If "doesn't scale" simply means "I need more proc/mem/disk" you can always throw more horsepower at the problem, but that shifts the solution to a question of how much money you have to spend on toys.
Let's give a concrete example to this. Consider a hypothetical 2 tier system with a thin client talking to a centralized database where all the business logic is being handled. This system gets deployed on a top-of-the-line Sun Fire server running Oracle. The system is successful and its usage grows rapidly. One day, that server can't keep up with the demand. The next stop is moving to DB2 on a mainframe.
Do you have any idea what the cost of a mainframe is? Your company has to be the size of a government in order to afford one. It would be cheaper to rewrite the application than to upgrade the hardware. When its cheaper to rewrite than to upgrade, then you don't have a scalable system.
If the business logic is handled on a separate app server and if the system is designed in order to be cluster aware, then all you have to do is buy another server of equal strength to the first server. Thus, the centralized database running all business logic is not scalable.
In particular, if you intend to port a proprietary (non-GPL'd) application using Cygwin, you will need the proprietary-use license for the Cygwin library.
The company, whom I work for, develops and sells closed source software. I contacted redhat for the details. The "buy out" license is prohibitively expensive. We ended up using a proprietary package because it was cheaper.
I use a lot of open source at work. cygwin, inkscape, Gantt Project, umlet, and dia to name just a few. But I use open source at work only as a consumer. I do not package any of the code in the company's products. At work, I use open source as a user, not a developer. Home is a different story. I code to plenty of open source there. None of that goes into work, however.
A new one will cost more than $100 but a TV cabinet is what I use. You have to get one where the compartment where the monitor goes is at the right height. The bottom compartment is where the rest of the equipment goes. I ended up getting one of those distressed, faux antique dressers so I had to drill a couple of holes in the back. It easily stores three PCs, router/firewall, KVM switch, 19 inch monitor, and speakers for everyone.
With the doors open, you're staring at a data closet. With the doors closed, a casual observer wouldn't even be able to tell you're a geek.
Just ignore the extra features. Take the defaults which are pretty much wide open. The only things you'll want to change is to not let anyone create an account and don't allow content to be searched by folks who aren't logged on.
The end user experience is, IMHO, not excessively complex given that content management is what you are truly looking for. You want to add a document/image/whatever? Click on "my folder," select document type, click add, fill out the form, click submit, you're done.
Well, it can be an informed and research driven opinion. There have been perception based studies on the usability of different GUI style/elements. I can't site the references right now. What I remember is that adults favor earth tone colors whereas children favor circus colors. Also, too much beveling is distracting.
ran out of patience for RPM hell...diving into Slackware...building everything myself
Let me see if I got this straight. The Redhat Package Manager doesn't always work so you moved to Slackware where you build everything yourself. Did I get that right? Why did you have to move to Slackware? Can't you build what you need on a RedHat distro?
I still use RedHat. Most of the time, the RPM works fine. When it doesn't, I install the source and build it. Sometimes I have to get into it to solve build errors but only very, very infrequently.
I know that version 3.6 supported writing to NTFS 5 (a.k.a. WinXP). It's a little bit of a pain because you have to run this tool and find the NTFS DLLs with it but it does work. It is my understanding that all versions of knoppix can write to NTFS 4 (i.e. Win2K) machines.
the market for tech skills is actually not that bad
More than that. I say that it is, once again, an employee's market. I am having a hard time filling two.NET developer positions in Birmingham, Alabama. My humble opinion is that all of the good developers are happily employed, thank you very much.
I don't give much credance to the "all of your development jobs are moving overseas" story. Now, I can see where I.T. development for large, mature corporations could find a good fit in moving their operations overseas. In that scenario, everything is well spec'd out or they don't mind rewriting the code over and over because that is precisely what they are already doing. Also, for I.T., development is a cost center so finding the cheapest labor makes sense. The offshore model doesn't work for small companies or I.S.V.s. Anytime you see one of those companies offshoring their development, you can be sure that it is a desparate last ditch effort to forstall bankruptcy proceedings.
Correct. From a CS perspective, the master's degree is good if you want to stay in academia. It neither helps nor hinders you in the business world. The co-op experience is good if you want to move to the business world.
Right. Future can be changed but not observed. Past can be observed but not changed. The so-called "new" model reported in this topic seems ill formed to me.
"You can go to the past to observe but not change anything" has two problems with it. "The past" is nothing more than a human consensus reality construct for all aggregated observational content. By definition, that which is observed is in "the past." To go back to the past in order to observe alters the construct, makes "the past" something that exists independently of observation.
Even if we could get past that problem, there's another one. It is not possible to go back to the past to observe but not change because, from a quantum perspective, the very act of observation makes change.
Anybody know if any major motion picture studios are using Blender?
After watching Merlin: Special Edition, I watched the "making of" special feature part of the DVD and recognized the blender GUI when the CG part of the movie was being discussed.
Re:Gamers never know what's good for them
on
A Gamer's Manifesto
·
· Score: 1
wants good AI's... immaculately aimed rail guns
Intelligence is not the ability to shoot with greater accuracy. Rather, it is the ability to reason about your situation. The bots in most games already shoot with unbelievably high accuracy.
Instead of going up against every larger numbers of bots with greater accuracy, wouldn't it be interesting to go up against only a handful with accuracy roughly equivalent to yours but also with the ability to play strategically and to change tactics in response to your tactics?
If you really suspect a web site as being a vector for attack, then use lynx and study the output before surfing to it with a modern browser. Alternatively, you could use an anonymous proxy.
Ah, I get it, the Postgresql object oriented SQL syntax. I missed that because I was interpreting the original post in the context of the increasing need for intelligent caching on the local hardware.
Okay, here's a question for you. If an RDBMS extends the SQL syntax, is it still an RDBMS? My take on that is this. If the mental paradigm is still set theory with predicate calculus and 2D tables with columns, then it is relational.
Re:Well, let's take a look at the highest profile
on
McVoy Strikes Back
·
· Score: 1
You're willing to give Firefox the benefit of the doubt due to tabbed browsing but you dismiss the GIMP as a knockoff of Photoshop? There are plenty of GUI innovations in the GIMP that are not found in Photoshop. You may not like them but they are still innovative.
There is no such animal. There is such a thing as object/relational technology. Some common examples of this are hibernate and EJB. These are not alternatives to relational databases. Rather, they serve to persist objects to and from relational databases. They are built on top of relational databases. They do not replace relational databases.
Alternatives to relational databases are LDAP databases such as openLDAP, OLAP databases such as Hyperion, and XML databases like exist. None of these technologies will replace the relational database. It's more about using the right tool for the job. Relational databases work best with operational data. OLAP works best for planning and forecasting. Think of LDAP as a distributed hierarchical database.
Various relational database products have proprietary extensions that may confuse you into thinking that they are alternatives to relational databases. For example, there are extensions to SQL Server that make them seem to act more like OLAP or XML databases.
refomat all your code into the "correct" tab/space arrangement
The problem with this approach is that when you do a diff with the previously checked in version, you will find that every line has changed which doesn't help you quickly identify the two or three lines that the previous coder really changed.
What's wrong with the Standard Template Library's string class is that it came too late to the party. For years, C++ was missing an object oriented way to handle character data. C++ software library builders had to roll their own. MFC has CComBstr and CString. Qt has QString. Xerces has XMLString. Try writing an application that combines different libraries and you quickly end up in string conversion hell.
Mod Parent Up. In the days of Qt/Java/Python/C#, why would I want to code directly to libc or libm for application development? Maybe LSB is for device driver writers but for the rest of us, it is DOA.
Here's why: What's the most important thing on your desktop? It's the data. If someone gets access to your libraries or whatever, who cares? Your data is the most precious thing on your computer. And whether you log in as root or log in as user, you have access to that data, technically anyone who's compromising your account has access to your data as well.
If you wish to run a program whose trustworthiness is uncertain, then log in as a different user than what you normally log in when accessing your precious data. Neither of these user accounts should be root.
I've found keeping it on a wiki helps to keep the design document more current than it otherwise would be.
Mod Parent Up. Wiki is a godsend to software development. The only problem with most Wiki technology is a lack of version control that is available in more sophisticated content management systems such as plone.
I think that this was a/. story but the ACM ran a humorous piece that depicted zealous overindulgance in all things UML as a disease to be cured.
Re:If you REALLY want to know yourself,...
on
Mapping the Mind
·
· Score: 1
Finally, our most unique and advanced feature -- consciousness -- is explained.
Oh, really.
Indeed. At first, I too, thought "incredible claims require incredible proof" but then I realized that this is a subtle form of hypnotism written by a reductionist for reductionists. The claim is that consciousness is explained but what does it mean to be explained? It means to make comprehensible. What does it mean to make something comprehensible? It means to make it easily understood. Notice that making something easily understood and fully understood are two entirely different goals. I could just as credibly state that consciousness is "The state or condition of being conscious" and I would have succeeded in explaining consciousness.
Does that answer big ticket questions like why are we here or what is truth, beauty, or love? No. According to reductionism, none of that is important.
Let's give a concrete example to this. Consider a hypothetical 2 tier system with a thin client talking to a centralized database where all the business logic is being handled. This system gets deployed on a top-of-the-line Sun Fire server running Oracle. The system is successful and its usage grows rapidly. One day, that server can't keep up with the demand. The next stop is moving to DB2 on a mainframe.
Do you have any idea what the cost of a mainframe is? Your company has to be the size of a government in order to afford one. It would be cheaper to rewrite the application than to upgrade the hardware. When its cheaper to rewrite than to upgrade, then you don't have a scalable system.
If the business logic is handled on a separate app server and if the system is designed in order to be cluster aware, then all you have to do is buy another server of equal strength to the first server. Thus, the centralized database running all business logic is not scalable.
Cygwin is not free. From http://cygwin.com/faq.html
In particular, if you intend to port a proprietary (non-GPL'd) application using Cygwin, you will need the proprietary-use license for the Cygwin library.The company, whom I work for, develops and sells closed source software. I contacted redhat for the details. The "buy out" license is prohibitively expensive. We ended up using a proprietary package because it was cheaper.
I use a lot of open source at work. cygwin, inkscape, Gantt Project, umlet, and dia to name just a few. But I use open source at work only as a consumer. I do not package any of the code in the company's products. At work, I use open source as a user, not a developer. Home is a different story. I code to plenty of open source there. None of that goes into work, however.
A new one will cost more than $100 but a TV cabinet is what I use. You have to get one where the compartment where the monitor goes is at the right height. The bottom compartment is where the rest of the equipment goes. I ended up getting one of those distressed, faux antique dressers so I had to drill a couple of holes in the back. It easily stores three PCs, router/firewall, KVM switch, 19 inch monitor, and speakers for everyone.
With the doors open, you're staring at a data closet. With the doors closed, a casual observer wouldn't even be able to tell you're a geek.
Just ignore the extra features. Take the defaults which are pretty much wide open. The only things you'll want to change is to not let anyone create an account and don't allow content to be searched by folks who aren't logged on.
The end user experience is, IMHO, not excessively complex given that content management is what you are truly looking for. You want to add a document/image/whatever? Click on "my folder," select document type, click add, fill out the form, click submit, you're done.
How do you upload photos using only a wiki?
Well, it can be an informed and research driven opinion. There have been perception based studies on the usability of different GUI style/elements. I can't site the references right now. What I remember is that adults favor earth tone colors whereas children favor circus colors. Also, too much beveling is distracting.
Let me see if I got this straight. The Redhat Package Manager doesn't always work so you moved to Slackware where you build everything yourself. Did I get that right? Why did you have to move to Slackware? Can't you build what you need on a RedHat distro?
I still use RedHat. Most of the time, the RPM works fine. When it doesn't, I install the source and build it. Sometimes I have to get into it to solve build errors but only very, very infrequently.
I know that version 3.6 supported writing to NTFS 5 (a.k.a. WinXP). It's a little bit of a pain because you have to run this tool and find the NTFS DLLs with it but it does work. It is my understanding that all versions of knoppix can write to NTFS 4 (i.e. Win2K) machines.
More than that. I say that it is, once again, an employee's market. I am having a hard time filling two .NET developer positions in Birmingham, Alabama. My humble opinion is that all of the good developers are happily employed, thank you very much.
I don't give much credance to the "all of your development jobs are moving overseas" story. Now, I can see where I.T. development for large, mature corporations could find a good fit in moving their operations overseas. In that scenario, everything is well spec'd out or they don't mind rewriting the code over and over because that is precisely what they are already doing. Also, for I.T., development is a cost center so finding the cheapest labor makes sense. The offshore model doesn't work for small companies or I.S.V.s. Anytime you see one of those companies offshoring their development, you can be sure that it is a desparate last ditch effort to forstall bankruptcy proceedings.
Correct. From a CS perspective, the master's degree is good if you want to stay in academia. It neither helps nor hinders you in the business world. The co-op experience is good if you want to move to the business world.
Right. Future can be changed but not observed. Past can be observed but not changed. The so-called "new" model reported in this topic seems ill formed to me.
"You can go to the past to observe but not change anything" has two problems with it. "The past" is nothing more than a human consensus reality construct for all aggregated observational content. By definition, that which is observed is in "the past." To go back to the past in order to observe alters the construct, makes "the past" something that exists independently of observation.
Even if we could get past that problem, there's another one. It is not possible to go back to the past to observe but not change because, from a quantum perspective, the very act of observation makes change.
After watching Merlin: Special Edition, I watched the "making of" special feature part of the DVD and recognized the blender GUI when the CG part of the movie was being discussed.
Intelligence is not the ability to shoot with greater accuracy. Rather, it is the ability to reason about your situation. The bots in most games already shoot with unbelievably high accuracy.
Instead of going up against every larger numbers of bots with greater accuracy, wouldn't it be interesting to go up against only a handful with accuracy roughly equivalent to yours but also with the ability to play strategically and to change tactics in response to your tactics?
If you really suspect a web site as being a vector for attack, then use lynx and study the output before surfing to it with a modern browser. Alternatively, you could use an anonymous proxy.
Ah, I get it, the Postgresql object oriented SQL syntax. I missed that because I was interpreting the original post in the context of the increasing need for intelligent caching on the local hardware.
Okay, here's a question for you. If an RDBMS extends the SQL syntax, is it still an RDBMS? My take on that is this. If the mental paradigm is still set theory with predicate calculus and 2D tables with columns, then it is relational.
You're willing to give Firefox the benefit of the doubt due to tabbed browsing but you dismiss the GIMP as a knockoff of Photoshop? There are plenty of GUI innovations in the GIMP that are not found in Photoshop. You may not like them but they are still innovative.
There is no such animal. There is such a thing as object/relational technology. Some common examples of this are hibernate and EJB. These are not alternatives to relational databases. Rather, they serve to persist objects to and from relational databases. They are built on top of relational databases. They do not replace relational databases.
Alternatives to relational databases are LDAP databases such as openLDAP, OLAP databases such as Hyperion, and XML databases like exist. None of these technologies will replace the relational database. It's more about using the right tool for the job. Relational databases work best with operational data. OLAP works best for planning and forecasting. Think of LDAP as a distributed hierarchical database.
Various relational database products have proprietary extensions that may confuse you into thinking that they are alternatives to relational databases. For example, there are extensions to SQL Server that make them seem to act more like OLAP or XML databases.
The problem with this approach is that when you do a diff with the previously checked in version, you will find that every line has changed which doesn't help you quickly identify the two or three lines that the previous coder really changed.
What's wrong with the Standard Template Library's string class is that it came too late to the party. For years, C++ was missing an object oriented way to handle character data. C++ software library builders had to roll their own. MFC has CComBstr and CString. Qt has QString. Xerces has XMLString. Try writing an application that combines different libraries and you quickly end up in string conversion hell.
Mod Parent Up. In the days of Qt/Java/Python/C#, why would I want to code directly to libc or libm for application development? Maybe LSB is for device driver writers but for the rest of us, it is DOA.
If you wish to run a program whose trustworthiness is uncertain, then log in as a different user than what you normally log in when accessing your precious data. Neither of these user accounts should be root.
Mod Parent Up. Wiki is a godsend to software development. The only problem with most Wiki technology is a lack of version control that is available in more sophisticated content management systems such as plone.
I think that this was a /. story but the ACM ran a humorous piece that depicted zealous overindulgance in all things UML as a disease to be cured.
Oh, really.
Indeed. At first, I too, thought "incredible claims require incredible proof" but then I realized that this is a subtle form of hypnotism written by a reductionist for reductionists. The claim is that consciousness is explained but what does it mean to be explained? It means to make comprehensible. What does it mean to make something comprehensible? It means to make it easily understood. Notice that making something easily understood and fully understood are two entirely different goals. I could just as credibly state that consciousness is "The state or condition of being conscious" and I would have succeeded in explaining consciousness.
Does that answer big ticket questions like why are we here or what is truth, beauty, or love? No. According to reductionism, none of that is important.
You also forgot about Mod Parent Up