July 5, 2009 at 01:45Joel test
The other day I read the book "Joel about programming" , which has a quality test of a team of software developers. I decided to publish his results for the company in which I currently work.
1. Do you use a source code version control system?
Not. This company is the second in which I work as a php programmer. For some reason I did not understand, neither the system nor the previous one used such systems, which is very sad. How much less problems would be. By the way, in this case, Joel himself advises the programmer to start using CVS himself and try to introduce other developers to this.
2. Can you complete the assembly of the product in one step?
No I can not. Although for some projects, there are scripts for automatically deploying a web application to a production server. But this is only for a few new projects.
3. Do you do daily compilation?
Oh no. Unfortunately, I don’t know how the team of Java programmers works.
4. Do you keep a database of errors in the program?
Not.
5. Do you fix errors before writing new code?
Not. Often errors are detected already at the stage of "industrial operation". And most often, error correction is entrusted to completely different people who before that had nothing to do with the project. A very sad situation.
6. Do you have an up-to-date work schedule?
Haha Yes, there is some kind of nonsense that is crammed into the calendar in the exchanger server, but it’s definitely not to be called relevant. And also there is no detail of the schedule that Joel “requires”.
7. Do you have specifications?
I would say no. Sometimes, some papers are written. If you're lucky, there will be screens to do. If less, then the text with a bunch of ambiguous texts. Often I just look at it, and then put it on the edge of the table, because they are all pretty useless.
8. Have you created calm conditions for the work of programmers?
This item upsets me more than anyone else, I guess. No, no and NO. Managers are sitting next to the programmers, phones are always ringing, someone is walking, talking. Half of the programmers have disgusting chairs, after an hour of sitting on which the ass becomes square. In general, I would say that the conditions are disgusting.
9. Do you try to use the best of existing tools for work?
Let's say yes. That is, I must say that we have different programmers. And different applications are written in different languages (C #, Java, PHP, Perl) and use different databases (Oracle, MySQL, PostgreSQL). True, I don’t know who is guided by what when choosing languages, tools and technologies.
10. Do you involve testers in your work?
Not. Also a terrible moment. All are tested by the programmers themselves. Then the manager can still test a little.
11. Do you offer job seekers to write a code during the interview?
Well, when I was interviewed, the second part was the completion of test tasks. The tasks were very simple, they could hardly appreciate me somehow. All that could be said on them, yes, I at least read the book "PHP in 24 hours". Spolsky himself advises checking out a little different. But set off.
12. Do you test usability on random people?
Not if the general liked it, then it will be so.
Total three points. What Joel writes about this: “Ideally, you should score 12 points. 11 points is a tolerable amount, and 10 or less indicate serious problems. In fact, most software organizations work with only two or three points, and they need serious improvements, because companies like Microsoft are constantly holding 12 points. ”
And how are you?
Update If interested, Joel Spolsky maintains a blog called “Joel on Software” , based on which books are published.
Update # 2 Original "Joel test" with explanations .
1. Do you use a source code version control system?
Not. This company is the second in which I work as a php programmer. For some reason I did not understand, neither the system nor the previous one used such systems, which is very sad. How much less problems would be. By the way, in this case, Joel himself advises the programmer to start using CVS himself and try to introduce other developers to this.
2. Can you complete the assembly of the product in one step?
No I can not. Although for some projects, there are scripts for automatically deploying a web application to a production server. But this is only for a few new projects.
3. Do you do daily compilation?
Oh no. Unfortunately, I don’t know how the team of Java programmers works.
4. Do you keep a database of errors in the program?
Not.
5. Do you fix errors before writing new code?
Not. Often errors are detected already at the stage of "industrial operation". And most often, error correction is entrusted to completely different people who before that had nothing to do with the project. A very sad situation.
6. Do you have an up-to-date work schedule?
Haha Yes, there is some kind of nonsense that is crammed into the calendar in the exchanger server, but it’s definitely not to be called relevant. And also there is no detail of the schedule that Joel “requires”.
7. Do you have specifications?
I would say no. Sometimes, some papers are written. If you're lucky, there will be screens to do. If less, then the text with a bunch of ambiguous texts. Often I just look at it, and then put it on the edge of the table, because they are all pretty useless.
8. Have you created calm conditions for the work of programmers?
This item upsets me more than anyone else, I guess. No, no and NO. Managers are sitting next to the programmers, phones are always ringing, someone is walking, talking. Half of the programmers have disgusting chairs, after an hour of sitting on which the ass becomes square. In general, I would say that the conditions are disgusting.
9. Do you try to use the best of existing tools for work?
Let's say yes. That is, I must say that we have different programmers. And different applications are written in different languages (C #, Java, PHP, Perl) and use different databases (Oracle, MySQL, PostgreSQL). True, I don’t know who is guided by what when choosing languages, tools and technologies.
10. Do you involve testers in your work?
Not. Also a terrible moment. All are tested by the programmers themselves. Then the manager can still test a little.
11. Do you offer job seekers to write a code during the interview?
Well, when I was interviewed, the second part was the completion of test tasks. The tasks were very simple, they could hardly appreciate me somehow. All that could be said on them, yes, I at least read the book "PHP in 24 hours". Spolsky himself advises checking out a little different. But set off.
12. Do you test usability on random people?
Not if the general liked it, then it will be so.
Total three points. What Joel writes about this: “Ideally, you should score 12 points. 11 points is a tolerable amount, and 10 or less indicate serious problems. In fact, most software organizations work with only two or three points, and they need serious improvements, because companies like Microsoft are constantly holding 12 points. ”
And how are you?
Update If interested, Joel Spolsky maintains a blog called “Joel on Software” , based on which books are published.
Update # 2 Original "Joel test" with explanations .