Builder: A New Development IDE Being Built For GNOME

[Akka]

Yes it does not have that, but my point would be it should be easier to add a gui builder to it, or even better make a api or framework to integrate that into many ides.

Its just like the kiddi that says hei I want to make game just for fun not so serious I want to start with a MMORPG its to big.

Just saying, btw if that is like qtdesigner only a tool to draw the widgets go ahead I have no problem with that, but a full ide makes no sense to me.

its like I want to framwork the world like in java or .net this just sucks... 1000 specialised projects with bindings to every language people use is much better. It scales better than one projekt that want to replace everything.

A gtk gui builder should be written in gtk. You can not import that into emacs (or vim).
Anjuta can already load glade so they can probably reuse that but hardly in emacs.

[jimbohale]

And yes u can target noobs that never used something better and u have some shiny thing thats ok for them because they dont know better. But is that really a good reason to start such projects, I mean having 1-2 of such projects is good too maybe, but there are so much editors or ides released lately.

You have to understand that there are very different kinds of developers. There are very highly skilled programmers who are much more knowledgeable than the less experienced programmers, but less experienced programmers are still very necessary for developing applications. Look at how Apple did it and take notes. They essentially made it easy for noobs to write programs that don't suck. That's what the goal should be, because if you can pull that off a lot more will people will be willing to write applications (albeit less than w/ apple because of lack of profits, but still some) where as right now people are generally turned off by the learning curve. They probably won't be the most impressive applications in the universe (i.e. not life-changing apps) but there isn't even a good note application on GNOME that can sync with my gmail account. I haven't complained, but I can guarantee if it was more newbie oriented someone somewhere would wind up developing it. Make it as easy as possible.

[matt_g]

I've only just got back to this topic - good to see my question has stirred up a lot of discussion. Sometimes I feel like a bit of a dinosaur when it comes to programming - I predominately write C, I use emacs, so it's interesting to here other peoples opinions on stuff.

I'm not sure I agree on making an environment for "newbies" to program in. I question the need for a newbie programmer to use an IDE at all.

At the risk of sounding even more like a dinosaur I first learned programming in High School by writing Pascal this was in an MS DOS environment the sorts of things I learned back then (if statements, loops, simple functions etc.) would still be relevent nowadays although I guess you'd use something modern like Python or Javascript to teach newbies. All you need is a very basic text editor, something like Gedit would suffice, if anything a full environment would be distracting and confusing.

I think the biggest opportunity lies in intermediate people like university students. I know when I started uni it was a huge cultural shock I grew up in the 90's during the Dos/Windows 95 era when I started uni the PC's in the computer lab weren't even PC's (they were Sun SPARC workstations - running some strange OS called Solaris and using CDE which is still ugliest desktop environment I've ever seen). I ended up learning Linux (anyone remember Mandrake - the Ubuntu of its day?) because I wanted to work on my assignments from home and needed a Unix environment. I'm sure people still do this today. If GNOME want to attract people they should be making an IDE aimed at this group - students new to linux if they like the environment well enough then they may be encouraged to contribute to the project.

[anda_skoa]

At the risk of sounding even more like a dinosaur I first learned programming in High School by writing Pascal this was in an MS DOS environment the sorts of things I learned back then (if statements, loops, simple functions etc.) would still be relevent nowadays although I guess you'd use something modern like Python or Javascript to teach newbies. All you need is a very basic text editor, something like Gedit would suffice, if anything a full environment would be distracting and confusing.

I have a very similar background so I can related to that and mostly agree. However, my understanding of the current situation is that students of any age nowadays need gratifaction really quickly otherwise they loose interest.

And gratification for them means good looking UI, not functionality.

So just an editor won't cut it anymore. Which is sad, learning to program was all about solving problems, automating things, etc., not looking at pretty pictures.

Cheers,
_

[blackiwid]

And gratification for them means good looking UI, not functionality.

I dont think tahts true, there are always only a few that use linux, its never the major thing, even in my technical university of applied science, most people did not like linux so well, they either only used it only on the desktop pcs (I think it was a suse linux) other some thin clients, or they used linux in virtual machine on their laptops.

And I dont think their problem was that visual c++ was such a good gui, its more like they needed to document it and they knew from their past how to use microsoft office and they dont wanted to invest time to get effective with something different.

but we should focus on this few guys and help them to be more effective. The geeks. And its a fact that developing something from 0 basicly is hyper-ineffective.

as example I have this tool written in python: mmailer.

https://github.com/spiderbit/mmailer

I think I chose gedit to do so, not that made me responsible for 1000 tasks taht I can fuck off like using always the same indent mechanism 2 or 4 spaces vs tabs, not only having a console beside it and pressing up + enter to check it again. with nosetests u can automate that saving -> tests runs. not only that using git through command line is not always maximal comfortable and good visual.

But a much simpler problem, the programm only (not tests) alone is 500 or 600 lines long. When I do the same integraded in emacs with lisp I can do it BETTER because I can use gnus for editing and managing mails in 20 or 30 lines.

So huch much time do u think would I have saved writing this 20-30 lines of code vs this 500-600.

Not to mention all the nice features like eval-region and so on from the ide.

Of course tahts not directly a feature of a ide, u could do that maybe too with not-integrated seperate tools use thunderbird etc. But u have then several seperate tools where u have to fiddle with 20 different apis.

Emacs devivers just the perfect integration between development and functionality. everything u can do there is a function name for it. there is no obvious similar way to get as easy the name of the function of the command to receive your news mails like that

in gnus
F1 k g (k for key)

【】 runs the command gnus-group-get-new-news, which is an interactive
compiled Lisp function

It is bound to【g】.

(gnus-group-get-new-news &optional ARG ONE-LEVEL)


in thunderbird. So u have to google 10 minutes around to find this same information and its most likely 30 lines of code to trigger taht from a pllugin or something.

So in other environments u have a total cut between development and usability, and u have a cut between specialised programms, back in the days that was even better in some way, before firefox came out netscape included more stuff.

so then devleopers have to write some interprozess code to get stuff together again to get functionality from different apps together.

So in my opinion, have a good backend maybe invest a bit to make emacs even better bring nonblocking stuff forward. And build some nicer looking guis for that would be the best solution feature wise.

Its a bit like gnome would say now they want to release a new office suite that will fight with ms office or libreoffice, they start now and in 1-2 years we have a msoffice and libreoffice killer. It not gonna happen.

Again I also understand the noob-ide argument, but why do u need 20 of such ides. so that people can write something in their resume I guess.

[blackiwid]

https://pypi.python.org/pypi/elisp/0.5

u could even use something like that to make it more accessable to others or if u really thing u cannot modernise emacs enough start on something similar written in python.

The Problem is to replace emacs u would need at least 10 years development from many people. I am not talking about the core stuff but about all its hundrets modules.

Even eclipse is there better its not so a small community than gnome-only developers.

But its java based so 100x slower uglier and worse than emacs especialy under linux where java is slow as hell.

[anda_skoa]

I dont think tahts true, there are always only a few that use linux, its never the major thing, even in my technical university of applied science, most people did not like linux so well, they either only used it only on the desktop pcs (I think it was a suse linux) other some thin clients, or they used linux in virtual machine on their laptops.

And I dont think their problem was that visual c++ was such a good gui, its more like they needed to document it and they knew from their past how to use microsoft office and they dont wanted to invest time to get effective with something different.

I don't see how that could possibly be related to the statement you were quoting.

as example I have this tool written in python: mmailer.

https://github.com/spiderbit/mmailer

Which does not seem to have any GUI, least a fancy one.
As I wrote before, the goal of people getting into programming seems to have shifted towards getting some fancy UI really quickly otherwise they loose interest.
Functionality comes after that, if at all.

Cheers,
_

[blackiwid]

I don't see how that could possibly be related to the statement you were quoting.


Which does not seem to have any GUI, least a fancy one.
As I wrote before, the goal of people getting into programming seems to have shifted towards getting some fancy UI really quickly otherwise they loose interest.
Functionality comes after that, if at all.

Cheers,
_

the mmailer example was to point out my efficiency argument, so if u can do the same in 1/10 or 1/20 amount of code with a integrated tool like emacs and u get better functionality, it would be more efficient to advertise emacs to more people. I am not saying that everybody has to use emacs, but every developer (except maybe web developers, but even there it cannot hurt) that develops also for linux should have advertised proactive how powerful emacs is. I had no idea till I was over 30.

It has not that good reputation it should have, it gets compared to vim, waht is basicly onle a editor, I thought emacs is a editor with some wierd shortcuts.

And even if u think emacs looks not nice enough (has no fance gui) or something else, that would be a point to target, to write a frontend people can at least relate more, make a shiny gnome client, modernise it give that client better defaults (shortcuts).

At least promote emacs as the ultimate professional linux development tool that nobody comes around if he wants to develop professionaly under linux heck even other oses.

or if something else cant be fixed and people dont like (but copy basicly) like lisp, start to write a successor of emacs.

And even if u think u need a shiny tool and u cant tune emacs to be that a bit, why do we need 20 such shiny ides everyone beeing not very good because the userbase is to small, because its easier to rewrite completly a tool in a other language and toolkit than integrate 2 toolkits and target them? U cant tollerate a qt programm to develop gnome apps and vise versa?

I get taht for a small text editor I get that for some video apps, even a browser taht uses the same backend (epiphany and konquerer), but I dont get that for a IDE.

[anda_skoa]

And even if u think emacs looks not nice enough (has no fance gui) or something else, that would be a point to target, to write a frontend people can at least relate more, make a shiny gnome client, modernise it give that client better defaults (shortcuts).

Ah, I see, you are not replying to my comments at all, but to some other discussion.

You could either use the "Reply to Thread" button or "Reply With Quote" on a comment that is part of that discussion..

My comment was not about editor/IDE UI at all, hence your comment looking so out of place.

Cheers,
_

[jimbohale]

Ah, I see, you are not replying to my comments at all, but to some other discussion.

You could either use the "Reply to Thread" button or "Reply With Quote" on a comment that is part of that discussion..

My comment was not about editor/IDE UI at all, hence your comment looking so out of place.

Cheers,
_

No idea where the guy who made that post made it, but I disagree with what he's saying. Trying to turn emacs into something it's not (a shiny simple editor that's very mouse oriented like the screenshot) will only cripple Builder and if they have to change emacs at all, cripple that. It has a very different goal for a very different audience, so it should not necessarily be an extension of emacs.