qooxlisp live - 3MB bloatware
in [JavaScript]
|
Prev: qooxlisp live
Next: ILC 2010 Call for Papers
From: Kenneth Tilton on 8 Jun 2010 12:34 Thomas 'PointedEars' Lahn wrote: > Kenneth Tilton wrote: > >> Scott Sauyet wrote: >>> Testing now at home over a mediocre >>> DSL, it's around 100 seconds on an empty cache. >> I sometimes see it loading the 300+ files 4/s, never wait around to see >> it finish. I think the Lisp server I use might be doing that. >> >> For me, resetting the browser gets it back to loading in 10+s. >> >> Anyway, I just did a "release build" and it makes one file, 989kb. >> Trying to get that working for a release 2.0. > > "kb" means "kilobit" to you too, yes? If no, forget it, you are not ready > for serious Web development. I think I'll forget you. I already did serious web development with qooxdoo a couple of years ago, but not with Cells. Client loved me for it. Then switched to some gentlemen in Viet Nam who were a tad cheaper. They used ExtJS, btw.
From: Scott Sauyet on 8 Jun 2010 13:01 Kenneth Tilton wrote: > Scott Sauyet wrote: >> Kenneth Tilton wrote: >>> RobG wrote: (RE: http://www.teamalgebra.com/) >>>> Enabling javascript, the page took nearly 4 minutes to load. >>> I get that, too, sometimes. Not sure what's going on. Possibly the >>> server is being silly, because other times it loads in 4-5s. > >> I tested with a very fast connection and fast computer, and it never >> loaded in less than 8 seconds. > > Yeah, I lied. I get 11s or so. The 4-5s was what I remembered from > loading from localhost. As others have pointed out, you need to look at overall size and see if the download time for that amount of data is likely to be acceptable to your target audience and their infrastructures. >> Testing now at home over a mediocre >> DSL, it's around 100 seconds on an empty cache. > > I sometimes see it loading the 300+ files 4/s, never wait around to see > it finish. I think the Lisp server I use might be doing that. If you can't wait around for your own demo to load, you might want to reconsider the demo. > For me, resetting the browser gets it back to loading in 10+s. What do you mean by "resetting the browser"? >>> Anyway, I just did a "release build" and it's one file, 989kb. Not >>> nothing, but should be even better. > >> There are regular complaints on this group about the size of a 70KB >> (unzipped) file. Is all that really necessary for this relatively >> simple page? > > I think it's like Lisp applications: even "Hello World" will end up with > most of Lisp in there. One could go crazy trying to have a build > procedure take out uneccessary code, but then (a) how much would one > save and (b) why bother? These frameworks are for RIAs, which will > indeed use many components of a framework. The only beneficiary would be > small demos, which I suspect is not worth the trouble. And libraries are > pyramids -- that "simple" demo uses a nice variety of widgets, including > a remote table with scrolling, movable and hidable columns, one column > data renderer, and probably reaches down into a lot of code. It's your call, of course. But it's commonly said that for web applications, smaller is better. I think there are good reasons for that. We all know that many websites have so many more bytes wrapped up in their images than in the JS, but every bit of savings that can be squeezed out is useful. And I didn't mean to dismiss the complexity of your demo. It wasn't until today when I tried it in Safari and Chrome that I really even saw the extent of it. 989K still sounds like a lot, but not nearly as badly so as what I saw not really doing anything in FF and IE. >>>>> I'll investigate and/or put the search button back in to beat the >>>>> thing into submission. >>>> You knew it was dysfunctional but posted a link anyway. Thanks. >>> I try. > >> You try to what, annoy the hell out of everyone here? > > I believe the people annoyed are the ones who hope to be both annoyed > and annoying, aka, mindlessly abusive of anyone not using raw HTML. Yes, > it gives me great pleasure to annoy them, since they are the ones being > bullies and they totally need to be laughed at, not listened to. Come on, posting a demo that doesn't work properly and then prickling at criticisms of it is not going to win you a Good Netizen award. Imagine this instead: |>>>> You knew it was dysfunctional but posted a link anyway. Thanks. |>>> Yeah, I'm sorry. I *thought* I'd checked it in all major recent |>>> browsers. I'll post a new version as soon as I work out this bug. That would not earn you particularly many rebukes. > You have a sick little cult in this NG, self-important and posturing, s/You/We. You're posting here too. > utterly convinced of themselves while the silent majority just rolls > their eyes at them and gets on with their work. Speaking of which... There is plenty of ego in this group, and it certainly has more than its share of pedantic nonsense. But you are posting your demo here for some reason. Is it just ego-gratification, do you simply want to show the world how smart you are? Or do you think there is pedagogical value in sharing this with other people interested in Javascript? Or are you here like most people asking for discussion, critiques, suggestions for your code? If it's the latter, then I think you need to be a little less defensive and to listen a little more carefully. >>>>> "The best Algebra tutorial program I have seen... in a class by itself." >>>>> Macworld >>>> Quotes without meaningful attribution make me suspicious. >>> You never heard of Macworld? > >> http://www.google.com/search?q=%22The+best+Algebra+tutorial+program+I+have+seen%22+site:macworld.com > >> returns no results... Do you have a more detailed reference? > > I have the whole review somewhere, but a digital version could be tough > to dig up. If you are seriously interested I'll look for it. I'm not. I am getting confused though between your demo, the software you say received such high praise and the software that -- is it web software that you're currently developing? Your quote, in this context, sounded like it was supposed to be related to the software relevant for this group. >> Are you telling us that this ran on a Mac web browser in 1991? > > No, it was written originally for the Mac, in C. The defter of the > intellects in this group might be able to figure out I have decided to > release it as a web app, hence qooxlisp. ie, plaintiff was confusing the > present with the past. The confusion is not surprising. You're here discussing some software that you're releasing, but seem to be promoting it with a quote that long predates this new software. -- Scott
From: Thomas 'PointedEars' Lahn on 8 Jun 2010 14:15 Kenneth Tilton wrote: > Thomas 'PointedEars' Lahn wrote: >> Kenneth Tilton wrote: >>> Anyway, I just did a "release build" and it makes one file, 989kb. >>> Trying to get that working for a release 2.0. >> "kb" means "kilobit" to you too, yes? If no, forget it, you are not >> ready for serious Web development. > > I think I'll forget you. I already did serious web development with > qooxdoo a couple of years ago, but not with Cells. Client loved me > for it. That must have been around 2008 (CE), when qooxdoo -1.0 was released.¹ (Impossible as it may seem, you are even a lousier liar than you are a developer.) > Then switched to some gentlemen in Viet Nam who were a tad cheaper. Yeah, sure. Must be great living in fairytale land. Ignorance is bliss, yes? > They used ExtJS, btw. Yet another piece of junk, reviewed here before. PointedEars ___________ ¹ First mention of qooxdoo was in c't Magazin 1/2009, version 1.0 was released on 2009-12-17. -- Prototype.js was written by people who don't know javascript for people who don't know javascript. People who don't know javascript are not the best source of advice on designing systems that use javascript. -- Richard Cornford, cljs, <f806at$ail$1$8300dec7(a)news.demon.co.uk>
From: David Mark on 8 Jun 2010 14:44 On Jun 8, 2:15 pm, Thomas 'PointedEars' Lahn <PointedE...(a)web.de> wrote: > Kenneth Tilton wrote: > > Thomas 'PointedEars' Lahn wrote: > >> Kenneth Tilton wrote: > >>> Anyway, I just did a "release build" and it makes one file, 989kb. > >>> Trying to get that working for a release 2.0. > >> "kb" means "kilobit" to you too, yes? If no, forget it, you are not > >> ready for serious Web development. > > > I think I'll forget you. I already did serious web development with > > qooxdoo a couple of years ago, but not with Cells. Client loved me > > for it. > > That must have been around 2008 (CE), when qooxdoo -1.0 was released.¹ > (Impossible as it may seem, you are even a lousier liar than you are a > developer.) > > > Then switched to some gentlemen in Viet Nam who were a tad cheaper. > > Yeah, sure. Must be great living in fairytale land. Ignorance is bliss, > yes? > He must be living in some sort of fantasy world as his "2.0" version (second one in two days) is a gray screen of death in the latest FF. Viewing the source yields an invalid XHTML 1.1 pretender. XHTML 1.1 on the Web. Go figure. NOSCRIPT element as well. It's a clinic on how not to write a Web page (as are so many sites these days).
From: Kenneth Tilton on 8 Jun 2010 16:00
Scott Sauyet wrote: > Kenneth Tilton wrote: >> Scott Sauyet wrote: >>> Kenneth Tilton wrote: >>>> RobG wrote: > > (RE: http://www.teamalgebra.com/) > >>>>> Enabling javascript, the page took nearly 4 minutes to load. >>>> I get that, too, sometimes. Not sure what's going on. Possibly the >>>> server is being silly, because other times it loads in 4-5s. >>> I tested with a very fast connection and fast computer, and it never >>> loaded in less than 8 seconds. >> Yeah, I lied. I get 11s or so. The 4-5s was what I remembered from >> loading from localhost. > > As others have pointed out, you need to look at overall size and see > if the download time for that amount of data is likely to be > acceptable to your target audience and their infrastructures. Sure. Right now she loads for me in a couple of seconds. My audience is students or tutors sitting down to learn/teach Algebra for 30-60min depending on how much fun they are having with who else is on line. I think a couple of seconds wait will be bearable. I had a spinning wheel animated gif going earlier but decided it was not worth it. > >>> Testing now at home over a mediocre >>> DSL, it's around 100 seconds on an empty cache. >> I sometimes see it loading the 300+ files 4/s, never wait around to see >> it finish. I think the Lisp server I use might be doing that. > > If you can't wait around for your own demo to load, you might want to > reconsider the demo. No, I am too smart to wait around for something obviously pathological (I can see the server feeding files one by one instead of in a blur). > > >> For me, resetting the browser gets it back to loading in 10+s. > > What do you mean by "resetting the browser"? > > >>>> Anyway, I just did a "release build" and it's one file, 989kb. Not >>>> nothing, but should be even better. >>> There are regular complaints on this group about the size of a 70KB >>> (unzipped) file. Is all that really necessary for this relatively >>> simple page? >> I think it's like Lisp applications: even "Hello World" will end up with >> most of Lisp in there. One could go crazy trying to have a build >> procedure take out uneccessary code, but then (a) how much would one >> save and (b) why bother? These frameworks are for RIAs, which will >> indeed use many components of a framework. The only beneficiary would be >> small demos, which I suspect is not worth the trouble. And libraries are >> pyramids -- that "simple" demo uses a nice variety of widgets, including >> a remote table with scrolling, movable and hidable columns, one column >> data renderer, and probably reaches down into a lot of code. > > It's your call, of course. But it's commonly said that for web > applications, smaller is better. I think there are good reasons for > that. We all know that many websites have so many more bytes wrapped > up in their images than in the JS, but every bit of savings that can > be squeezed out is useful. Sure, but considering the application I think I can get away with a few second load. To be honest, they'll be waiting longer for the ads if I go that route. > > And I didn't mean to dismiss the complexity of your demo. It wasn't > until today when I tried it in Safari and Chrome that I really even > saw the extent of it. 989K still sounds like a lot, but not nearly as > badly so as what I saw not really doing anything in FF and IE. David did not like my point, but I think it valid: with big frameworks the first bite is ineluctably huge. Libraries are pyramidal, and a simple button can pull in a lot of the library code. The good news is they grow slowly thereafter, so I know now pretty much how big my full rollout will be. > > >>>>>> I'll investigate and/or put the search button back in to beat the >>>>>> thing into submission. >>>>> You knew it was dysfunctional but posted a link anyway. Thanks. >>>> I try. >>> You try to what, annoy the hell out of everyone here? >> I believe the people annoyed are the ones who hope to be both annoyed >> and annoying, aka, mindlessly abusive of anyone not using raw HTML. Yes, >> it gives me great pleasure to annoy them, since they are the ones being >> bullies and they totally need to be laughed at, not listened to. > > Come on, posting a demo that doesn't work properly and then prickling > at criticisms of it is not going to win you a Good Netizen award. > Imagine this instead: > > |>>>> You knew it was dysfunctional but posted a link anyway. Thanks. > |>>> Yeah, I'm sorry. I *thought* I'd checked it in all major recent > |>>> browsers. I'll post a new version as soon as I work out this > bug. > > That would not earn you particularly many rebukes. > > >> You have a sick little cult in this NG, self-important and posturing, > > s/You/We. > > You're posting here too. > >> utterly convinced of themselves while the silent majority just rolls >> their eyes at them and gets on with their work. Speaking of which... > > There is plenty of ego in this group, and it certainly has more than > its share of pedantic nonsense. But you are posting your demo here > for some reason. Is it just ego-gratification, do you simply want to > show the world how smart you are? Or do you think there is > pedagogical value in sharing this with other people interested in > Javascript? Or are you here like most people asking for discussion, > critiques, suggestions for your code? > > If it's the latter, then I think you need to be a little less > defensive and to listen a little more carefully. No, I am incredibly busy with real work and just taking a moment to share something with other people excited about technology. The anti-library crowd in here is just how I get the publicity. I like sharing cool technology, and qooxlisp is that. I know well, however, that doing o/s increases the effort of developing something by a factor of 27, possibly 81. Me not go there, got Algebra to do. > > > >>>>>> "The best Algebra tutorial program I have seen... in a class by itself." >>>>>> Macworld >>>>> Quotes without meaningful attribution make me suspicious. >>>> You never heard of Macworld? >>> http://www.google.com/search?q=%22The+best+Algebra+tutorial+program+I+have+seen%22+site:macworld.com >>> returns no results... Do you have a more detailed reference? >> I have the whole review somewhere, but a digital version could be tough >> to dig up. If you are seriously interested I'll look for it. > > I'm not. > > I am getting confused though between your demo, the software you say > received such high praise and the software that -- is it web software > that you're currently developing? Your quote, in this context, > sounded like it was supposed to be related to the software relevant > for this group. It was in my sig. It mentioned Algebra clearly both times. I have been talking about and demoing a Lisp+qooxdoo framework, and did not say a word about Algebra until one of the hyenas started chewing on it for no reason other than to have an opportunity for more abuse. I owe the hyena a beer for the excuse to spam that, tho. > > >>> Are you telling us that this ran on a Mac web browser in 1991? >> No, it was written originally for the Mac, in C. The defter of the >> intellects in this group might be able to figure out I have decided to >> release it as a web app, hence qooxlisp. ie, plaintiff was confusing the >> present with the past. > > The confusion is not surprising. You're here discussing some software > that you're releasing, but seem to be promoting it with a quote that > long predates this new software. See above. My other sig quotes Elwood P. Dowd about the importance of being pleasant. Does anyone think I am selling DVDs of Harvey when I post about JS above that sig? kt |