Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [ide-dev] IDE working group questions

First let me introduce myself, I'm the current maintainer of PyDev (which I believe is currently the de-facto plugin for editing Python in Eclipse).

I must say that I share some of your views (so much that I did a crowdfunding for PyDev and an Eclipse-based product which I called LiClipse: The crowdfunding ( was successful enough to keep PyDev going and to create LiClipse, but didn't reach the higher levels which I had set in order to make the code to LiClipse open source (i.e.: in LiClipse adding a new language is mostly specifying a single file with enough information for syntax highlighting, brackets matching, outline, etc.).

I believe some of my main points are exactly the ones being discussed (1. get funds, 2. improve some fundamental things in Eclipse -- in my case, having an improved dark theme and default editors).

So, I'd like to share a bit of my views on the subject:

First, the Eclipse platform as is covers a whole range of things from RCP applications to the IDE, so, there are lots (and LOTS) of different things it must support and definitely means different things for different people.

Basically, I don't think the basic platform must cover all cases. That's the beauty of its plugin system. Yes, I also have my own pet-peeves there (i.e.: currently each plugin must re-create the spelling engine, it's not possible to style some things because they don't have ids or are not reachable, swt table selection colors is not customizable, etc. -- so, some things can't really be done out of the platform, and I think this is where the effort should be geared here, for most others, additional plugins can usually be installed to do what you want). 

In the particular case of having a default editor the platform does give the proper support for anyone that wants to do that (which LiClipse does) -- but I can also see others wanting the current behavior, so, maybe having a separate plugin for that is better -- which also probably avoids a ton of other related option in the platform :)

Now, I'm not saying that all should happen outside of the platform: it does need to go on and get more developers and make a stable/feature rich platform, but from an outer point of view and if someone is asked to help there, it'd be interesting to delineate what is being talked about. I.e.:
- on what exactly would the money be spent?
- is it a collection of issues that should be solved to make it go forward? 
- who decides which are those issues?

I'm also not really aware how things work now on the eclipse foundation -- all devs that work come from some external company? (i.e.: how was the decision on eclipse 4 taken and where did the resources come from?).

Personally, I think that a crowdfunding could be a choice to make things better on the platform, but it may be really hard to delineate what it'd mean when it comes to Eclipse (and resonate with a community, especially because I think Eclipse has a strong enterprise-like image and in my own crowdfunding experience, it's harder to get companies to contribute to a vision than individuals).

Also, in my case, just by explicitly showing that I needed help, more people appeared and now there are more regular contributors in PyDev. In the case of Eclipse, it could be helpful to ask for development effort on specific areas to get developers there (and only after having that explicit, then it may make sense to ask for funding) -- having things specified by a committee later may easily not give people what they want.

Just my 2c.

Fabio Zadrozny

On Mon, Oct 21, 2013 at 2:09 PM, Eugene Ostroukhov <eostroukhov@xxxxxxxxxx> wrote:
I see no point in voting or opening new bugs. No number of votes will fix 7 y.o. bug. Eclipse bug tracker is becoming not unlike those pointless online petition sites.

1. Inability to handle unknown file types. General-purpose editors (e.g. Sublime) will open any file type as text file and provide fairly easy means (e.g. not requiring PDE experience) to setup syntax highlight, brackets matching, etc. Eclipse simply pops open a "system editor" (e.g. Xcode, web browser or Visual Studio) which is incredibly irritating and slow.
You should add your comment and your vote to that bug: , and also create a new bug to ask a property to select behaviour for unknow file types: open system editor, or open default text editor in Eclipse.

Default text editor needs to be augmented to compete with popular lightweight editors. Or should become pluggable. Neither can happen as it requires changes in Platform.

2. Inability to work with non-workspace files. The only way in Eclipse is by using that menu item - while other editors can "open" a folder and then provide features like Eclipse "Open Resource…". In Eclipse it is not possible to use Git integration, Open Resource and such.
You should either open a bug to request that "Import directory as project" or "New Project from existing directory". Please come back with a link when you're done, that's actually something that I miss too, and that is probably not-that-difficult to implement and get merged.

At the very least, .project should not be put into that folder, there are probably other issues as well. It is possible to implement (I wouldn't be surprised if some Eclipse-based products already have that) - but impossible to get accepted to the mainline. Opening a bug will not solve this.

Best regards,

This email message is for the sole use of the intended recipient(s) and may contain confidential information.  Any unauthorized review, use, disclosure or distribution is prohibited.  If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message.

ide-dev mailing list

Back to the top