An historical observation
-
As I approach my 22nd anniversary working for my current company, I just noticed something. In all that time, the automated builds have taken roughly the same amount of time. In 1990, I was building a 'suite' of MS-DOS applications, written in 'C'. The automated build process was a batch file, running on a 12 MHz Pentium 60 MHz machine with 4 MB of RAM. It took from 45 to 75 minutes to build the suite. In the early 2000's, we were building products made up of a single application and several COM servers, all written in C++. The automated build process was a native mode application setting up and running a VBScript. The products took from 50 to 75 minutes to build, depending on the specifics. Today, I am building a product made up of multiple applications and Windows services, written in C# and C++. The automated build process is a combination of native mode and .NET applications, VBScript, multiple batch files, and an assortment of compilers. It takes 50 to 90 minutes to build the product, depending on options. Am I Sisyphus[^] reborn?
Software Zen:
delete this; -
As I approach my 22nd anniversary working for my current company, I just noticed something. In all that time, the automated builds have taken roughly the same amount of time. In 1990, I was building a 'suite' of MS-DOS applications, written in 'C'. The automated build process was a batch file, running on a 12 MHz Pentium 60 MHz machine with 4 MB of RAM. It took from 45 to 75 minutes to build the suite. In the early 2000's, we were building products made up of a single application and several COM servers, all written in C++. The automated build process was a native mode application setting up and running a VBScript. The products took from 50 to 75 minutes to build, depending on the specifics. Today, I am building a product made up of multiple applications and Windows services, written in C# and C++. The automated build process is a combination of native mode and .NET applications, VBScript, multiple batch files, and an assortment of compilers. It takes 50 to 90 minutes to build the product, depending on options. Am I Sisyphus[^] reborn?
Software Zen:
delete this;Do you have a distributed build system ?
Nihil obstat
-
Do you have a distributed build system ?
Nihil obstat
Sort of. We have two servers capable of running builds. Each one can run a couple builds at once without slowing down the build process too badly. We've never gone with a distributed build process because the time required for the build really isn't objectionable. At our worst, we don't do more than three or four builds in a day.
Software Zen:
delete this; -
As I approach my 22nd anniversary working for my current company, I just noticed something. In all that time, the automated builds have taken roughly the same amount of time. In 1990, I was building a 'suite' of MS-DOS applications, written in 'C'. The automated build process was a batch file, running on a 12 MHz Pentium 60 MHz machine with 4 MB of RAM. It took from 45 to 75 minutes to build the suite. In the early 2000's, we were building products made up of a single application and several COM servers, all written in C++. The automated build process was a native mode application setting up and running a VBScript. The products took from 50 to 75 minutes to build, depending on the specifics. Today, I am building a product made up of multiple applications and Windows services, written in C# and C++. The automated build process is a combination of native mode and .NET applications, VBScript, multiple batch files, and an assortment of compilers. It takes 50 to 90 minutes to build the product, depending on options. Am I Sisyphus[^] reborn?
Software Zen:
delete this;I've always commented that it's the same for operating systems. It used to take the same amount of time to load DOS 5 (from floppies) as it does to load Windows now. The speed of computer hardware is equally and oppositely balanced by the bloat of its operating system. I think I'll call that Wallace's Law.
I wanna be a eunuchs developer! Pass me a bread knife!
-
I've always commented that it's the same for operating systems. It used to take the same amount of time to load DOS 5 (from floppies) as it does to load Windows now. The speed of computer hardware is equally and oppositely balanced by the bloat of its operating system. I think I'll call that Wallace's Law.
I wanna be a eunuchs developer! Pass me a bread knife!
I think your law outdoes mine: Wheeler's Law: "Typing proficiency is inversely proportional to the number of people watching you do it."
Software Zen:
delete this; -
As I approach my 22nd anniversary working for my current company, I just noticed something. In all that time, the automated builds have taken roughly the same amount of time. In 1990, I was building a 'suite' of MS-DOS applications, written in 'C'. The automated build process was a batch file, running on a 12 MHz Pentium 60 MHz machine with 4 MB of RAM. It took from 45 to 75 minutes to build the suite. In the early 2000's, we were building products made up of a single application and several COM servers, all written in C++. The automated build process was a native mode application setting up and running a VBScript. The products took from 50 to 75 minutes to build, depending on the specifics. Today, I am building a product made up of multiple applications and Windows services, written in C# and C++. The automated build process is a combination of native mode and .NET applications, VBScript, multiple batch files, and an assortment of compilers. It takes 50 to 90 minutes to build the product, depending on options. Am I Sisyphus[^] reborn?
Software Zen:
delete this;I think that's because that kind of time is the longest that's acceptable, so the complexity of systems grows such that the build time runs up against that ceiling. So instead of today's systems being much faster than those 20 years ago, they are instead 10 or 100 times more complex. Also, bollocks to official grammar rules, when it's a hard H it should be 'a historical'. 'An historical' and the like sound incredibly awkward. H is a consonant and should be treated as such.
-
I think that's because that kind of time is the longest that's acceptable, so the complexity of systems grows such that the build time runs up against that ceiling. So instead of today's systems being much faster than those 20 years ago, they are instead 10 or 100 times more complex. Also, bollocks to official grammar rules, when it's a hard H it should be 'a historical'. 'An historical' and the like sound incredibly awkward. H is a consonant and should be treated as such.
BobJanova wrote:
Also, bollocks to official grammar rules, when it's a hard H it should be 'a historical'. 'An historical' and the like sound incredibly awkward. H is a consonant and should be treated as such.
In written communications, I tend use "an h----", while verbally I use "a h----". I'm old enough I had teachers who believed in formal English grammar, punctuation, and so on. None of this "express yourself any way you like" dreck.
Software Zen:
delete this; -
I think that's because that kind of time is the longest that's acceptable, so the complexity of systems grows such that the build time runs up against that ceiling. So instead of today's systems being much faster than those 20 years ago, they are instead 10 or 100 times more complex. Also, bollocks to official grammar rules, when it's a hard H it should be 'a historical'. 'An historical' and the like sound incredibly awkward. H is a consonant and should be treated as such.
BobJanova wrote:
it should be 'a historical'
Go West, young man. Join your brothers in the continents of the Americas, who also don't speak English.
I wanna be a eunuchs developer! Pass me a bread knife!
-
BobJanova wrote:
it should be 'a historical'
Go West, young man. Join your brothers in the continents of the Americas, who also don't speak English.
I wanna be a eunuchs developer! Pass me a bread knife!
Somebody wake Mick up; we've got the start of another Anglican pissing contest.
Software Zen:
delete this; -
I think your law outdoes mine: Wheeler's Law: "Typing proficiency is inversely proportional to the number of people watching you do it."
Software Zen:
delete this;Why do yopu say taht to me in praticuler?
Everyone would think I were an idiot without a spell-checker. But I have a spell-checker, so everyone just thinks I'm an idiot.
-
BobJanova wrote:
it should be 'a historical'
Go West, young man. Join your brothers in the continents of the Americas, who also don't speak English.
I wanna be a eunuchs developer! Pass me a bread knife!
It saddens me to say it but on this particular issue they have a point. Go on, have a little experiment. Ask someone if they are an happy person, and see if they give you a weird look compared to asking if they are a happy person. Or 'I've got to find an hockey club'. It just doesn't work at all.
-
It saddens me to say it but on this particular issue they have a point. Go on, have a little experiment. Ask someone if they are an happy person, and see if they give you a weird look compared to asking if they are a happy person. Or 'I've got to find an hockey club'. It just doesn't work at all.
In [American] English, the use of 'an' before words starting with 'h' varies with the word. "An historical", "an hyperbola", and so on. "An hockey"? Not on your life. I work with a guy who plays hockey, and he would high-stick you if he heard you say something like that.
Software Zen:
delete this; -
BobJanova wrote:
Also, bollocks to official grammar rules, when it's a hard H it should be 'a historical'. 'An historical' and the like sound incredibly awkward. H is a consonant and should be treated as such.
In written communications, I tend use "an h----", while verbally I use "a h----". I'm old enough I had teachers who believed in formal English grammar, punctuation, and so on. None of this "express yourself any way you like" dreck.
Software Zen:
delete this; -
It saddens me to say it but on this particular issue they have a point. Go on, have a little experiment. Ask someone if they are an happy person, and see if they give you a weird look compared to asking if they are a happy person. Or 'I've got to find an hockey club'. It just doesn't work at all.
Dear me. Next, you'll be telling us how you spent your holiday in "a hotel". tut tut.
I wanna be a eunuchs developer! Pass me a bread knife!
-
Oh, I had teachers that believed in 'proper English', as do I. It's just that when a rule is objectively stupid, I ignore it.
:thumbsup:
-
Dear me. Next, you'll be telling us how you spent your holiday in "a hotel". tut tut.
I wanna be a eunuchs developer! Pass me a bread knife!
-
As I approach my 22nd anniversary working for my current company, I just noticed something. In all that time, the automated builds have taken roughly the same amount of time. In 1990, I was building a 'suite' of MS-DOS applications, written in 'C'. The automated build process was a batch file, running on a 12 MHz Pentium 60 MHz machine with 4 MB of RAM. It took from 45 to 75 minutes to build the suite. In the early 2000's, we were building products made up of a single application and several COM servers, all written in C++. The automated build process was a native mode application setting up and running a VBScript. The products took from 50 to 75 minutes to build, depending on the specifics. Today, I am building a product made up of multiple applications and Windows services, written in C# and C++. The automated build process is a combination of native mode and .NET applications, VBScript, multiple batch files, and an assortment of compilers. It takes 50 to 90 minutes to build the product, depending on options. Am I Sisyphus[^] reborn?
Software Zen:
delete this; -
Gary Wheeler wrote:
made up of multiple applications ... It takes 50 to 90 minutes to build the product, depending on options.
So why don't you just build one at a time?
The individual applications are compiled one at a time, but the product is a coordinated set. That's only part of the build process, however. Once all of the applications are compiled, the installer is built, an image is constructed of the distribution media, and the entire mess gets archived into a DVD iso that will eventually get burned to disc.
Software Zen:
delete this; -
Oh, I had teachers that believed in 'proper English', as do I. It's just that when a rule is objectively stupid, I ignore it.
Quote:
when a rule is objectively stupid, I ignore it
That's what I did but the policeman that gave me the speeding ticket didn't sympathise with me at all!
- Life in the fast lane is only fun if you live in a country with no speed limits. - Of all the things I have lost, it is my mind that I miss the most. - I vaguely remember having a good memory...
-
As I approach my 22nd anniversary working for my current company, I just noticed something. In all that time, the automated builds have taken roughly the same amount of time. In 1990, I was building a 'suite' of MS-DOS applications, written in 'C'. The automated build process was a batch file, running on a 12 MHz Pentium 60 MHz machine with 4 MB of RAM. It took from 45 to 75 minutes to build the suite. In the early 2000's, we were building products made up of a single application and several COM servers, all written in C++. The automated build process was a native mode application setting up and running a VBScript. The products took from 50 to 75 minutes to build, depending on the specifics. Today, I am building a product made up of multiple applications and Windows services, written in C# and C++. The automated build process is a combination of native mode and .NET applications, VBScript, multiple batch files, and an assortment of compilers. It takes 50 to 90 minutes to build the product, depending on options. Am I Sisyphus[^] reborn?
Software Zen:
delete this;Just quickly .... SSD ...
I need a 32 bit unsigned value just to hold the number of coding WTF I see in a day …