Posted: Fri Apr 28, 2006 12:48 pm Post subject: How will the beta be different?
In previous versions, the beta was simply the alpha with the latest patch.
Similarly, the "release" was identical to the beta with the latest patch.
Now we have the EAP program. By the time we're ready for a beta, everyone and their brother will have the alpha (with the latest patch), so there needs to be something new in the beta, to justify the EAP for people.
What is it going to be? More native dialogs? Some non-disruptive bug-fixes?
IMHO, you are getting way ahead of things. One step at a time. I still have to fix the multitude of bugs that are in Neo 2.0 Alpha and then I need to push an Intel version of the same code out through an EAP release. By then, it will probably be the middle of summer. Plan more than three months in advance is very silly considering the bootstrap funding that we work with.
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Fri Apr 28, 2006 1:26 pm Post subject:
Currently, I hope the native widget code will be finiehd in time for beta. It definitely won't be ready for alpha or any of its patches and isn't on those branches in CVS.
That shouldn't be a reason for folk to want to participate in EAP, though. Again, the EAP stuff really (to me) is a fund drive and is something that's totally optional for everyone. If eventually the alpha gets good enough for folks to use, then I'm all for them continuing to use it. After all, the beta will still be freely available in binaries after a month or so and folks can just grab it then.
I guess it's all a mindset thing. Some folks (like me) are crazy enough to consider dropping boucha dough on WWDC just to get our hands on a buggy copy of 10.5 in a legitimate way a couple of months early. In a way I guess that too helps Apple to continue to fight the good fight and, unlike EAP, it comes with a "free" t-shirt and other swag too
I guess it's all a mindset thing. Some folks (like me) are crazy enough to consider dropping boucha dough on WWDC just to get our hands on a buggy copy of 10.5 in a legitimate way a couple of months early. In a way I guess that too helps Apple to continue to fight the good fight and, unlike EAP, it comes with a "free" t-shirt and other swag too
Also remember that in the Neo/J 1.1 cycle, going from Alpha to Beta and from Beta to RC also meant switching OOo codebases and getting those bugfixes (and let's hope there are a lot of them, esp. in Base for poor fabrizio )
Smokey _________________ "[...] whether the duck drinks hot chocolate or coffee is irrelevant." -- ovvldc and sardisson in the NeoWiki
Joined: Apr 25, 2006 Posts: 2315 Location: Montpellier, France
Posted: Sun Apr 30, 2006 10:45 am Post subject:
Quote:
IMHO, you are getting way ahead of things. One step at a time. I still have to fix the multitude of bugs that are in Neo 2.0 Alpha and then I need to push an Intel version of the same code out through an EAP release. By then, it will probably be the middle of summer. Plan more than three months in advance is very silly considering the bootstrap funding that we work with.
So the next step is NeoOffice 2.0 Alpha for Intel ??? I thought we'd be getting 2.0 Final for PPC before Alpha x86 ...
So the next step is NeoOffice 2.0 Alpha for Intel ??? I thought we'd be getting 2.0 Final for PPC before Alpha x86 ...
Final for PowerPC will be at least 6 months away. Since Neo relies on community testing and bug fixing is slow and tedious (i.e. costly), we don't rush out a final release. For example, when we did Neo 1.2.1 final, it had been in Alpha and Beta for only 3 months and, although no one was reporting any bugs, a whole bunch of bugs got reported a month after we released the final. This made me realize that 3 months is too short of time for a full release cycle so I will put at least 2 months between each release on each platform e.g. Neo 2.0 Beta PowerPC will be at least two months after Neo 2.0 Alpha) so that there is adequate time for users to find bugs and for me to fix them.
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Sun Apr 30, 2006 11:15 am Post subject:
Hopefully with longer test cycles we can weed out more of these issues. Already in 2.0 alpha there's enough changes and, in light of more planned, there's going to be a lot more issues that will crop up than in past releases.
Another issue that may have complicated the 1.x cycles earlier is that there may not have been enough reason for more folks to check the application out during the testing cycle. Once there's a Final version, a lot of folks (espeically corporate deployments) will sitck with Final instead of deploying a prerelease for users. I wonder with the level of changes if some more of these types might be more prone to try and integrate the prereleases into production environments (which always provides the best type of test).
The other thing that is going to complicate and slow down 2.0 is that there is also a larger amount of "inherited" bugs from mainline OOo that folks have already encountered. This will definitely clutter the process of not only identifying where the issue is but also if and how we address it (e.g. is it something we refer on or something platform specific we want to tackle in our own patches).
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Sun Apr 30, 2006 11:21 am Post subject:
Samwise wrote:
I see. Too bad, but as long as it finally comes and is stable ...
I peronally pride stability and quality above all else. Slow and steady may not win the race, but it definitly provides a more rock-solid product in the end. Sometimes the pressure to release something "cool" results in a disaster not only for the developers but also for end users and a "brand" (see the Apeture fiasco). The one thing I love about open source is that, when done right, all of the "pressure" to release simply doesn't exist. You can put quality above quantity without being badgered by marketing drones, office politics, or shareholders. I think that the emphasis on quality has been something that our users have definitely appreciated and benefitted from.
Joined: Apr 25, 2006 Posts: 2315 Location: Montpellier, France
Posted: Sun Apr 30, 2006 1:31 pm Post subject:
Quote:
I think that the emphasis on quality has been something that our users have definitely appreciated and benefitted from.
I totally agree. I just find frustrating that we can't get the same quality with a faster release cycle, but obviously it's not possible, especially in the current situation.
BTW I haven't found that there were so many bugs in the 1.2 releases (1.2 Alpha worked fine for me), but it's probably due to the fact that I'm only an occasional NeoOffice user...
I totally agree. I just find frustrating that we can't get the same quality with a faster release cycle, but obviously it's not possible, especially in the current situation.
This is one of the downsides of open source software: users see the entire development process. In contrast, all that is hidden with commercial software. But just because it is hidden doesn't mean that things are faster. In fact, I would guess that Microsoft spends at least US$50 or US$100 million on the Mac Office release alone. Sound high? Not really. All that development and all that testing, marketing, and support staffing takes a lot of employees.
Compare that to what Ed and I accomplish by me working full time and Ed working in his spare hours. Basically, we may be slow, but we run NeoOffice on less than US$100K per year and provide downloads to nearly 2 million users per year. Granted, I am still funding more than half of that amount (the EAP revenues are definitely bridge that gap), but if users were not a part of the painful development and testing process, there would be no NeoOffice as I don't know anyone crazy enough to fund the couple million that would have been required to fund a full development, test, and support staff.
Joined: Apr 25, 2006 Posts: 2315 Location: Montpellier, France
Posted: Sun Apr 30, 2006 2:05 pm Post subject:
Quote:
I just find frustrating that we can't get the same quality with a faster release cycle, but obviously it's not possible, especially in the current situation.
Quote:
In fact, I would guess that Microsoft spends at least US$50 or US$100 million on the Mac Office release alone.
Quote:
Basically, we may be slow, but we run NeoOffice on less than US$100K per year and provide downloads to nearly 2 million users per year.
With Intel processors we get the best performance-per-watt, and with Ed & Patrick we get the best software-per-buck !
Good job guys. I've wanted to learn programming for quite a while, but my skills are still minimal and it would take many years before I could be of any real help (to NeoOffice). BTW I have a question for you two, since you've been to University in the US : did you study computer sicence or did your interest for coding come afterwards ? And if you did study computer science, what do you think are the prequisites for the best possible ride through undergraduate studies in C.S. ?
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Sun Apr 30, 2006 2:25 pm Post subject:
I did study CS (and physics) but had been passionate as a kid about programming (my father was teaching me BASIC and Pascal when I was in 3rd grade). In terms of what helps going into it, it depends what kind of program. For programs that are really theoretical (like the dept. at my college), you definitely should take some statistics and combinatorics courses that some schools offer. I didn't have any and it really bit me in the a** when it came time to start proving runtimes of algorithms and proving correctness of graph algorithms. Not all CS programs go that much into detail requiring proofs, however, and I definitely have never used any of those techniques as actual "skills" in my real jobs. There's definitely a difference between "academic" CS programs and "applied" CS programs.
For any type of CS program, it's really great to have some type of computer programming exposure in high school, although most schools don't even offer that now and instaed are doing more "vocational" computer courses (learn to use Excel, etc.). Another thing that I think is really useful is geometry. It's a personal thing, but I found introductory programming was really similar to doing geometry proofs. In a proof you need to break larger concepts down into step by step logical progressions. The way that you start with simple proofs and then use those concepts as building blocks in more complex proofs is really similar to procedural programming.
Granted, both my parents were math teachers, so I guess I might be biased
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You cannot download files in this forum
)
