关于cache数据库的一些思考！

Cacheman · 发表于 2005-8-17 21:24:03

<div class='quotetop'>引用

jdli @ 2005-08-16 23:19)</div><div class='quotemain'>Oracle, SQL Server, Caché 在HIS INDUSTRY均有成功的公司. Oracle, SQL Server, Caché, 甚至Hibernate 各有千秋. 最终成功与否, 主要取决于对不同数据管理系统的了解和掌握程度.有资深的ORACLE专家在手, 我看不出公司有什么理由非要用Caché System而舍弃ORACLE.
<br>另外,公司的商业宣传总是强调产品好的一面,有意无意忽略产品的薄弱面, 用户应该对此有所准备.
<br>美国的 Department of Veterans Affairs 可以说是Caché的最大用户( >130 VA Hospitals cross nation), 但其核心管理系统是自主开发的FILEMAN而不是Caché. 作为20多年的Cache和MUMPS支持者,VA目前正在进行美国规模最大的re-engineering, 把现在使用的MUMPS(CACHE)/DELPHI client/server application 转化为Java/Oracle multi-tier browser/server application, 而且VA的联邦医学数据中心(National Health Data Repository)采用的是ORACLE而不是CACHE.
<br>想说的是: 任何主流数据库都各有所长,最终选择那种取决于诸多因素 (database expertise, available development resource, dbms vender's financial stability...)
<br>再有, CACHE 作为 Healthcare 数据管理系统的详尽优劣分析报告(FROM VHA)曾在HL7 TOOLING COMMITTEE 发布过, 整理后我会贴在这儿.
<br>AHRRR...CAN'T BELIEVE IT TOOK ME 40MINUTES TO TYPE THESE LINES IN CHINESE...NEED MORE PRACTICE  <img src='images/smiles/embarassed.gif' smilietext=':embarassed:' border='0' style='vertical-align:middle' alt=':embarassed:' /> </div>
<br>
<br>1）“"我看不出公司有什么理由非要用Caché System而舍弃ORACLE"，27维度的数据挖掘和分析，Oracle怎么实现？而没有挖掘功能的电子病历就是数据垃圾。
<br>2)VA: 请查阅<a href="http://www.cms.hhs.gov/quality/VistAQsAs.pdf" target="_blank">http://www.cms.hhs.gov/quality/VistAQsAs.pdf</a>. 注意13点.这是VA最新的 EHR。如果你有其他更新的内容，我很愿意看一下。
<br>

jdli · 发表于 2005-8-17 22:17:40

<div class='quotetop'>引用

Cacheman @ 2005-08-17 23:24)</div><div class='quotemain'>1）“"我看不出公司有什么理由非要用Caché System而舍弃ORACLE"，27维度的数据挖掘和分析，Oracle怎么实现？而没有挖掘功能的电子病历就是数据垃圾。  <br>2)VA: 请查阅<a href="http://www.cms.hhs.gov/quality/VistAQsAs.pdf" target="_blank">http://www.cms.hhs.gov/quality/VistAQsAs.pdf</a>. 注意13点.这是VA最新的 EHR。如果你有其他更新的内容，我很愿意看一下。</div>
<br>Cacheman:
<br> Thank you for the link.
<br> In past, VA's VISTA are based on two systems: VMS/MUMPS and Cache/MUMPS. For the hospitals using Cache/MUMPS system, they have to buy user licenses for user access. Since early this year, all VMS platform are converted to Cache platform, but VISTA did almost not use any Cache self's function components and management tools. VA uses MUMPS routines to read and write data, using self-developed FileMan (database management system) to do data management, using self-developed KIDS ( Kernel Installation and Distribution System) to do patch distribution.
<br>  Please be aware, I am not saying which main stream DBMS is better than others, what I am saying is that <b>any DBMS has itself's advantage and drawback</b>. Personally, I like CACHE, and have been woked on VMS/CACHE/MUMPS more than 5 years, but just because I like it, do I have to say CACHE is the best system for healthcare industry, "no".
<br>  I am not sure about the so called 27 dimension data mining? What that mean? What's the definition of 27 dimension data mining? I know one of the best US HIMS system (IHC system) fulfilled fine granularity data mining and data analysis, as well as fine granularity level decision support logic module management system, and IHC does not use CACHE, also think about the key market player like 3M system, IDX system, EPIC system, do you think they are all "...数据垃圾" because they are not using CACHE and not able to fulfill 27 Dimension data mining?
<br>  Finally, I'd like to emphasize that no DBMS is perfect, to go with what DBMS, is based upon your business needs, based upon your expertise knowledgement, based upon available development work force you can find from market...

Cacheman · 发表于 2005-8-17 23:57:07

<div class='quotetop'>引用

jdli @ 2005-08-17 14:17)</div><div class='quotemain'>Cacheman:
<br>  Thank you for the link.
<br>  In past, VA's VISTA are based on two systems: VMS/MUMPS and Cache/MUMPS. For the hospitals using Cache/MUMPS system, they have to buy user licenses for user access. Since early this year, all VMS platform are converted to Cache platform, but VISTA did almost not use any Cache self's function components and management tools. VA uses MUMPS routines to read and write data, using self-developed FileMan (database management system) to do data management, using self-developed KIDS ( Kernel Installation and Distribution System) to do patch distribution.
<br>  Please be aware, I am not saying which main stream DBMS is better than others, what I am saying is that <b>any DBMS has itself's advantage and drawback</b>. Personally, I like CACHE, and have been woked on VMS/CACHE/MUMPS more than 5 years, but just because I like it, do I have to say CACHE is the best system for healthcare industry, "no".  <br>  I am not sure about the so called 27 dimension data mining? What that mean? What's the definition of 27 dimension data mining? I know one of the best US HIMS system (IHC system) fulfilled fine granularity data mining and data analysis, as well as fine granularity level decision support logic module management system, and IHC does not use CACHE, also think about the key market player like 3M system, IDX system, EPIC system, do you think they are all "...数据垃圾" because they are not using CACHE and not able to fulfill 27 Dimension data mining?
<br>  Finally, I'd like to emphasize that no DBMS is perfect, to go with what DBMS, is based upon your business needs, based upon your expertise knowledgement, based upon available development work force you can find from market...</div>
<br>
<br>you do not understand China
<br>

Cacheman · 发表于 2005-8-18 00:06:19

I agree any DBMS has its advantage and disadvantage. 每个人都承认.
<br>let me ask you a question, why is Cache so strong in healthcare industry? Oracle and Microsoft are strong enough,they have enough ressource, in terms of money and developer. why can ISC exist? Form the business side, ISC has no strength, so ISC and Cache must have something which sholuld be very interesting for AP and users.
<br>So you are long time in USA/ Cah you tell me what it is?
<br>

jdli · 发表于 2005-8-18 00:16:00

<div class='quotetop'>引用

Cacheman @ 2005-08-18 01:57)</div><div class='quotemain'>you do not understand China</div>
<br>I don't like it. We are talking about pure technical topics, right? What's the matter with whether I understand China?

Cacheman · 发表于 2005-8-18 00:17:46

<div class='quotetop'>引用

jdli @ 2005-08-17 16:16)</div><div class='quotemain'>I don't like it. We are talking about pure technical topics, right? What's the matter with whether I understand China?</div>
<br>let me ask you a question, why is Cache so strong in healthcare industry? Oracle and Microsoft are strong enough,they have enough ressource, in terms of money and developer. why can ISC exist? Form the business side, ISC has no strength, so ISC and Cache must have something which sholuld be very interesting for AP and users.
<br>So you are long time in USA/ Cah you tell me what it is?
<br>

jdli · 发表于 2005-8-18 01:02:39

<div class='quotetop'>引用

Cacheman @ 2005-08-18 02:17)</div><div class='quotemain'>let me ask you a question, why is Cache so strong in healthcare industry? Oracle and Microsoft are strong enough,they have enough ressource, in terms of money and developer. why can ISC exist? Form the business side, ISC has no strength, so ISC and Cache must have something which sholuld be very interesting for AP and users.
<br>So you are long time in USA/ Cah you tell me what it is?</div>
<br>
<br>Cache, like any other DBMS, have advantages and drawbacks. Depending upon business needs, available IT resources, size of DB system, marketing strategy, different companies and organizations have different flavors to choose systems they think can fulfill their needs. That's why many healthcare institutions favored CACHE, also many companies and organizations choose other systems. What I try to say is: CACHE is a good product, but is it the best product fit into Healthcare information industry under different business scenarios? I doubt it.
<br>
<br>You keep saying "Cache is so strong in healthcare industry", if possible, I'd like to see some statistical analysis from you regarding US market, at what market percentage do the companies using cache share?

Cacheman · 发表于 2005-8-18 01:12:57

<div class='quotetop'>引用

jdli @ 2005-08-17 17:02)</div><div class='quotemain'>Cache, like any other DBMS, have advantages and drawbacks. Depending upon business needs, available IT resources, size of DB system, marketing strategy, different companies and organizations have different flavors to choose systems they think can fulfill their needs. That's why many healthcare institutions favored CACHE, also many companies and organizations choose other systems. What I try to say is: CACHE is a good product, but is it the best product fit into Healthcare information industry under different business scenarios? I doubt it.  <br>You keep saying "Cache is so strong in healthcare industry", if possible, I'd like to see some statistical analysis from you regarding US market, at what market percentage do the companies using cache share?</div>
<br>
<br>Very easy: there are top 100 IT companies in US healthcare industry., It is not difficult to find the liste of these top 100. Since you know US better than me, I let you to get the liste.
<br>If u get it, you will find ISC is at number 37, or 38.
<br>Among the top 60 there are no other DBMS companies there.
<br>My result: Cache is the best DBMS and No. 1 in US healthcare.
<br>

jdli · 发表于 2005-8-18 03:23:03

<div class='quotetop'>引用

Cacheman @ 2005-08-18 03:12)</div><div class='quotemain'>Very easy: there are top 100 IT companies in US healthcare industry., It is not difficult to find the liste of these top 100. Since you know US better than me, I let you to get the liste.
<br>If u get it, you will find ISC is at number 37, or 38.
<br>Among the top 60 there are no other DBMS companies there.  <br>My result: Cache is the best DBMS and No. 1 in US healthcare.</div>
<br>
<br>Searched over Google with different keyword combination, but not able to find the top 100 companies in health care industry.
<br>
<br>Don't want to argue anymore, it's understandable you speak for CACHE.

hykionliu · 发表于 2005-8-18 09:04:05

Cacheman：
<br>
<br> 您好像被别人引入了一个误区！
<br>
<br>要知道世界之大，无奇不有！MIforum是一个藏龙卧虎的地方！有时候不要刻意去说CACHE的advantages，您如果看过我转贴的芙蓉姐姐的话，那您应该知道，我们都像别人一样，都喜欢那些在完美中带有一些缺点的东西，那个断臂女神不就是引起无数人的遐想和崇拜吗？
<br>
<br>太完美的东西在现实世界中是不存在的，那可能是上帝、神仙，但不会存在于现实中的！
<br>
<br>我们需要做的是，没有最好，只有更好！
<br>
<br>我非常钦佩您的敬业精神，其实我们大家都没有说CACHE不好，只是想通过这种讨论，获取对CACHE更全面、更真实的了解！毕竟在中国，这种新东西肯定会引起很多人的兴趣、好奇、质疑！我所接触的情况也基本是这样！
<br>
<br>所以，我们做的就是让更多人更全面、更真实的了解CACHE！因为有了这种质疑，我们才会成长、CACHE才会能更好地满足大家的期望！

Cacheman · 发表于 2005-8-18 09:19:58

<div class='quotetop'>引用

hykionliu @ 2005-08-18 01:04)</div><div class='quotemain'>Cacheman：
<br>  <br> 您好像被别人引入了一个误区！
<br>要知道世界之大，无奇不有！MIforum是一个藏龙卧虎的地方！有时候不要刻意去说CACHE的advantages，您如果看过我转贴的芙蓉姐姐的话，那您应该知道，我们都像别人一样，都喜欢那些在完美中带有一些缺点的东西，那个断臂女神不就是引起无数人的遐想和崇拜吗？
<br>太完美的东西在现实世界中是不存在的，那可能是上帝、神仙，但不会存在于现实中的！
<br>我们需要做的是，没有最好，只有更好！
<br>我非常钦佩您的敬业精神，其实我们大家都没有说CACHE不好，只是想通过这种讨论，获取对CACHE更全面、更真实的了解！毕竟在中国，这种新东西肯定会引起很多人的兴趣、好奇、质疑！我所接触的情况也基本是这样！
<br>所以，我们做的就是让更多人更全面、更真实的了解CACHE！因为有了这种质疑，我们才会成长、CACHE才会能更好地满足大家的期望！</div>
<br>
<br>我的观点是:
<br>1)面向对象的数据库要比关系型数据库好.
<br>2)尤其在电子病历和挖掘方面, Cache的树状结构非常适合。
<br>3）要目前将　Ｃａｃｈｅ和　Ｏｒａｃｌｅ就公司的资源，历史，市场份额，知名度站在同一个起跑线上来比，是不公平的。
<br>我也希望大家更了解Cache, 所以一直参加辩论。把事情说得更清楚一点。
<br>
<br>刚刚得到美国总部的关于　ＶＡ的事情：
<br>原文公布：
<br>Cache' is not being replaced by Oracle. Cache' will remain as the local data store for each facility as a long term solution. What is in debate is a Health Data Record (Data Warehouse for patient data through out the VA) Oracle has been targeted for the solution for this, but it is still at the design and discussion point and no final decision has been made. We are still working to bring Cache' into this function as well
<br>

Cacheman · 发表于 2005-8-18 12:22:26

<div class='quotetop'>引用

jdli @ 2005-08-17 19:23)</div><div class='quotemain'>Searched over Google with different keyword combination, but not able to find the top 100 companies in health care industry.  <br>Don't want to argue anymore, it's understandable you speak for CACHE.</div>
<br>
<br>

lease have a look at <a href="http://www.healthcare-informatics.com/issues/2004/06_04/100_2004.pdf" target="_blank">http://www.healthcare-informatics.com/i...04/100_2004.pdf</a>

Cacheman · 发表于 2005-8-18 12:47:47

<div class='quotetop'>引用

hykionliu @ 2005-08-18 01:04)</div><div class='quotemain'>Cacheman：
<br>  <br> 您好像被别人引入了一个误区！
<br>要知道世界之大，无奇不有！MIforum是一个藏龙卧虎的地方！有时候不要刻意去说CACHE的advantages，您如果看过我转贴的芙蓉姐姐的话，那您应该知道，我们都像别人一样，都喜欢那些在完美中带有一些缺点的东西，那个断臂女神不就是引起无数人的遐想和崇拜吗？
<br>太完美的东西在现实世界中是不存在的，那可能是上帝、神仙，但不会存在于现实中的！
<br>我们需要做的是，没有最好，只有更好！
<br>我非常钦佩您的敬业精神，其实我们大家都没有说CACHE不好，只是想通过这种讨论，获取对CACHE更全面、更真实的了解！毕竟在中国，这种新东西肯定会引起很多人的兴趣、好奇、质疑！我所接触的情况也基本是这样！
<br>所以，我们做的就是让更多人更全面、更真实的了解CACHE！因为有了这种质疑，我们才会成长、CACHE才会能更好地满足大家的期望！</div>
<br>1)我完全同意你的意见, 除了"您好像被别人引入了一个误区！".
<br>2) "MIforum是一个藏龙卧虎的地方！"
<br>这里高手如云,我学了很多.
<br>3)"芙蓉姐姐的话，那您应该知道，我们都像别人一样，都喜欢那些在完美中带有一些缺点的东西，那个断臂女神不就是引起无数人的遐想和崇拜吗？"
<br>不好意思，我从没有听过芙蓉姐姐的话。我不喜欢带有一些缺点的东西，除非实在没办法。我们的工作是能做多好就做多好。最好不给客户留带有一些缺点的东西。当然ISC不是我的公司，我只能将Ｃａｃｈｅ最好地展示在中国　ＨＩＳ行业。
<br>4）“我们做的就是让更多人更全面、更真实的了解CACHE！因为有了这种质疑，我们才会成长、CACHE才会能更好地满足大家的期望！”
<br>完全同意。我回答我知道的东西。说实在的，在目前InterSystems合作伙伴中，有投机的公司，跟着市场说他们的系统也基于Cache, 其实根本不了解Cache,从签订合作伙伴协约，没有做过开发。这种公司的宣传我不支持。另外一些公司真正在投入，在开发。我们支持这样的公司，因为他们的系统才可以让终端用户了解　Ｃａｃｈｅ。
<br>５）"您好像被别人引入了一个误区！".
<br>我坚持我认为对的东西。关系型数据库确实是一项伟大的发明。但时代过去了，又很多应用很难用关系型数据库用表结构表达。特别是电子病历和他的查询挖掘分析，这种倒树状的结构是最适合Cache的。
<br>我不认为这是个误区。
<br>当然我承认每个数据库都有自己的独到之处，但我也坚信Cache在电子病历和他的查询挖掘分析，特别在海量数据方面是最佳的选择。而电子病历肯定会是将来医院的核心。
<br>至于一些小型的简单的数据存储和查询，Cache 数据库的优势体现不出来，这当然是事实，但不能因为这个，贬低了Cache 数据库的优势。
<br>
<br>得罪之处，多多见谅。
<br>
<br>最好地了解 Cache，还是得通过培训的方式，我们正在和我们好的合作伙伴一起努力。
<br>
<br>
<br>
<br>

Cacheman · 发表于 2005-8-18 13:08:30

<div class='quotetop'>引用

learning @ 2005-08-18 01:49)</div><div class='quotemain'>前天看到纽约时报上一篇文章上讲, ALIBABA的马元是一个Master of Selling. 非常佩服CACHEMAN的信念和知识. 看来中国不缺马元.  <br>查看了04年HIT公司的排名, ISC从39跃升到28. 不过不知道HEALTHCAREITNEWS这个排名权不权威, 上面讲是公司自己填REVENUE, 方法学上虽然有缺陷, 但应该是相信老美的, 就是奇怪怎么没有其他数据库公司榜上有名?  <br>问了好多朋友在HIT工作的, FINANCIAL FIRM做数据库的, 学校做数据库研究的大都不知道CACHE. 有人跟我说那个排名不完全可信, 有没有CACHE的市场分额介绍, 以及典型公司介绍, 很多人都不相信CACHE在美国的大CIS, EHR, 等占最大分额, 看来需要进一步学习, 对行业了解太少</div>
<br>谢谢夸奖。只是在做本职工作，尽应尽的力量。
<br>排名是否权威，我也不知道，但至少是我目前所知道的。总比什么都没有强吧。有些大公司写的关于Cache的报告，你可以看看
<br><a href="http://www.intersystems.com/cache/analysts/analyst-reviews.html" target="_blank">http://www.intersystems.com/cache/analy...st-reviews.html</a>
<br>FINANCIAL FIRM做数据库知道 Cache的比较少。相对于关系型数据库来说， Cache确实声音很小。Cache在法语的愿意是后面，也就是不站在前面，不直接面对终端客户。这也符合公司的商业模式。

hykionliu · 发表于 2005-8-18 16:39:07

您可能没有理解我说的误区是什么！
<br>我是说当别人越是质疑的时候，您越是强调cache的好处，而您越是这样努力，有的人因为性格原因和文化背景，他就越会认为cache 肯定有一些不尽人意的地方，自然就会引起他们的研究兴趣，去研究cache 不足之处，从长远来看对cache 是有利的！但您却只强调一面，在cache 的不足方面却极少提及，而这也刺激了我们一些专家的研究兴趣！
<br>
<br>您觉得呢？
<br>
<br>作为贵公司的合作伙伴，我们非常希望cache 在中国能够得到广泛应用，但在具体应用的时候可能会由于软件开发、数据库本身以及其他因素的影响而使cache 不能体现出他的优越性能！
<br>
<br>因此，您作为非常了解cache的专家！我们需要您帮助的就是cache 在什么样的应用环境中会发挥它的卓越性能，反之在什么应用环境中则不能！
<br>
<br>凡事强调正反两个方面，通过对比分析，相信大家对cache 一定能有一个比较全面、准确、真实的了解！

		自动登录	找回密码
密码			欢迎注册

关于cache数据库的一些思考！