If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#31
|
|||
|
|||
I was told "Fields are expensive, records are cheap"
=?Utf-8?B?RGVubmlz?= wrote in
: As repeated stated, and repeatedly missed the question concern disk access speed over the network and the speed of ACCESS to extra values from a big record. Maybe when people aren't answering the question you wish was being answered, it's a clue that the problem is not with all the people who are offering what you consider unacceptable answers. -- David W. Fenton http://www.dfenton.com/ usenet at dfenton dot com http://www.dfenton.com/DFA/ |
#32
|
|||
|
|||
I was told "Fields are expensive, records are cheap"
=?Utf-8?B?RGVubmlz?= wrote in
: It has been my experience that posting something in the Database group is a measure of last resort. This question is a prime example. People want to talk about something that is not part of the question. I would say that this is your fault for the way in which you cast your question. You were responding to a rule of thumb that many database developers use and you went off on a red herring about something that hasn't really been relevant in schema design for about 20 years. That those who responded to your question didn't follow you down that garden path is not an indication of a problem with the newsgroups -- it's an indicator that you weren't asking the right question, or you were doing so in a manner that was unprofitable. -- David W. Fenton http://www.dfenton.com/ usenet at dfenton dot com http://www.dfenton.com/DFA/ |
#33
|
|||
|
|||
I was told "Fields are expensive, records are cheap"
Dennis
Your comment; Please re-read my response. It starts with the word "If". I was describing the workload/maintenance of an overly-simplified design, and pointing out that adding fields is expensive, in terms of the maintenance it requires to all affected objects. My response: Opps, you are absolutely correct. My bad. Your comments: That would depend on what data elements the user & I agreed were needed. My response: Hmm, you obviously have better educated users than I do. Normally, the users do not request a data field. I guess they sort of do. They specify the output they want. From the output, I determine which fields the need to input. I’ve yet been able to figure out how to include a field on some output (display, report, data feed, etc.) that was never entered or calculated within the system. So for me, I don’t believe it is for me to decide if a fields is required or not. The system output requirements determine the need for a field. If the user decides that the want to pay for that output item (display, report, data feed) then the fields is required. If not, then the field is not required. I once had a customer require that I add about 5 new data fields to the input screen just so that they *only* showed on a single display screen / form. I thought this was a waste, so I asked them why wanted this enhancement? They explain that the addition of those 5 fields enabled the customer combined with the information already on the form enabled the customer service rep to completely answer about 75% of the calls on the first call received (versus the 50% pre-enhancement). In addition, they did not have to pull the policy file to obtain these 5 little pieces of information. Nor did they have to re-file those policies folders. This little change increased the customer service group’s performance and drastically reduced the load on the filing staff. I have asked why this had not been done sooner. The response was the previous developer did not consider this an important upgrade and never did it. I believe that I should advice the client on the cost of a field, but only they can decide if it worth the cost. Therefore, we will have to agree to disagree on whether or not the developer should be involved in the decision as to which fields should be included or not. Your comment: If you are looking for other folks ideas, to compare them with yours and decide what "balanced" approach would work best for you, let us know. You asked for an assessment of John V.'s statement. My response: No, that is not quite right either. Maybe in a round about way it is. I misunderstood his remark to mean that it was better for Access to have two smaller records in a 1:1 relationship than it was to have one large record. My response to this was obviously John knows something about Access that I don’t and I need resolve this issue so I can create efficient designs. Hence, the disk access speed versus Access extracting variables from large records. It was all of the other respondents who have brought up the data structure issues. At *no time* did I bring up the data structure issue. (I only responded to those who brought up the issue.) As I’ve repeatedly stated through out my responses, the data structure was *never* the question. I’ve done it on other issues, but not this one. I really don’t know how much clearer I could have been on this issue. Your comment: Maintenance is maintenance, whether on one object or several. My comments were intended to offer the option of a design that would require NO additional maintenance, since the table would grow longer, not wider. Response: Ok, now you have my interest! I agree that maintenance is maintenance. I am being very serious when I’m saying that I would love to know how to capture additional data and utilize that data on new pre-defined / formatted forms without require any additional maintenance. What is the secret to this? Please I am being serious. How can I add new fields and create new reports without any additional maintenance? Your comment: Response: Again, I wish you had read my introduction where I state I have 30 years of relational database design experience. As I stated in my response to Rick, this was something I learned NOT to do over 30 years ago. That you've learned not to do this was not evident in your original post. My Response: I guess there is a two part answer to this question. First, as I have repeatedly stated the data structure was NEVER part of the question. Given that it was never part of the question, why should I include anything to do with a question that was never asked? That’s like saying why did I not include the weather conditions at the time? It is because it had nothing to do with my original question. Per my original post ” I've always been taught the exact opposite - that "Fields are cheap, records are expensive" *since going to disk is so slow versus accessing data in memory*.” People started focusing on a non-issue as far as I was concerned and ignored the original question. I thought I stated my question pretty clearly about disk access speed versus accessing data in memory. Also, you have to remember that I interpreted John’s original comment to mean something that it did not. But I did not know that at the time, hence my question. The question was ONLY about disk access speed versus Access’ speed of extracting variables from a long record. NEVER in my wildest imagination did I EVER think it would become about data structures because that was NEVER the question. Secondly, as I stated in another response, if a doctor was to tell you that he had been a doctor for 30 years, would you ask him if he could read a thermometer? Of course not. When I stated I had 30 years, I figured that would tell people that I had some experience under my belt and was asking a more detailed and in depth question that a less experienced person would not know to ask. Your comment: If you want detailed suggestions/ideas/approaches, provide detailed descriptions. Response: In this question, I did *not* want suggestions, ideas, or approaches. I wanted the answer to the question is there something about Access that causes it to be very slow with large records? As I stated before, the really sad part of all of this is that NO ONE in this series of questions has answered the original question of disk access speed versus Access extracting data from a large record. I don’t count David’s response concerning disk access speeds. That is because his comment on this discussion was made AFTER he responded to Allen Browne’s answer to my re-statement of my question in this forum. Your Comment: Responding as you have could be interpreted as 'baiting' folks, offering an incomplete description and then criticizing folks for not inferring or knowing your detailed situation. My Response: I take EXTREME INSULT (caps meant) at your statement that “Responding as you have could be interpreted as 'baiting' folks”. I am a lone developer and I believe that this forum is a GREAT PRIVILEDGE! I have FREE access to some of the greatest Access minds in the world. What a GREAT PRIVILEDGE! The access to such knowledge is something to be highly valued and respected! There have been so many people in this forum that have helped me slowly climb the Access learning cliff that I don’t know how to or even where to begin to thank them. Even though I’m still somewhere near the bottom of the Access learning cliff, I could not have gotten as high as I have without the assistance of the knowledge people on the Access forums. I try very hard to always thank the people who help me. As a way of show my appreciation to them, I try to answer those question on the forum for which I am qualified to answer. I figure that if I answer the ones I know, that provides more time to the more knowledge people to answer the harder questions and help just a few more people. I have to say that I’ve learned more from the members on this forum than I have from all of the Access books that I have read. I go to great length to avoid wasting anyone’s time. Before I post a question on this forum, I try reading my books, searching for the answer via Google, or searching the different forums. ONLY after I have read, and searched, and could not find the answer or did not understand the answers I found do I post my question. Yes, I take GREAT OFFENCE (caps meant) to you accusing me of wasting the time of the people who are so kind to respond to question. Even though all of the respondent totally misunderstood the question and went off on irrelevant tangent, I actually respect and appreciate their intentions and their passion. I am very passionate about my work and I am always glad to associate with people of equal passion. Even if I disagree with them or they don’t understand me. Regarding your “incomplete description” comment. I guess you are right on that one. It was an incomplete description because it had *nothing* do with my disk access speed versus memory access speed. That is like criticizing me for not telling you that it was blues skies and 75 degrees Fahrenheit outside when I wrote the original question. It had *nothing* to do with my question. If I had had *any* idea that people would take an irrelevant comment and blow it up into a huge mountain, I would *never* have included it in my original question. If I have any criticism, it is of the folks (including you) that even after I have repeated and repeatedly stated the fields was *not* an issue, they would not drop that issue. I understand that issue and I fully agree with that issue. Even you last comment “Definite expensive” address an issue I never asked about. And as I said before, the truly sad thing is, no one in this discussion (excluding David because read and responded to the other discussion before he repeated his answer in this forum) has addressed my original question. Even *after* I have repeatedly stated that my question had to do with disk access speed. NO ONE responded to my question. I saw one person stated that there were very experienced people willing to help. If that is so, why is that that not one of those experience people ever answered the question? Even after I repeated the question? Fortunately, Allen Browne, and John Vinson, and David Fenton were kind enough to answer this same question quite completely in another discussion on this forum. I really can not believe that over 32 responses were generated regarding something that was repeatedly stated was *not* an issue and the original question was *never* answered. Dennis |
#34
|
|||
|
|||
I was told "Fields are expensive, records are cheap"
=?Utf-8?B?RGVubmlz?= wrote in
: I misunderstood his remark to mean that it was better for Access to have two smaller records in a 1:1 relationship than it was to have one large record. That was a pretty perverse reading of it. As I think you now understand, the statement was about normalization, that a large record very often has repeating data where each instance of the repeating field should be a record in a separate table rather than a field in the main table. And the reason for this is not as much data retrieval performance (though it is often more efficient to retrieve data from a single field in another table than it is to query multiple fields in a single table; not to mention multiple index maintenance vs. a single index to maintain) as it was the cost in terms of front-end design. I don't think it was reasonable at any point to think the 1:1 design was being promoted by anyone. -- David W. Fenton http://www.dfenton.com/ usenet at dfenton dot com http://www.dfenton.com/DFA/ |
#35
|
|||
|
|||
I was told "Fields are expensive, records are cheap"
=?Utf-8?B?RGVubmlz?= wrote in
: I really can not believe that over 32 responses were generated regarding something that was repeatedly stated was *not* an issue and the original question was *never* answered. You're very close to landing in my killfile, and I'm one of the 3 you single out as praiseworthy. Make of that what you will, but you might consider it evidence that just because I'm still engaging in the discussion does not mean I think you've behaved admirably throughout. You've been at times combative, dismissive, obtuse and stubborn. You are lucky that anybody is still reading your posts. -- David W. Fenton http://www.dfenton.com/ usenet at dfenton dot com http://www.dfenton.com/DFA/ |
Thread Tools | |
Display Modes | |
|
|