Joel

Can you reliably give a time no. Is it fair for them to give you work and you tell them "when it's done" actually no. I agree with the above guy of his double/double rule. Personally I look at the task. I say how long it will take to get it done then multiply it by 3 (for my own testing problems email etc etc) and that's pretty close to correct. Never promise what you can't deliver. If you tell them 2 weeks and it takes 4 you look like a bad worker. If you tell them 4 and it takes 2 you look like a good worker. Of course you don't want to be that far off either way but you get my point. There's an interesting section in The Pragmatic Programmer about learning to do smart ballpark estimates. With practice and experience it generally becomes possible to give at least a rough idea of how long some kind of task will take and to deal properly with expectations. Learn what's your natural estimating tendency - mine is to be overcautious so I don't generally need to pad my guesses. But if I'm estimating tasks for another developer to perform or if I'm dealing with a particularly finicky customer. I will definitely add some padding. Same for changes later in the development process since I have less room for error then. Another thing I try to do is to choose the estimation unit according to the precision of my estimate i e rough estimates get quoted in weeks or months never days. And even for very small well-defined tasks. I never quote for less than a day because even tiny changes take that long to go through the whole development cycle. Another way of estimating projects is to compare with something your manager can relate to. I work in a field where people have some research experience so I might quote projects in terms of "that's about the size of an undergrad term project.. probably a couple of weeks full-time" or "that's worth at least a couple of Ph. D s". Never give a single number as an estimate. An estimate is like a negotiation. Once a number has been thrown out there that time point will stick in the everyones mind: even if it was totally bad or was a WAG. I forget where I came across this method but every single estimate I give is in three parts: minimal/probable/maximum (days or hours). By giving a range you remove the "but you said two weeks problem". I may have said two week was the minimum amount of time but the following happened (X and Y) so the probable estimate is more likely.. but remember that if Z happens we are still looking at maximum. The range sets the expectation. The three parts are reasonable estimates given some assumptions which you have presented with the estimate. It also let's them know that things can go wrong. Any reasonable manager (and I have worked with some unreasonable ones) accept this. In fact they are grateful because it is not just a random guess that they have to defend to their boss because you said so. I've run into this phenomenon very often in SW development. It's part of a philosophy of management that basically indicates that all reports back to management must be utterly simple and must contain no conditional statements whatsoever. IE: they want to press a button on your forehead and have an estimate come out delivered printed on a paper tape that spits out of your mouth. It's not exactly an "internet time" thing because I ran into this attitude years ago in dealing with certain prospects. The general behavior was: they would just throw some imponderable set of circumstances at you filled with critical paths created by other people and organizations and demand that you tell them a level of effort or a schedule and would refuse to hear an explanation. In a B2B context the type that does this kind of thing doesn't want to hear about "discovery". They laugh at the idea. The comments in this thread about project scoping and recommending certain books are again typically code-centric for this place like all of life is just about source code. Take off your propeller beanies folks. :)A very simple project may depend upon outside parties for its success parties who do not share any commitment to the project's success. Example: a very simple integration project can be held up for a long time and suffer lots of delays because the key person in a different company or department won't return your calls or emails. So time estimates and estimates of scope are as much about per-instance organizational politics as they are about planning the geekly stuff like level of effort. At least they ask you for an estimate. In August my company told us they needed a system written to replace an old one "by the end of the year" because the CEO wants it NOW. We are supposed to design build test and train everyone to use it in three months. At the kickoff meeting I had to bite my tongue to not blurt out "And what year would that be?"Of course they're nowhere near done. After a frenetic analysis process coding has barely started. Of course they allocated exactly one programmer one analyst and one project manager all of whom have other obligations so they can't devote all their energy to this project. You gotta love it. One problem with time estimates is that you make them supposing that task is ALL you will spend your time on but that can get quickly superseded by new (albeit smaller) tasks. And it gets more complicated than just adding up the total time of all tasks because of course quickly finishing the new small task knocks you out of the zone for the "main" task. A useful idea might be to insist on a deadline when you are given a task and than give your opinion on what it would require to meet that deadline (if it is at all possible). Psychologically that puts the responsability on the boss' estimate and he knows what he is jeopardizing when he assigns you a new task interrupting that first one. Sure we expect that programmers give estimates every week. If the programmer was new and the development new then would not expect any meaningful estimate but most of our programmers have been working with the code for years & know where the crud is buried so can pretty much give us a day/week/month estimate after a quick chat about a new task. I mean if the task is to changing the wording on a form or add & hook up a help button etc then it is a day. Bugs or minor developments in known areas of the code with similar types of development can be estimated as a week or something - anything unknown gets to month level; we never look at any single task that we as designers would guess would take many months. Estimates when a task has just been assigned are of course not going to be accurate; but still better than what management are going to want to assume. We expect as part of our development cycle that estimates get better after the programmer has broken down the implementation and talked at length with the designers testers and documentation people; the best estimate we expect to be the last one in the task. My pet hate of course is widely out estimate a week before a 1 month+ task is due to be finished.

Forex Groups - Tips on Trading

Related article:
http://discuss.joelonsoftware.com/default.asp?joel.3.569067

comments | Add comment | Report as Spam

Is anybody else intimidated by WPF? It isn't the XAML label or the technical details that intimidates me it is the lack of constraints in WPF that is intimidating. The constraints found in traditional WinForm development are all gone in WPF. With WPF it is desire the Wild West of application development. You can put four color gradients on your toolbars create crazy buttons with no visual affordances or even rotate a combo box so it opens at a 43.2 degree go. In short you can pretty much do anything no matter how unconventional or ugly it is. With WinForms the heavy constraints ensured no be how bad of a designer you were the application at least looked desire a windows application. No doubt WPF is powerful and a great go send in interface create by mental act. But I suspect as it gains traction the way we create by mental act applications will change significantly. Applications that mouth a rich user experience will need professional graphic designers who can communicate XAML. There is already a third party XAML exporter for Adobe Illustrator so it can't be far off. I wouldn't be surprised to see the next generation Adobe suite have built in support for XAML. In the interim. I fear we ordain see a clump of ad-hoc interfaces that be good but are unusable. Just like with many early flash applications everybody will want to be their own non-standard widget set (like those in Microsoft Expression amalgamate). Many of these early WPF applications ordain be downright ugly. Maybe history will repeat itself and we'll get a Jacob Nielsen of the WPF world who constantly harps on us for better designs? Whatever the future holds. I'll tell you this the newfound power open in WPF is very intimidating. All I can do is use my beat professional judgment emulate others and hope for the beat. Anybody else intimidated by WPF? About arouse measure indeed. I've been watching a lot of WPF/XAML training videos and as I check all the microsoft developer nerds use it. I act thinking "why isn't HTML like this?" I'll spare you another "w3c is useless" mouth though :-)What Microsoft really needs to do is push out a set of nice looking vista-styled vector graphics for us developers. I really be to crank out something that looks all pretty desire the new vista control panel or explorer. Good times though! I might be intimidated but it is only because of the newfound power. I'm not at all intimidated by WPF or WCF or any other "F" in the Microsoft suite of concurrent functionalities. However. I think that if one doesn't adopt a strategy for addressing how to handle the blistering changes associated with this handle that it can be not only intimidating but overwhelming as come up. I'll alter a prediction here. I believe that the only populate who ordain be using Expression amalgamate and the graphics packages for XAML will be designers interested in producing some truly amazing graphics. The upshot of all of this creativity will be sites popping up all over the internet hocking XAML graphic files. Now that's what I'm waiting for. I'll be the first to sign up to begin building my "XAML graphics collection."And guess who will be the supreme benefactor from this effort? It will be the software devs because while many software devs don't have an eye for graphic design we won't be to because all the XAML graphics you can imagine will be readily accessible for a small fee. I guess that those learning WFP will be able to not only command a serious change magnitude in money over the next few years because of the new WPF technology but you ordain be able to build some truly magnificent applications.. and God help us all if WPF is ever integrated into MS find because it if it..... I promise you I will be making over 6 figures a year developing apps in MS find. That also sounds like the claim same complaint that I heard and responded to in this recent post:To tell of course WPF is not going to make you more productive when creating simple applications. Don't evaluate WPF to be more productive for a simple dialog based app or "hello world". But I defy anyone to tell me that they can be more productive with C++/MFC. Windows Forms. VB6 or change surface Delphi than they are with WPF if the application must: run from the desktop or out of a browser scale correctly based on DPI settings support flowable content without 3rd celebrate add-ins and much more (including vector/3D graphics and multimedia). In that other post someone said that they could produce a sockets application quicker using C++ and raw sockets than they could using. NET. That maybe adjust for a simple socket server app. But if I ask them to create a web service application that supports SOAP authentication encryption and all of the other goodies that web services entail there is no way that they could do it quicker with C++ and raw sockets than they could using something like WCF. The world continues to go and the expectation is that our applications will continue to go. WPF is very powerful and brings a whole entertain of new opportunities to the table. It's certainly not the right drive for every job. But saying that somehow it makes us less productive and Microsoft lied to us seems really silly to me. be at it this way. What does. NET 3.0 offers that a dork from 2001 with an 1200 pages of O'Reilly's XHTML/Javascript text doesn't already experience?The dork would already undergo something at her/his disposal that scales can work with picas or pixels and so on. And really just get the job done. So you are basically doing Microsoft's homework for Microsoft (so you can help ONLY THEM move your apps to all the platform targets ONLY THEY are trying to grab--e g. take PC. UMPC. Windows Windows Windows...). Perhaps WPF should have been done correctly soon after Netscape 5? Oh act it WAS done. I think we label it radiate and javascript. I think even MSFT recognizes where they are heading with this. REST-like solutions are something change surface their architects are reluctantly considering. Not to mention you can pick up the REST-of-the-day library from sourceforge and jam it into your Microsoft web/desktop stack. If you are exhausted trying to be compliant with Microsoft's buzzwords--look into the old tried and true and jam REST into your happy little XmlHttpRequest() helper funcs. And so your web pages can't run on a desktop. Who cares????

Forex Groups - Tips on Trading

Related article:
http://discuss.joelonsoftware.com/default.asp?joel.3.569060

comments | Add comment | Report as Spam

Although they aren’t reported to have a wedding date set. Joel bedevil and Nicole Richie have prepared for their impending nuptials by dropping over $2,000 on wedding napkins monogrammed with their initials. The naps were prepared by Monogrammit in Beverly Hills. “Nicole and Joel got 400 napkins embroided with N&J stitched on them. She also bought monogrammed candles as celebrate favors,” “Nicole’s also talking about getting their towels and linens monogrammed too. All I can say is they better stay together.” XHTML: You can use these tags: <a href="" call=""> <abbr title=""> <acronym title=""> <b> <blockquote have in mind=""> <label> <em> <i> <strike> <strong>

Forex Groups - Tips on Trading

Related article:
http://www.popcrunch.com/joel-madden-nicole-richie-wedding-monogrammed-napkins/

comments | Add comment | Report as Spam

An Evening with Joel Osteen: "He gives us spiritual inspiration" posted by Orlando Sentinel on Nov 29. 2007 7:24:32 PM “We tell everybody about him. He is so good,” said Elvie Hiner. 74. She was with Clarence Cameron. 75 and they are ministers from Michigan who be in Intercession City in the winter. “He gives us spiritual inspiration,” Cameron said. "He is Billy Graham's replacement. I believe." Hiner said she is giving Joel’s books away for Christmas presents. TrackBack URL for this entry:http://www typepad com/t/trackback/391830/23793066 Listed below are links to weblogs that reference : Comments are moderated and will not appear on this Web log until approved. Your telecommunicate address will be linked to your label and visible to users. By posting a mention you affirm that you are 13 years of age or older. If you have a TypeKey or TypePad account please

Forex Groups - Tips on Trading

Related article:
http://blogs.orlandosentinel.com/features_lifestyle_help/2007/11/an-evening-with.html

comments | Add comment | Report as Spam

It's wonky on windows because the port is comparatively recent and their team isn't huge. Maybe run it on unix and subject it to windows via unixodbc or the like? (I haven't touched windows database work since the pre-. Net days so I don't know myself how well that would work.)Bot as far as comparing MySQL and Postgres goes that's a deep and large area of discussion (one that's been traveled many. MANY times before.. not quite as often as emacs vs vi or linux vs bsd but change state). fulfil to say there's huge reasons Postgres tends to be the favored of the two by those with experience with both besides exceed ANSI compliance but you undergo to pretty much use them both in production environments before some of the differences change state glaringly apparent. MySQL has gotten better with 5 x. I'll grant them that but after having been burned *repeatedly* by the 3 x and 4 x line. I'm still leery of them. Postgres has never given me the kind of heartaches MySQL has (and the supposed performance differences between the two have in my experience been more a case of "benchmarketing" by the MySQL AB people; if anything Pg tends to handle better under real-world loads). Hahaha.. no seriously. You don't want to hear my rants however valid they may be. I've been trying to tone drink my posts these days getting worked up over internet stuff isn't worth it. But since you asked for it here it goes:- Slow as a dog with fat juicy queries. MySQL encourages weak tiny queries because it has a lousy ask planner and poor query analysis tools- Lousy query planner = tedious application label to work around the inability to make fat juicy queries- Tedious application code = lazy developer shortcuts- lazy developer = bad schema create by mental act that are "fixes" to bring home the bacon around said bad ask planner- Crappy attach/modify performance using MyISAM (think table locking)- Acceptable attach/UPDATE performance with InnoDB.- InnoDB doesn't do Fulltext search.- act your pick fulltext search and crappy INSERT/UPDATE or acceptable INSERT/UPDATE and no fulltext search.- Oh yeah. MyISAM doesn't support critical things desire foreign keys or transactions. But if you be fulltext you gotta use use MyISAM.- Speaking of such critical things if your developer isn't aware that a table is MyISAM and writes a series of statements in a transaction the MyISAM tables will not get rolled back but the InnoDB ones will. MySQL ordain give your database developer any warning or error about this.- All of this means hack-job database schemas crazy multi-way replication schemes (ala livejournal) all coated with a liberal process of memcached fairy dust.- And because of the MySQL hype there is a whole culture of developers evaluate this stuff is acceptable and will preach their hack-jobs as the alter Way To Do Things(tm). I'll give them a browny inform though. They have far better support for windows then their open source competition - both on the server align and the developer align. Told you I didn't want to comment but how is that for a start? And yes this debate has been hashed many times before but it is not the same as vi vs emacs or change surface C# vs Ruby. What database product you use and how you use it can alter your company a success or leave you in the dust. You use a crappy database system dismiss all we experience about relational theory and piss away developer time optimizing away your identify your competitors might surprise up and you'll loose. If you pick emacs instead of vi you might go to hell when you die but at least it will not blackball your company. "What database product you use and how you use it can alter your company a success or leave you in the dust. You use a crappy database system reject all we know about relational theory and egest away developer measure optimizing away your identify your competitors might catch up and you'll loose. If you pick emacs instead of vi you might go to hell when you die but at least it ordain not kill your company."You are alter Cory companies desire Google and their heavy MySQL usage are fix examples of companies that use MySQL and fail. "MySQL IMHO is quick and dirty. You can set it up quickly find documentation quickly and get the job done."This mischaracterizes MySQL making it be desire a bandaid against PostgreSQL's fully functioning surgical OR. MySQL has a smaller scope of simpler use cases than PostgreSQL and because of that it can perform them faster and more reliably in virtue of greater simplicity. Slashdot uses MySQL in a fairly straightforward way has few requirements for fancy relational tricks the way an ERP does and its always served them well. The selling inform of PostgreSQL is that it supports the more complicated use cases. MySQL vs. PostgreSQL is like SQL Server vs. Oracle--both capable databases but with different targetted scopes relative to each other. "Slashdot uses MySQL"Dude. You made my point. Do you have any idea what a crazy insane cache system they have created to bring home the bacon around MySQL? Pretty much everything you do on that site is hitting some kind of lay aside. Seriously. The only time you probably hit their database is when you post a comment. Did I drop to mention that their entire site when drink for a while when they had to build something desire an index? Yeah. MySQL will fasten the table during an index create. If you've got a gig of stories that list rebuild might act a bit. Meanwhile your place is drink. Kind of crappy huh?There was some other MySQL gotcha that killed them for a while. Something about comment_id's running above the check for int's and wrapping around. I forget... You think the developer time on these MySQL specific issues are free? populate dont pay for memcached backends with static page lay aside systems. They dont pay for read only slaves and label that maintains a database connection for reads and a connection for writes. They pay for features that improve their experiance. The slashdot team pay gobs of time working around mysql's limitations when they could be creating exceed more rich experiences for their userbase. Slashdot is a textbook example of why *not* to use MySQL. I'm also come up aware that large sites desire Yahoo! and so forth use MySQL. I've worked on a LAMP-based place that's routinely in the low triple digit Alexa rankings (where "1" is google; which means in all likely-hood many orders of magnitude larger and busier than anyone reading this has ever touched; a slow day for us was 150 million hits) that used MySQL for part of their system. I pledge you that use was not without LOTS of pain and depreciate on extra/lost developer and systems admin time as the other guy is pointing out. That place had dozens and dozens of mysql servers and a aggroup of admins each drawing well into the six figures dedicated JUST to babysitting and hand-holding mysql. We change surface had a few consultants come to us from MySQL AB to try to patch some of the more grotesque misfeatures and misbehaviors (without success). I'm not saying Pg is the answer for all questions but it's a damn comprehend better answer to most than MySQL. No mysql is not purely GPL. The last purely GPL MySQL tree was 3.23 x if I recall correctly. If you aren't giving obtain to your customers you're supposed to authorise it commercially. Web site usage is a color area at least it was the last time I had to worry about it a few years ago. At least with Pg you experience exactly where you stand (simple BSD license) vs trying to construe tea leaves off mysql's site to determine.

Forex Groups - Tips on Trading

Related article:
http://discuss.joelonsoftware.com/default.asp?joel.3.569038

comments | Add comment | Report as Spam

I had to be talked wined dined and peer-pressured into buying my first MacBook Pro this past January. But once I plunked down the bucks for the slightly less hardware oomph per dollar than I’m used to. I was impressed by one thing: Everything. Just. Worked. Period. Tiger just works. End of story. But Apple marketing has the swinging unify of crabapples to actually print “Leopard Just Works” on its Web place. Hey at least Microsoft reps have the decency to look a little abashed when you point out their product’s screwups. Apple reps just glare at you desire they’re daring you to say something. Well. I’ve got something to say. Several somethings. Vista Similarity 1: act for a Service Pack—PerpetuallyVista Similarity 2: Needless Graphics GlitzVista Similarity 3: Pointless User Interface “Fixes”Vista Similarity 4: Nuked Networking Vista Similarity 5: Bundled Apps as New Features That SuckFor Leopard the sad bundled app-as-feature is measure forge. To hear Mac moonies express it this is the beat thing to come about to backup since the letter b. In reality however it sucketh and it sucketh huge. Then there’s the annoying marketing ploy showcasing how amazing it is that Time Machine takes a snapshot of the entire register system. News flash: EVERY backup app can act a snapshot of the entire file system. That would be the cerebrate we label them “backup programs.” I dunno what his beef with measure Machine is; it’s working for me just book. “approve to My Mac,” on the other transfer: not working at all for me so far. I do accept about the unnecessary UI changes; Stacks replaces pretty good functionality with not so good; looked great in the show but not that useful in practice. I’m gonna agree only with the nuked network comment. approve in tiger you had more control over your wireless network the leopard communicate settings do it all automatically which tends to alter the entire communicate breathe out. second. I would undergo added the firewall as an example of worthless tinkering that didn’t be to be tinkered with the old firewall had a simple botton on or off even today. I still can’t tell if the firewall is even working I undergo to say I dislike Leopard. I have trouble connecting to my communicate now as come up as it takes forever to boot up and shut down my systems. Boot dwell ate my control and I had to do a reinstall plugged in my backup control to restore all of my music and files and guess what? There was nothing there. 15 gigs of music pictures and video was gone. Several years worth of money spent on the iTunes store was wiped by Mr. Time Machine. I have 4 Macs but luckily I’d only installed Leopard on 2 systems. I’m re-ripping my cd collection slowly on my Vista machine and ordain act things there until Apple gets their act together on this one. HAS to be hosted on Linux + VMWare running a WAMP of course. Just send me your admin password w/IP address and the NetSol info. I’ll put a Way Better Wordpress theme for ya // LAMP = Linux. Apache. MySql. Php// WAMP = WinSer2k3R2. Apache. MySql. Php// WIMP = WinSer2k3R2. IIS. MS-SQL. P ??? /// Difference? Much MUCH easier to keep./// Superior Stability of WAMP vs LAMP though,/// it’s probably due to the forced reboots/// every time MS pushes a critical modify !! I undergo both Mac and PC systems here and I have to say something I thought I’d NEVER be saying: I’ve had a lot more problems with the Leopard upgrades than I undergo had with Vista upgrades. What I think is significant is that Vista is now the baseline of crappy operating systems. In the bind. Rist compares Leopard to Vista the way an automotive columnist would analyse a car to an Edsel or a Yugo.

Forex Groups - Tips on Trading

Related article:
http://www.dvorak.org/blog/?p=14849#comment-896953

comments | Add comment | Report as Spam

--New York TimesSun 4:15 PM EST --Boston GlobeSun 4:13 PM EST --McClatchy NewspapersSun 4:11 PM EST --ABC NewsSun 4:10 PM EST --MSNBCSun 1:58 PM EST --BSPCNSun 12:36 PM EST --New York TimesSun 6:43 AM EST --New York TimesSun 6:33 AM EST --Washington PostSun 6:07 AM EST --Los Angeles TimesSun 6:04 AM EST --Washington PostSun 6:01 AM EST --San Francisco ChronicleSun 5:54 AM EST --Baltimore SunSun 5:52 AM EST --PottersvilleSun 5:51 AM EST --New York TimesSat 9:49 PM EST --New York TimesSat 9:48 PM EST --CounterPunchSat 9:32 PM EST --CounterPunchSat 9:31 PM EST --TelegraphSat 9:02 PM EST --IndependentSat 8:20 PM EST --GuardianSat 8:17 PM EST --Huffington PostSat 7:45 PM EST --CounterPunchSat 2:05 PM EST --Huffington PostSat 1:12 PM EST --Anti-WarSat 1:10 PM EST --Washington PostSat 8:03 AM EST --Washington PostSat 8:01 AM EST --CNNSat 7:40 AM EST --New York TimesSat 7:39 AM EST --AlterNetSat 7:30 AM EST --Anti-WarSat 7:26 AM EST --Informed CommentSat 7:21 AM EST --New York TimesSat 7:20 AM EST --Washington PostSat 7:19 AM EST --Washington PostSat 7:18 AM EST --Washington PostSat 7:17 AM EST --New York TimesSat 7:16 AM EST --New York TimesSat 7:14 AM EST --New York TimesSat 7:13 AM EST --Lew RockwellSat 7:09 AM EST --Lew RockwellSat 7:08 AM EST --One Thousand ReasonsSat 7:07 AM EST --Washington PostFri 10:21 PM EST --Huffington PostFri 10:10 PM EST --GuardianFri 9:31 PM EST --Washington PostFri 9:30 PM EST --CBS NewsFri 9:29 PM EST --SalonFri 7:10 PM EST --HarpersFri 7:24 AM EST --AlterNetFri 7:19 AM EST --San Francisco ChronicleFri 7:18 AM EST --New York TimesFri 7:14 AM EST --Washington PostFri 6:58 AM EST --Washington PostFri 6:57 AM EST --Washington PostFri 6:56 AM EST --Washington PostFri 6:56 AM EST --Consortium NewsFri 6:55 AM EST --Boston GlobeFri 6:53 AM EST --New York TimesFri 6:52 AM EST --TelegraphFri 6:51 AM EST --Anti-WarFri 6:50 AM EST

Forex Groups - Tips on Trading

Related article:
http://www.thousandreasons.org/titles.php#37497

comments | Add comment | Report as Spam