At 11:27 Uhr +0100 19.02.2003, David wrote:
Exposure of "Fink", yes. Exposure of "fink" (the package manager), no. Most people don't even realize it as being one separate part of the big whole which is Fink - and IMHO that's good this way. Nothing to be gained here. Also note, this release is a release to the unstable tree, so again for the average user, announcing it big would only lead to lots of emails which ask why they can't seem to be able to get the update, since a big part of Fink users runs the bindist or stable-cvs.-----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160On Mittwoch, Februar 19, 2003, at 09:32 Uhr, Max Horn wrote: <snip>?But I do. Exposure is the one key issue, which will, amongst other things, keep Fink alive and kicking.
Yes. First off, the package manager never was announced widely, and there is no reason to do so with this release. I am not sure if we should announce it *anywhere*, but IMHO at most on the mailing lists and at our own web site.
I give a damn on being on the front page of freshmeat, to be honest.
The fact, that Fink really is a two part effort (package manager + all the packages) has never really been shown to the better of the users.True. The question is, do we want to change this? Personally I don't see any need for it... but that's maybe just me.
Freshmeat is often visited by developers, much less by users, at least that is what I could learn in my personal experience over the years. Someone might be interested in simply developing for the package manager without installing the whole "distribution".This is possibly, but I believe the number of such people is tiny. And while I am interested in promoting Fink, there is nothing I see to be gained by promoting fink. I won't object if people want to use the fink package manager on its own, but there is no motivation to encourage it either, is there?
However, I do not quite see why we could not release it as a separate project. It shows, that we are gaining popularity by the moment we are exposing ourselves. Reading over the stats for the project that shows clearly.Maybe, maybe not, hard to tell from the raw stat alone where those people come from. It has been the case in the past (before any special PR efforts where made) that our popularity increased with each release. A nowadays lot of blogs, articles, installation instructions now mention Fink casually - it seems to have become a common good among Unix interested mac users. E.g. regulary I see boosts in the fink stats after certain /. articles which have high ranked comments that reference to Fink. And when we released a 10.2 release, quite naturally a lot of people got highly interested again.
That said, the added news exposure certainly is good and helped a lot, I don't want to doubt that in the least, I just think it is not alone responsible :-)
But anyway, see above for my stance on this: I see an interest to expose "Fink", the project, but "fink", the package manager? Hum.
I don't understand how your example shows that we gain something. I do see that it restricts the way we work, but besides discomfort for the developers, I have to repeat my question: what would be gained by this?Furthermore, I see no point in making daily snapshots. What would be gained by this over just injecting a cvs checkout?That you have a time frame. For example "do not check in a highly volatile patch before 10pm, because it might break the whole app. The snap is made at 10:30pm thus it ensures, that the daily snap is "stable".
Anyways, it is not necessary, it was merely a suggestion.
Daily builds make sense for projects like Chimera or Eclipse etc., because the only way to get them testes by a broad audience is by providing precompiled binaries - not everybody can be expected to install developer tools, get a source snapshot, compile it, etc. But the situation is different for fink, once you obtained checkout, you need a single command to install the new package manager in a matter of seconds.
The only possible gain I see is for people that do not have access to CVS. For those we have cvs-proxy. And for those who can't use that one either (yes I know this is the case for some people), well, tough luck, but it's not really as if it's required to get these builds; normal mortals don't need to get them anyway, they should just wait for a point release instead of risking their installs anyway - which renders the whole discussion about daily snapshot moot in my eyes. However, if you see any convincing arguments for dailys, please feel free to explicitly list them :-)
Well, for a new Fink release, I could understand this to a degree (I don't necessarily agree; my experience is not like yours, it seems, but well <shrug>). But for the package manager? It is completly unimportant if everybody besides the core team misses the announcement... in fact as I said, I don't plan a big announcement. We didn't announce PM releases in the past besides on fink-devel and maybe fink-users in a few exceptional cases.In fact, I see no reason to wait till friday with the release as you requested <shrug>. I can release today, or tomorrow, or saturday, not friday, but where is the point in waiting?Wendsday and Thursday are never good release days, because they are in the middle of the week and people tend to get very busy. The news editors and news site also accumulate a lot of news from the weekend to about Tuesday, Wendsday. There are also various acclaimed studies which seem to show that middle of the week announcements are treated less interested than end of the week or start of the week announcements. Thus, out of a very old habit I try to keep it that way.
The second and much more interesting reason is, that I was away from Thursday to Tuesday on a business Trip. I did not bring my Computer, so I need to catch up. Both on my work accounts and Fink.Hm, I don't want to be rude, either, but like Finlay, I have to say, why does this matter? Don't get me wrong, I don't say "I give a damn for your opinion, so why should I wait" - this is not true, I do mind your opinion like that of any other core member. Rather, I wouldn't wait for drm or RangerRick, nor would I expect them to wait for me, if one of us had been away for a few days prior to a PM release, unless there was an imminent pressing reason why that particular one of us has to perform last minute checks. But usually, if two of three project leads give their OK on the code base, and several of the other core members double checked everything, too, then normally there is no need to wait for one more project lead or core memeber. Things are different of course if somebody shouts "WAIT! I think I found a severe problem", then we have to investigate.
So it's not really a matter of not being curteous to you, and more a matter of treating you like other core memebers/project leads :-)
Of course this would be different if we said that we want a full PR coverage of the PM, too. And in additiona, let me emphasize that it's fine by me to wait for you upon a full Fink release, where we want to make announcements everywhere possible, and which is a much larger undertaking, i.e. we plan it over several weeks not just one, so we can easily modify the release schedule to accomodate you/drm/me/etc.
Cheers,
Max
-------------------------------------------------------
This SF.net email is sponsored by: SlickEdit Inc. Develop an edge.
The most comprehensive and flexible code editor you can use.
Code faster. C/C++, C#, Java, HTML, XML, many more. FREE 30-Day Trial.
www.slickedit.com/sourceforge
_______________________________________________
Fink-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/fink-devel
