What would you like written today?

Posted 9 Jun 2003 at 18:25 UTC by Stevey Share This

Free Software You Want?

Several meta-repositories of open source software exist, ranging from the popular indexes provided by Freshmeat, to actual developers sites such as SourceForge, Savannah, and Alioth.

Looking over these sites it's clear that there are projects covering a wide array of topics, and different levels of complexity. For example there must be a million and one different editors in existance. (I sometimes wonder if I'm the only developer who's never released a text editor ..)

Some people see this duplication as a bad thing, others agree that diversity is a good thing and don't worry too much about the wasting of effort as long as the programmers involved are scratching their personal itches.

Every now and again somebody will step forward and say "we" need application X if we want to achive world domination(tm) and usually these statements are rightfully ignored. It's not possible to force somebody to develop code for free upon their own time just because you want it.

I have personally duplicated effort by creating software tools which already exist in the wild, and I'm sure I'll do so again if the circumstances warrant it.

What Do We Want?

But while I am happy to write applications and random software for amusement, and to settle immediate needs I do recognise that there are things that I'd like to see created and released under an open license.

I hesistate to say that we need them because I'm nobody special and I see no reason to force my views upon others.

For example there are a million and one email clients in existance, and there are a large number of mail servers which are also available - but "we" really are missing an open replacement for Microsofts Exchange Server.

I would be enormously greatful to any group that produced such a thing, and I'd happily donate them gifts, code, documentation, or even hard cash.

I know that the protocols used are sufficiently complex that either a small team or an extremely dedicated, focussed developer would be necessary to kick it of. (I certainly believe that the effort commitment and attention to detail necessary are beyond my present skillset)

But I am a reasonably competent coder, I've worked on some large projects, I've created at least one popular application from scratch, so it seems to me that I do have the ability to make a difference. To give people something they'd want. What I'd like to suggest with this article is that people should list the tools, toys, and applications that they would love to see. If there were such a list of desired software I'm sure that developers could look at this, and eventually a pairing would occur.

A developer could see something within the list and say "I could do that, hell I'd even use that!".

Essentially I think it would be a great way of kickstarting a developer who was bored and unfocussed - somebody who wanted to code but lacked inspiration. (This discribes my current state to a T!)

Infrastructure

I'm not currently aware of a good "idea repository" for storing these ideas, I believe from thinking it through a little, and discussing it with friends that such a thing would be useful.

Whilst I don't believe it would ever be as popular as Freshmeat, or SourceForge, being a resource primarily aimed at developers not users, I do think it could come to nicely compliment such sites as the Unmaintained Free Software index. (Formerly hosted at SourceForge, then bero.org - now unmaintained itself?)

I'm posting this article in part to solicate opinions of whether such a site would be useful, and in part to encourage users to post ideas for coding they'd like to see implemented.

As a kickstarter I remember being solicated out of the blue to produce a simple console application that would copy the contents of a text file to the clipboard - as the only existing such program was shareware (Considering the ammount of time it took to write I was appalled that it was $20 to register this "application" - and this was going back a good few years).

This is a perfect example of my being able to produce something which somebody else really appreciated (although in this case I'd be hard pressed to say they needed).

I'm sure I'm not alone in wanting to repay the great debt of software I've been fortunate enough to receive - from the GNU project, The Apache Foundation and hundreds of other groups and individual producers of the free software I use every day.


GNU Task List, posted 9 Jun 2003 at 19:30 UTC by yeupou » (Master)

The meta-project GNU Task List address this question "What Do We Want?"

The meta-project is now using Savannah as is only starting, but plans to be an "idea repository".

https://savannah.nongnu.org/projects/tasklist

Re: GNU Task List, posted 9 Jun 2003 at 21:00 UTC by Stevey » (Master)

With all due respect the GNU Task List represents what the GNU Project wants, not what you or I want.

The two may well coincide, but I would not assume so. Looking at the list there are too many arbitary or openended things included:

New frontends for GCC, Some games wanted, Documentation.

These may well be fine things to have, but they're not things I've ever wanted, nor are they things I believe random users would care about.

Ultimately this is a list of things for developers to implement required by developers - not by users which is where my idea was really focussed.

I can easily imagine users having desires which couldn't be achieved such as "a single unified userinterface, and common help system across all GUI's - but these are the kind of things the GNU list covers.

Ultimately I was envisioning users/developers asking for a tool to do "something", or some glue between "this" and "that". Perhaps trivial items to the correct implementor.

Exchange, posted 9 Jun 2003 at 21:57 UTC by lkcl » (Master)

To get from here to there, i've outlined a number of times what is needed to do an Exchange Server.

To make the job easier, the path that needs to be followed, with timescales, is as follows, where i _do_ know someone who has completed some of these tasks, already (and is looking for buyers of their code):

1) add modules into FreeDCE to support NT 4.0 Security Model (4 months, using Samba TNG codebase modules as the beginnings, to save time)

2) create the 5 or so IDL interfaces and implement the server-side of the required services: remember with FreeDCE if you have the IDL files you get the client-side calls "for free" (about 8 months)

3) network-reverse-engineer MAPI and use the appropriate IDL interface to "proxy" it over DCE/RPC (about 6 months).

these are full-time tasks. you're looking at about 18 man-months of effort.

a further 4-6 months if you want client-side Kerberos/LDAP functionality added in (for Windows 2000 security) as well.

one particular issue that you now have to be careful of is that nobody inside the United States, except those who work for Internet Security or Anti-Virus companies, are legally allowed to work on this project, because of the DMCA.

i _did_ warn you about this one, and you [you know who you are] refused to listen.

Exchange is the one major project which would make all the difference in taking away Microsoft corporate domination.

in the same way that Samba quietly and unobtrusively makes Unix appear in the Network Neighbourhood, Exchange for Unix would make it possible to bring back the $30,000 part-time Unix Admins instead of having to employ TWO full-time $40,000 dumb-ass NT administrators.

_and_ companies would save a stupid amount on license fees.

and if MS complained about _that_, there's always the option of writing an open-source Win32 / Cygwin version of Outlook!

p.s, posted 9 Jun 2003 at 22:06 UTC by lkcl » (Master)

if you do not believe that exchange is important, or more widely, that full microsoft compatibility is important, you have not been in the real business world.

real businesses communicate with microsoft tools and programs and protocols, because all businesses do. if you are not compatible with microsoft, your business will fail.

simple.

inalienable.

fact.

just because it's "that madman" telling you doesn't give you an excuse to ignore harsh realities.

and the reality is "open source alternatives are not important to businesses unless they are microsoft-compatible. ONCE they are microsoft compatible and you have enough market share, you can THEN think about extending beyond to better and more convenient things".

... it's because of people like myself that the marketeers in microsoft hate open source.

Exchange, posted 10 Jun 2003 at 00:33 UTC by djm » (Master)

First, thanks Stevey for writing a good article - the front page has been barren of these for a while!

Rather than implement the MS legacy protocols, why not just implement the parts that we are currently lacking: the calendar and meeting arrangement system. We already have contacts (LDAP), mail (SMTP, IMAP) w/ server side filtering (CMU Sieve) and shared folders. Because the extant parts are implemented using sane, mostly IETF-derived protocols they already include TLS and Kerberos support. I'd be willing to bet a bottle of good Australian wine that an implementation of the missing parts of Exchange using IETF protocols would take a lot less time than the 24 months outlined above. If support for Outlook was desired, it could probably be done using plug-ins similar to HP's (now-dead) Exchange replacement product. However, I suspect that this is getting away from the real point of the article...

I don't think that an "Ideas Repository" would help matters. As you observe, there are many good ideas already listed on SourceForget and its cousins. The problem is not so much the idea generation, but the implementation - most of these projects are completely stillborn: some have a little code, some are just idea's with a webpage and grandiose plans. It seems that for a project to be successful, it needs 1) To be a good idea and 2) For a basic implementation to be done.

That second requirement is the real killer - most of the low-hanging fruit have already been picked and, as you observe, most of the remaining wants are _BIG_ (e.g. Exchange replacements). The burden this places on the initial development team is too large. Perhaps this could be overcome by the visible involvement of known and respected personalities. This may attract a critical mass early to see it through the initial implementation. Academia or corporate sponsors may be able to help here too.

Another problem is NIH. Even with a list of good ideas that everyone wants, people are going to have differences of opinion. Disagreements over design, license, personality, language, name or code formatting style would likely result in multiple projects to implement the same ideas. To a certain extent such diversity is desirable, but I don't think it is clear that the "wasted effort" that they encompass would be reduced by a central source of ideas.

For example, I neglected to search the net for similar products to my Softflowd before I started developing it. Afterwards I discovered nProbe. Beyond my suprise that the apparent similarity of commandline options, I decided that working on Softflowd was still worthwhile because of license issues (mine is BSD, their's is GPL demoware - whatever that means). I don't think a list of ideas which mentioned "Software NetFlow generator" (which is unlikely to begin with) would have prevented me from doing it anyway

What I want, posted 10 Jun 2003 at 03:16 UTC by tk » (Observer)

First, this is an inalienable fact: I don't care about world domination, I don't care about doing business, and I don't care about marketeers in Microsoft. When will people acknowledge this simple truth?

So what do I want? One thing that I'd like to see is an online service for notifying people of changes in web pages, similar in spirit to Netmind (which, IIRC, is now a paying service). Currently I track changes by running a home-brewed script upon each logon, but that doesn't look very neat.

Second, a PowerPoint-like presentation creator which is WYSIWYG (I'm not sure if it's been done already though).

Last but not least, some sort of software that can somehow put a complete end to SARS, because SARS sucks rocks!!!!!!

Yeah..., posted 10 Jun 2003 at 05:20 UTC by pphaneuf » (Journeyer)

An Exchange replacement would be cool...

djm: by the way, MAPI (and thus Exchange/Outlook) is surprisingly more general than the average IETF protocol. Believe it or not, IMAP is closer to being usable for re-implementing the Outlook contacts than LDAP, for example. And Exchange is massively more complex and powerful than it looks at the surface. This is groupware, think Lotus Notes application if you've seen them. For example, you could do a bug tracker in Exchange that would be accessible by non-modified Outlook clients.

good for you., posted 10 Jun 2003 at 08:24 UTC by lkcl » (Master)

tk, if there were enough people like yourself around, then it would be microsoft (and consequently businesses) having to follow suit, not the other way round.

Some responses.., posted 10 Jun 2003 at 09:27 UTC by Stevey » (Master)

tk Personally I don't care about domination either. I use Debian at home and would continue to do so as long as I could. I'm lucky that I maintain a server room full of Linux machines at work (alongside Solaris, and some evil SCO machines ;) but ulimately if I had to work with Microsoft Servers I would do so. I need the money!

djm: Thanks! If I were to start an Exchange replacement I would start trying to bolt on a simple iCal server to the top of an existing IMAP server. I believe that would be a credible way to start and should allow functionality to grow in a measurable way.

lkcl: Yes I agree. In my former position I was involved with a cluster of Exchange (5.5) servers and I know how important it is seen to business. It's the integration that is the compelling factor, the combination of mail and appointments.

After seeing Outlook + Exchange not many businesses would be happy with another mail server and a web based calendaring system.

As an aside I know that SuSe produce a server they call open exchange but I don't believe it's free software.

minor details, posted 10 Jun 2003 at 10:11 UTC by yeupou » (Master)

Stevey, if I correctly understand Toby Cabot, that manage the GNU Task List on Savannah, sure it will list items that the GNU Project think important, but anybody will be free to propose things. It will be, so, an idea repository. Do you want an idea repository without management and control? It can be good too, a wiki could do the job, but it can be messy...

djm, The term "cousins" suppose a familial background. I cannot tell for Alioth, but Savannah is definitely no longer in the family SourceForge family since SourceForge is a proprietary software, despite a relatively similar purpose and it's common old codebase.

I like the idea, posted 10 Jun 2003 at 12:33 UTC by kai » (Journeyer)

Thanks for this cool article!

Some items on my wishlist:

  • An outline font editor.
  • DTP software like Adobe Pagemaker or Quark XPress.
  • A config snippet for XEmacs that binds Ctrl-Del and Ctrl-Backspace like, say, KWrite (or Microsoft Word).
  • A text marking feature in Konqueror, Mozilla, etc.
  • Free Software SVG and Flash plugins.
  • Code-folding for XEmacs
  • A treemap file system viewer (du) for Unix. (Could be a plugin view for Nautilus or Konqueror)

I think that an "ideas repository" would be very useful. What I would like to see is a repository not only of "grande" ideas like requests for an Exchange Server, PageMager, or Maya clone but also for "small" ideas like small scripts, enhancements to existing software or other small details that could be of use to someone.

(Most) individuals can be more creative in a group-setting rather then in isolation, a potential such a repository should aim to harness. (Edward DeBono, Wikis, ...)

djm: Many projects listed on SourceForge" I wouldn't even call ideas, have no webpage nor any plans, let alone grandiose ones.

Sites like SourceForge or freshmeat.net contain so much information that it becomes almost impossible to find anything but the most popular, active, or recently announced packages. Keeping track of things (overview, navigation) is one of the central aspects an ideas repository has to adress to be successful.

Such a site should allow to add comments, mockups (for GUI applications), and links to related software or articles. Other important tasks include structuring and organizing ideas that start out widely unrelated, as well as evaluation of ideas and highlighting "popular" ones.

I have a lot of ideas (from past discussions with other people) I intend to put on the web (one by one) but using a site which makes it possible to put things in the right context and allow other people to find them easily (to embrace or reject) would be even more interesting :^)

I found an outline font editor, posted 10 Jun 2003 at 14:16 UTC by Omnifarious » (Journeyer)

And I used it to modify a TruType monospace font I liked to have a slashed 0, which I consider essential for programming. But, now I can remember what it's called. But, there is a libre program out there.

so much to ask, posted 10 Jun 2003 at 14:45 UTC by rillian » (Master)

What I want written is a video editor. A really good one, as powerful as Final Cut Pro, but easier to use. As sleek and compelling as the GIMP was when it came out. One that holds video to film standards of quality, and always does the right thing to preserve the quality of the source whatever it's format. One that would actually help filmmakers switch to linux.

I want someone to replace make+autotools with something which is as portable, but isn't actually a pain to use. That one's been on the list for a long, long time. Autotools works, of course, but it's frustrating to deal with, and usually there's only one way to do things for reasons that aren't at all obvious from the naive perspect. I think that's why all the projects I know have a few people who've spent to time to learn the details and do all the build system and work, and some people who flatly refuse to touch it.

I also find myself trying to implement expert system behavior in configure macros. If these three things are available, and this other is called whatsit, test for these things; if they're not available unwind, dropping two of the three and fallback, advising the user of what's missing. Shell isn't the best language for that sort of coding, so usually I settle for half measures.

I actually have some hope in this direction. CVS is finally developing some replacements and that's been on my wishlist for about as long as I've been involved with open source. Now if only GNOME would stabilize...

kai: you do know about pfaedit? It's not pretty, but it does work.

I wish people would get excited about GNUstep and finish it. I used NeXT machines for years and would still prefer that as my desktop. It has a real imaging model and could give us resolution-independent displays right now. Apple adopting NeXTSTEP for MacOS X is a tremendous opportunity, and being able to write code that would run on any *nix, including a major consumer platform would totally, totally rock. Especially with the updated ui features Apple has added. GNUmail is really the only application I know that's tried to run on both.

As usual though, there are a lot more ideas than time to do them, but excercises like this are certainly fun. Hopefully it also helps with inspiration. I've been feeling we dropped the ball lately; like no one's picked up the slack from the explosion of the community to actually do all the work. I don't feel like I have much of a mentoring relationship with less experienced developers, and frankly for a long time I wasn't all that inspired to volunteer my time for coding. That seems to be changing at last.

&quoText marking&quo in Mozilla, posted 10 Jun 2003 at 16:24 UTC by gerv » (Master)

A text marking feature in Konqueror, Mozilla, etc.

I don't know what that really means, but try loading Mozilla and pressing F7. Is that (Caret Browsing) what you are after?

Gerv

Free Software Needs Marketing, posted 10 Jun 2003 at 19:37 UTC by glyph » (Master)

... and I don't mean sales.

Discussions like this can be energizing, but at the end of the day, unless someone has done the legwork to do polls and collect statistics and ask the right people the right questions, we don't really have any idea if these things that "we need" will really drive adoption.

I do think that the culture of free software is detrimental to the establishment of such a project. First, we all tend to regard marketing staff as cretins. Either they're often muddled up in the sales staff, or they're very nontechnical and don't have an appreciation of the actual products they're trying to do analysis on.

However, every time I see a request these days like "we need an exchange replacement" or "we need an IIS-compatible webserver", I wonder, "says who?".

selection highlighting, posted 10 Jun 2003 at 19:38 UTC by kai » (Journeyer)

gerv: Seems to be called selection highlighting and related is keyword highlighting like in google or XEmacs. A small Windows utility which appears to have it. It should be possible to select from multiple highlight colors, maybe even add sticky notes. Ideally, pages with highlighted sections should be stored locally so comments are not lost when a page changes or gets removed.

marketing?, posted 10 Jun 2003 at 21:11 UTC by mslicker » (Journeyer)

If free software was meant to be the same in all respects to corporate produced proprietary software (with the exception of certain freedoms), what is the point?

The primary benefit I see in free software is to break all the existing rules of corporate driven software development and address needs directly. There is so much ineffiency and waste in the corporate development method that can be avoided entirely with free software methods. It is time we fully took advantage of this freedom, freedom from the proprietary method of software development.

most needs have already been addressed, posted 11 Jun 2003 at 09:00 UTC by lkcl » (Master)

the needs of the majority of businesses in the world have already been addressed: by microsoft.

therefore, to make yourself heard, and important, the first priority is to get in on their act.

if you (plural) manage to achieve the goal of open source becoming important to businesses by any other way, i will be delighted - and surprised.

priorities, posted 11 Jun 2003 at 09:08 UTC by lkcl » (Master)

i think it is clear that there are enough source code repositories out there to drown in.

hence, the interest by the author in raising the question "what's missing, because i sure can't find what i believe is needed".

i believe that there needs to be agreed and then communicated to others what the PRIORITY projects are - the ones that take up man-years of time to even get started.

it is unfortunate that the projects on which the majority of businesses now critically depend are multi-man-year (and in some cases multi-man-decade) development efforts.

also, i don't see the open source community having the time or intellectual resources to spend on such projects without having full-time people to help guide them on such projects.

consequently, a priority short-list of critical projects would help people who want to invest in open source's future to direct their $$$ wisely.

at present, there is a risk of large corporates who wish to invest in linux and open source from burying $$$ into inappropriate projects.

lkcl, are working on GNOME :) ?, posted 11 Jun 2003 at 09:48 UTC by yeupou » (Master)

Sure, lkcl, Microsoft addressed the needs of the majority of businesses.

But instead of cloning, what Microsoft done in the past, why not doing what we think good? Are we just so stupid that we cannot invent something?

For instance, Microsoft Project addressed a need of the industry. Mr Project clone Microsoft Project. But I know several industry that cannot really use Microsoft Project for several design issue. Why writing another Microsoft Project, recreating the same issue - while, eventually, Microsoft enhances his Microsoft Project to make it really different (what we would maybe have done at first), to fit the need of real industry?

Ugh..., posted 11 Jun 2003 at 11:27 UTC by tk » (Observer)

...I forgot to mention that I'd also like an electronic mail system which is relatively resistant to server outages.

[mslicker] The primary benefit I see in free software is to break all the existing rules of corporate driven software development and address needs directly.

glyph might not have phrased it well, but he was indeed talking about addressing needs. The point is, in order to address people's needs, one must first find out what those needs are. Thus some sort of `market survey' machinery is needed -- even if it's just a quick survey, or an article like this one.

Just as there are ways to find out users' actual needs, there are also ways not to find them. These include saying things like "I know that an Exchange clone is in great demand", or "Anyone who doesn't want feel the need for an Exchange clone is a moron", or "99% of the world uses Exchange, therefore an Exchange clone is in great demand" (this last one depends on so many bogus assumptions that I don't know where to start debunking it).

[lkcl] the needs of the majority of businesses in the world have already been addressed: by microsoft.

But of course! One should care only about businesses, not people!

Marketing is Questioning, posted 11 Jun 2003 at 16:11 UTC by glyph » (Master)

mslicker, I think you're misunderstanding what I'm saying. I'm certainly not saying that developers shouldn't scratch their own itches, or that they should be required to have a marketing department to oversee their work once they have.

Let me be a bit more precise. F/OSS developers have differing motivations for writing the software they do. Some do it for principle, some do it "just for fun", but almost everybody has mixed motivations. We want users to use our software, but we don't want them to sap too much of our time. We think the principles of the community are important, though we differ over which principles and how much.

One common theme is that developers want users to use their software and enjoy it. However, there is very little actual information about what users, in the large, want out of free software. How do you prioritize your features if you don't know what your user-base (and potential user-base) want? Almost by definition, you don't get feedback from potential users, only users who have already used your thing, so how do you know what a different audience might want?

Answering these questions effectively is the real definition of the term "Marketing". It is discovering what markets exist for your "products" and understanding what they want. Lying to people to get them to buy your products is "Sales" (or as we say here in freedom-land, "evangelism"), and should be as separate as possible from marketing.

Good marketing can prioritize bug lists and feature requests so that old users stay happy and new users are impressed. I think that a project which did this for free software developers would be of great use.

surveys, posted 11 Jun 2003 at 16:45 UTC by mslicker » (Journeyer)

tk, I'm not sure surveys are apropriate for designing good software. I imagine surveys work quite well to sell a lot of things, but again, this is the corporate model of product development.

What is needed is analysis of the tasks. Indeed, you must work with other people to find out what the tasks are and how computers can help. A good example of this aproach seems to be the Presto music composition system. I discovered this paper when researching pen based input systems.

I think the free software aproach can be much more personal and individual in adressing needs. Mass production, mass marketing is very impersonal by nature. This also has relavance for free software economics. If we want to create demand for our skills, we need to demonstrate the benefit of the very personal aproach that can be offered.

newer presto link, posted 11 Jun 2003 at 16:55 UTC by mslicker » (Journeyer)

The document I meant to link to is here

Free Software and World Domination, posted 11 Jun 2003 at 17:13 UTC by kai » (Journeyer)

Personally, I don't care for world domination, either. Nevertheless, since the freedom promised by Free Software is severely put in danger by recent developments like the DMCA, world-wide proliferation of software patents, hardware companies closing out Open Source (NVIDIA, etc.), and de-facto monopolies of Microsoft, SAP, et. al. in certain areas it is going to be increasingly important for Free Software to maintain critical mass in order not to loose momentum. Business-acceptance may play a role in this, but I'm not convinced it's the only road.

Anyway, regarding Exchange Server, the kroupware project may be of interest.

The source repository chaos

lkcl points out that finding relevant projects is often difficult: the huge amount of information on sites like freshmeat.net is overwhelming (or even SourceForge, with so many dead-on-arrival projects.)

Maybe more places like atai's GUI toolkits and frameworks page, where insightful information about Free Software projects can be found, are needed. SourceForge foundries are going in the right direction but they suffer from the same problem as the rest of SourceForge: plenty of technical infrastructure with no relation to actual content available, such as project evaluation, feature comparisons, or an organized knowledge base of the field (over-engineering ?).

most-wanted software

The idea of an unmaintained projects list, the Savannah task list, or a Free Software "hot-list" sounds useful. Who decides what's hot, though ? How should market-research techniques be applied to a field where market features are not apparent ? How can a list containing, say, a "dance notation software" as well as an "Exchange Server clone" be organized in a useful and scalable way ? If it's true that the low hanging fruits have been picked how can the process of getting from a project idea to a healthy open source project be organized for highly complex software systems ?

Free Software and World Domination, posted 11 Jun 2003 at 17:13 UTC by kai » (Journeyer)

Personally, I don't care for world domination, either. Nevertheless, since the freedom promised by Free Software is severely put in danger by recent developments like the DMCA, world-wide proliferation of software patents, hardware companies closing out Open Source (NVIDIA, etc.), and de-facto monopolies of Microsoft, SAP, et. al. in certain areas it is going to be increasingly important for Free Software to maintain critical mass in order not to loose momentum. Business-acceptance may play a role in this, but I'm not convinced it's the only road.

Anyway, regarding Exchange Server, the kroupware project may be of interest.

The source repository chaos

lkcl points out that finding relevant projects is often difficult: the huge amount of information on sites like freshmeat.net is overwhelming (or even SourceForge, with so many dead-on-arrival projects.)

Maybe more places like atai's GUI toolkits and frameworks page, where insightful information about Free Software projects can be found, are needed. SourceForge foundries are going in the right direction but they suffer from the same problem as the rest of SourceForge: plenty of technical infrastructure with no relation to actual content available, such as project evaluation, feature comparisons, or an organized knowledge base of the field (over-engineering ?).

most-wanted software

The idea of an unmaintained projects list, the Savannah task list, or a Free Software "hot-list" sounds useful. Who decides what's hot, though ? How should market-research techniques be applied to a field where market features are not apparent ? How can a list containing, say, a "dance notation software" as well as an "Exchange Server clone" be organized in a useful and scalable way ? If it's true that the low hanging fruits have been picked how can the process of getting from a project idea to a healthy open source project be organized for highly complex software systems ?

Free Software and World Domination, posted 11 Jun 2003 at 17:13 UTC by kai » (Journeyer)

Personally, I don't care for world domination, either. Nevertheless, since the freedom promised by Free Software is severely put in danger by recent developments like the DMCA, world-wide proliferation of software patents, hardware companies closing out Open Source (NVIDIA, etc.), and de-facto monopolies of Microsoft, SAP, et. al. in certain areas it is going to be increasingly important for Free Software to maintain critical mass in order not to loose momentum. Business-acceptance may play a role in this, but I'm not convinced it's the only road.

Anyway, regarding Exchange Server, the kroupware project may be of interest.

The source repository chaos

lkcl points out that finding relevant projects is often difficult: the huge amount of information on sites like freshmeat.net is overwhelming (or even SourceForge, with so many dead-on-arrival projects.)

Maybe more places like atai's GUI toolkits and frameworks page, where insightful information about Free Software projects can be found, are needed. SourceForge foundries are going in the right direction but they suffer from the same problem as the rest of SourceForge: plenty of technical infrastructure with no relation to actual content available, such as project evaluation, feature comparisons, or an organized knowledge base of the field (over-engineering ?).

most-wanted software

The idea of an unmaintained projects list, the Savannah task list, or a Free Software "hot-list" sounds useful. Who decides what's hot, though ? How should market-research techniques be applied to a field where market features are not apparent ? How can a list containing, say, a "dance notation software" as well as an "Exchange Server clone" be organized in a useful and scalable way ? If it's true that the low hanging fruits have been picked how can the process of getting from a project idea to a healthy open source project be organized for highly complex software systems ?

Free Software and World Domination, posted 11 Jun 2003 at 17:13 UTC by kai » (Journeyer)

Personally, I don't care for world domination, either. Nevertheless, since the freedom promised by Free Software is severely put in danger by recent developments like the DMCA, world-wide proliferation of software patents, hardware companies closing out Open Source (NVIDIA, etc.), and de-facto monopolies of Microsoft, SAP, et. al. in certain areas it is going to be increasingly important for Free Software to maintain critical mass in order not to loose momentum. Business-acceptance may play a role in this, but I'm not convinced it's the only road.

Anyway, regarding Exchange Server, the kroupware project may be of interest.

The source repository chaos

lkcl points out that finding relevant projects is often difficult: the huge amount of information on sites like freshmeat.net is overwhelming (or even SourceForge, with so many dead-on-arrival projects.)

Maybe more places like atai's GUI toolkits and frameworks page, where insightful information about Free Software projects can be found, are needed. SourceForge foundries are going in the right direction but they suffer from the same problem as the rest of SourceForge: plenty of technical infrastructure with no relation to actual content available, such as project evaluation, feature comparisons, or an organized knowledge base of the field (over-engineering ?).

most-wanted software

The idea of an unmaintained projects list, the Savannah task list, or a Free Software "hot-list" sounds useful. Who decides what's hot, though ? How should market-research techniques be applied to a field where market features are not apparent ? How can a list containing, say, a "dance notation software" as well as an "Exchange Server clone" be organized in a useful and scalable way ? If it's true that the low hanging fruits have been picked how can the process of getting from a project idea to a healthy open source project be organized for highly complex software systems ?

Free Software and World Domination, posted 11 Jun 2003 at 17:13 UTC by kai » (Journeyer)

Personally, I don't care for world domination, either. Nevertheless, since the freedom promised by Free Software is severely put in danger by recent developments like the DMCA, world-wide proliferation of software patents, hardware companies closing out Open Source (NVIDIA, etc.), and de-facto monopolies of Microsoft, SAP, et. al. in certain areas it is going to be increasingly important for Free Software to maintain critical mass in order not to loose momentum. Business-acceptance may play a role in this, but I'm not convinced it's the only road.

Anyway, regarding Exchange Server, the kroupware project may be of interest.

The source repository chaos

lkcl points out that finding relevant projects is often difficult: the huge amount of information on sites like freshmeat.net is overwhelming (or even SourceForge, with so many dead-on-arrival projects.)

Maybe more places like atai's GUI toolkits and frameworks page, where insightful information about Free Software projects can be found, are needed. SourceForge foundries are going in the right direction but they suffer from the same problem as the rest of SourceForge: plenty of technical infrastructure with no relation to actual content available, such as project evaluation, feature comparisons, or an organized knowledge base of the field (over-engineering ?).

most-wanted software

The idea of an unmaintained projects list, the Savannah task list, or a Free Software "hot-list" sounds useful. Who decides what's hot, though ? How should market-research techniques be applied to a field where market features are not apparent ? How can a list containing, say, a "dance notation software" as well as an "Exchange Server clone" be organized in a useful and scalable way ? If it's true that the low hanging fruits have been picked how can the process of getting from a project idea to a healthy open source project be organized for highly complex software systems ?

Different sides of the same coin, posted 11 Jun 2003 at 17:24 UTC by tk » (Observer)

I think both survey and analysis are important. A survey gives a general view of the landscape, while analysis fleshes out the details of a particular task. It'll be foolish to start writing a game without the faintest idea of what people look for in a game, even if one knows that games are in great demand. Similarly, it'll be silly (in fact impossible) to do a full-blown analysis for a development environment for the Papaboolaboolah programming language, if the demand for such a thing is zero.

(And, the "demand" for a project doesn't include the number of people who believe there's demand...)

marketing, posted 11 Jun 2003 at 17:41 UTC by mslicker » (Journeyer)

glyph, I associate "marketing" with selling products, perhaps many others do as well. It may be that your concept of marketing, and my concept of addressing needs are the same but it does not apear that way to me now. I think creating truely useful software requires deep analysis of the tasks and their context, and perhaps a good deal of interaction with the people who carry out these tasks. This not the equivalent of polls, stastastics, or simple feedback/sugestions which to me are associated with "marketing" driven approach.

Groupware, Exchange, most wanted etc, posted 12 Jun 2003 at 11:05 UTC by Malx » (Journeyer)

..you could do a bug tracker in Exchange that would be accessible by non-modified Outlook clients

Same you could do with Mozilla and XUL - you could create ready-to run from URL application (see http://www.xulplanet.com/).

http://groupware.openoffice.org/ - yet another groupware project.

There is lot's of groupwares, which are Web based (Proiektor for Zope is an example). I think they are not widely used becouse of web-interface used :( But you could get sources and algorithms there.

And about "most wanted". IMHO there should be site to collect ideas, keep list of current projects (it's features actually, but in standartized form), post requests for images, designs, translations etc.
And for me "most wanted" is solution for education (local and remote, for kids, schools and universities, tools for easy and fastcreation of portable multimedia presentations and learning books/models/environments). Something like Asymmetrix Toolbook, Macromedia Authorware, Flash(creating exe files and autorun CDs) etc.

IMHO - education is a basis for long-term goals :)

project assessement site, posted 13 Jun 2003 at 09:02 UTC by lkcl » (Master)

okay, how about this (or does this sound utterly mad, or something that's been tried already and gone horribly wrong)

a site is set up that allows projects to be strategically assessed.

we understand that freshmeat.net with its "ratings" doesn't work because the "rating" isn't specific enough.

we understand that sourceforge.net with its "activity" doesn't work because, as already stated, a project has to remain at the top of the league (by doing automated daily cvs commits?) in order to get noticed.

we understand that actually _finding_ anything on sourceforge takes time: search by keyword is your best hit-and-miss simplistic approach, followed by doing time-consuming categorisation, which most people won't want to bothe with.

advogato's "project" system, which was very basic to begin with? don't go there :)

... but the sort of projects we are talking about here are "nice to have" or "strategically important". such projects fit at the BOTTOM of all of the "ratings" systems i can possibly think of.

so, is it a waste of time to consider doing a "better" project assessement site?

is it worthwhile rating projects by the _type_ of interest the rater has in the project?

is it worthwhile recording on the web site the criteria that the searcher uses most often, to help regular searchers save time?

etc.

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!

X
Share this page