The cause of the degradation of programming
This is page 2 of the discussion. Read the opening post
Web Development
This is page 2 of the discussion. Read the opening post
by jaqui
which it what should be happening.
a good example of bad codeing practices is the linux kernel. :)
some of the least readable code around.
with the poorest commenting.
written by old school hackers for old school hackers.
by jaqui
a cell phone is a good example of this, yet nokia is going with embedded linux, and has chosen to use khtml as the ui for thier phones.
( the developer team just joined the khtml developers email list )
by jaqui
concern I have with it is the size.
the linux kernel is huge, so very limited devices, very limited fs type, this is what's done for embedded.
but khtml is designed to run on top of kde, which is a huge item.
using the core of khtml as a foundation makes sense, I'll have to wait until they haver had time to see what they do with it.
by rayjeff
also did 6809 and Z80.
As I shifted into high level languages I dropped the machine code. I could probably go back to it if I had to, but I'd have to do a lot of reading up on current architecture, we've progressed a bit since then.
by jaqui
the printout is the base lfs system.
all gui elements are in blfs, which is a download you would have to get.
so it's a full day of work to get a functional workstation.
set it up to your liking, but you only have one day.
by apotheon
I'm glad to hear that was of some use.
I'm a consultant, and part of that job entails being able to offer technology analyses for tasks that need to be filled for clients. That means I spend a lot of time learning about various information technologies and their comparative characteristics, so when someone asks a question about an area within the realm of what I've investigated, I tend to have a lot more information to offer than might be expected. As I'll tell any client that wants to hear it, though, only you know what exactly you need, so you should investigate possible solutions yourself where possible.
by tagmarkman
"a good example of bad codeing practices is the linux kernel. :)
some of the least readable code around.
with the poorest commenting."
I'll agree with you there... to make matters even more difficult the only book that I've read of O'Riley's that I hate is the Linux Kernel one... There are just soo many things in it that are blatantly incorrect.
by tagmarkman
"Must confess I rarely bother."
I don't use them for everything...
"For event driven aplications they aren't even suitable."
Yes and no. I use different "types" of diagrams for representing differnt types of data. Normally I use a combination between UML diagrams and XP coding methods unless it's a one off or I'm doing the coding just for me (then the code itself is the design).
"I'd rather spend time making sure my code is readable"
I do a lot of work in very large organization that often have a whole lot of stakeholders. It's critical for everyone to be on the same page. Almost every large company I have come into I can poke holes all over their business processes, procedures, and uncover huge issues in their code and business model. This is often because each person works in a vaccum or just within their group, they tend to forget about the communications end in favor of the code.
I don't think it's appropriate in all situations, it can be a waist of time and even counter productive in many cases, but in most of my situations it is a life (and project) (and Company) saver.
i'm totally with you stan. or maybe some of us here feel threatened by the new wave of macro-wielding, gui-laden development coders of our time. They cannot and will not be able to replace pure developers. The new coders can be productive in their own little enclave of macros but the real dough lands into the hands of those guys who can make those macros more efficient and I suppose this is the point of the whole thread. In other words there is an increasing load of responsibility laying into the hands of those who make the underlying code as computing power increases.
If you are just a programmer nowadays then you have very little options.
ps.. linux user 216679
by jaqui
even though modular in design. has code for evey filesystem type, for evey device on a system.
( maybe not newest device support but support for evey type of devcie ) this creates a huge pile of source code for it. a generic config wheen compiling it adds support for crap that you don't need on your system.
or worse yet, a distro uses a device type that your bios doesn't recognise, so when the device type gets loaded during boot it locks the system up.
( my system with scsi drivers actually, makes suse useless for me )
by tagmarkman
"even though modular in design. has code for evey filesystem type, for evey device on a system."
Yes..
"adds support for crap that you don't need on your system." .. "a distro uses a device type that your bios doesn't recognise"
Completely true.
I have mixed feelings about this issue. On one hand it can make the run of the mill user more likely to get their products up and running at least to the point to where they can upgrade their drivers. On the other hand, it can be a real mess.
MS handles device drivers significantly different. It allows the adding of a driver and developing for it easier, however, the way that it hooks into the ring 0 makes it easy for developer to muck it up when distributed thereby causing system crashes far more often than one would expect. It also tends to be more abused for speeding up products that should be run in User mode instead of Kernel mode.
Kernel development is nasty stuff no matter what OS you are looking at... Working in it too much will cause you to get old way before your prime. ;)
by kjamal
This may a very simplistic view but programming is going through what carpentry and fine furniture making has gone through.
Once carpentry was considered an elite skill and furniture cost a fortune. Then the masses began demanding lower cost goods. So the engineers brought in the assembly line, standard fit fixtrures, prefab ricated items, etc. Then the masses asked for even cheaper furniture, so the industrialists shipped production off-shores. But still, the masses demanded even cheaper goods.So the engineers created "assemble your own" stuff. And today, carpentry is largely a hobby.
That is what is happening with programming and many other once "elite" skills. Any commercialized activity (vs. artistic) is subject to the market's desire for more goods at cheaper prices. The more complex a task the less able it can satisfy this need. So it creates an opportunity to creat tools and processes that simplify the task and make production cheaper by expanding the labor pool.
Today, programming is often nothing more than "slapping together" pieces of re-usable code. There is very little real innovation.
Unfortunately, for the purist, this is continuing trend. Business want more for less and they want more people to be able to do it with less.
by hal 9000
With carpentry the way they they got around a lot of the costs involved was to make wood out of rubbish by mulching it then gluing it back together as a board hence Chip Board which incidentally falls to bits when it gets wet enough long enough.
Programing on the other hand might still be largely tool based but a lot of it while being moved offshore has not got any cheaper. I think I paid something like $80.00 AU for DOS and currently I'm paying $242.00 AU for an OEM version of XP Pro the same applies for the different Office programs that we used to use I think that Word Perfect 3 - 5 was somewhere around the $120.00 AU mark and the current price of MS Office is $385.00 AU for the SBE and $495.00 AU for the Pro Version both OEM. In its hay day I know I was paying about $140.00 For the Complete Lotus Suite and something like $200.00 for the original Word Perfect Suite when Novel had it. Now the latest WP Suite which has a lot less is around the $1,200 AU mark about the same as the Retail copy of MS Office and it is no better and in some things a lot worse. When you add to that price the extra costs involved in hardware which really has got a lot cheaper over the years the software has got more expensive in comparison. We used to think that paying $100.00 AU per MEG of Storage or RAM was a great price now $100.00 AU will buy a 512 MEG Stick of Cosair SDRAM or a 80 GIG HDD.
At a slightly latter time we where paying $25.00 AU per blank CDR which we can now buy for only a few cents even a DVDR is less than 50 cents AU.
The Carpentry Analogy may be correct when applied to hardware but not when applied to software which if anything has got more expensive while at the same time becoming less reliable.
Col ]:)
by tagmarkman
Lets say that it is "like" funiture making.
I have several antiques in my home. They are of outstanding quality. They survived me moving, if they get a scratch they can be buffed out, I always have got my moneys worth out of good funiture.
However, the assembly line funiture doesn't survive moves very well, scratchs are pretty much permenant, and they are not very pleasing to the eye (most of them at least).
If we are talking quality... we really screwed up with the funiture industry... if we relate that to software... well, I suppose you can expect the products not to be high quality... You can expect more bugs, more security issues, slower, bloated, unfriendly, and less productive.
If we are talking cost then part of it is the cost of the materials but it is also about the automation and the assemble your own as you stated above and distribution and a whole lot of other factors.
Software doesn't fall neatly into this catagory because it has no "physical" inventory. The production is really about brain power and "residual" sales. A lot of entrepenures then treat the intellectual labor pool as a commodity. In other words they treat it like inventory or an expense instead of the best asset the company could have...
by jaqui
programming and wodd/materials for furniture
old way, use REAL WOOD
new way
use glued together bits that will break easily!!!!
sounds like and excellent reason to stick large objects up the boss' hind end for suggesting using new ways.
by fatfinger
To Whom it May Concern,
Jaqui and whoever else this may apply to; there is a steady decline in the level of knowledge that your average "programmer" or "developer" has because of people like you who lord your knowledge over the less knowledgeable and spend their time pining over "the good 'ole days," which were never really all that great. That resentment, mistrust, animosity, or whatever you want to call it that you harbor, made evident by your comments untoward others whom you do not know and cannot possibly pass judgment on with regards to their overall skill-level regarding computers and/or programming languages, is the same feeling that others have for you. I gather that you have never truly served anyone else in a selfless capacity. Either way, if you really want to prove to the rest of the world how knowledgeable you are in comparison to us mere mortals, devise a way to prevent the rest of us from wasting any more of your precious time (those people who pay your salary can?t possibly be worth it, now can they?). Not everyone is a scientist and I honestly don't believe that it is entirely necessary to the task of programming. By the way, I am currently a computer science student and interestingly enough, all of the "real" programmers are to busy to teach the rest of us peons how to program. I may "only" write Macros at work, but interestingly enough, those programmers come to me more often than not because even though it is "only" a Macro, typing out 20,000 - 50,000 lines of clean, thoroughly-tested code is a little bit of a chore and they have better things to do. Although I don't consider this to be any great accomplishment, I see few other people there, save programmers, doing anything similar or more complex. Now that I've dispensed with the pleasantries regarding your post, I'll say this about the programming languages and open source/proprietary nits.
As far as the superiority of the products from a proprietary source to those from an open source is concerned, I tend to think that the amount of effort that goes into a product will be greater when someone's job is at stake. I realize that this is not always the case, but open source just means less documentation and questionable issue remediation with the vendor because there are fewer users and fewer paying users.
C++ is a wonderful tool that will allow you to solve any problem that you can precisely define. I am particularly fond of it because some of the things I do wouldn't be possible without it. Is it faster, cleaner, more precise? Absolutely. It is also much harder to make use of with an acceptable level of proficiency. For everything else, there is a scripting language of some flavor. My severely limited understanding of computer science notwithstanding, I seem to recall that the whole point of using a programming language with a high level of abstraction is to make it simpler, reduce errors, and increase productivity. Is there something other than pride that reinventing the wheel garners for one's self? I know how to build a working radio from junk, more or less, but the radio I get from Walmart has been put together by a team of electrical engineers that know more about electricity, radios, and testing methods than I, as one person, could ever hope to know. The reason companies like Microsoft invented tools like Macros is to make programming, development, general office staff work, or whatever else you wish to describe it as more uniform, with more predictable results, and faster. It may be perfectly acceptable to spend a month or so writing code for an application that will have a team of experienced programmers involved, but for the more mundane work that most of us peons do, Macros get the job done quicker. Every minute you have to spend actually producing the code is a minute you could have used for testing or modification. I don't pretend that anything I do will ever be as fast or as clean as anything written in C, but then again, I didn't have to waste 30,000 - 40,000 dollars (as a side note - a mistake I intend to commit to, fully) going to college to learn how to write a program to do something that my ham-handed VBA does without me knowing a blessed thing about computer science. Speed of execution is relative, too. I have replaced programs written in VBA that were well over a hundred times slower than mine, so speed is relative. We're not trying to put a rocket into orbit, just processing and manipulating data. As far as computer science is concerned, if you really want optimal performance from your database, then you're going to have to know a little bit about RAM, data structures, and database design. I guess the level of knowledge that you perceive others to have is only relative to your own. I don't expect many people to even attempt to learn what I freely spend all of my off-hours on. Mankind put forth the investment in knowledge and labor so they wouldn't have to. When I do become a "real" programmer, I'll create a bevy of fat, feature packed link libraries, chock-full of seldom-used functions, in your memory, Jaqui. You're right, as far as the title "inventor and scientist" is concerned, the rest of us may not definitively be programmers, but as far as regular users go, we're not all exceedingly dull, either.
by jaqui
our comments.
we were slamming the schools for not even trying to teach anything other than macro scripting more than the people that got stuck with the courses.
the people we were slamming were only those who refuse to accept the idea that they do have more to learn, even from those of us that never got a degree.
nowhere did I see any one of us saying we are not willing to teach those that are willing to learn.
you admit throughout the entire post you don't know what you are talking about,so how do you know what we are?
We aren't taling about the apparent productivity gains from using scripting/macro tools. What we are saying is that they require less technical skill, therefore the more complex/unsuitable a task they are used for the worse qualitatively the result.
Quality will always count, it's never a waste of time or resources as long as you accept the fact that it costs initially, the amount you save later on in a product with some reasonable longevity is enormous. There are no long term gains for a lack of quality, long terms costs is nearer.
Scripts can be and are the right tool many times, however if they are the only one available, they will be used, by those who do not understand their limitiations with very poor results.
by fatfinger
by hal 9000
I'll take you to task over another issue that you raised about the differences in documentation between Open Source and Proprietary products.
If you want free software you can download it and use it without documentation but if you are willing to pay for the Professional Product you do get documentation example SUSE has a massive amount of documentation as well as whatever Mandrake is now as they have just absorbed Lycoris and I'm not exactly sure what name they are going under this week. The retail Version of RedHat 9 has a 1,000 + page book with it and it sold for under $40.00 AU here with all the applications you could ever want. These places employ programmers and developers to write their code and make books for their products as well and they are really closer to the Uniform OS UNIX than anything MS has so far put out. It was MS who has attempted to reinvent the wheel not Linux as that is just another form of Unix which I might add in your computer Science course you should have been taught had the very first GUI developed by Xerox on Unix and then dropped when Xerox decided that there was no future in PC's. Every GUI since has been an attempt to reinvent that OS and GUI and it still is not up to the standard that Xerox had all those years ago on what would now be considered as antiquated hardware. 1,000 PC's networked sharing not only files but CPU, RAM and HDD space as required without the artificial limits imposed by places like MS for their own financial benefit.
Now that programming has become more mainstream it is no longer considered as a "Black Art" only practiced by those un-kept people who where never allowed out of their dungeons.
I also am at a loss about your misrepresented statement that we are unwilling to teach others this is so wrong as to be laughable the people that I used to associate with welcomed the chance to bring new blood into their ranks but back in the days when I was punching code it was the new guys who where unwilling to learn what we did and how we did it they where fresh out of whatever education course that had done and "Knew It All" and could not be told a single thing, I know from bitter experience that this was the case as I ran into it more times than I actually care to even think about now days.
A perfect example was what started this entire thread off back in those days we where making engine management systems and only had 64K available to write the entire code and then we always needed a bit left over for changes when I gave a new guy a subroutine to rewrite he came back with it complete and only using a little over 120K the fact that it was only a small part of a much bigger program and that it by itself would not fit in the available space was to him immaterial as that was how he had been taught to write code while at school and any different way was just not going to happen as long as the sun shone.
It is people like this who do not want to continue learning who will be the chaff split on the wayside when things get tough as there will come a time when Code Bloat will become unacceptable and instead of tinkering around the edges to make something work better we will have to rewrite the entire thing and not consume clock cycles at an alarming rate unnecessarily by shoddily written code that does it all with no necessary help from a programmer it just is nothing more than a super compiler where you enter the required items for the program and it then compiles the entire thing for ease of use and then maybe thinks about some form of security which again most likely will be another add on consuming even more system resources with a much bigger footprint and higher overheads so effectively there will be very little left to actually run the program let alone enter data into it and more importantly manipulate that data so you get the required answers in a short time. Computers where supposed to make life simpler and remove the hard work that was required before they where around and freely available so people would not need to spend weeks with a "Slide Rule" doing calculations only to find out that they had made a slight mistake and then had to redo all the calculations again from the point of the original mistake. It was possible on a Z 80 CPU machine to do these CAD calculations in a matter of hours that would otherwise taken weeks to perform. But now days with much faster hardware the processing time is again rising instead of decreasing. Working IT doesn't just involve working with Data Bases there are many different areas all specialized in their own right which not only need addressing in a professional manner but deserve this as well.
Despite what you may have been taught MS doesn't rule in the computer industry and they actually do not make a lot of products that address many areas of the market. Data Bases are also not the mainstay of the industry either they are a useful tool for a specific application nothing more the fact that a lot of places actually use them in one form or another only means that they are useful and that it is important to have them work as well as possible instead of the way that things are going now packaging more items into them at the expense of performance. The reality is that todays programs on much faster hardware doesn't execute much faster if any than the older stuff on much slower and far more expensive hardware.
Now why do you think this is that way?
Col ]:)
by tagmarkman
I would venture to say that most of us here do teach to some degree or at least mentor. If they didn't I doubt they would be professing here regularly :) I've taught at Universities as a visiting expert, I have mentored and lectured at companies, and I've giving lectures at hardware and software conferences. Our knowledge is more valuable when other people in the industry understand.
That being said, I've never saw a developer, or tester, or architect, or administrator, or any high tech person worth their salt that didn't continue to learn "outside" of the University.
Still, I don't think anyone has disagreed with you. We all start knowing very little. I remember being dumbfounded when I first saw a language without line numbers, I couldn't wrap my head around it. I laugh at this thought today but then again I was in my teens.
We don't have a problem with productivity of Drag and Drop, we have a problem with the watering down of the developers because it has helped contribute to. Now I have to interegate my new hires to find out if they are good at assembling code through drag and drop or if they are a software engineer that can fix the bug found "inside" the component.
I'll hire both of them because if they are smart and good at what they do and they show a desire to be better but one is going to get better pay and I'll cater education to make them better developers but I need to know "where" they are in the technological ladder for me push them in the right direction.
Knowing what each person is really good at lets me cater their workload that challanges them without drowning them. It also allow me to ask the right person the right question so that I can continue to learn. I'm not going to ask a senior engineer about the column structure on a C# grid control when I can get that from a less experienced developer and I'm not going to ask a junior developer about a C++ theading issue that tends to randomly crash an application when I'd be better off going to the person that has dealt with that in the past.
You understand the difference between the levels of experience, the level of skill, and the deeper understanding knowledge between developer. However, many in the corporate world don't (which is our real complaint). Often, to them, a developer is a developer get it out the door so we can make money, but then they are stunned when they learn about security issues, poor performing products, loss of revenue, a poor bottom line, and difficult to use products.
You should be concerned too. When I was fresh out of college, I could command the equivilant in todays cash of over $80K yr. before bonus. Now, fresh graduates make a fraction of that and yet there is more demand than ever for developers.
by jaqui
by hal 9000
The problems start to arise when people do not know their limitations and blindly rush in to do a job that they are just not capable of doing with tools that are not suited to the job in hand even though they may be totally suitable for another application.
Where the main problems start to arise is with the Non Technical people who think that because you are capable of doing one application relating to computers you must be capable of doing other things so you are handed a brief and told to write something with your current tool kit. Is is possible Yes!
But that really is not the question what the question should be is it the best way to do a particular job that requires doing?
I've seen numerous Macro writers who seem to think that because they are capable of writing a reasonably good Macro they are capable of doing anything at all from writing that Macro to developing a whole new application, it is these people who are responsible for Dumbing Down a lot of the industry and what is even worse the accountants who now generally run the companies believe that because they have someone capable of doing one thing very well they must be capable of doing other things just as well only on a much bigger scale.
I've seen far too many people just dropped in it way over their heads and with the build up that they have had over the years are under the mistaken belief that they are capable of doing whatever it is that is asked of them. These are the ones that I worry about.
Col ]:)
is the english comment.
The problem isn't that people use 'quick tools', I do a lot of Drag & Drop myself, it's when they believe that ability makes them capable of doing wih a scripted language something it's wholly unsuited to.
I use scripting languages, drag & drop a lot. I'm capable enough to avoid some of the more expensive but quick ways of doing things with them, but there are many tasks I wouldn't even try to use them for.
Unfortunately I've seen a lot of designs where the limits of these tools was not appreciated or was ignored and the resultant mess tended to get extremey expensive. I've seen people with enough knowledge to dangerous address inherrent inefficiencies in scripted environments and absolutely crap all over their design in order to alleviate the problem.
You can stretch your own limits and expand your skill, you can chuck more memory and processor at a problem, you won't ever make a script that's more efficient than well thought out compiled code though.
That's only one quality issue though, quality of the deliverable, quality of the code is related but not exactly the same thing. You can in theory at least get a good quality deliverable out of poor code, the next version will cost you though.
A simple thing like variable declaration and typing can make a massive difference to code quality.
As for the more complex structures, concepts and algorithms we use, when you need them you need them.
by marathoner
It's the 90/10 proportion mentioned earlier in this discussion. An office clerk can use the GUI produce a query like
"SELECT * FROM customers WHERE customer_id = 42"
and so he thinks he's an IT pro and he's a nice guy so he calls himself an IT pro and wears a tie and gets a bigger desk and everybody thinks of him as the magic query guy. Meanwhile his access database is not relational at all and he's running circles going thru all kinds of manual labor to do something that would be handled easily by a simple relational join query, but it does work, and look what a hard worker he is, everybody knows that IT is hard work and the system grinds away. Bottom line: it works and nobody knows any better.
by tagmarkman
He was always busy. Nobody questioned him because he would always move down the hall very quickly with a notepad in his hand.
He'd be given an assignment to do and he'd sometime take days to get it done. When I took over the department, I made it a requirement to include the query with the results not because of him but because I felt it necessary to understand HOW a statistic was created.
That's when we found out he was running single queries putting the data in excel fiddling with it, etc to come out with the numbers...
He was a good guy just didn't have the knowledge, training, etc to make his life efficent... I didn't get upset at him... instead, I send him to SQL training and he became much more effective... He was also very aware that he was not a developer. I can't say that about everyone I've seen in that situation.
by tagmarkman
You are not the problem. In fact, if more developers were like you, we wouldn't have the problem that we are discussing (aside possibly from lower pay).
The key word here is willingness to learn. I truely believe that many people can be good developers but they have to have passion. I know people in the field that are there simply because they can get a paycheck. These are some of the people that have turned the computer field from engineering to a trade.
A person truely passionate about computers want to know more than the surface. They want more than the bare minimum to perform their job. If they use components, they want to know more about the components (how to they work, how can it be changed, what are the flaws, etc.). If it's hardware they want to know more than just how to swap out a video card.
By making things easy (although I do agree with making things easy to use), has allowed a massive group of people into the profession that do not have that passion for technology.
A regular person usually does not have the ability to tell the difference between an engineer that does have that drive and some hack that doesn't. Hence, they get lumped into the same category and a persons experience with them are decided by that interaction... sadly.. there are a hell of a lot more below par devs than they are of good engineers. So our odds of giving them a good impression first is fairly low.
reason behind uml, flow charts, ERDs etc was to communicate a design in a non proprietry ( as in not code) format. You were either communicating to clients, the non-technical or the one I feel most important a project team. Pissing about with a stencil for a month before you get your first semi-colon in the source when the only person who needs the docs is a waste of time. Your design and it's implmentation will diverge farther apart dependant of the detail of the document. Equally when you have to teach the client the notation in order for them to understand the documentation, the plot was missed in it's entirety.
I'm more than happy to do some rough draft scribbles, even to jazz the result up for a draft high level specification.
So I do do it when the idea is to communicate a design without actually implementing, as in a high level design for a junior devloper, of for a rough costing for a quote.
Equally I love doing this sort of thing as a joint sort of brainstorming exercise in a team. If I was producing it as a deliverable though, the documention would be done from the implementation, as doing that sort of thing twice sets my teeth on edge.
by kjamal
thanks to reusable code, development tools, and greater processing power, the big mass of boom time developers that these fetched in have to take naff salaries because the market is glutted with them which effectively de-valued the entire industry.
Quality is coming back though, so those laurels you are sitting on are about to become brambles.
by tagmarkman
I don't think I completely agree with you. I understand what you are trying to say (today is about reusing what developer made for them yesterday). I agree with that and that has degraded this engineering field.
There is no separtation between skill sets. A construction worker that builds framing for a house does not get lumped into the same category of the architect that designed it. or even the material engineers for that matter. They are distinct. In development... the assembly line developer is lumped into the same category as the innovative software engineer.
Rates in 2000 was not because programmers were good. The waters were well mucky by the time the bubble got that big.
I also don't think that the lower cost of development was because of resuable code, the development tools, or processing power. A lot of it has to do with the market. More developers are willing to work for less.
Do you get the same output... from 2000 probably... from 1992 I doubt it. It might be the same (numbers) that are produced but not the same innovation. If the majority of development today had the innovation of the late 80's and early 90's (and sooner), I would expect a lot more from development today.
What I see improvements recently have been in are: patterns, processes, diagrams, usability, and open source (which is not coding) and questionably (security and error reporting). I'm sure there are others but that is what was on the top of my head. Note that none of those are achieved through drag and drop.
by kjamal
by tagmarkman
Correct... I was having this very conversation on another article. Programmers did it to themselves. It happened because of a number of reasons an you just listed one of the biggest.
Developer put a very low value on their work because they wanted "experience" they wanted "recognition" they wanted "immediate" financial satisfation and a low price seemed just find for their current lifestyle and didn't think too far ahead or how it would effect he economy if eveyrone followed suit. Then there was foriegn visas and outsourcing that slammed the nail in...
"Just make it work" cost us dearly but the labor market is what really did us in and we did nothing to protect ourselves... I tried a long time ago but I eventually threw up my hands after talking to deaf ears and concentrated on my work.
And BTW: I think I hired most of them too :) cause I don't see many of them anymore. At least I know you picked them up and they didn't become waiters.
by jaqui
I do not.
( work for you )
;)
but, since I only support professional quality oerating systems, any company that has support for m$ products will never get me.
( making assumption based on the 90% of companies focussed on m$ crap )
by tagmarkman
by tagmarkman
UML is not a hammer... too many people use it as that. I have two types of clients: external to my company and internal to my company.
External client rarely if ever see the UML. They get pretty pictures you see in magazines or cool presentations or even visual mockups. Funny thing is that most of it is so generic, we can use the same diagrams over and over again and from client to client.
Internal to my company, every stakeholder knows basic UML notation for the diagrams that they would use. For example, most managers wouldn't care less about a class diagram but is critical with the use case diagram. My testers don't care much for the use case diagrams but want the activity diagrams. and so on..
"communicate a design in a non proprietry ( as in not code) format."
It was design to visually represent a problem and solution in order to communicate more effectively. I've put out specification, and had no engineer point out problems with it and I gave them more than enough time and incentive. I sent out the same problem visually with UML and I got TONs of feedback both critical and mundane.
By putting out those diagrams on project compaired to the ones that I didn't, I found over a 40% decrease in bugs. That saved my company a huge amount of money, and spared my developers a lot of grief.
If it's a small shop, I don't think it's worth the effort (aside from a basic one here or there or to really think through a confusing part).
"Your design and it's implmentation will diverge farther apart dependant of the detail of the document."
This doesn't not have to be the case. In fact, I use it to help understand "what" was different from the original specification and how to improve and prioritize designs later. For example, if I have a class diagram that an architect creates and the software engineer agree on... The engineer codes the product and deploys to test, I'll run a reverse engineering UML product and compare the design with the implementation. I will almost always find differences. Then, I can ask why? Usually, there are good reasons and the architect can take that into consideration on the next project but I also find that sometimes the engineer just didn't understand something and implemented it incorrectly although it "seemed" like it worked.
"Equally when you have to teach the client the notation in order for them to understand the documentation, the plot was missed in it's entirety."
My external clients get no UML.
"I'm more than happy to do some rough draft scribbles, even to jazz the result up for a draft high level specification."
Usually, that is all that is necessary, but as the program become more complex and more people are invovled it become more difficult without being a little more detailed. Still, I must say that many people tend to over do UML and argue the the tiny things instead of saying... "do you understand what I am trying to say..."... now lets get on with it... We get paid to deliver....
"Equally I love doing this sort of thing as a joint sort of brainstorming exercise in a team."
Good stuff for brainstorming... I tend to use the XP card method.
"If I was producing it as a deliverable though, the documention would be done from the implementation"
My company codes to specification. but I would NEVER have a developer both spec and develop (aside from input), or develop and UI (aside from influence) or develop and document (aside from self document code), or develop and test (aside from peer testing).
Everytime a developer is force to do double duty you can expect a lower quality product or at least a slipped project date.
by davbrook
Once upon a time at the beginning of our marriage a question came up about which one of us was 'smarter'. I said, 'you think you are as smart as I am?' to which she replied 'I believe everyone is smart in their own way.'
I felt my blood pressure go up and was about to launch a heated reply when (and I like to think God stepped in at this moment) the phone rang.
She answered it and it gave me a moment to think. I realized there could not be anyway this conversation could have a happy ending. In fact, I could think of no way anything could be said which would not be hurtful to either her or myself.
So I dropped it.
However, those words about how 'everyone is smart in their own way' has stuck with me these many years.
-----------------------
I think programming is much like that. A programmer is one who uses a computer to solve problems by writing instructions to direct the computer to do some task(s).
Having said that I should point out it really covers a large amount of different skill sets. It covers anything from writing macros to writing multidimensional applications. I can write programs is several different languages but do I want to spend my time writing queries or macros? No. I can but it is not something I want to do. Having someone around who is very good at a limited skill set is a benefit to me because it lets me go do the things I good at.
I wish I could tell you there is someone out there whose skill set allows them to comfortably sit down and program in any language or any operating enviroment. He or she doesn't exist. I can tell you there are people out there with lesser skills than you have but there is always someone out there who has better skills than you have. That is a truism you can bet and a truly health perspective.
-------------
Your employer generally has a different set of standards than you do. How can you expect him to see your job the same way you do?
by garyklinecc
Programming languages have been constantly evolving and will continue to do so. There's at least one university lab, if not others, that is working on a natural language system that writes programming code based on a conversation between it and a human user.
Programming will continue to become easier, as more knowledge is 'bound up' within the compilers. And yet, good programming design is still a high level skill.
There was a time when people who were skilled in writing applications using Lotus 1-2-3's macro language were high-tech MIS experts. To a certain extent, that level of 'macro-programming' is still beyond the typical user.
If macro-programming and SQL really programming? Heck yes! Is it easier for 'typical users' to dabble in it? You betcha!
But that makes them 'power users', people who have gone beyond the basic functions of the applications and are now unleashing the full potential of their apps.
People with the analytical skills and training to breakdown a problem, and construct a solid solution will continue to be needed. Yes, we're seeing the 'power users' replace the 'inhouse programmer'. But to be truly effective, such power users will need to learn good programming skills.
by jaqui
"There was a time when people who were skilled in writing applications using Lotus 1-2-3's macro language were high-tech MIS experts. To a certain extent, that level of 'macro-programming' is still beyond the typical user."
it is not application programming.
it is business logic programming.
appliucation programming is writing the code for lotus 123 itself.
if the end result is not runnig straight through the kernel, no other app required, then it is business logic programming, not application programming.
so your application specific macro writers do not do application programming.
by tagmarkman
I don't think anybody here disagrees with you are "ever one being smart in their own way."
I've always belived that (almost) everyone was smart. However, I also belive that most people tend to be lazy or don't have the passion for their work.
What we were talking about was not if one developer was smarter than another but more of do these tools allow more people (of lower skill in this subject) build programs of inferior quality? And what effect is that having on the software development community?
My issue is that public opionion is that a computer person is a computer person.. it's difficult for them to see the difference and we've done a poor job at educating people of this. We tend to just zip our lips and roll our eyes.
I do expect my employeer to see my job the way I do. I certainly try to see my employees from their perspective. How else can you be a truly effective leader? Any manager that is that disconnected from their subordinates shouldn't be in that position.
by apotheon
You say you don't think anyone here disagrees with the notion of everyone being smart in their own way. You're wrong, unfortunately. I disagree.
I disagree primarily on the level of the sentiment. Oh, sure, you can make a case for the notion being "technically correct" in some way, and I'm good at performing such verbal and logical gymnastics, but at the real heart of the matter it's just not true, no matter how much my mother wants it to be.
Let's take that example of programming. As applied to the programming discussion, this "everyone's smart" thing is being used to justify calling people who play with macros as part of their job descriptions "programmers". Tell me something, then: If forced to write macros, how well do you think a real programmer would do? I rather suspect that the majority of Programmers would kick ass at the task. If you reverse the direction of migration, though, and start dumping "power users" into integrated development environments, you'll get crap code if you get code that'll run at all. Even if you give them a stack of manuals and a bucket of good source code, then stand back and give them a couple years to get up to speed, the majority of them will at best achieve the worst programs you can imagine. A few might excel, proving that some people are just smarter than others.
Take these same people, those who excel at programming and those who suck no matter how much work they put into their skills. Drop them into a literature and writing class. Leave them there for a couple years. Make their lives depend on their skill levels when they come out. Get them to write some short stories for you. There will be a few people that were bad at programming who will excel at this, but mostly the people who are incapable of programming effectively will also be incapable of writing effectively.
Here's another one for you: Take a bunch of physicists and historians. Give each of them two years to learn the others' trades. Anyone that doesn't make significant progress dies. Watch the physicists succeed where the historians, for the most part, fail.
There are [b]some people[/b] who are "intelligent in their own way", and there are a great many people who are more intelligent in one area than another, but for the most part people either are intelligent, or aren't. That's just the way it is. The fact that I prefer working with computers over, say, sewing doesn't mean I couldn't be a whiz-bang tailor if I really wanted to. I'd like to see you turn a random crocheting luddite hippie into a unix sysadmin and programmer, though. People tend to end up doing what they do in large part because it suits their intelligence level, preferences, and dedication. It typically has nothing to do with any "type of intelligence".
I've noticed that people who are truly incompetent in any way are typically incompetent in every way, and it drives me up the friggin' wall on a daily basis.
kicking at your place then. I've been doing double duty for so long I'd forgotten there were two duties.
LOL
What am I on about anyway, three duties I do the database design as well, and the optimisation.
It's certainly a reason approach to solit, if you can avoid the big picture problem, but to be quite honest being ajust a coder would n't satisfy me and not being able to implement at least some of my design would also be downer.
As for over use of UML, I couldn't agree more.
When I was talking about divergence from design to implementation, I was think more practical difficulties, such as denormalising, defferred instantiation etc. As I do do double duty I do understand the design I'm meant to be implementing as it's mine. I have to agree about the skill factor though a good business analyst is worth their weight in platinum, as it's not my strongest skill, also theres the learning curve in every new business environment to surmount.
Still could be the Vienna Developement Method instead.
by tagmarkman
"I've been doing double duty for so long I'd forgotten there were two duties. LOL"
Ain't that the truth... Waaaay to many company get caught up in that. Sometimes it necessary... other times... not so... :(
"but to be quite honest being ajust a coder would n't satisfy me and not being able to implement at least some of my design would also be downer."
I agree with you. It would frustrate most engineers but I'm pretty strong in the belief that developer not get attached to their code. That is one reason among several that I don't have designers implement their code. But here is the cool part, on one project you might be a designer, on another project you might be a coder, and so on (just not on the same project). It gives engineers the inside and flexiblity to do several jobs and understand how to work with each other better. The tricky part for me is to make sure that the person doing the work has the right skill set for the job and that can be very tough to coordinate and as a result some people tend to do more of the same thing than others.
"divergence from design to implementation"
Very true... It often doesn match perfectly... and that's ok... To keep it as closely matches as possible I try to keep projects contained in a 2 day, 2 week, 2 month cycle. It tends to keep every one in sync and I know when to choose low priorites projects (anything that falls in Dec).
"Still could be the Vienna Developement Method instead."
Awww common... there was always Z notation :)
by jaqui
here the only unix only shops are mainframe shops.
the others are *x webservers with windows desktops
rarely *x servers with windows desktops.
so when they are looking for someone to support thier networks, specifying *x skills, then they say also win2k/xp/nt I start laughing as I walk out.
sorry but windows support is from microsh|t, not from me. if m$ won't support thier product why should I?
( an m$ has no real customer support, even for business customers )
by tagmarkman
"here the only unix only shops are mainframe shops.
the others are *x webservers with windows desktops
rarely *x servers with windows desktops."
Very true...
We have 95% Unix Servers (FreeBSD, IRIX, and SUN) we even have a Vax (gasp) and about 5% Windows Servers (a necessary evil in our work).
Client side we have about 80% Windows and 20% full mix of IRIX, SUN, Linux, FreeBSD, Mac, and a few others include WinRiver and such.
"an m$ has no real customer support, even for business customers"
Incorrect. Corporate support is alive and kicking. A lot of effort goes into this especially with the OEMs and large business partners. End-User support is pretty limited which is what you are probably basing your judgement on.
by tagmarkman
The moment you start believing that someone else is not smart, you are underestimating your opponent. I also did not say "everyone" I said "almost" everyone is smart. Also, smart does not mean genius. It means that most people are smart about something. This is not logical gymnastics.
For example, my mother has a 8th grade education and I've always considered her well... dumb or uneducated or ignorant but very hard working. But later in life I found that she was absolutely brilliant when it came to understanding relationships... To this day, I've met no one that can match her in this.
In your example, you are making level comparisons of people. I'll agree with you. There are cetainly levels of understanding. And I agree that some poeple are "smarter" in general than others.
When I was in Africa, I was given an assistant. I stayed in a hut but the bugs were horrible. I was given a netting for my bed. It was rectangle. I was going to put it up myself and it would have only taken me a moment but my native assistant insisted to do that work.
After it was done, he appologized that it was too small for the bed. I almost laughed because when I look at it the short ends were aligned with the long ends. I could lay on the bed and have my feet and head exposed and stretch out my arms and it still be covered.
My initial reaction was, "It's official, I have just met the dumbest person in the world." But I changed my mind during the next few weeks. He had such a grasp of verbal communication that it baffeled me. He certainly had a better grasp of it than almost anyone I've met in the states.
If you are speaking "smarts" in the ability to learn and assimalate quickly, then yes, I'll agree with you. If you are talking about somebody ability to be intelligent about "something" then I have to disagree.
by apotheon
You're not talking about intelligence. You're talking about knowledge within given fields. That's an entirely separate concept.
Frankly, the average human being spends a significant percentage of its life learning a skillset that the most-intelligent two percent could master to an equal or greater degree in a matter of months, given motivation and a minute shift in perspective, while the opposite is not true. Yes, a given person of average intelligence who has spent his life in the field of "sanitation engineering" may well be better at unclogging toilets with a plunger than any low-energy particle physicist, but with a rearrangement of priorities that physicist could probably come up with a better way to unclog those toilets with the same available resources in a few days.
I guess that what you're actually talking about is that everybody can offer a contribution that is uniquely suited to that person, though in some cases that involves unclogging toilets because low-energy particle physicists' time is better spent examining the building blocks of our universe, and forcing physicists to clean the toilets in the lab is a waste of resources.
by jaqui
getting them to clean the toilets is good for controling thier egomania. ;)
after all if they have to scrub crap out of toilets even with thier skillset they aren't any better than the janitors.
( can ya tell I'm from a school of opinion where everyone does every job, just so they have an appreciation of what the others go through )
by tagmarkman
Yes, I'm talking about knowledge and skills within a given field.
We are talking different... I'm talking EQ and your talking IQ.
Once again... We agree without knowing it :)
by jaqui
so how much a month on top of the thosands a month ( pro rated ) for the software licensing?
good money after bad to use windows. ;)
I was actually referring to m$ response to a service call for a full install of windows.
" it's an oem, you have to talk to hardware supplier "
me: kay, it's a home built system with a storebought full install, you are the supplier whose product is failing, you are the correct person to be calling.
we can't help you, cause it's not an upgrade version.
so the guarantee that is printed right here means that microshaft willingly chooses to go to court over a breach of contract lawsuit?
no
then you will support this product and get this danged copy of crashware activated before the lawsuit is filed!!!!!!
on 10 separate occasions, dixing people's systems.
by tagmarkman
"so how much a month on top of the thosands a month ( pro rated ) for the software licensing?"
The support is without charge for the business that I dealt with, but I don't know of the cost for small shops.
"I was actually referring to m$ response to a service call for a full install of windows."
Does ANY company do that that is not a VAR?
"it's a home built system with a storebought full install, you are the supplier whose product is failing, you are the correct person to be calling."
Yep... limited free support (robots) which in my eyes are close to zero. This is something I disagree with but (in their opinion) the store that sells it (according to the reseller agreement should support this).
by apotheon
I'm pretty good at cleaning toilets. I'm not such a good cook, though.
C'est la vie.
by hal 9000
But here any VAR should not Activate any M$ product as if they do the end user can not read the License agreement and agree with it.
I've run into this problem as well with mass produced systems from places like HP, Dell and the like when they require reloading because the partial install of XP dies or some more hardware is added MS goes through exactly the same thing and tries very hard to make you buy new products preferable Retail not OEM.
The last time that I went through this was with a HP something or other which was all of 3 weeks old and had no AV product installed the system wouldn't even boot and was badly affected by Viruses so the owner wanted a "Clean Install" performed and since they where paying for my time I did exactly this after first cleaning up the system in about 10 minutes. I backed up their data and wiped the HDD reinstalled the software from the Recovery CD's and when I rang MS for product activation they spent at least 20 minutes insisting that it was Pirate Software. I just replied take it up with HP and sorry I do not know where is was bought from as I didn't sell it or buy it I'm only fixing it with what looks like a Genuine COA and the Original HP CD's. I then went on to insist that even the shop that sold it was not responsible as they where only selling what HP supplied so to take it up with HP as if anyone was doing something wrong it had to be HP and not the poor retailer who was selling the HP product. :)
I used to go through the same thing on my test station where I tested new hardware as I would need to reactivate 17 different MS products that actually where only on there for testing purposes and not actually used for anything other than making sure that a bit of hardware didn't interfere with their working. I latter switched that install to a Volume License copy to avoid having to constantly ringing MS for activation codes. It was just plain and simple painful. ;)
Col ]:)
by tagmarkman
"But here any VAR should not Activate any M$ product as if they do the end user can not read the License agreement and agree with it."
VARs don't activate M$ product here either as far as I'm aware.
Activation has bugged me from day one. It uses hardware identification of several hardware components. Change a few components and it's downright annoying let alone if you buy a few of those components though Ebay...
by fstallwo
First, bear with me as I eloborate on the obvious fact that the technology ranks have now been infiltrated and filled by many whom have never ever had even the slightest inclination toward technology and its inner workings--much less, in programming. This was as apparent during the analog age as it was during the advent of the digital age. And, it is apparent now to more than myself in the information age, as it was during the foregoing precursors I've just mentioned.
One must come to simply understand that historically, we can readily observe this group of infiltrators is the same group whom have always regarded, or viewed, something as simple as setting up the various options of a VCR as something akin to programming, e.g., channel designations and time of day recording timers.
We can further observe historically--beyond the home environment and wisked into that of the office setting--the same view, or perspective, holds true when acknowledging the fact that this very same group was predominately assigned the typical function of PBX operator--a task that more frequently than not involved setting up the PBX's various telephone extensions and mailboxes due to many a company's policy of of utilizing employees to the fullest capacity and benefit. Clearly, it is no tall task to observe there is no more programming here in this particular job function than there is in the setting up of a satellite television receiver, but alas, so too has it been called programming. At best, any of the examples given thus far can be called configuration. Nothing more, although perhaps something less, namely, setup.
So now, in this present day and age of informaion technology, where anything akin to using a Querty style keyboard or other conglomeration for the purpose of inputting and/or extracting datum, data, variables, parameters, information and the like, this group of whom I've spoken will invariably seek to infuse its ego (a gender agnostic quality) with splendor and also to raise itself to the heights of professional programmer, and in severe cases even to equate itself to such indeed.
I for one have seen these infiltrators--imposters that they are. I have seen them slithering into the ranks of the professional installers for years upon years and I am here to tell you that I agree: there is a formidable amount of degradation that has occurred within the ranks of professional programming. However such degradation has not come to pass by the hand of the professional installer. The professional installer knows that setup and configuration are one and the same. The infiltrators, imposters that they are, sees such activities as programming. This due to their innate character flaw: perpetual self-admiration. The goal of which is always personal gain--or gold digging-- through any means, and at any and all costs.
Such has it been since the beginning of time in human existence.
by jaqui
how about the people that call installing an os and additional software programming?
I hear people saying that so and so is programming the computer, look at so and so, and they are installing software.
by hal 9000
To most people just being able to push a key on a keyboard or move a mouse around makes them a "Computer Expert!" :^O
It's only fair that someone who can install a copy of windows with out the OPK kit should consider themselves a programmer. After all with out what they installed the thing doesn't work right does it? " :D
Col ]:)
by fstallwo
Point made and well taken. I totally agree, and I must confess: I don't get it. What planet do those folks come from?
While programming isn't my forte, I must say I certainly know the difference between programming, which involves significantly more than just the act of coding.
Programming occurs long before a single line of code is written -- and well before a single com, exe, dll, msi, or other computer input has been developed, input to a computer, witten to media, and distributed.
The inference that properly setting up a machine -- any machine, that is -- which consists of installing, configuring and verifying functional/operational readiness is programming is . . . to put it in Mike Tyson's words . . . absolutely ludicrous.
So yes, Jaqui, I'm with you here: they don't have a clue.
by jaqui
talking about installing windows.
it's not like you need to be able to think, ms has removed that requirement. ;)
by hal 9000
MS have put that 25 character string in there that you have to type in correctly to get windows to finish installing.
But at least they have removed the need to tie a Key to a CD so now you can use any CD from that family to install the product but there is still that Product Key that has to be got past. Now that requires some brain power if only to read & type rubbish at the same time. :D
But if you really want to talk about a total joke I stuffed up good this week I have an old 200 MMX with a 1.2 GIG HDD with Woody loaded on it. Well I've had a copy of the Sarge install Cd's around for quite a while so I thought I'll have a play so I wiped the HDD that Woody was on and proceeded to install Sarge all went great until I got to the stage of downloading available updates for the Kernel before the install was done as I was told that there wasn't enough HDD Space available so I started again and this time chose a minimal install and got exactly the same message again. Next time I chose to use the supplied install packages and got exactly the same message again. Well it's newer so I thought it must be bigger so I went to put Woody back onto the unit and got exactly the same message again. Now I was getting confused as I already had Woody installed on this machine and I was now being told that there wasn't enough room to install it. So I wiped the HDD and started the install process again but got the same results. so I pulled the box out of the rack that it is in with all the other Linux Boxes opened it up to find that it had 2 HDD's in it one being the 1.2 GIG and a 124 MEG drive that I had been using for a swap file and I'd been trying to install both installations onto the 124 MEG drive no wonder it wouldn't fit. :)
Actually there is a lot to be said for actually looking at the install instructions and not just treating it as a side issue when you are doing other work. I could have saved myself so much time if only I actually took the time to look at what I was telling the poor thing to do. ;)
But whats it matter I was only killing time during installs of a couple of Windows Boxes so I wasn't even really paying attention to the poor thing. :p
Col ]:)
by jaqui
though when I'm getting paid to fix or install windows I usually pull out my knitting.
yes, knitting. literally.
for some reason this really bothers people.. to see a tech knitting while doing something as "complex" as installing windows.
by apotheon
by jaqui
that's why I do it.
when they ask about it I simply tell them that since ms products can't be ignored and require input during installation I have to sit here anyway, this is multitasking and keeping my stress level down.
they are fine with it then.
by hal 9000
When I was working mainframes and downtime was a no no when I got called into a place where something had failed we had a test program which we could load run and diagnose where the problem was. Of course this relied on the thing actually working which most of the time it was but falling over due to some hardware problem while crunching numbers. So we used this diagnostic utility which was great but!
The people who where paying for my time and counting all the lost income would invariably get hot under the collar when they walked is and saw me playing what looked like "Space Invaders" on their Multi Million Dollar piece of hardware which was bad enough but then I was wasting time as well and they was what really hurt. One guy in particular went berserk and demanded that I do the job that I was called in to do and no matter how much I tried to explain that is exactly what I was doing he seemed to think that pulling covers and digging at the internals was the way to fix things. Once he had destroyed the 11 inch floppy I did as he asked and spent 3 days finding the problem that I could have found in under 15 minutes and spent another 30 minutes fixing.
Seems that it was perfectly OK to waste several days in his books than to spend 10 -15 minutes fault finding. But what did I care he was the one paying the bill and he had to account for the lost processing time. The next time that went there I was allowed to run the diagnostic and had the thing up and running in under 1 hour. :D
I suppose it just might have been the explanation for the long repair time that got the job done right the next time. :)
Col ]:)
by fstallwo
And that's just the kind of thinking management seeks to retain since industry's objectives are: increased reliability, availability and performance.
All of which translates into machines requiring little to no human intervention.
So nowadays, all anyone, er, the "programmer" that is, has to do is know how to stick his memory stick into the little hole on the front of that baby! Whoo! Hoo! Yeah, baby! Yeah!
Right?! :-D
by fstallwo
What if you were paying an attorney to prepare your case and that attorney, while waiting for research materials, e.g. caselaw, to download from the internet, pulled out some knitting and began knitting instead of doing something you're paying oodles and oodles for--such as preparing an interrogatory or subpoena for your case.
Would you be happy?? Or, would you too be bothered??
There's a boat load of constructive things a tech can do while software is loading. Please refrain for diminishing the profession.
Give the customer the entire TIME they are paying for. Everyone in the entire world understands the concept of time = money.
They're paying the MONEY, you are professionally obligated to MAXIMIZE THE PRODUCTIVE USE OF THEIR MONEY TO THEIR BENEFIT -- not the other way around -- i.e., to your benefit. Anything less is simply UNPROFESSIONAL.
I hope this basic lesson in business 101 helps in the future.
You're welcome.
by srikantg
This discussion is closed to new replies.
Sort by Trending