Re: [Pharo-project] [Bug Tracker] Issues tagged 1.3
I think this attitude is what will continue to hold Pharo back from wider adoption. I've found 1.4 to be unusable for daily development. And yet there's no interest in fixing problems in 1.3.
There is. But we need to be reasonable... we can not back-port everything. Even fixing bugs changes the system so that it leads to incompatibility and new problems.
And lest you think I don't want to contribute, I'll point you to
http://code.google.com/p/pharo/issues/detail?id=4549&can=1&q=test%20runner&c...
where I a) submitted a bug report b) included a video demonstrating the bug c) included a patch fixing the bug
Yes, and I only understood the problem last week. Because I use a mouse, and it works with a mouse with right-click... it even works when your track-pad is configured with right click. It only does not work if not.
responses were - "not a showstopper -> move it to 1.4" - "for me the bug entry is not really relevant"
Then complain! There are 450 open reports in the database... we closed nearly 5000 issues. There will be mistakes. The thing is that we could have not released 1.3 until we fixed all bugs.... then we would never have released it. So we go through all reports and decide: what needs to be really really fixed? Yes, we make mistakes. I do not know how to not do mistakes, sorry.
seven months later it finally made its way in.
Because *I* pushed. Ben is using a mouse, so he just did not see the problem. In general, there are just not enough people to look at bugs... if nobody helps, it takes a looong time.
I can see saying, "move it to 1.4" if 1.4 is right around the corner.
We could have blocked the release of 1.3 and then we would have not released 1.3 until now. (keep in mind there are 450 other open issues besides the one you are talking about).
As long as 1.3 is the prime time release - i.e. the release suitable for active development - you have to support it and care about the people using it. This is not some backward compatibility thing. It's a daily usability thing. You can't tell people, "yeah that is fixed, but only in 1.4, but 1.4 isn't really ready for you to use yet." And if I submit a patch for 1.3, then accept the damn patch for 1.3. Don't tell me "it'll go into 1.4" when I can't even use 1.4 myself. That's how you alienate users and keep the community tiny.
We can not backport everything... especially as the amount of people looking at stuff is very small. yes, we should release more often... E.g. since I asked for help on 1.3 issues: http://code.google.com/p/pharo/issues/list?can=2&q=Milestone%3D1.3 There was *no reaction at all*. Nothing. Zero. Sometimes I think we should just stop and do something else. This is clearly not working. Marcus -- Marcus Denker -- http://marcusdenker.de
Dear Marcus, Am 01.03.12 23:35, schrieb Marcus Denker:
I think this attitude is what will continue to hold Pharo back from wider adoption. I've found 1.4 to be unusable for daily development. And yet there's no interest in fixing problems in 1.3.
There is. But we need to be reasonable... we can not back-port everything. Even fixing bugs changes the system so that it leads to incompatibility and new problems.
Personally, I am convinced that the Pharo core team is willing to support users with a stable environment which is in our case also 1.3. But I need to say that Pats statement sound familar to feedback I heard from my dev team too. Reading the comments of some issues and being one month on this list now I suppose the root of the problem to be completely on the layer of communication, nothing else. Actually we all "love smalltalk and pharo and really want to see it win and receive wider adoption" but IMHO this does not work as long as there are mutual recriminations and no appreciation. Let's get back to work together with united forces, with understanding for the counterpart, helpfulness and constructiveness! And fun will return for sure...
We can not backport everything... especially as the amount of people looking at stuff is very small. yes, we should release more often...
E.g. since I asked for help on 1.3 issues:
http://code.google.com/p/pharo/issues/list?can=2&q=Milestone%3D1.3
There was *no reaction at all*. Nothing. Zero.
Sometimes I think we should just stop and do something else. This is clearly not working.
I understand your personal feelings very well but don't take it too seriously. You (and all other contributors too) are doing a great job and all of the users appreciate your work very much otherwise nobody would actually use Pharo at all. Keep it going! We (to be more precise: Andrea) tried to deliver input for at #4617 and #5045 and will continue to support you too as much as we can. Just let us now how we can help you with additional input for this issue. We will definitely send you virtual flowers and offer some original Swiss chocolate if you find a way out of this problem... :) Good night! Chris P.S: I don't intend to be instructive and I am aware it to be off-topic... just would like to help in the area where I am clearly better than as a Smalltalk developer.
Marcus, On 01 Mar 2012, at 23:35, Marcus Denker wrote:
We can not backport everything... especially as the amount of people looking at stuff is very small. yes, we should release more often...
E.g. since I asked for help on 1.3 issues:
http://code.google.com/p/pharo/issues/list?can=2&q=Milestone%3D1.3
There was *no reaction at all*. Nothing. Zero.
We are all guilty, me too. I know. I try to keep my main contributions (Zinc HTTP Components & Zodiac) compatible with both 1.3 and 1.4 and so far that works. But of course, some deeper issues are only resolved on 1.4, perfectly normal. The work that you do, including the backporting to 1.3 is much appreciated: it is _you_ who makes that happen.
Sometimes I think we should just stop and do something else. This is clearly not working.
The current strategy is good, IMHO, given the resources available. The current stable version, 1.3, is a very good version: it was developed by a whole community for a long time, and we all worked and lived with it for a long time, a lot still do. Today, 1.4 is moving fast and is very promising, for many of us, it is our daily workhorse. Maybe one of the tasks of the consortium could be to dedicate some extra effort to maintaining the current stable version. Sven
participants (3)
-
Christoph Wysseier -
Marcus Denker -
Sven Van Caekenberghe