home *** CD-ROM | disk | FTP | other *** search
- Path: mujibur.inmind.com!usenet
- From: mfinney@inmind.com
- Newsgroups: comp.lang.c++
- Subject: Re: Hungarian notation
- Date: 2 Jan 1996 03:50:14 GMT
- Organization: In Mind, Inc.
- Message-ID: <4caa1m$1vc@mujibur.inmind.com>
- References: <30CBAA22.791B@erols.com> <4agus6$qqc@news1.usa.pipeline.com> <30CE4E36.5190@erols.com> <30CFD9CD.3D09@zeta.org.au> <4c3ncl$gm4@nnrp1.news.primenet.com>
- Reply-To: mfinney@inmind.com
- NNTP-Posting-Host: finneyman.inmind.com
- X-Newsreader: IBM NewsReader/2 v1.2
-
- In <4c3ncl$gm4@nnrp1.news.primenet.com>, jrice@primenet.com (Jer) writes:
- >The above is "nice". But, when the boss walks in and says, "I want
- >this out now!", 'standards' go to the wind, relative to a pay check.
-
- No. Not if you are professional. At worst case, you will use a
- personal standard if there is no other standard available. Nor
- is there any such thing as a "one shot" program. No compromises
- should be made when writing a program just because you are
- told it is a "one shot" program. I know of far too many programs
- of this nature that have been shipped out the door. If somebody
- tells me that something is a one shot program then I write it, run
- it and erase it. Period. And even then it is written according to
- a standard.
-
- >Secondly, standards are only helpful if the "replacement" programmers
- >have been trained in The 'standard', otherwise they have to learn the
- >'standard' before proceding, which again puts them in jeopardy of the
- >first comment I made.
-
- Yes, but how long does that take? Five minutes, ten? Surely not
- more than an hour! You can afford an hour if it has to come out
- of your own time. And if the company is not willing to give a new
- hire an hour (or a day for that matter) to "come up to speed" on
- issues like this, then I strongly suggest looking for another position.
- After all, it is not something that (normally) changes on a day to
- day basis!
-
- My companies have even required short term contractors to program
- to the local standard. No quarter in this area!
-
- > Most of the managers that I have worked for
- >don't give a hoot about standards, all they care about is when!!
-
- But *YOU* as a professional programmer do! If the employer of a
- civil engineer felt that safety margins were not "necessary" for a
- bridge and the engineer omitted them, he could be sued for
- unprofessional behaviour. There are simply somethings you don't
- compromise on.
-
- Writing good code is one of them.
-
- Even under time pressure. Even if lives are depending on it.
- Because doing so is more likely to save those lives by *NOT*
- having a bug in the code than taking the time to be careful
- in the first place. And even that time is almost always paid
- back before you are done.
-
- Not that adherring to standards takes one second longer than
- writing random code. I suppose you could argue that indentation
- takes longer -- but I would suggest that even there, learning to
- type (or to type faster) would be far more significant. Not to
- mention using something besides EDLIN to enter you program!
- And the data entry of a program is rarely the limiting factor in
- writing code. Understanding is. And *THAT* is where adherrance
- to standards helps.
-
- In other words, expediency is *NOT* justification for throwing
- professionalism out the window! Have a set of personal standards
- which are the best you can make them which you can use if there
- is no overt standard or which you can use to fill in the holes of
- any existing standard. If you are modifying an existing program
- use the style and standard that program uses. If done properly,
- the original author should not be able to tell that he did not write
- the code after you alter it (at least without direct comparison to
- the original code). If writing a new program use the standards
- for the company. If there aren't any then use yours. They may
- well become the default standard. But *USE* one -- or get
- another profession.
-
-
- Michael Lee Finney
-
-