WTF
Everyone tells me I should give up programming because it's a discipline too easily outsourced to third world countries, however, the more I work with a certain group of folks in a certain tiny Jewish Crusader State the more convinced I am that I won't have any trouble keeping food on the table.
Today was release day for the latest software build. That's always a hectic day, but today was worse than normal. Why was it worse than normal? Because when they typed "make" this morning the project spat out a ton of errors and refused to build. And whose fault was that do you suppose? I bet you can guess. And who got bitched at? I bet you can guess that too. I don't remember seeing "Whipping Boy for Israeli Programming Team" in my job description, but that's basically what it amounts to since the higher ups can't exactly go over there and yell at them. Especially since they're on the other side of the world, work crazy hours, and have obscure religious holidays on release day.
The good news is that I isolated and fixed the problem in about five minutes with nothing to go on but the compiler error messages. That's the other reason why they pay me. However, it would really improve my morale (and the morale of everyone else who works even remotely with health management) if they would just let me do the actual development. I would gladly single-handedly take over the workload of the entire AI team just so I wouldn't have to deal with the absurdities. Like, dudes, the release version management system is NOT a freaking generic code repository. Don't randomly check files back in when you go to the bathroom. Thanks.
Anyway, what did they do today, you might ask?
Well they modified some header files that are used ALL OVER the project !!!WITHOUT TELLING ANYONE!!!. Yes, on release day. And yes with less than a month to go on a multi-year project.
They had a datatype (we'll call it fooBarTux) which was in camelCase even though the project coding guidelines specify old C style type names. That should have been caught and fixed IMMEDIATELY, but it wasn't. It's been in there for years. Possibly since 777. And since its been there for years there is a ton of code that references it.
Well this morning they went in and changed it to foo_bar_tux_t which is how it should have been to begin with, except doing that broke ALL THE CODE IN THE WHOLE SYSTEM that had been written to use fooBarTux. Are they THAT stupid that they don't know not to do something like that without warning people?
I typedefed it to the old name and then sent out a note telling people what had been changed and why and asking them to update their code before the next release. Was that so hard!?
Blargh.
Today was release day for the latest software build. That's always a hectic day, but today was worse than normal. Why was it worse than normal? Because when they typed "make" this morning the project spat out a ton of errors and refused to build. And whose fault was that do you suppose? I bet you can guess. And who got bitched at? I bet you can guess that too. I don't remember seeing "Whipping Boy for Israeli Programming Team" in my job description, but that's basically what it amounts to since the higher ups can't exactly go over there and yell at them. Especially since they're on the other side of the world, work crazy hours, and have obscure religious holidays on release day.
The good news is that I isolated and fixed the problem in about five minutes with nothing to go on but the compiler error messages. That's the other reason why they pay me. However, it would really improve my morale (and the morale of everyone else who works even remotely with health management) if they would just let me do the actual development. I would gladly single-handedly take over the workload of the entire AI team just so I wouldn't have to deal with the absurdities. Like, dudes, the release version management system is NOT a freaking generic code repository. Don't randomly check files back in when you go to the bathroom. Thanks.
Anyway, what did they do today, you might ask?
Well they modified some header files that are used ALL OVER the project !!!WITHOUT TELLING ANYONE!!!. Yes, on release day. And yes with less than a month to go on a multi-year project.
They had a datatype (we'll call it fooBarTux) which was in camelCase even though the project coding guidelines specify old C style type names. That should have been caught and fixed IMMEDIATELY, but it wasn't. It's been in there for years. Possibly since 777. And since its been there for years there is a ton of code that references it.
Well this morning they went in and changed it to foo_bar_tux_t which is how it should have been to begin with, except doing that broke ALL THE CODE IN THE WHOLE SYSTEM that had been written to use fooBarTux. Are they THAT stupid that they don't know not to do something like that without warning people?
I typedefed it to the old name and then sent out a note telling people what had been changed and why and asking them to update their code before the next release. Was that so hard!?
Blargh.