An open source project requires an active and large community of both users and contributors to keep on evolving. Let us consider the Linux ecosystem. By providing easy and ready to used binary distributions, Canonical (via Ubuntu) or Redhat (via Fedora) have played a significant role toward adoption of Linux as a largely diffused desktop solution. Among this community of users growing exponentially, some got more involved into this system they were familiar with. They acquired great expertise, and are today major contributors to all kind of open source projects, from kernel, Gnome, Firefox to all those "small" applications which contribute in making user experience getting richer. Of course, periodically, some raise polemics about e.g. Canonical not contributing patches to kernel, to glibc, or including some non free drivers or applications (e.g. flash player) in their distribution. That's just non sense. Vendor primary mission (often without profit) is to provide an operating system that end users want to use, and make Linux popular. If they succeed (and they do), they contribute in making community larger and, transitively, to core developments.

We believe that tomorrow's contributors (both to Tcl core and new extensions) will be found among today's new Tcl users. eTcl aims to promote Tcl usage today, on as many platforms as possible, to make Tcl community larger and stronger tomorrow. Not all newcomers want to start their Tcl experience with compiling their own interpreter. This is even more true when target architecture is one of those emergent mobile platforms, where Tcl is offering a unique and elegant solution for rapid and portable development. Since 4 years now, that philosophy motivates our public (and free) release of eTcl binaries, which aim to offer, through a a ready to use and easy to deploy solution, a way to lower barrier to entry the great Tcl universe. On most platforms, eTcl is built on top of plain vanilla sources. A couple of closed modules providing optional features (e.g. pixane to script image transformations, wce and s60 to access platform-specific native API on Windows Mobile and Symbian respectively) were packaged in, in the hope to improve user experience. Some platforms required minimal patch to Tcl core, which was sent to anyone asking for it. Actually, nearly all value added by Evolane comes from its expertise in cross-compiling for a large set of target architectures. We shared result from this expertise in the hope to promote Tcl adoption among population which were used in only platform specific, non portable, and often complex to set up, technologies (anybody who has ever tried to do some development on Windows Mobile or Symbian S60 platforms knows what I mean). Tcl has been adopted by a large number of mobile developers who never considered dynamic languages before. Some of them contributed very nice piece of codes back to the community.

eTcl was delivered as a community courtesy, without profit. Unfortunatly, whereas we received encouragement from final users who enjoyed our contribution, this philosophy is obviously not shared with some of the most influent Tclers. Very negative (if not libellous) opinions have been periodically expressed publicly, during conferences or in forums, often supported with false assertions (and no intention for their authors to verify them). Not only providing eTcl is not considered a valuable contribution, but it is said to be contrary to "community convention", prejudicial to Tcl, and as a consequence, is deserving contempt and no respect. If they are right and reflect a dominant opinion among the community, then eTcl must be withdrawn immediatly, and we (as a commercial company) have to stop consuming lot of our resources on something offering no profit but damaging our public image. Maybe, since 15 years now, we missed the Tcl philosophy. Maybe we misunderstood Dr. Ousterhout intention to have Tcl both opened and meant to be used commercially, making many companies rely on Tcl for their services and products. Maybe Tcl should be released under a weak (LGPL) or strong (GPL) copyleft license, which would make the so-called "community convention" match license constraints. Or … maybe eTcl was actually a fair and valuable contribution dedicated to "silent masses", but some activists speak louder? If so, either let's call on silent masses to make their voices heard, or let's withdraw eTcl as a public service, and wish good luck for others to mantain and deliver all the binaries distributions Tcl deserves.