|
a pm who hates meetings - the only thing you need for tolerable standups
|
# ? Mar 19, 2016 23:16 |
|
|
# ? May 11, 2024 09:32 |
|
i have two different standups in the morning, thats how agile i am, like a cat
|
# ? Mar 19, 2016 23:26 |
|
we used to have one standup in the morning that lasted like 20 minutes and people complained so now everyone has like 3 separate standups that last like an hour or so total because no one can really get going until 10:30 or so
|
# ? Mar 19, 2016 23:53 |
|
our scrum leader is amazing and we're agile kanbsn or something. i dont really care what it is but basically our ba's make my job really easy so i try to pay them back by making their job easy. it's by far the best management situation ive encountered and I basically have total loyalty to my managers
|
# ? Mar 19, 2016 23:54 |
|
we do scrum and it's fine i guess? i'm currently the tech lead and de facto project owner for a team of three people (including me) while the fourth is on paternity leave. i feel like i'm bad at running things but also that little damage can be done as long as we are good at code quality and testing, which we're not, but are getting better at also another team in my org has been continually awful since their tech lead quit a couple years ago, they've had rotating leadership and org charts and goals and it never seems to get better. now they're trying to work on an our-company specific spin of an electron.js program that's being primarily written by another part of the company that's largely web people and it's going awfully. the real reason i brought this up was to complain that their problems meant everybody got scrum training, again, but this time it was hours and hours of scrum training on the day after somebody's going away party that was on st patricks day. i'm a goddamn hero for making it through
|
# ? Mar 20, 2016 00:10 |
|
lol if you do any work in the mornings, or indeed at all
|
# ? Mar 20, 2016 00:19 |
|
Soricidus posted:lol if you do any work in the mornings, or indeed at all yeah, by the time my standups are done its almost 11am, and then you dont really want to start much before lunch at 12 and then its 1pm and well gently caress almost time to go home
|
# ? Mar 20, 2016 00:22 |
|
agile seems like a decent idea that was ruined when business people got ahold of it
|
# ? Mar 20, 2016 00:30 |
|
Valeyard posted:yeah, by the time my standups are done its almost 11am, and then you dont really want to start much before lunch at 12 and then its 1pm and well gently caress almost time to go home every friday
|
# ? Mar 20, 2016 00:30 |
|
our standups are surprisingly decent, we're usually done in 20 minutes even when theres a huge tangent. but our end of sprint meetings, oh boy. 3 hours of sprint demos across 3 teams in different locations, then 5 hours of sprint planning. oh and this is on friday so get ready for another 5 hours of sprint planning on monday when we rehash everything we did friday
|
# ? Mar 20, 2016 00:47 |
|
our problem is that our team owns like 5 things and we should really be split up but that's an org level problem. i think we handle it well given our resources
|
# ? Mar 20, 2016 01:09 |
|
triple sulk posted:tdd is great in theory but your most useful tests are the ones written against edge case bugs you don't come up with until they show themselves test your tests with mutation testing
|
# ? Mar 20, 2016 02:25 |
|
all roads lead to boosting coverage
|
# ? Mar 20, 2016 02:29 |
|
Arcsech posted:but our end of sprint meetings, oh boy. 3 hours of sprint demos across 3 teams in different locations, then 5 hours of sprint planning. oh and this is on friday so get ready for another 5 hours of sprint planning on monday when we rehash everything we did friday at my last job, the first time we had a sprint planning meeting on a friday afternoon, it went on until 6:30 or so, which did a pretty good job of convincing everyone involved that either we should speed those up, or at the very least not hold them on a friday afternoon. i suppose it was rather productive in that way
|
# ? Mar 20, 2016 02:30 |
|
Valeyard posted:all roads lead to boosting coverage if you use mutation testing to make your metrics you end up with less coverage because you see how useless coverage is
|
# ? Mar 20, 2016 02:33 |
|
YeOldeButchere posted:at my last job, the first time we had a sprint planning meeting on a friday afternoon, it went on until 6:30 or so, which did a pretty good job of convincing everyone involved that either we should speed those up, or at the very least not hold them on a friday afternoon. i suppose it was rather productive in that way The answer is that all meetings should be held at 4:30pm on a Friday afternoon to convince people to stop wasting time in meetings.
|
# ? Mar 20, 2016 02:50 |
|
Brain Candy posted:test your tests with mutation testing Someone needs to adapt Trusting Trust to Testing Tests.
|
# ? Mar 20, 2016 02:51 |
|
hackbunny posted:NextStep, OS X's pre-transition deadname isn't the real joke something about how Obj-C doesn't have namespaces
|
# ? Mar 20, 2016 03:56 |
|
lord funk posted:just spent 20 minutes working on a subclass and trying to figure out why it wasn't working install R# and then you can lol in a code review about someone who spend 2 days refactoring methods that aren't used anywhere
|
# ? Mar 20, 2016 03:58 |
|
triple sulk posted:tdd is great in theory but your most useful tests are the ones written against edge case bugs you don't come up with until they show themselves [quote="http://research.microsoft.com/apps/mobile/news.aspx?post=/en-us/news/features/nagappan-100609.aspx"] The study and its results were published in a paper entitled Realizing quality improvement through test driven development: results and experiences of four industrial teams, by Nagappan and research colleagues E. Michael Maximilien of the IBM Almaden Research Center; Thirumalesh Bhat, principal software-development lead at Microsoft; and Laurie Williams of North Carolina State University. What the research team found was that the TDD teams produced code that was 60 to 90 percent better in terms of defect density than non-TDD teams. They also discovered that TDD teams took longer to complete their projects—15 to 35 percent longer. “Over a development cycle of 12 months, 35 percent is another four months, which is huge,” Nagappan says. “However, the tradeoff is that you reduce post-release maintenance costs significantly, since code quality is so much better. Again, these are decisions that managers have to make—where should they take the hit? But now, they actually have quantified data for making those decisions.” [/quote]
|
# ? Mar 20, 2016 04:16 |
|
The entire study was performed against one 9 person team in IBM and three ~7 person teams in Microsoft. All of the teams were adding onto a legacy system that was already in production. The time increase was spitballed by just asking the managers how much time they thought the additional testing added. The study doesn't specify how that question was asked, so who knows whether they hosed up the whole study by framing the question poorly. Even by their own admission: "the results should not be taken to be valid in the general sense, only in the context in which the case study was conducted." So lol who knows what any of this means, we just figured we'd waste everyone's time
|
# ? Mar 20, 2016 05:00 |
|
coveralls failed my last pull request because i decreased coverage by 0.009% because i deleted a bunch of code lol
|
# ? Mar 20, 2016 05:01 |
|
Progressive JPEG posted:So lol who knows what any of this means, we just figured we'd waste everyone's time isnt that basically the case with 'all' software development studies
|
# ? Mar 20, 2016 14:45 |
|
Blinkz0rz posted:coveralls failed my last pull request because i decreased coverage by 0.009% because i deleted a bunch of code lol yeah coveralls sucks
|
# ? Mar 20, 2016 15:24 |
|
software metrics: more lines of code means more bad code i cannot recall the study i found but a large amount of code complexity metrics didn't really make much more noise than a rough count of lines i would say that using static analysis where available would be time better spent. A lot of metrics create busywork
|
# ? Mar 20, 2016 16:45 |
|
Progressive JPEG posted:The entire study was performed against one 9 person team in IBM and three ~7 person teams in Microsoft. All of the teams were adding onto a legacy system that was already in production. The time increase was spitballed by just asking the managers how much time they thought the additional testing added. The study doesn't specify how that question was asked, so who knows whether they hosed up the whole study by framing the question poorly. i think the point is more the measured decrease in defects than the subjectively rated increase in development time. sulk was saying tdd doesn't capture your weird edge cases, which is true but doesnt mean theres no value to it. uncurable mlady posted:so how do I learn how to do tdd or w/e the only really effective way ive seen is pairing with someone who knows it, but theres a couple of good books. TDD By Example is the original kent beck book which shows a style that works really well if youre doing nice functional code with minimal side effects. Growing Object-Oriented Code Guided By Tests is also really good, but is very OO focused and very mockist which is a dealbreaker for some folks.
|
# ? Mar 20, 2016 16:45 |
|
like the only thing metrics wise that i remember is that "number of local variables is fine, add more if you need em, but if you have too many function arguments you have probably missed one" which i have lived through time and time again with minor api changes requiring a configuration detail to be passed through and somehow i end up with a 20 arg constructor
|
# ? Mar 20, 2016 16:46 |
|
tef posted:software metrics: more lines of code means more bad code your article went around my team and myself and another guy loved it and are actively trying to reduce the size and complexity of a bunch of our tools
|
# ? Mar 20, 2016 17:15 |
|
tef posted:software metrics: more lines of code means more bad code Greg Wilson covered a lot of it in this talk: https://vimeo.com/9270320
|
# ? Mar 20, 2016 17:52 |
|
TDD is stupid for a lot of things but it's at least simple to actually do step 1. write an empty function with some name that consumes some arguments and returns some type step 2. write unit tests for the function that fail step 3. implement the function the trouble is that it assumes all software dev problems are "I know what the input looks like, I know what the output looks like and I just need to figure out what to do in the middle" one of the better things about TDD probably is that it forces you toward more reasonable designs, specifically designs that make you isolate nondeterministic parts (90% of the time, this refers to your POS CRUD web app going to the DB) from deterministic parts (that elusive stuff called "logic" we are always trying to find a place for)
|
# ? Mar 20, 2016 18:52 |
|
tef posted:software metrics: more lines of code means more bad code this is exactly why i'm happy about mutation testing the tooling mutates your code, by flipping conditionals, etc. and then runs your test suite if the tests failed, congrats, you've got good tests if it passes, it can guide you to which tests to delete
|
# ? Mar 20, 2016 19:00 |
|
MrMoo posted:Greg Wilson covered a lot of it in this talk: https://vimeo.com/9270320 goddamn
|
# ? Mar 20, 2016 19:38 |
|
has there been any studies if people like "whoops, something went wrong, lol" -type errors more than "database error, error code 0xdeadbeef, please contact your system administrator" -type errors? because the former aggravate the gently caress out of me. but then i'm a programmer, maybe actual people like that
|
# ? Mar 20, 2016 20:50 |
|
Wheany posted:has there been any studies if people like "whoops, something went wrong, lol" -type errors more than "database error, error code 0xdeadbeef, please contact your system administrator" -type errors? when writing an internal tool consumed by non-programmers i tend write a little "please contact the tech team" and then include the exception message, as they tend to be both an actual sentence while still useful to me
|
# ? Mar 20, 2016 21:10 |
|
yeah i think it really depends on the domain.
|
# ? Mar 20, 2016 21:11 |
|
MALE SHOEGAZE posted:yeah i think it really depends on the domain. it helps that absolutely any problem that somebody has with the service (there's only like 3 users lol startips) i can personally deal with. there's no expectation that somebody could fix their own errors based on the message (which is when it becomes really important to be include useful information while still being accessible.
|
# ? Mar 20, 2016 21:14 |
|
tef posted:software metrics: more lines of code means more bad code quote:Conclusion: Apart from size, surrogate maintainability measures may not reflect future maintenance effort. Surrogates need to be evaluated in the contexts for which they will be used. While traditional metrics are used to identify problematic areas in the code, the improvements of the worst areas may, inadvertently, lead to more problems for the entire system. Our results suggest that local improvements should be accompanied by an evaluation at the system level.
|
# ? Mar 20, 2016 23:10 |
|
Blinkz0rz posted:your article went around my team and myself and another guy loved it and are actively trying to reduce the size and complexity of a bunch of our tools christ
|
# ? Mar 21, 2016 00:58 |
|
Wheany posted:goddamn yep although one of the books he cites defends the 10x talk (actually it's a real good talk)
|
# ? Mar 21, 2016 00:58 |
|
|
# ? May 11, 2024 09:32 |
|
Wheany posted:has there been any studies if people like "whoops, something went wrong, lol" -type errors more than "database error, error code 0xdeadbeef, please contact your system administrator" -type errors? in before mononcqc http://www.hpl.hp.com/techreports/tandem/TR-85.7.pdf
|
# ? Mar 21, 2016 01:02 |