community code sprint

[Nicolas]

We all know some aspects or have one idea that we'd like to do and that could be of value to 'the community'.

But more often than not, as individuals, we are missing something : wether it is math skills, html5, design talent, the ability to take a fresh look on documentation we did or simply unsure if it is worthwhile.

If we could pool together a group of few people, that would make overcoming those obstacles much easier.

The idea is to have some F# programmers of different levels teaming up together to work on one thing.

I imagine that starting with a physical meeting provides the highest practical bandwidth.

So, FSharp Open Source Community, I wonder :

What would be the closest thing to kickstart this ?

Doing this on top of the already scheduled events in London ?

Having some meetings in London (Cambridge? Elsewhere ?)

What's your take ?

[Eric]

Hi Nicolas,

I think just promoting project(s) as open for others to jump in is
enough. Be it here, at a user meeting, or on a forum on a Q&A site, in
the readme as a Github project, etc.

If we had an ongoing thread or web page that was updated monthly of
projects open for others to join, that would be better.

> --
> --
> You received this message because you are subscribed to the Google
> Groups "FSharp Open Source Community" group.
> To post to this group, send email to fsharp-o...@googlegroups.com
> To unsubscribe from this group, send email to
> fsharp-opensou...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/fsharp-opensource?hl=en
>
>
>

[Nicolas]

That is always possible, and best for long term contribution.
but I feel like this is also inefficient for certain smaller commitment :

-People can stump on very small things, upon installing a project for instance.
With a few experts providing guidance on the spot, those problems 'go away'.
Real world has a much higher bandwidth

-Collaboration can be of short length, say 3 hours or just an evening, and for those short term collaborations the overhead might simply be too much. In 3 hours one can do a lot of things, for instance testing installation on different platforms and reporting bugs. Things that, without some social context or additional interest, would be quite a pain. ( especially if the original author of a library has done a significant effort already on non trivial things!)

From a community perspective it is also less intimidating for beginners than diving into a hard subject by themselves (who said 'handbook of automated reasoning' ? :) ) and promotes an active attitude over tutorials.

--

I like the idea of having a list of open projects as this makes the information flow from project to contributors.
I wonder what is the best form for this idea, as it might be not dynamic enough.

May be integrating :
- a web page 'community sprint' with a theme being displayed in advance which leads to
- a physical meeting
- plus a chat (or google hangout), integrated on the github pages of the project being worked on, for people not physically present but who want to contribute as well

[Eric]

Hi Nicolas,

I'm glad you mentioned projects of a few hours, I was not thinking in
that direction.

While I too prefer in-person meetings, if the people willing to help
are not local then one has to weigh the pros and cons of forgoing
in-person meetings over the benefit of help from others with
complementary skills and availability regardless of the burden of the
communication channel.

I am not sure how much this idea dove-tails with The F# Software
Foundation but I would not mind hearing from them on this thread.

While I am not a user of remote control software for another computer,
this might be something worth thinking about. If one is concerned
about security, then one could sand box the system needing help in a
virtual machine before allowing someone to take control of it.

Thinking of virtual machines, maybe a virtual appliance for F# based
on Virtual Box, Linux, Mono and the open F# could be created.

[Robin Neatherway]

I think it's a great idea to try and get more people involved if they
are perhaps waiting on the fringes but unsure how to get into the
code. Physical meetings are probably best for smaller projects of a
few hours or working out some design decisions and directions. For
existing contributors, Google Hangout or similar would probably be
almost as good. Location-wise for me, Oxford is the best, then London,
then Cambridge.

[Nicolas]

I threw out some pages here :

http://fsharpcommunitycodesprint.github.com/

As of now, since I do not have any idea of what the audience is and how to mix the ingredients so that everyone is happy, it will probably just be me and some friends .... ! :)

The point is to get a feel of the surroundings (literally actually, having landed in England 6 days ago)

Hopefully there will be a way to make it worthwhile for all the audience involved.

I'll convene with a bearded pillar member of the foundation on the 12th of feb at skillsmatter, so we'll of course coordinate with them in the future.

[Nicolas]

The idea here would be at first to test ride, count the forces in presence, and get the communication flowing between the skilled F# people in a real world environment.

Then to really try to open up toward the audience (outside | on the fringe | new to)  F# ... and hopefully give them the practical means and guidelines to become active contributors.

[Tomas Petricek]

Hi Nicolas,

Thanks for organizing this! It sounds like an excellent initiative and I would definitely love to come (I think meeting in person when collaborating on projects is really useful). London sounds good to me – there are many F#ers in London and it is easy to get there from Cambridge, Oxford (and Ely :-)) too. I will be away on the first weekend of Feb, but otherwise I’m free (filled the Doodle) and would love to come.

 

I think there is a lot of interesting projects that could be done around Open Source Editor/Integration. This includes working on MonoDevelop and emacs integration, but also making this available as a web service (for tools like www.fssnip.net and www.fsnotebook.net), which is something I discussed briefly with Tim (author of F# Notebook) and Phil.

 

Another (related) project would be creating hostable F# interactive, which is a something that quite a few people ask about and should not be too hard (again, this could also wrapped in an easy to use web service). Tim even designed a mockup for the possible API: http://docs.fsnotebook.apiary.io  

 

Another project I had a lot of fun with is https://github.com/ZachBray/FunScript. We did some coding on this with Zach (while not listening at a F# user group meeting J) and I think it is very clean project and a one that is easy to contribute to.

 

T.

[Nicolas]

Those projects are definitely interesting, and this is great that you provide your overview on which one might be more suitable for this kind of contribution.

I added those choices to the doodle.

Although having a enlightened dictator might be better, explicit choices are not of null value either [insert pun on null here]

------

I personally would be quite excited to see FsNotebook grow further:

IPython showed us a great deal at what is feasible, and a serious F# version might go a long way. 

Apart from ease of contribution, my personal criteria for selecting projects are :

- absolute necessity, like good mono integration

- potential for reaching out audience : we need tool geared at reaching out and exposing the advantages of using F#.

In that last category, the web is mandatory and static html is a simple, easy to host, approach.

Having FunScript seems perfect to sidestep the downside of 'static hosting' and give it 'dynamic features' (did not look so much into it though)

So I would not mind trying to bend Funscript with an eye on reaching out/simplicity

And a complete FsNotebook would of course be a good showcase and useful communication medium for scientific related problem.

-----

[Robin Neatherway]

I'm keen to keep pushing forward on the editor integration (hopefully
supporting other editors eventually as well). I'm interested in the
fsnotebook too. I'm happy to come and see what other people are
working on and bounce my ideas off them -- I think that alone will be
very useful.

I'm interested in how fsnotebook compares to tryfsharp. It works on
Linux so it obviously isn't using Silverlight. Is the F# process
running on the server?

[Tim Robinson]

Yes, your code runs on a pair of Azure servers that I have. This means fsnotebook can't do the same kind of Silverlight integration that tryfsharp does, but it can get away with using plain HTML.

[Adam Granicz]

You guys are also welcome to look at and contribute to WebSharper, I
could come over to London for some coding if anyone is interested.
There are relatively few F# web programmers, so it's important to
understand what frameworks are out there and what are their main
benefits/drawbacks. BTW, has anyone compared WebSharper/Pit/FunScript
in any way? It would be useful for those of us working in
F#+web/mobiles to understand each better.

Cheers,
Adam.

Adam Granicz, IntelliFactory
www.intellifactory.com

[Nicolas]

Added Websharper to the list.

We definitely need to have at least one Mister Websharper based in London at some point.

I am not yet so 'webish' so I defer to you guys to have judgement on which project to prioritize.

From an ecosystem pov, that would be in order to not reinvent existing wheel (within the AGPL limit, which I have not explored), but also so that we can put you forward as a biz when possible.

-----

The community needs fruitful companies, and independently of the technical solution, websharper offers a unique feature among what is available : this is backed by an actual company, so customers know they can contract / rely on you. This creates trust, and trust is (primarily?) what they pay for. 

So I think the community should have you in their head, and have the ability to pinpoint a good showcase work of yours to start giving confidence to prospective clients.

[Tomas Petricek]

I think the two main active projects for JavaScript at the moment are WebSharper and FunScript (AFAIK, there is a plan to integrate the interesting parts from Pit into FunScript).

My view is that it is really good to have both –

1. During my recent talks, I used FunScript, but I always said that there are two projects and those who want commercial support can look at WebSharper (which has – or I guess will soon have ☺ – pretty much all the features of FunScript). FunScript has some exciting features that are nice to demo, but those interested in enterprise development (and have money to spend) will be more interested in what WebSharper offers.

2. I think it is good to have two independent project as it opens room for more inovations – e.g. I think FunScript integrates nicely with some of the F# Data type providers while WebSharper has more advanced libraries for GUI and client/server programming and they can both learn from the other and try different approaches.

3. I think the audience is different – my impression is that WebSharper is aimed at more professional and/or enterprise developers, while FunScript aims at "hipsters".

From a community perspective, it is nice to have an Apache-like licensed project. It means that if I contribute to it, I will be able to use it (and my work) for both open-source and commercial purposes without restrictions. AFAIK, with AGPL licensed code, I would not be able to use the project (with my contributions!) for closed-source project and that is a major limitation for me.

Anyway - I guess we can focus at playing with compiler components at the first meeting and perhaps try to do web-development at some future meeting.

T.

[Adam Granicz]

On Tue, Jan 22, 2013 at 1:27 AM, Tomas Petricek <to...@tomasp.net> wrote:
> I think the two main active projects for JavaScript at the moment are WebSharper and FunScript (AFAIK, there is a plan to integrate the interesting parts from Pit into FunScript).
>
> My view is that it is really good to have both –
>
> 1. During my recent talks, I used FunScript, but I always said that there are two projects and those who want commercial support can look at WebSharper (which has – or I guess will soon have ☺ – pretty much all the features of FunScript). FunScript has some exciting features that are nice to demo, but those interested in enterprise development (and have money to spend) will be more interested in what WebSharper offers.

I am going to dwell on this a bit, because by such misdirection you
are doing a lot of harm not to just WebSharper and FunScript, but also
to anyone who is seriously looking at F# for web development. While
we are indeed more interested in a proven, enterprise-ready framework,
we can't fully get there without open source users, some of whom just
want to play with web programming.

I can see two features of FunScript that are not yet available in
WebSharper, and this is perfectly normal to me (and don't forget the
dozens of features the other way around - by the virtue of WebSharper
and FunScript aiming to be different things). These are:

1) Being able to compile specific quotations into JavaScript - giving
you control over what gets translated. WebSharper takes a holistic
approach and gives you not a single script, but an entire web
application for your F# code, with all pieces in place. These pieces
include client and server code, and no manual hacking is needed to get
the individual scripts in place where they are needed. Indeed, with
applications with tens of thousands of LOC, this is a must.

As you indicated, we are providing a low-level API for this sort of
manual, low-level use (it's on a branch ready to release to support
customers) - as indeed there are some useful situations where this is
desirable. Our upcoming Cloud IDE supplies one such case, where you
can run F# client-side code in interactive mode on the fly.

2) Being able to interface with external JavaScript libraries through
a type provider for TypeScript definitions. Indeed, this is massively
helpful, and we have been working on a robust implementation for
WebSharper for quite some time now. There are some important caveats
of a TypeScript type provider, which are more apparent in larger
bodies of TypeScript definitions such as Metro JS, one of the case
studies we are currently exploring with our implementation. The
FunScript TP is likely to hit similar issues when it matures and
becomes more complete.

All in all, both differences raise significant points about the
difficulty in covering these features - and leaves a lot to be
explored. I can't express enough how useful FunScript is in helping
to understand these areas better, and bringing new ideas on how they
can be explored and utilized.

> 2. I think it is good to have two independent project as it opens room for more inovations – e.g. I think FunScript integrates nicely with some of the F# Data type providers while WebSharper has more advanced libraries for GUI and client/server programming and they can both learn from the other and try different approaches.

Absolutely, in fact I would welcome even more - they are all helping
each other by bringing attention to functional web development. Can
you explain what you mean by better integrating with data type
providers? Here is our Cloud IDE doing interactive data
visualizations using type providers:

http://vimeo.com/57399434

This is running F# code without Silverlight, an order of complexity
beyond running data visualizations from a web app (WebSharper or not).
WebSharper helped us tremendously to get this carried through, and
was quite a pleasure to work with.

> 3. I think the audience is different – my impression is that WebSharper is aimed at more professional and/or enterprise developers, while FunScript aims at "hipsters".
> From a community perspective, it is nice to have an Apache-like licensed project. It means that if I contribute to it, I will be able to use it (and my work) for both open-source and commercial purposes without restrictions. AFAIK, with AGPL licensed code, I would not be able to use the project (with my contributions!) for closed-source project and that is a major limitation for me.

If you mean "hipsters" who don't want to pay for a commercial license
to use an open source project for closed source use - sure. And that's
totally fine. Our users want to get things done and they are willing
to take the bumpy ride to learn about something, with the hopes that
it can help them in the future. But I don't think "hipsters" are any
different in the long run.

> Anyway - I guess we can focus at playing with compiler components at the first meeting and perhaps try to do web-development at some future meeting.

Seriously - my top priority at these code meetings would be to extend
and enhance language coverage to/on new platforms. Use what's out
there, and improve on things that need improvement. Don't create
inner competition, and above all artificial competition.

You guys rock! Looking forward to do some of these code runs...

[Adam Granicz]

Hi Nicolas,

It should be fairly easy to find WebSharper gurus-to-be in and around
London - I know of several people working with it on various projects.
But yes, the more the better and the merrier :)

BTW - should we at some point arrange some WebSharper training in
London? Anyone else interested in spreading the word?

[Nicolas]

Awesome video. I can imagine how powerful this editor is.

I imagine that open many path for custom applications.

My two cents on Licence : 

In the end, what really cost in putting a software at work is integration.

This is the major cost, in time and money, hence a billing opportunity.

People are ready to pay to avoid that cost (and uncertainty, which is where trust comes in)

The other thing that business pay for is service.

- I own a garage, and I need to handle a calendar for my clients and me, to make appointments.

- I am a bank and I pay for Bloomberg access for my employee.

- I am a taxi driver, and I refer to a call network to dispatch me customers

Here they pay the service, because it is (ideally completely) orthogonal to their trade, and they think they are best just taking care of their client and outsource one functionality

I am not sure Licence, by itself, is an effective monetizing strategy.

But I guess the discussion belongs to another thread, so lets continue it somewhere else...

[Nicolas]

We'll probably have to print out a recommended 'tool stack' to streamline dev.

I'd imagine

- mono as target platform, as we try to integrate into OSS

- git for source control

- nuget for dependency management

- fake for builds

- Formatting for documentation

does it make sense ?

If so, I'll add it to the website along with some instructions to get the piece working.

[Robin Neatherway]

On Tue, Jan 22, 2013 at 7:48 AM, Nicolas <nicolas...@gmail.com> wrote:
> We'll probably have to print out a recommended 'tool stack' to streamline
> dev.

Generally a good idea, I think, although being *too* prescriptive can
put people off. At least having good recommendations and how to get
them working can be helpful.

> I'd imagine
> - mono as target platform, as we try to integrate into OSS
> - git for source control

Seems pretty standard to use github at the moment.

> - nuget for dependency management

I haven't used nuget, does it work on Mono/OSX? My googling says no at
the moment.

> - fake for builds

I haven't used fake either, but would like to see what it can do. At
the moment, `xbuild project.fsproj` does the job for me, as I don't
have any more complex requirements. I saw that it can integrate calls
to msbuild/xbuild.

> - Formatting for documentation

I think the Formatting output is very nice. As we all know, people
tend to prefer coding to documenting, but if Formatting makes it
easier to produce documentation that can only be a good thing (I think
there should be documentation, just to be clear).

[Nicolas]

On Tuesday, January 22, 2013 11:41:56 AM UTC, Robin Neatherway wrote:

On Tue, Jan 22, 2013 at 7:48 AM, Nicolas <nicolas...@gmail.com> wrote:
> We'll probably have to print out a recommended 'tool stack' to streamline
> dev.

Generally a good idea, I think, although being *too* prescriptive can
put people off. At least having good recommendations and how to get
them working can be helpful.

Yes, it would be meant as a guideline, for people in search of guidance on where to start.

That could be important to avoid unnecessary overhead in 'short' sessions.

The more advanced can adapt and use their own toolchain.

 

> I'd imagine
> - mono as target platform, as we try to integrate into OSS
> - git for source control

Seems pretty standard to use github at the moment. 

Yup, github account is added on 

> - nuget for dependency management

I haven't used nuget, does it work on Mono/OSX? My googling says no at
the moment.

Good news : it does work ! 

Personally, I can't not use it, I think life is too short for not using dependency tool, like maven, bundler, brew, whatever.  "nuget install packages.config -o Packages", and you are good to go ! 

I'll add the few instructions on the website

 

> - fake for builds

I haven't used fake either, but would like to see what it can do. At
the moment, `xbuild project.fsproj` does the job for me, as I don't
have any more complex requirements. I saw that it can integrate calls
to msbuild/xbuild.

I used it only to compile Funscript, and since the good guys are using it.   

We can have many 'recommended' toolchain, (actually it can be good to test out to have different reference setup), the point is to have an tool for that step.

 

> - Formatting for documentation

I think the Formatting output is very nice. As we all know, people
tend to prefer coding to documenting, but if Formatting makes it
easier to produce documentation that can only be a good thing (I think
there should be documentation, just to be clear).

I am a fanboy of Formatting. And I think literate programming can go way further than that.

Btw, if you need a simple workflow to use the project, that is already packaged in a smooth CloneOnce® experience :

https://github.com/nrolland/ghnotes 

[Nicolas]

I added a better stack description 

[Robin Neatherway]

Wow, for me NuGet now works very nicely. This is great news.

If I download from http://nuget.codeplex.com/releases there are two choices:

1. NuGet Bootstrapper 2.0 (June 2012)

If I run this with `mono NuGet.exe` then I get:

Package restore is disabled by default. To give consent, open the
Visual Studio Options dialog, click on Package Manager node and check
'Allow NuGet to download missing packages during build.' You can also
give consent by setting the environment variable
'EnableNuGetPackageRestore' to 'true'.

If I run again `EnableNuGetPackageRestore=true mono NuGet.exe` then I get

ApplicationName='/Users/<user>/.local/share/NuGet/NuGet.exe',
CommandLine='', CurrentDirectory='/Users/<user>/Downloads', Native
error= Cannot find the specified file

but an up to date NuGet.exe has appeared in ~/.local/share/NuGet, from
which I can install packages from the web fine.

2. NuGet.exe command line (July 2011)

This replaces itself in-place without any complaints and the new
binary is the same version as that downloaded by the above. Then
everything just works. So this is easier, although you have to put
NuGet.exe in ~/.local/share/NuGet if that's where you want it.

[Nicolas]

Good to know.

I experienced the same issues but did not look into the differences between the 2

[Nicolas]

So as the date is approaching, and the majority is available on saturday the 9th, I will ask Google if they are fine with that date. 

Tentatively, we could do 2 hours of hacking at 10am then lunch, then everyone goes its way.

We could of course do more hacking that day but that would depend if we found a way to be productive together / have the will / time etc..

We can arrange the format, especially for the first session, but that might not be 

---------

For future events, does anyone knows a really super good punchy web designer ?

Ideally, we'd do something close to code of course, github hosted etc.. but a nicer communication of our goals and a sleak web might be a good contributing factor for the unitiated who normally just watches F# 

"we help you get up and running", "get real and into the action", "the enabler bridge" could be some taglines...

-----

In the 'meta discussion' during lunch, we have to think about ways to reach out and the 'pipeline' of incorporations of newcomers at the different phases : before the event through the web, at arrival at the event, during, after + the associated metrics.

Each phase might be analysed with respect to incentive / tracking etc..

For instance post-event, there must be somewhere a "badge system" a la stackoverflow for github project contribution... Everyone like F# badges. People even have mugs with F# on it !

[Nicolas]

Hi guys, here is an update I posted on the web site :

If you know any place that could host us in London, I would not mind a pull request :) 

What is the next sprint about ?
The majority has spoken !
The next sprint, will be held on Saturday 9th of february 2013 starting from 10am ! It will focus on helping out with
Open source editor integration
This is the occasion for you to get real and hands-on with F#, learn from the bests, and be enabled to author and contribute to F#

On small glitch is that Google's place is already booked that date !
Please watch in here or on the open source f# google group for update on the place.
We are learning and asked if they can supply a updated calendar in the future. Meanwhile we are looking for a place for around 5-10 people (the first sprint will be limited but everyone is welcome) ideally halfway between Padington and Kings Cross, in order to accomodate for all. Suggestion/Booking welcome !

[Nicolas]

Hi everyone,

 

A quick update on the location issue.

Most places I asked so far are closed on Saturdays for non permanent members.

 

We always have the option of joining in a café with internet and finding a better location for later sessions.

I doubt those will be very noisy in a Saturday morning anyway !

 

I am thinking the deadline for specifying the location would be by the end of the week.

I'll keep you updated and until then, happy coding.

[Nicolas]

Hi everyone,

 

So we have a proper venue, thanks to Tim ! who also sent a pull request for the website !

You can therefore now see here all the details about this code sprint !

 

---------

 

In order to make the session productive, I would suggest that you look into creating "issues" here and there in projects, so that we have stuff ready in a bag to solve straight away. Resist the urge of actually fixing them to keep them hot and ready.

 

For people not 'in the know' of FSharpbinding, I would suggest

- to install the emacs mode from Robin and try to make FSharpbinding tests pass

 

For those who already contributed to FSharpbinding, what would be the best entry point in your opinion ?

 

 

---------

 

Let's see how our push pipeline works together and may a thousand pull request bloom on our way !

[Robin Neatherway]

Thanks Tim!

A general update on the Emacs integration:

The main repo under fsharp has completion and tooltips. In my repo on the intellisense branch I have implemented error underlining and jump to definition, but it isn't ready for release yet. I need to do some rationalisation, which will hopefully happen this weekend, with another release to MELPA. This involves changes to FSharp.AutoComplete to keep the code maintainable.

For the code sprint, I have a few suggestions:

1. More tests! All three of FSharp.AutoComplete, FSharp.CompilerBinding and the Emacs fsharp-mode need more tests. I have a few for AutoComplete and just one for fsharp-mode. However, this is not very glamorous!

2. More editors. If someone is coming who knows about extending another editor such as Vim or Sublime Text, I will be able to help get you up and running quickly.

3. Elisp skills. If someone is coming who knows plenty about extending Emacs there are plenty of things I would like to improve in the fsharp-mode. Examples are: making the FSI interface set output from FSI to be read-only, allow up/down cursor keys for prev/next in FSI.

4. Command-line debugger integration with Emacs (see discussion about mdbg at https://github.com/rneatherway/fsharpbinding/pull/1). This would be a great addition.

I'd be keen to hear other people's ideas too.

Robin

To unsubscribe from this group and stop receiving emails from it, send an email to fsharp-opensou...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Tim Robinson]

I'm quite excited about the venue - it's central (close to Holborn tube), we'll have the place to ourselves, and we can stay all day. There's power, wifi, blackboards and a kitchen, but of course we'll need bring our own laptops.

I was thinking of looking into Vim integration based on the FSharp.AutoComplete code, as Vim is the editor I use most for F# apart from Visual Studio. I don't know a lot about Vim scripting but it would be fun to learn on the day...

[Matthew Ellis]

Hi folks. Is this an open house? I'd very much like to come along and pick some brains on the compiler integration. I'm working on the FSharper ReSharper plugin, and I'd love to know if it could make use of the compiler bindings. 

--

[Tim Robinson]

It certainly is an open house: anyone with an interest in F# is welcome - experts, novices or just spectators. It's helpful if people can reply in this thread to give us an idea of numbers though.

I'm optimistic of getting some useful results out of next Saturday's sprint, but as it's the first one, I think part of it is going to be working out how to make the most of the format, and finding out how the different coders fit together.

[Matthew Ellis]

Brilliant, looking forward to it!

 

Thanks

Matt

 

From: Tim Robinson <tim.g.r...@gmail.com>
Sent: 02 February 2013 11:17 am
To: fsharp-o...@googlegroups.com
Subject: Re: community code sprint

 

[Robin Neatherway]

Vim integration sounds like a great idea. What do you use at the moment? Is there a canonical version of an fsharp mode for Vim? A quick google found a few options:

https://github.com/DrTom/fsharp-vim

https://github.com/kongo2002/fsharp-vim

[Tim Robinson]

I've been using the kongo2002 bundle. I don't think I saw the DrTom version, but it looks to have been forked from the same place originally.

[Robin Neatherway]

I have a large pull request outstanding at the moment, so if people want to look at the code I recommend checking out my repo.

For Vim/other editors, to take advantage of the facilities in FSharp.AutoComplete we will want to know how to do the following:

1. Start a background process and communicate over a pipe.

2. Display tooltips

3. How to complete a word under cursor from a list of possibilities (preferably displaying a menu or something). Omnicompletion for Vim looks nice, not sure how to plug into it though.

4. Highlight errors

5. Schedule actions to take place when the editor is idle (no user input for some period)

On Sat, Feb 2, 2013 at 5:50 PM, Nicolas <nicolas...@gmail.com> wrote:

Fyi, the code sprint now displayed under the Functional Londoners meetup page

 

---

 

Thanks Robin for the directions on fsharpbinding :

That participates to focusing and streamlining our efforts.

 

Everyone who already knows about the library / wants to share directions / his experience is welcome to contribute on this matter.

To post to this group, send email to fsharp-opensource@googlegroups.com

To unsubscribe from this group, send email to

fsharp-opensource+unsubscribe@googlegroups.com

For more options, visit this group at
http://groups.google.com/group/fsharp-opensource?hl=en
 
---
You received this message because you are subscribed to the Google Groups "FSharp Open Source Community" group.

To unsubscribe from this group and stop receiving emails from it, send an email to fsharp-opensource+unsubscribe@googlegroups.com.