Re: TDD vs. Big Ball of Mud (was Re: fusebox vs model glue)
I like that you've been thinking about thinking while writing your framework, Isaac. I dig that. :-) The way I'm currently tackling the tests, is to try to work them into my routine. That hard part is, as been mentioned, not having to go back a lot to keep it all in sync. I wonder if this ties into keeping branches synced up with trunk and whatnot with SVN... heh. Somehow similar... Anyways, where I'd, before, just keep doing stuff in a scribble file, sorta testing, building, testing (but not saving any tests), I now try to save the tests, sorta. It is hard to write tests that are easy to refactor, but it's getting easier as time goes by. There are sweeping changes sometimes, and I'm a sweeping changer, for all my not much changing... but it really does save time having those old tests, of the various bits. If just in having that history/memory refresher... I'm still working on the testing of the behaviors, I guess, because I'm quite often including stuff in later tests that are tested themselves, but should be stubs (instead of the real things), just to verify the behavior, but that's still sorta beyond me. Basically I can see cascading fails, where I don't think you'd have that with stubs and whatnot. It still works, but I could see how it would work better to keep it strictly to interactions, vs. creating and whatnot. If that makes sense. Probably didn't explain it too clear since I'm still twirling the ideas around in my head, sorta. I get the failure aspect of overthinking, or overengineering, and I battle those twin demons (or maybe it's just one demon with two heads), and other, more Rube Goldberg-ish impulses, but I have actually seen some pretty good return on tests, even with poor style, as it were. What sells me, is projects like the javascript toolkit "dojo". That thing is verbose, but put together really well. It's a snap to read the sources, and there are some good standards, and *various* types of tests, from your more literal "it passed!" to your, "here a few random examples". I've had really good luck refactoring and keeping up-to date with it, and I attribute the majority there to well written code. It has been really interesting to watch it evolve. Same with projects like Eclipse and whatnot. Apache, hell, it's been ages. I think that part of the reason open source is kicking ass, is because it's really tackled the problems of the future, you know? There's a lot of history, too, that we can learn from, all the various flavors of OSes.. Mozilla... from sources open to closed to a mix of the two... but, heh, the battle of big balls of mud... Maybe it's fundamental. Part of nature. Something about entropy or whatnot, ya know? Chaos and order, a spinning yin-yang of seasons. :] Balance. I'm trying to hold it, but it still sorta feels like jumping from one side to the other, versus just sorta maintaining. Not exactly smooth... but not too rough. Hmmm... maybe it will be easier if I go faster! HaHA! -- There is no such thing as a 'self-made' man. We are made up of thousands of others. George Matthew Adams ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311897 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
Re: fusebox vs model glue
They all have their ups and downs, I would imagine. I can vouch for the fact that well-written MG apps, as I assume is the same for the other frameworks, are pretty easy to debug and whatnot. Poorly written, however, sucks donkey balls. And I've written some code that deserves to be flushed. I'd love to see it do that little circle... well, guess it's more of a spiral... :-) Seems like that's just how it goes. A lot depends on your background, strengths, weakness, etc., as to which you'd be able to get up to speed with faster. And there are some cool choices out there. It ain't just FB and MG, and yes, it all takes time, but ideas can be worth, like, millions-- and everyone seems to say that trying new things gives them "new" ideas. I know, I know... time is money, and there is a bottom line, and that cold hard logic we all must at least pay some sort of tribute to... But there's more to it than just the... machine mentality, or whatever. Time is funky-- sometimes a little investment frees up more than you'd think, later on... whatever you choose to invest in.. or maybe it's sorta like art, and some kinds are more appealing, just cause. Eh, I'm not really sure where I'm going with this. Heh. It can all be crap, and it can all be roses, and somehow there's a relation? Bah. :-) :D -- There is no such thing as a 'self-made' man. We are made up of thousands of others. George Matthew Adams On Mon, Sep 1, 2008 at 2:59 PM, Eric Roberts wrote: > Which I have run into with MG...most of what we are doing is made overly > complex by using MG, which in turns made it overly difficult to debug when > we had probs. I can see where MG would be very useful for a very complex > application...but I don't think it is the solution for everyone. I > personally prefer the ease of use and flow of FB (or more specifically, my > bastardized version of FB ;-) which based on a combination of my personal > coding style that is very close to a fusebox-esque style and a version that > we were forced to use due to draconian server restrictions at > AT&T ). > > Eric ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311896 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
Re: kinda OT : sqlite "statement is not executing "
On Mon, Sep 1, 2008 at 8:39 AM, Greg Morphis wrote: > Okay, odd even further... > I remove the CFQueryparams and the row gets added, no error.. > > So why would cfqueryparam be causing a "statement is not executing"? Random guess: you're binding it to the wrong datatype? Or something like that. Is the database telling you anything? Nothing in the logs, eh? -- There is no such thing as a 'self-made' man. We are made up of thousands of others. George Matthew Adams ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311895 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
Re: How can I unzip a password-protected zip file with cfzip?
Rick Faircloth wrote: > I couldn't find anything about using a password, either. > Surely that'll be added next version. core java zip (java.util.zip) class doesn't support passwords. it's been an open bug/enhancement for the last 8 years (yes, that *eight* years) so kind of doubtful cf9 will include it: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=4347142 if it's really that important now you might look at: https://truezip.dev.java.net/ and maybe prod the cf team about it. or if you like chilkat & don't mind paying a bit: http://www.chilkatsoft.com/purchase2.asp ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311894 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
Re: How can I unzip a password-protected zip file with cfzip?
if you are on your own server, you can also install winzip or winrar or 7zip and use cfexecute to run the command-line interface with necessary parameters. Azadi Saryev Sabai-dee.com http://www.sabai-dee.com/ Rick Faircloth wrote: > I have the password and can unzip a file manually, but > is there a way to use the password with cfzip action="unzip" ? > > Thanks, > > Rick > ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311893 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
RE: How can I unzip a password-protected zip file with cfzip?
Thanks, Justin. I couldn't find anything about using a password, either. Surely that'll be added next version. Thanks for the tip, too, on the CFX_Zip tag! Rick > -Original Message- > From: Justin D. Scott [mailto:[EMAIL PROTECTED] > Sent: Monday, September 01, 2008 11:57 PM > To: CF-Talk > Subject: RE: How can I unzip a password-protected zip file with cfzip? > > > I have the password and can unzip a file manually, but > > is there a way to use the password with cfzip > > action="unzip" ? > > Hi Rick, I haven't actually used the new CFZIP tag yet, but I don't see > anything in the documentation that indicates support for password protection > in saving or opening ZIP files. We are using the CFX_Zip tag from EmTek > Systems (http://www.emteksys.com/products/cfxzip/). We've been using that > for a couple of years without any problems for password protected real > estate data feeds. > > > -Justin Scott > > > ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311892 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
RE: How can I unzip a password-protected zip file with cfzip?
> I have the password and can unzip a file manually, but > is there a way to use the password with cfzip > action="unzip" ? Hi Rick, I haven't actually used the new CFZIP tag yet, but I don't see anything in the documentation that indicates support for password protection in saving or opening ZIP files. We are using the CFX_Zip tag from EmTek Systems (http://www.emteksys.com/products/cfxzip/). We've been using that for a couple of years without any problems for password protected real estate data feeds. -Justin Scott ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311891 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
Re: TDD vs. Big Ball of Mud (was Re: fusebox vs model glue)
> Whoa, hold it right there. Show me the TDD advocate who promotes > writing all tests before writing code and I'll personally have their > card revoked! TDD means write a test, then write some code, then write > another testby the time you write your last test, you've written > all your code bar that last function. You will actually rewrite less > of your stuff than otherwise, because writing tests first really > brings a poorly factored design into the spotlight much earlier than > would otherwise happen. That's probably true some of the time... I'm not necessarily convinced it's true most of the time... And I'm going to delve a bit into the land of personal anecdote (yeah yeah) to comment on this one here, rather than citing other cognitive studies. My own experience at various companies has been that "big picture" changes to the system often have to be made mid-project. It seems to happen because in the initial planning stages there's a certain kind of inherent myopia that we all seem to fall prey to where we're focusing on getting the design fleshed out for a small part of the system and because we're focused on that small "unit" we don't notice a larger interoperability issue. On my last project we had clients who were probably fairly typical in wanting the application to read their minds. (And I'll be the first to say I love really good intuitive designs that seem like they're actually doing that.) Except on this project mind reading had been promised up front. So we've got this application with a multi-page form and two things are demanded. 1) never under any circumstances should the application ever forget anything that a person has typed into any field, even if they abandon that page of the form 2) never under any circumstances should the application ever display information in a form field that the user doesn't want in that field. So right off the bat, if the user changes their mind, hits back then goes forward again, the system is supposed to magically know that they changed their mind. I'm not just griping to gripe, there's a point to this. :) So this multi-page form actually has a 2nd multi-page form embedded in the middle of it. It's a custom project management app and each project has phases. Up to this point, the user enters the info on a particular page, that gets saved to the database (if valid) before they go back to a previous tab or before they go forward to the next tab. In a normal, intelligent design you would think the name of the phase should be somewhere on the first page of the multi-page form to enter phases, so that, if nothing else there is at least a label to identify it. Not so. It's part of the spec that the name of the phase is entered on page 2. And I'm told in no uncertain terms that this decision is from the client and therefore set in stone. It's also part of the spec that they should be able to go "back" to previous pages of the form and then return and as previously mentioned, the application should never ever forget anything under any circumstances... So since the data on page 1 of the form is saved to the database and has no label, I add the label "unnamed" when they're looking at the list of phases to indicate that a particular phase was abandoned halfway. This maintains 1) remembering the information they entered and 2) gives them the opportunity to delete the item if they really had decided they didn't want it. Then the problem -- "it should never say 'unnamed' in the list of phases... if they change their mind about entering a phase, it should just throw it away"... ummm... this violates the first few principles of the design. That it should always remember everything and it should allow them to go "back" to previous pages in the form without forgetting. And of course nobody noticed this major design flaw because everyone was focused on small units of the design and didn't notice the big picture. We went round and round trying to get the thing settled. And had I written behavioral tests for these things I'm sure a lot of them would have had to be rewritten because of mid-project design changes that were a result of thinking about things as isolated "units" in the early part of the design. (Which seems to catch us all off-guard, myself included.) Now I'm not necessarily saying that's the majority of projects or even that it's the way things should be. It's just been my experience that these kinds of things happen ... well they've happened to just about every project I can remember having worked on... More of the ubiquitous forces behind that "big ball of mud" design pattern that's so popular. When I worked for SiteManageware in Ft Lauderdale we were working on an overhaul of the pricing system and I think we ended up essentially going back to scratch on the technique two or three times. I don't even necessarily mind the idea of "thinking outloud into a test case"... but to be the devil's advocate, I'm not convinced it results
RE: Coldfusion 8 Standard Edition & multiple sites
Other than the throttles on SMS, and cfdocument etc, standard itself won't provide any specific performance improvements. You may be considering running enterprise and using multiple instances of ColdFusion for each site. Other than being able to allocate more memory to each instance, you won't neccessarily get any performance out of it if they are all running on the same server. What you will get is separation of one site from another (One instance may go down, but the others won't) and redundancy if you are using more than once instance for each site in a cluster (Through JRrun's load balancing). http://www.adobe.com/products/coldfusion/editions/ ~Brad ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311889 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
Re: TDD vs. Big Ball of Mud (was Re: fusebox vs model glue)
> TDD advocate here... I really think folks are going into TDD with > this mindset that it's just write a test before you write code (so yes > your are somewhat right). Doing this is going to result in frustration > and an eventual failure and abandonment of TDD. Focusing on just > writing a test is really the wrong approach The fact is some > very smart folks have realized TDD was a horrible name for a great > idea as it put us into this mind set of testing a unit. This is > why many folks (mostly outside of the CF community) are "evolving" TDD > into BDD (behaviour driven development). I could ramble on more > probably but I think I've said enough for now, maybe this deserves a > blog post? Thanks Adam, I'd be interested in reading more about that. :) I hadn't heard about the shift in thinking from TDD to BDD. There's an old apocryphal story I like about how language helps to shape thought. The story goes that the people in a particular area (let's say it was an island) were having a problem in which people were falling into unexplained comas and they feared people were being burried alive. They wanted to resolve this problem, so they assigned the task of devising a solution to two teams and would then choose the best of the two. One team produced a solution that was complicated, time consuming, labor intensive, questionably effective and expensive. The other team produced a solution that was unquestionably fast, cheap and effective (you might call this an "elegant solution"). The complicated solution was to burry each person with a small amount of food and water and raise a pipe from the coffin up to ground level so that if the person woke from their coma they could call up to someone monitoring the grave sites and help would rush to pull them out. After a few days they could be reasonably assured the person had died and fill in the pipe. Problems involved rain flooding the coffin, etc. The question posed to the team who produced this solution: how to we keep from burrying people who are still alive? The "elegant" solution was to mount a 6-inch spike to the inside of the coffin lid at roughly chest-height. When you closed the lid of the coffin, you would remove all doubt that the person could still be alive. The question posed to the team who produced this solution: how to we ensure that everyone we bury is dead? I've been trying to come up with a good way of explaining a similar kind of shift in thinking with regard to the notion of a "front controller" and I'm not sure how well I've done in that regard. It's posted on the onTap framework wiki as "The Rear Controller and You". The basic premise of my calling this a new design pattern (as opposed to simply describing it as a sub-set of the front controller) is to flip the thinking involved in the "front controller" away from forcing everything to deal with the challenges of the front controller (the application says "I INSIST you speak to me THIS WAY") and instead allow the controller to do the work more seamlessly by saying something to the effect of "I don't know X but if you hum a few bars I can fake it". :P The system accomodates both the traditional front controller index.cfm?event=company.about style urls as well as more traditional /company/about.cfm style urls which makes it a trivial matter to move pages into or out of the framework without changing any existing URLs. For anyone interested in the Rear Controller design pattern, you can find more information here: http://ontap.wikispaces.com/The+Rear+Controller+and+You -- s. isaac dealey ^ new epoch isn't it time for a change? ph: 781.769.0723 http://onTap.riaforge.org/blog ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311888 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
RE: TDD vs. Big Ball of Mud (was Re: fusebox vs model glue)
> -Original Message- > From: s. isaac dealey [mailto:[EMAIL PROTECTED] > Sent: Tuesday, 2 September 2008 3:06 AM > To: CF-Talk > Subject: TDD vs. Big Ball of Mud (was Re: fusebox vs model glue) > > And so it's an assumed that the TDD advocates who talk about writing all > your tests before writing your code will be overconfident about their > design. Whoa, hold it right there. Show me the TDD advocate who promotes writing all tests before writing code and I'll personally have their card revoked! TDD means write a test, then write some code, then write another testby the time you write your last test, you've written all your code bar that last function. You will actually rewrite less of your stuff than otherwise, because writing tests first really brings a poorly factored design into the spotlight much earlier than would otherwise happen. The assertion is that if you can't write a test you don't *have* a design, or indeed a clear intention. And if your tests are hard to write, it's because you have a bad design. So given all of this, it's hard to argue that the right thing to do in this situation is to go ahead and write some code anyway. "Thinking out loud" in code is fine, but thinking out loud into a test case is even better. @Adam: Agreed that "behaviour" is a better word. I also like the last "D" to stand for "design". "Test driven design" takes away some of the quality-control mindset that's introduced by the word "test". Jaime ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311887 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
Re: fusebox vs model glue
> Eric sometime we should talk about these draconian restrictions and > what you've had to do with Fusebox, drop me a line sometime. I imagine you were thinking something like I was... What was it in Fusebox (of all things) that would be anathema to the server managers at AT&T? FB's always struck me as being pretty non-invasive from the standpoint of server requirements, irrespective of the fact that it's generally not my tool of choice. -- s. isaac dealey ^ new epoch isn't it time for a change? ph: 781.769.0723 http://onTap.riaforge.org/blog ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311886 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
How can I unzip a password-protected zip file with cfzip?
I have the password and can unzip a file manually, but is there a way to use the password with cfzip action="unzip" ? Thanks, Rick ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311885 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
Re: Coldfusion 8 Standard Edition & multiple sites
In CF 8 Standard, the SMS gateway is single threaded; it will only run in one request at a time. If your sites are low volume, this may be fine. If not, you'll need Enterprise. If you need to connect to Oracle, the drivers supplied with CF Enterprise have more features than the Oracle thin driver. On Tue, Sep 2, 2008 at 7:17 AM, Sebastian Powell <[EMAIL PROTECTED]> wrote: > I'm new to coldfusion and have a few developed sites which I plan to host on > a dedicated server. In process of working out which version of CF I need. > Wanted to know if Standard edition can handle multiple sites and what the > impact performance wise would be? or anything else I should consider? > > As context the sites utilise the following services: > > SMS gateway > Webservices > Scheduled Tasks > Logging > multiple databases -- mxAjax / CFAjax docs and other useful articles: http://www.bifrost.com.au/blog/ ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311884 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
Coldfusion 8 Standard Edition & multiple sites
I'm new to coldfusion and have a few developed sites which I plan to host on a dedicated server. In process of working out which version of CF I need. Wanted to know if Standard edition can handle multiple sites and what the impact performance wise would be? or anything else I should consider? As context the sites utilise the following services: SMS gateway Webservices Scheduled Tasks Logging multiple databases Thanks, Bas. ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311883 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
Re: TDD vs. Big Ball of Mud (was Re: fusebox vs model glue)
TDD advocate here... I really think folks are going into TDD with this mindset that it's just write a test before you write code (so yes your are somewhat right). Doing this is going to result in frustration and an eventual failure and abandonment of TDD. Focusing on just writing a test is really the wrong approach, hell the fact you get some tests out of TDD is just a nicity. Use TDD or do not the value to me is the thought process involved this is the indispensable part of TDD. The problem comes in to the fact that how do I teach a group of people good TDD in a 1 hour (hell 2 hour) session at a conference? I am attempting it at bFusion, hopefully it will turn out well and I can take it on the road so to speak ;) Here's my short synopsis of how we stand with TDD: Proper TDD seeks to break down that 80% failure by focusing on putting more thought into your *design*. We can either write, rewrite and rewrite production code, or we can write and rewrite test cases until we are more confident with the design. Then we may only write and refactor production code once or twice. I'd much rather see that happen. The fact is some very smart folks have realized TDD was a horrible name for a great idea as it put us into this mind set of testing a unit. With this mindset we focus on asserting simple values and string, or looking at booleans of a unit. What the hell is a unit anyway? A CFC? A Method? The whole concept of a unit is based off its context. Really what TDD should be about is testing the behavior and (more importantly in many systems) the interaction of objects in the system. This is why many folks (mostly outside of the CF community) are "evolving" TDD into BDD (behaviour driven development). It's the same thing just worded differently. Instead of using a word like testing which someone assumes oh just write tests first, this name sort of smacks you in the face. FOCUS ON BEHAVIOUR, yet if you read up on BDD its all about validating your design (fancy way of saying test ;) ). I could ramble on more probably but I think I've said enough for now, maybe this deserves a blog post? Adam On Mon, Sep 1, 2008 at 1:05 PM, s. isaac dealey <[EMAIL PROTECTED]> wrote: > > It's actually switching over, to where the tests come first, that's > > the hard part, for me. Due to a lot of the reasons listed in that > > article about big balls of mud. ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311882 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
Re: fusebox vs model glue
Eric sometime we should talk about these draconian restrictions and what you've had to do with Fusebox, drop me a line sometime. Adam On Mon, Sep 1, 2008 at 4:59 PM, Eric Roberts < [EMAIL PROTECTED]> wrote: > Which I have run into with MG...most of what we are doing is made > overly > complex by using MG, which in turns made it overly difficult to debug when > we had probs. I can see where MG would be very useful for a very complex > application...but I don't think it is the solution for everyone. I > personally prefer the ease of use and flow of FB (or more specifically, my > bastardized version of FB ;-) which based on a combination of my personal > coding style that is very close to a fusebox-esque style and a version that > we were forced to use due to draconian server restrictions at > AT&T ). > > Eric > > /*-Original Message- > /*From: Larry Lyons [mailto:[EMAIL PROTECTED] > /*Sent: Monday, September 01, 2008 11:01 AM > /*To: CF-Talk > /*Subject: Re: fusebox vs model glue > /* > /*>denstar wrote: > /*>> Sounds like with FB you could end up with a Pretty Entertaining > /*Environment. > /*> > /*>*groans* > /*> > /*>> Is there a down side to all the flexibility? :-) > /*> > /*>Yes. It means that no 2 application developers will develop websites the > /*>same way. Though IMHO, that's not much of a downside. > /* > /*It could also mean that you end up with it going down the TOtally > /*Ineffective Ludicrous Entertaining Technology or TOILET. > /* > /*ok its been a long weekend. > /* > /* > > ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311881 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
(ot) A Wee Dram - CF Day in London UK
Sorry for the Off Topic Post: Just in case some haven't heard the sad news that CFDEVCON has been cancelled. To help compensate for this loss the Scotch on the Rocks team have brought together some of the US Speakers already committed to flying over for a low cost, one day event. More details & registration at http://aweedram.com Date: Thursday 25th September 2008 Time: 10am to 6pm Cost: £10 on the Door [Cash Only] (Includes 2 Free Drinks + 10% off SOTR09) Location: Square Pig 30 - 32 Proctor Street Holborn London WC1R 4QG Schedule 10:00 - Key Note - From Adam Lehman & Claude Englebert 11:00 - Break 11:05 - Coldspring by Mark Drew 12:05 - Break 12:10 - RAD Object Oriented CF Development by Peter Bell 13:10 - Lunch 13:40 - High Availability: Clustering ColdFusion Applications by Mike Brunt 14:40 - Break 14:45 - Subversion - better living through branches by Sean Corfield 15:45 - Break 15:50 - Setting up a Solid Level Local Dev Environment by Kurt Wiersma 16:50 - Break 17:00 - Railo 3.1 The Open Source Story by Gert Franz 18:00 - Thank Yous & Raffle 18:15 - Networking in the bar -- Big Mad Kev email: [EMAIL PROTECTED] blog: http://inner-rhythm.co.uk/blog -- ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311880 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
RE: fusebox vs model glue
Which I have run into with MG...most of what we are doing is made overly complex by using MG, which in turns made it overly difficult to debug when we had probs. I can see where MG would be very useful for a very complex application...but I don't think it is the solution for everyone. I personally prefer the ease of use and flow of FB (or more specifically, my bastardized version of FB ;-) which based on a combination of my personal coding style that is very close to a fusebox-esque style and a version that we were forced to use due to draconian server restrictions at AT&T ). Eric /*-Original Message- /*From: Larry Lyons [mailto:[EMAIL PROTECTED] /*Sent: Monday, September 01, 2008 11:01 AM /*To: CF-Talk /*Subject: Re: fusebox vs model glue /* /*>denstar wrote: /*>> Sounds like with FB you could end up with a Pretty Entertaining /*Environment. /*> /*>*groans* /*> /*>> Is there a down side to all the flexibility? :-) /*> /*>Yes. It means that no 2 application developers will develop websites the /*>same way. Though IMHO, that's not much of a downside. /* /*It could also mean that you end up with it going down the TOtally /*Ineffective Ludicrous Entertaining Technology or TOILET. /* /*ok its been a long weekend. /* /* ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311879 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
TDD vs. Big Ball of Mud (was Re: fusebox vs model glue)
> It's actually switching over, to where the tests come first, that's > the hard part, for me. Due to a lot of the reasons listed in that > article about big balls of mud. :] I'm not entirely convinced that writing the test before writing the code is the best strategy... or at least that it's the best strategy for everyone. Development work by its nature means doing things that (we hope anyway) haven't been done before. And while you may have some ideas about how to accomplish those things ahead of time, the human brain is really just not equipped to map out large numbers of variables accurately in advance. I realize of course as soon as having said this a lot of TDD advocates will say "but you're not mapping out large numbers of variables in advance, you're mapping out small individual 'units'". Which is again what I think Hermes Conrad would call "technically correct... the best kind of correct". Even though you might be writing tests for a very small "unit" like an individual function, the reality is that you're still basically trying to start from E=MC2 and reverse engineer three gigantic chalk-boards full of equations to get to that solution. It works yes, I'm just not convinced it's the most efficient method. The reason I'm not convinced it's efficient is because I know there's a reason for the 80% failure rate in R&D. (Which not enough programmers are familiar with - good start on Wikipedia.) And don't treat programming as an R&D discipline like it should be treated. It's human nature to assume that things are easier than they are -- to be overconfident. You can't get around that, it just is. And humans are predisposed to be wrong about everything due to confirmation bias. The Wason Selection task proved that. And so it's an assumed that the TDD advocates who talk about writing all your tests before writing your code will be overconfident about their design. Over the course of development it will be necessary to rewrite about 80% of the initial design. And consequently about 80% of the tests... but if you write them in advance and you continue to rewrite them when the design changes, it's not just 80%, it will end up that in the best circumstances, you've (re)written something like 1000 lines of test code for every 20 lines of functional code. These are also forces that contribute significantly to the big ball of mud in spite of the fact that they're not described in Foote and Yoder's treatment of it. It's a shame that more of us in the IT industry haven't read up on cognitive bias. I'm not an expert, just saying based on what I've read it seems *really* important and the IT industry as a whole is almost completely unaware of any of it. I wanted to actually point to an example I know I saw where Matt Woodward had posted a comment on the Mach-II wiki about developers underestimating the challenges involved in maintaining code being a reason for one of their new features in 1.5. I'm having difficulty locating that particular comment at the moment. Anyway without using the term, he was describing precisely the effect of overconfidence bias. Showing really that you'll be affected by these things whether you're aware of them or not. Better imo to have at least an awareness of the research into cognitive bias even if you can't be an expert. > Maintaining the tests is what's hard for me. I don't do it right, so > I have to update them, instead of the other way around. So given what I've said above, I don't think this is just you. :) I think it's the approach to TDD. I think there's great promise in TDD though I'm not convinced the most effective approach to TDD is what's widely practiced (yet) in much the same way that most sold software in practice is in some kind of "big ball of mud" in spite of the fact that we have all this really good design available to us. > I'm sorta thinking that there should be some ways to automate a lot of > test creation. Or a way of flipping the problem on it's head... > something like introspection combined with a metasploit sorta deal... > hmm... sometimes the things I come up with are just pie in the sky > (or, not well thought out, conversely), See "overconfidence bias". :) > so I feel bad writing tests and stuff before I know the idea will work, > but I should stop that. Freaking throwaway code! But hey, maybe I'm > just planning on tossing at least the first version of everything. =] I'm thinking personally it may be better to work the tests into the middle, once you know it works and before it goes into maintenance mode. Not on an entire large application, but on a mid-sized logical bundle like a handful of CFCs that manage X. Though that leads to another problem in that once you're at that stage, there's a real risk of not getting around to writing the tests because either your boss sees what you've accomplished and insists that you immediately get on to the next section of the project without stopping to write the tests (because he doesn't u
Re: fusebox vs model glue
>denstar wrote: >> Sounds like with FB you could end up with a Pretty Entertaining Environment. > >*groans* > >> Is there a down side to all the flexibility? :-) > >Yes. It means that no 2 application developers will develop websites the >same way. Though IMHO, that's not much of a downside. It could also mean that you end up with it going down the TOtally Ineffective Ludicrous Entertaining Technology or TOILET. ok its been a long weekend. ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311877 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=11502.10531.4
Re: kinda OT : sqlite "statement is not executing "
Okay, odd even further... I remove the CFQueryparams and the row gets added, no error.. So why would cfqueryparam be causing a "statement is not executing"? On Mon, Sep 1, 2008 at 1:30 AM, denstar <[EMAIL PROTECTED]> wrote: > What do the database logs say? Anything? Can you put it into verbose > mode, if not? > :D > > -- > A cheerful frame of mind, reinforced by relaxation... is the medicine > that puts all ghosts of fear on the run. > George Matthew Adams > > > On Mon, Sep 1, 2008 at 12:01 AM, Greg Morphis wrote: >> I was actually thinking it was related to this >> >> http://www.mail-archive.com/[EMAIL PROTECTED]/msg00743.html >> >> Odd how the row gets added to the DB... >> The logs only show : >> 08/31 18:00:10 Error [web-2] - Error Executing Database >> Query.statement is not executing The specific sequence of files >> included or processed is: >> C:\JRun4\servers\cfusion\cfusion-ear\cfusion-war\testCF\create.cfm, >> line: 92 >> 08/31 18:18:00 Error [web-0] - Error Executing Database >> Query.statement is not executing The specific sequence of files >> included or processed is: >> C:\JRun4\servers\cfusion\cfusion-ear\cfusion-war\testCF\create.cfm, >> line: 92 >> >> >> On Mon, Sep 1, 2008 at 12:37 AM, denstar wrote: >>> You checked the DB logs? Something somewhere probably says sorta >>> what's going on. > > ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311876 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
Re: Pre-filling FileField Values
Rick Faircloth wrote: > However, if I want to take responsibility to designate an entire folder > of files for upload, I should be able to do that, too. Not just one file > at a time, but choose the folder and all its contents. With the extensions of RFC 2388 the HTTP protocol is perfectly capable of allowing a user-agent to upload an entire directory in one HTTP post. The choice not to implement that functionality in browsers is one made by browser vendors. Any inquiries as to why they made that choice should be directed to them. Jochem ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311875 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
Re: cf8 enterprise failover
Matthew Williams wrote: > The only complaint I've had for our public facing servers in > the last few months is an issue with our front end balancer. If a > CFHTTP request gets routed back to the same server making the request, > the load balancer fails the request. This only applies to non-CFM > related requests, so calls to XML files and images would fail > intermittently. The workaround for that is to use things like CFFILE > instead, but I have no control over the front end load balancer. The simple solution typically is to add all the hostnames you are using to the hosts file of your CF server with the 127.0.0.1 IP address. (You need to restart your CF instance after that because it caches DNS lookups.) Jochem ~| Adobe® ColdFusion® 8 software 8 is the most important and dramatic release to date Get the Free Trial http://ad.doubleclick.net/clk;203748912;27390454;j Archive: http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:311874 Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
