I am pretty new in builders and absolutely newbie in FinalBuilder.
It is a great app, very intuitive, well documented with lots of features.
But I can't find one thing:
How can I determine who is responsible for the broken build?
This is a notoriously hard problem to tackle, and I'd be surprised if FB server got it right. That's not to say that there isn't any function to get some information out of it, but consider this:
If two commits are done within a short timespan, and the server is currently occupied, by the time it picks up the changes and runs, it will run for the last one. How do you know which of the two changesets are really responsible?
Also, if I remove a method, and you call it, which of the two are in error? Me because I removed the wrong method, or you because you call a method we're not supposed to be using?
TeamCity has some provisions for this, but it frequently gets things wrong, in particular, consider the following sequence of events:
Person A checks in something, which breaks the build. Person A gets an email about this.
Person B checks in something, build still not running correctly, person B does not get any email
Person C checks in something, that by itself would also break the build. Build is, however, still not running, but now only partly because person A still hasn't fixed the problem he introduced.
Person A checks in a fix to his problem, build still not running because of person C's change, which still haven't been fixed.
It gets complicated knowing "who broke the build", at most you can figure out who "started a sequence of breaking the build". I know TeamCity gives you that, I don't know if FB server does.
It depends how you define "who is responsible for the broken build" If you define it as author of the last commit then you should retrieve the information from source control system. In case of subversion you can use svn info task to get the author.
Related
Suppose for an application which will never receive internet connection during its lifetime, how can you prevent the piracy of the software?
There cannot be a single product key requirement during installation because, once installed legitimately anybody can copy the installation and re-distribute it.
So every time the application runs it should check for something and crash if the check fails.
Now what could it possibly check?
Initially I thought keeping an encrypted binary file will do the job, but as answered here, that seems a negligible prevention.
Any hacker can modify the executable so that instead of crashing when the check fails it should continue running.
So no matter how difficult the check is, the cracked application will always run.
Now I cannot see any possible solution to this problem.
PS: I am a single independent developer who is developing productivity software with very low charge. Seeing this question I believe I just have to let it go. Sigh....
EDIT: I would like to thank all the contributors in this discussion in letting me know the grim reality...
What I understand now is that you are indirectly submitting the source code of your application in the form of the target executable. Its source code can be modified by anybody using a debugger, thus ANY method of preventing piracy through source code of your application is useless. The only possible solution to this problem is to keep your legitimate customers happy by providing them services (apart from the software) and keep your price below their expectations.
I was think of solving this problem for past 3 days and now all seems worthwhile but still learnt a lot in this process, which I wouldn't have otherwise...
I ha
The only standalone thing I've seen that is semi-effective is hardware keys that come with the boxed software. They used to attach to a parallel port or a serial port and get checked when you started the program.
AutoCad and similar programs used to do this, but it is a BIG PAIN for your customers. Any time it doesn't read it, or a key goes bad, customer productivity suffers. It hurts your legitimate customers far more than those who end up pirating it anyway, and a sufficiently motivated pirate can make a VM that will overcome this. Modern versions of this use USB.
My recommendation is to trust people. Upon install, make them click a "I promise I paid for this" button and be done with it. If they click "I didn't pay for this" show them a small paragraph about how to help keep good software coming and prevent customer-harming DRM schemes by simply contributing to the success of good software authors.
You could generate a unique copy for each user, create a database, and check it agents copies you find online if you like playing the biggest game of wack-a-mole ever.
I've been looking into using Mercurial for version management and wanted to try it out. I'm planning on doing a project on my own to help me learn C++ coming from Java and thought it would be a good idea to try version management with it. It would probably be a game or some sort of simple application, not sure with a GUI or not.
The problem is I don't know much about version management. When in a project is a good time to make the repository? Should I just do it right away? How often should I make commits?
I'll be using Netbeans for my development since it seems to be pretty good, also has Mercurial controls built in, but what should I set in the .hgignore for my project? I assume the repository would be the whole folder Netbeans creates but what should I make it ignore?
Bonus questions: I'll be using Bitbucket for hosting/backup, should I make the project public? Why/why not? Also would that make it open source? Should I bother with attaching a license? Because I have no idea how the licenses work and such. Then again, I sincerely doubt I'll be making anything anyone will want to copy too badly.
On repos:
You should use a VCS repository as soon as possible. Not only does it allow for off-site backups, but it allows you to track changes. To find out when a bug was introduced.
Having a VCS, particularly a distributed VCS like Mercurial, will change the way you code. You won't have to be careful not to break things, because you always have the old version that you can revert to. If, after a few weeks or months, you decide that a particular course of development was a bad idea, you can rewind all the way back to some prior point.
You should generally commit either every day, or every time you finish a task. Like, you might have a 4-hour task of "write this class and get it working". You commit after that. When you're done for the day, you commit. I wouldn't suggest committing in a non-compiling state, though, so you should try to stop for the day only when everything still builds.
As for ignoring, you should ignore things you don't intend to commit. Things that you wouldn't want in the repo. Generated files, temporary files, the stuff that you wouldn't need.
What should be in the repo is 100% of the information necessary to build the project from scratch (minus external dependencies).
On licenses:
I would say that it is very rude to put up a public repository without some kind of license attached. Without a license, that means that anyone even pulling from the repo (which is what you're inviting by making it public) is a violation of your copyright without direct, explicit permission from you.
So either look at how copyright works and pick a license to release your stuff under, or keep your repos private.
The instant you conceive of it, usually. There is no penalty for putting everything you have into a VCS right away, and who knows, you may want that super-duper prototype version later on down the road. Unless you're going to be checking in large media files (don't), my answer is: right now.
As for how often to commit, that's a personal preference. I use git, and as far as I'm concerned there is no problem with checking in every time you make some unit of significant change. The more snapshots along the way you have, the easier it becomes to track down bugs later on.
Your .hgignore is up to you. You need to decide what to version and what not to. As a rule of thumb, any files generated during compilation probably shouldn't be checked in.
Also, if you release it to the public, please attach a license. It doesn't matter what it is, but it makes all the bosses and lawyers in the world happy when they know what they can and can't do with your code. Finally, please don't release something to the public unless you think someone else will benefit from it. The internet already suffers from information overload.
Commit early, and commit often.
With distributed VCS like hg or git there's really no reason not to start a repo before you even write any code. If you decide to scrap your project, you can always delete your repo before you post it to whatever hosting site you're using. The more often you commit, the easier it will be to hone in on where you messed up later on (you will).
Stuff to put in your ignore file would be build files and anything that your IDE generates periodically, like user preferences and tag caches. You don't want to foist your preferences on someone else and you don't want to have to create a commit every time you change the position of a window in the IDE. What's in the repo should be the minimum that someone would need to be able to work on your project.
If you're just doing a small project to learn, then I wouldn't worry about licenses or having a private repo. You can always add a license later. Alternatively, just pick a permissive license and go with it.
Based on my experience with git, I would recommend the following:
Create repository along with project.
Commit as often as you can with detailed comments
When you want to test a new idea - create a branch. If the idea is successful, merge your code to the master branch. If the idea is not successful, you can just discard the branch. Probably this is the most important idea.
Put the names of object files, executables and editor backup files in the ignore list.
We've had problems recently where developers commit code to SVN that doesn't pass unit tests, fails to compile on all platforms, or even fails to compile on their own platform. While this is all picked up by our CI server (Cruise Control), and we've instituted processes to try to stop it from happening, we'd really like to be able to stop the rogue commits from happening in the first place.
Based on a few other questions around here, it seems to be a Bad Idea™ to force this as a pre-commit hook on the server side mostly due to the length of time required to build + run the tests. I did some Googling and found this (all devs use TortoiseSVN):
http://cf-bill.blogspot.com/2010/03/pre-commit-force-unit-tests-without.html
Which would solve at least two of the problems (it wouldn't build on Unix), but it doesn't reject the commit if it fails. So my questions:
Is there a way to make a pre-commit hook in TortoiseSVN cause the commit to fail?
Is there a better way to do what I'm trying to do in general?
There is absolutely no reason why your pre-commit hook can't run the Unit tests! All your pre-commit hook has to do is:
Checkout the code to a working directory
Compile everything
Run all the unit tests
Then fail the hook if the unit tests fail.
It's completely possible to do. And, afterwords, everyone in your development shop will hate your guts.
Remember that in a pre-commit hook, the entire hook has to complete before it can allow the commit to take place and control can be returned to the user.
How long does it take to do a build and run through the unit tests? 10 minutes? Imagine doing a commit and sitting there for 10 minutes waiting for your commit to take place. That's the reason why you're told not to do it.
Your continuous integration server is a great place to do your unit testing. I prefer Hudson or Jenkins over CruiseControl. They're easier to setup, and their webpage are more user friendly. Even better they have a variety of plugins that can help.
Developers don't like it to be known that they broke the build. Imagine if everyone in your group got an email stating you committed bad code. Wouldn't you make sure your code was good before you committed it?
Hudson/Jenkins have some nice graphs that show you the results of the unit testing, so you can see from the webpage what tests passed and failed, so it's very clear exactly what happened. (CruiseControl's webpage is harder for the average eye to parse, so these things aren't as obvious).
One of my favorite Hudson/Jenkins plugin is the Continuous Integration Game. In this plugin, users are given points for good builds, fixing unit tests, and creating more passed unit tests. They lose points for bad builds and breaking unit tests. There's a scoreboard that shows all the developer's points.
I was surprised how seriously developers took to it. Once they realized that their CI game scores were public, they became very competitive. They would complain when the build server itself failed for some odd reason, and they lost 10 points for a bad build. However, the number of failed unit tests dropped way, way down, and the number of unit tests that were written soared.
There are two approaches:
Discipline
Tools
In my experience, #1 can only get you so far.
So the solution is probably tools. In your case, the obstacle is Subversion. Replace it with a DVCS like Mercurial or Git. That will allow every developer to work on their own branch without the merge nightmares of Subversion.
Every once in a while, a developer will mark a feature or branch as "complete". That is the time to merge the feature branch into the main branch. Push that into a "staging" repository which your CI server watches. The CI server can then pull the last commit(s), compile and test them and only if this passes, push them to the main repository.
So the loop is: main repo -> developer -> staging -> main.
There are many answers here which give you the details. Start here: Mercurial workflow for ~15 developers - Should we use named branches?
[EDIT] So you say you don't have the time to solve the major problems in your development process ... I'll let you guess how that sounds to anyone... ;-)
Anyway ... Use hg convert to get a Mercurial repo out of your Subversion tree. If you have a standard setup, that shouldn't take much of your time (it will just need a lot of time on your computer but it's automatic).
Clone that repo to get a work repo. The process works like this:
Develop in your second clone. Create feature branches for that.
If you need changes from someone, convert into the first clone. Pull from that into your second clone (that way, you always have a "clean" copy from subversion just in case you mess up).
Now merge the Subversion branch (default) and your feature branch. That should work much better than with Subversion.
When the merge is OK (all the tests run for you), create a patch from a diff between the two branches.
Apply the patch to a local checkout from Subversion. It should apply without problems. If it doesn't, you can clean your local checkout and repeat. No chance to lose work here.
Commit the changes in subversion, convert them back into repo #1 and pull into repo #2.
This sounds like a lot of work but within a week, you'll come up with a script or two to do most of the work.
When you notice someone broke the build (tests aren't running for you anymore), undo the merge (hg clean -C) and continue to work on your working feature branch.
When your colleagues complain that someone broke the build, tell them that you don't have a problem. When people start to notice that your productivity is much better despite all the hoops that you've got to jump, mention "it would be much more simple if we would scratch SVN".
The best thing to do is to work to improve the culture of your team, so that each developer feels enough of a commitment to the process that they'd be ashamed to check in without making sure it works properly, in whatever ways you've all agreed.
What methods are available for preventing sloppy developers from breaking builds.
Are there any version control systems which have a system of preventing check-in of code which breaks the build.
Thanks
Microsoft TFS Build has something called "gated check-ins" which provides this, by performing a private check-in (called Shelving) which is promoted to a normal check-in if the build succeeds.
http://blogs.msdn.com/b/patcarna/archive/2009/06/29/an-introduction-to-gated-check-in.aspx
TeamCity has the concept of "delayed commit"
http://www.jetbrains.com/teamcity/features/delayed_commit.html
I can wholeheartedly recommend TeamCity!
Get a LART and beat developers who break the build.
Have the build status on a great big screen where everyone can see it with the a red/green background and the name of the last person to commit.
Have the build server send an e-mail to the whole dev team fingering the developer who broke the build.
Honestly... why do people get so hung up on "making developers do X". Tell them it's the process, and fire them if they don't follow it.
EDIT: because the following was too long for a comment.
I work on a team with 12 or so developers. Some people consider that large, some consider it small
We have a large screen (32" flat screen TV on a 6 foot stand) that everyone can see that tells us all sorts of information - including (in the biggest box on the screen) the state of the "commit build".
Our process is to update from SVN and run the commit build locally (about 2-3 mins) before committing. If it passes, ship it. If not, fix it locally and repeat. Since we do TDD, this generally only happens if something you pulled out of SVN broke something you were working on.
If it fails in CI you either ignored the process or your commit collided with someone else's in a bad way. The screen goes red, someone shouts at you, you fix it and move on. This generally only happens to us about once a week or so; mostly it goes red because people try and cut corners ;-)
Nobody needs to "force developers" to do anything. We're creative, artistic individuals who are generally adult and professional enough to follow a process if that process makes sense. In this case, build locally and only commit if that passes.
Who cares if the build breaks in CI, as long as it's fixed quickly so it doesn't prevent the team from working?
During the process of building software applications, you would start testing what you have built in stages even before it is complete and you could start seeing issues/bugs. How do you track them, Do you use your regular bug tracking tool to add them as issues(waste of time - since it is a work in progress), just have them in your head to fix later, or have a simple text list.
What would be an efficient way to make sure that whatever you have found is eventually fixed as development progresses? Are there any tiny tools to do that?
What I usually do is the following:
Gauge the size of the bug/issue
If it's too big, create an issue in the bug tracker.
If it's small enough, write a failing unit test and then come back to it after I've finished the original functionality.
I've found that the simplest and most efficient way to track tasks of all types (todos, work items, bugs, etc ...) is to use a single system. Typically a bug tracking system. This allows you to see all of the work remaining on your project in a single place.
Having multiple tracking systems almost always results in lost data. People eventually pick different systems, don't tell people about the system they are on, lose the piece of paper which has the list of work items, etc ...
Most bug tracking systems allow you to categorize your bugs so it's easy to distinguish the type of work remaining as well.
Make sure you CI tools such as CruiseControl.NET run unit tests as part of the build. This will cause the build shown as broken when unit test fails and the person who last checked in will be responsible for fixing it.