gig of DDR400 ram and a GeForceFX 5900 running WinXP also. I could give two points of view from my end and I'm sure there are others that can do the same or even more. Obviously I'm lacking in the Linux and Mac department, lol.
You need people that have proven, or are willing to prove that they can handle such tasks as making or testing patches.
You need people that are active in the community.
And most of all, you'll need people that WANT to be a part of it, although thats the easy part, lol.
@ Britton - "I'd like to suggest that the more managers the better."
While in certain circumstances I would never argue that, in this one I would. To many people in your code can be a disaster waiting to happen. If nothing else my suggestion would be to have a maximum of 4 maintainers from each platform(3 would be my target personally). That should be plenty to ensure that all patches and issues can be addressed in a very timely manner. And you would still have the benifit of your current staff to review the maintainers work and progress.
02/19/2004 (12:51 pm)
"EDIT: The biggest problem we have is that testing, integrating, making sure the patch is done correctly, etc. takes a lot of time. If we can come up with a community certificatoin program, then things could progress as a much more rapid pace. Any suggestions?"My only suggestions are along the obvious lines....
You need a diverse group that can cover all platforms (Win32, Linux, Mac) and with diverse systems if possible.
I myself have 3 PC's ranging from a Windows 98 machine that would probably choke on Torque, to a well built P3 system with Geforce2 GTS, 512 MB's of ram and Windows XP, to a brand new hand built P4 HyperThreaded with 800 Mhz bus 1 Gig of DDR400 ram and a GeForceFX 5900 running WinXP also. I could give two points of view from my end and I'm sure there are others that can do the same or even more. Obviously I'm lacking in the Linux and Mac department, lol.
You need people that have proven, or are willing to prove that they can handle such tasks as making or testing patches.
You need people that are active in the community.
And most of all, you'll need people that WANT to be a part of it, although thats the easy part, lol.
@ Britton - "I'd like to suggest that the more managers the better."
While in certain circumstances I would never argue that, in this one I would. To many people in your code can be a disaster waiting to happen. If nothing else my suggestion would be to have a maximum of 4 maintainers from each platform(3 would be my target personally). That should be plenty to ensure that all patches and issues can be addressed in a very timely manner. And you would still have the benifit of your current staff to review the maintainers work and progress.
#10
I think it's a high ideal and one that shouldn't be given up on, but I don't think there's a solution to getting bugs fixed and tested. Maybe there are more ideas out there for helping with this?
One crazy idea, allowing anyone to check-in to CVS. GG reserves the right to deny anyone the ability to check in or out. "Stable" builds can still be released (1.2 for instance) by Garage Games. Anything else is subject to verification by the whole community. If, for example, I update a "bug fix" that breaks something else, the person that finds it breaking can fix it themself or let me know. If a malicious user decides to throw in obvious bugs or damaging code, GG could rollback their change and ban them from CVS updates (or worse).
This is similar to the Wiki idea. I've never actually used Wiki, but from what I hear, it's working fairly well. And yes, I started with "crazy idea".
-Nerseus
02/19/2004 (4:35 pm)
While I love Torque, its community, and GG especially, I don't think it's realistic to get ALL of the known bugs fixed. Garage Games just isn't that big of a company - surely not large enough to employ a full QA team that can do regression testing as bugs are fixed. Having the community help is a GREAT idea, but I doubt there are many that are knowledgable enough with Torque to verify bug fixes and have enough time/desire to do it, especially for free.I think it's a high ideal and one that shouldn't be given up on, but I don't think there's a solution to getting bugs fixed and tested. Maybe there are more ideas out there for helping with this?
One crazy idea, allowing anyone to check-in to CVS. GG reserves the right to deny anyone the ability to check in or out. "Stable" builds can still be released (1.2 for instance) by Garage Games. Anything else is subject to verification by the whole community. If, for example, I update a "bug fix" that breaks something else, the person that finds it breaking can fix it themself or let me know. If a malicious user decides to throw in obvious bugs or damaging code, GG could rollback their change and ban them from CVS updates (or worse).
This is similar to the Wiki idea. I've never actually used Wiki, but from what I hear, it's working fairly well. And yes, I started with "crazy idea".
-Nerseus
#11
02/20/2004 (12:37 pm)
Is GG officially working on a v2 of Torque? I havn't come across anything other than member's saying that they are.
#12
@ Dan....
"I don't think it's realistic to get ALL of the known bugs fixed"
Had you said "I don't think it's realistic to get ALL of the bugs fixed". I might have agreed with you, however, what reason makes sense for not being able to fix all "known" issues? There certainly are enough programmers here to make it happen. In fact, if we could get 50 people to commit to it for one week, I'll bet good money that 90-95% of it would be cured in that week. Heck, a lot of it has already been cured and just needs to be wiped from the list.
I won't beat this issue to death though, I just don't buy into the "Can't be done" mentallity. Where there is a will, there is almost certainly a way.
"One crazy idea, allowing anyone to check-in to CVS."
I'll agree with that, that is one CRAZY idea, lol. (Just laughing with you dude)
I have every confidence that GG will find a way to satisfy this matter. The guys at GG have proved time and time again that they are a step ahead of the rest. Torque is all the proof you should need of that fact.
BTW, I'm claiming the first use of the term, "Tv2" untill someone else can prove they said it first, lol. (don't take that to mean legal rights, simply bragging rights)
02/20/2004 (6:26 pm)
I'm curious about this myself. If Tv2 is a considerable improvement over the current engine, then we would really have a hell of an engine to work with.@ Dan....
"I don't think it's realistic to get ALL of the known bugs fixed"
Had you said "I don't think it's realistic to get ALL of the bugs fixed". I might have agreed with you, however, what reason makes sense for not being able to fix all "known" issues? There certainly are enough programmers here to make it happen. In fact, if we could get 50 people to commit to it for one week, I'll bet good money that 90-95% of it would be cured in that week. Heck, a lot of it has already been cured and just needs to be wiped from the list.
I won't beat this issue to death though, I just don't buy into the "Can't be done" mentallity. Where there is a will, there is almost certainly a way.
"One crazy idea, allowing anyone to check-in to CVS."
I'll agree with that, that is one CRAZY idea, lol. (Just laughing with you dude)
I have every confidence that GG will find a way to satisfy this matter. The guys at GG have proved time and time again that they are a step ahead of the rest. Torque is all the proof you should need of that fact.
BTW, I'm claiming the first use of the term, "Tv2" untill someone else can prove they said it first, lol. (don't take that to mean legal rights, simply bragging rights)
Torque Owner Gonzo T. Clown
And I would like to add, that any of you that haven't at least looked at the issues to see if you could solve just one of them aren't being fair IMO. All of us are enjoying improvements that others have submitted to the engine, so in a way we each owe it to the community to at least try to solve one problem if we have the skills to do so. It wouldn't take but about 50-60 of us to do this ONE TIME to cure all of them. So please, if you have some free time, lets clean this puppy up for everyones benifit.
[endRantAgain]