Would a "Windows toolserver" require an own physical server, or have you planned running it virtual?
On Feb 7, 2008 5:35 PM, Simetrical Simetrical+wikilist@gmail.com wrote:
On Feb 7, 2008 10:18 AM, Danny B. Wikipedia.Danny.B@email.cz wrote:
I've seen many cases of intolerance to minorities (because of race, religion, sexual oritentation etc.) and this seems to me it becomes to be the same. I wonder what makes Windows programmers worse than Linux guys. And what makes Linux guys pontificaly reject the equal rights and chances for Windows people. I thought this is community project with equal opportunities to contribute to everybody who's interested. Or are we all equal but some of us are more equal than others?
There are those of us who do believe that reliance upon proprietary software is in fact damaging, contrary to Wikimedia's mission, and perhaps even wrong. That would indeed make us reject "equal rights" for Windows users, just as it would make us reject equal rights for those who don't want to GFDL their Wikipedia contributions. This is not, however, an issue on the toolservers, because neither those who own them nor those who run them are sympathetic to that point of view. Either way, that question is explicitly off-topic here, although I wouldn't mind debating it with you elsewhere.
The question is whether getting a Windows server is worth the cost. So far we have one person who has concretely said he would be interested in moving his application to a Windows toolserver. That project is already running on a private Windows server and will presumably continue to do so even if a Windows toolserver is not provided (or at least, we were told nothing to the contrary). I'm not clear, either, on whether that code would run anyway under Mono with no modifications -- the fifth post in this thread was by River, saying that Windows vs. Mono "shouldn't make any difference as far as i know" for running C#/.NET. The base cost is at least $400, for a one-CPU license -- this allows multicore, by the way, so that's unlikely to be a big problem. At $6/user, the per-user licensing cost seems likely to run to at least one or two hundred dollars more per year, unless it's very unpopular (in which case that's an argument against bothering in itself).
So that means at least $500-600/year, if I'm correct in assuming that the $400 base is also per year. (If it's one-time, that seems like fairly remarkable pricing, I'm pretty sure lower than a copy of Vista Home Premium and Office, so I think I'm correct.) It might be higher if some cost was forgotten or misunderstood, but it's maybe also possible to get a lower price from some other supplier. If this figure is correct, I think it's fair to say that to justify the cost, we would need at least one useful thing that would be run on the toolserver that we are fairly certain would not be run otherwise. That's just as a minimum, to justify the expense and the effort. Preferably you'd think there should be more than one thing. But as far as I can tell, nobody has yet come up with even one thing, so based on the response so far, it doesn't seem that there's anything that's been suggested yet that would justify this.
Is there any disagreement on the last two paragraphs?
Toolserver-l mailing list Toolserver-l@lists.wikimedia.org http://lists.wikimedia.org/mailman/listinfo/toolserver-l