TIOBE Programming Community Index for August 2011
Locked
12
12
 
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
On Sat, 2011-09-03 at 06:29 +1000, Paul King wrote:
[ . . . ] > Yes, I think more discussion and action in regard to the documentation > and the site is welcome. I know that SpringSource/VMWare have also > been looking at that space but the more input the better I would > think. Does anyone know anything about the algorithm that TIOBE use? I know it is secret like Google's algorithms, but some high-level details would help. What actually is it that gets analysed? Clearly it isn't actual use -- simply looking at the top 50 list and applying prejudice gets you that far. My suspicion is that there is a huge amount of historical data included in each months figures, creating a systematic bias in favour of older languages. Also there seems to be a bias in favour of languages used in universities. There is also a huge non-linear amplification effect going on that is great for marketing, but not so good for truth. Compare Java, Ada, Lua, F#, Python in the August 2011 figures currently at http://www.tiobe.com/index.php/content/paperinfo/tpci/index.html -- The "binning" / "siloing" effect is remarkable. Java up 1.4% F# up 0.3% yet F# is a rising superstar of green, and Java is a mere static thing of yellow -- simply because the headline is about relative position not about actual facts. -- The disconnection from reality is clear. Ada up 0.3% (huge green) Lua up +0.8% (huge green) Python down 0.8% (small red) No way does this mirror a change in usage in the real world, actual or otherwise. Moreover we are seeing the "binning" effect, and also an issue of dealing with masses of small numbers. All of these factor come together to create huge swings that create headlines. Python has managed to milk this twice to become the annual champion. Of course some of the trends have clear echos in reality. The rise of Objective-C and F# are reflected by observation of the real world. Lisp's rise may actually be an issue with Clojure -- why Lisp is separated as a category is not clear, there are so many of them: Common Lisp, X-Lisp, E-Lisp, Clojure, Scheme, . . . As to how Q, ABAP, MOO, Alice, etc. are in the top 50 but Groovy is not must clearly lead to some indication as to what to do to get recognition. What is needed is some TEO (as opposed to SEO), but for this we need insight into their algorithm, not speculation as above. -- Russel. ============================================================================= Dr Russel Winder t: +44 20 7585 2200 voip: sip:[hidden email] 41 Buckmaster Road m: +44 7770 465 077 xmpp: [hidden email] London SW11 1EN, UK w: www.russel.org.uk skype: russel_winder |
signature.asc (205 bytes)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
In reply to this post by Ronny Løvtangen
On Sat, 2011-09-03 at 22:47 +0200, Ronny Løvtangen wrote:
> On Sep 3, 2011, at 9:44 PM, Bill Stephens wrote: > > Scrolling: Our web usability guru says that "a lot of scrolling is > > bad" [ . . . ] > > Scrolling isn't as bad as it used to be (according to Jacob Nielsen). > For articles at least, scrolling is much better than paging. The issue here is matching the user interface activity to the user activity and task model. I suspect the point here is that scrolling a big text document that is being read linearly is better than paginating it since the user interface activity interferes less with the process of reading -- despite what we do with paper novels. On the other hand having single page chunks with good links make for more easily browsed content. The single most important thing though is navigability: can the user arrive at the site and find the material they want in 1..3 clicks, and if they want something else can it be got in 1..3 clicks from wherever they are. This is simplifying the issue for the purpose of this email not becomming a novella. There are folk out there (I know I used to be involved with the human--computer interaction and psychology of programming communities) for whom all these navigability issues are well understood and easily implementable. Oh, and how big is a page? Actually that seemingly humorous question enters into the realms of much serious activity in terms of usability. Remember that default font size is set by the writer, but actual font size is set by the reader. > But I agree that the menu could be restructured. And that there should > be more attention to what is above the fold. Now the content above the > fold is mostly commercials (about Typesafe Scala Training ;) and > headers. > Well here we come to the crux of the matter. Scala now has lots of people with financial vested interest, Grails and Gradle likewise. Thus there is generally some funding for sales and marketing activity. From what I can tell the only commercial funding of Groovy is, I believe, VMWare paying Guillaume and Jochen -- who, I think they would admit themselves, are not user interaction experts, nor being paid to do user experience development on Groovy marketing. Also of course, not to forget, Andrew and Andy who are doing work on Eclipse support for Groovy and Grails. So the question is how does a language get commercial funding to support it's marketing? Well Typesafe is following the same path as G2One tried. If G2One were still here then the banner advertising would be from them rather than Typesafe. As it is VMware/SpringSource clearly see no commercial advantage in supporting marketing of Groovy -- why should they is is not an income generating thing, that is Grails, Roo, etc. Clearly this is a bit short-sighted since without Groovy development Grails stagnates. > The Front page contains too much information. > * details about what is new in Groovy 1.8 could be removed. Too much > noise on a front page. Keep just a link to download latest version. Agreed. > * JAX innovation award 2007. Please remove. It's intention is probably > "look how cool Groovy is", but what I read is "not much has happened > since 2007". Indeed, and agreed. > * Samples: this is nice to have on the front page. Developers loves > samples. But needs better formatting (newlines, colors), and maybe > some more examples of what makes Groovy groovy (e.g. collection > methods) Th D programming Language website comes in for criticism for having an example on the front page! http://d-programming-language.org/ > * Documentation: this could be removed, available under the menu > "Documentation" > * Modules: move to menu, or under Documentation. > Whilst it can, and will, be improved, the GPars site is better at the issues of scale [ . . . ] > > Tutorials / Documentation: > > I'm a big believer in tutorials and I write them for my products and > > struggle with them. > > It's down to resource though. Volunteers come and go as they want, starting activities and dropping them unfinished if they so desire. If Groovy is to have quality materials guaranteed then some quality technical authors need to be given contracts to provide quality finished material approved by Guillaume. > > > > I find the Cookbook Examples to be useful. Some of the other > > documentation seems to have been quickly written and poorly > > formatted. > > See above! > > > > Example: I have to click through 4 pages to get to > > XMLSlurper /XMLParser documentation. When I get there, I'm not > > exactly thrilled with the content. I get a few text boxes and a some > > comments. The box containing the Groovy parsing examples could be > > broken out into smaller, clearly commented sections. > > See above! [ . . . ] -- Russel. ============================================================================= Dr Russel Winder t: +44 20 7585 2200 voip: sip:[hidden email] 41 Buckmaster Road m: +44 7770 465 077 xmpp: [hidden email] London SW11 1EN, UK w: www.russel.org.uk skype: russel_winder |
Loading...
| Powered by Nabble | See how NAML generates this page |
