Why Never Follow Schedule?

Status
Not open for further replies.
fonz said:
Yes they can, just add two or three months to the initial target date.

An IT truism is that a project will always take twice as long as estimated... even if you expect that and allow for it in the estimate.
 
drhowarddrfine said:
I'm face palming myself to any mention of missing release dates as this happens all the time in all industries and some, here, are acting like it's a FreeBSD only thing. What a childish and immature position to take. It sounds like little kids stomping their feet on the ground and yelling, "But you promised!!!".

Those with ANY experience in manufacturing, software, or...a job of any sort...realizes that such ESTIMATES are ESTIMATES!!! I told the story somewhere on this board of the time I was asked to give a date when I would finish a project and I said three months, my boss put down six months, and I finished in a six weeks. Was I being dishonest? Why did any of us bother to write down a schedule if we didn't plan to follow it?!

Jeez. Get a life. Learn how the world works. Get a real job.

OK no need to take an offensive position in this, as we are just discussing.

About estimates... I am a fresh software engineer myself, so I know some things about that. Usually, you will get a client with a request/project (you know they always do). And most of the time they are demanding a lot. They will say "I need this, this and this. AND I need it in 4 weeks time." They don't care about how hard the project might be, or about anything else. If they say 4 weeks, you know you will have to deliver before the time runs up. You can't say "yes, we were expecting a 4 week release, but we decided we would do some more testing, so come back in a fortnight. Sorry for delaying but we were just estimating".

Maybe this is not how it goes with a FreeBSD release, but if they promise the release is on 12th of November, then the release must be on that day. I might be thinking wrong, but just like I promise my clients about a delivery, if FreeBSD devs promise us something, then they should stick to it.

The simple way to go about this is to either not give any estimates, or say that it is tentative. Still not very professional, but whatever...

Just my thoughts...
 
19758320.jpg
 
Majorix said:
OK no need to take an offensive position in this, as we are just discussing.

About estimates... I am a fresh software engineer myself, so I know some things about that. Usually, you will get a client with a request/project (you know they always do). And most of the time they are demanding a lot. They will say "I need this, this and this. AND I need it in 4 weeks time." They don't care about how hard the project might be, or about anything else. If they say 4 weeks, you know you will have to deliver before the time runs up. You can't say "yes, we were expecting a 4 week release, but we decided we would do some more testing, so come back in a fortnight. Sorry for delaying but we were just estimating".

Maybe this is not how it goes with a FreeBSD release, but if they promise the release is on 12th of November, then the release must be on that day. I might be thinking wrong, but just like I promise my clients about a delivery, if FreeBSD devs promise us something, then they should stick to it.
The developers at FreeBSD work for free. Your not entitled to anything.

Majorix said:
The simple way to go about this is to either not give any estimates, or say that it is tentative. Still not very professional, but whatever...

Just my thoughts...
ten·ta·tive: unsure; uncertain; not definite or positive; hesitant: a tentative smile on his face.
I thought it was implied? Why else would we have "Expected" and "Actual" dates?
 
Majorix said:
OK no need to take an offensive position in this, as we are just discussing.

About estimates... I am a fresh software engineer myself, so I know some things about that. Usually, you will get a client with a request/project (you know they always do). And most of the time they are demanding a lot. They will say "I need this, this and this. AND I need it in 4 weeks time." They don't care about how hard the project might be, or about anything else. If they say 4 weeks, you know you will have to deliver before the time runs up. You can't say "yes, we were expecting a 4 week release, but we decided we would do some more testing, so come back in a fortnight. Sorry for delaying but we were just estimating".

Maybe this is not how it goes with a FreeBSD release, but if they promise the release is on 12th of November, then the release must be on that day. I might be thinking wrong, but just like I promise my clients about a delivery, if FreeBSD devs promise us something, then they should stick to it.

The simple way to go about this is to either not give any estimates, or say that it is tentative. Still not very professional, but whatever...

Just my thoughts...

You're forgetting one thing... FreeBSD teams are VOLUNTEERS and they do not get paid to program, build and test the software. Show some appreciations for their hard work on FreeBSD projects.
 
Yes, they are not paid to do develop FreeBSD, and I really, really appreciate that they take their time to bring us a very solid operating system. However, it's still a project that they maintain, and I feel they have the responsibility to iron out its problems and bring a release in time.
 
Majorix said:
but if they [red]promise[/red] the release is on 12th of November, then the release [red]must[/red] be on that day.
You still don't seem to get it: there is no "promise" and FreeBSD "must" not anything.

Fonz
 
Dates do count because they affect planning. Anybody who was ever involved in software development is aware of this.

So, please don't start jumping on someone who poses a legitimate question. And be kind enough to avoid personal insults.

@Majorix,

I understand your frustration mainly because I often have to explain why I follow STABLE. It has also been suggested by others to introduce a more extensive errata policy, using official patches.
 
The recent security incident also caused further slippage because the infrastructure needed to be rebuilt. Shit happens, I'm comforted by knowing that when the -RELEASE announcement is finally posted I can immediately upgrade my production systems without things breaking, because the FreeBSD team delivers quality. Consistently. That's the reason FreeBSD is my default general-purpose OS choice for mission-critical production systems
 
So, Majorix! 30+ posts later, will you sleep better at night if the table said 12 March 2013 instead of 12 November 2012 and the team released 9.1 during December 2012? :p
To me this looks like a cosmetic matter.

Majorix said:
Maybe this is not how it goes with a FreeBSD release, but if they promise the release is on 12th of November, then the release must be on that day.
Well, that's the thing. The developers never promised anything.
A promise is an oath or a vow; it is, or ought to be, unbreakable. An expectation on the other hand is something that is *meant* to happen, it's a *supposed* event, an *estimation*, a *rough guess*...

Majorix said:
just like I promise my clients
The FreeBSD Project and Foundation are not our suppliers, and we're not their clients or shareholders.
You get paid to satisfy some client in the market. OSS developers have their own jobs and lives; they code during their free time for fun and/or to improve their skills and/or to add something interesting to their CV.
 
OK, I think everyone has their opinions here. Being a person who respects others' thoughts, I won't continue discussing this.

I think I made my point, and by the replies of some others, I would say I am not completely alone on the matter.

I explained myself, and made all the points I could present. So I will just refrain from posting on this thread anymore, particularly because I don't want to look like a troll or anything.

Peace!
 
gkontos said:
Dates do count because they affect planning. Anybody who was ever involved in software development is aware of this.
And that IS my point. The date is for planning but the date is NOT the goal nor a requirement for success.
 
drhowarddrfine said:
And that IS my point. The date is for planning but the date is NOT the goal nor a requirement for success.

Yup! I remember FreeBSD 5. That was alot of fun. Don't want to ever run into that situation again.
 
Majorix said:
So the questions are:
1. Why formally declare release dates if you won't follow them?
2. Why need 6 months of testing to just release a minor version?

They are guidelines, not hard release dates.

It is FAR better to miss a release date and have a more stable release.
 
drhowarddrfine said:
And that IS my point. The date is for planning but the date is NOT the goal nor a requirement for success.

Aside from FreeBSD.

When we plan on delivering a project, we put project dead lines and milestones. If team (a) can not deliver within that time frame then team (b) can not pick up. As a result I have resources that are being wasted and a product that delays significant.

This affects the success of my project / product.

There is a very important factor here that most project managers and business developer managers often don't see. The most important and critical job here is the job of the person or people who are responsible to coordinate. And in the end of the day you will see that well bonded teams who work under the supervision of capable and dedicated project managers always deliver within a certain time frame with quality.
 
Majorix said:
Usually, you will get a client with a request/project (you know they always do). And most of the time they are demanding a lot. They will say "I need this, this and this. AND I need it in 4 weeks time." They don't care about how hard the project might be, or about anything else. If they say 4 weeks, you know you will have to deliver before the time runs up. You can't say "yes, we were expecting a 4 week release, but we decided we would do some more testing, so come back in a fortnight. Sorry for delaying but we were just estimating".

And this is the biggest problem with the software industry today. As above, if you want FreeBSD team to ship broken crap, there's nothing to stop you downloading and building from -CURRENT yourself. Shipping broken crap on-time is fine if you're a game developer, or web "developer". If you're shipping an OS that is expected to be stable and able to deal with threats from internet facing interfaces, a little more quality control is required.

Maybe this is not how it goes with a FreeBSD release, but if they promise the release is on 12th of November, then the release must be on that day. I might be thinking wrong, but just like I promise my clients about a delivery, if FreeBSD devs promise us something, then they should stick to it.

You keep using this word "promise".

No one has promised anything. The project is staffed by volunteers for the most part.

And of course there was the recent security incident to deal with.


If you want to see a real missed target, see Duke Nukem Forever (peeps finished college, got married and had kids who then grew up to be old enough to play it during the delay - it was expected initially in 1996 or 1997 from memory), and the features Microsoft has still to ship from "Cairo", which was "promised" in 1993.
 
Vista was delayed somewhat so Microsoft could push XP SP2 out the door to fix the collossal security disaster that previous XP versions were.

However, even before then:

Windows 95 was delayed..

Windows NT 5 was delayed at least a year (ended up being called Windows 2000).

Running late, in 1998? it came out in December 1999.

Windows 2000 Sp1 was delayed as well. In fact, most operating systems seem to miss their initial estimated ship date.


If Microsoft can't ship on time (even though they're clearly willing to ship with bugs), what hope does a band of volunteers have?
 
I think the argument "they're volunteers" isn't at all relevant for two reasons

1. The Foundation collects money to assist in FreeBSD development. If they're accepting funds then they don't get to also claim to be doing this solely out of the goodness of their hearts.

2. Whether a person is a volunteer or not should not, in any way, affect the quality of their work. When the FreeBSD team puts up a schedule they should stick to it. To let a release slip so far past its target release date looks sloppy and unprofessional.

If FreeBSD wants to be taken serious as a platform for serious developers, admins and professionals then they should act in the same manner.
 
NewGuy said:
I think the argument "they're volunteers" isn't at all relevant for two reasons

1. The Foundation collects money to assist in FreeBSD development. If they're accepting funds then they don't get to also claim to be doing this solely out of the goodness of their hearts.

2. Whether a person is a volunteer or not should not, in any way, affect the quality of their work. When the FreeBSD team puts up a schedule they should stick to it. To let a release slip so far past its target release date looks sloppy and unprofessional.

If FreeBSD wants to be taken serious as a platform for serious developers, admins and professionals then they should act in the same manner.

Back your point up. Where in the hell have you seen word "promise", or "We promise to release FreeBSD bla.... bla... bla on day.... bla bla bla". If you can't find it sit tight and wait for it to happen. It will happen sooner, or later (better late then never)


DutchDaemon, how about you close this thread for good.
Everything is said over and over and over already.
 
NewGuy said:
1. The Foundation collects money to assist in FreeBSD development. If they're accepting funds then they don't get to also claim to be doing this solely out of the goodness of their hearts.

They are not on a salary if that is what you imply.

NewGuy said:
2. Whether a person is a volunteer or not should not, in any way, affect the quality of their work. When the FreeBSD team puts up a schedule they should stick to it. To let a release slip so far past its target release date looks sloppy and unprofessional.

Generalizing in software development is a recipe for disaster. Also don't judge people who have no affect on when and if a RELEASE will be available as scheduled.

NewGuy said:
If FreeBSD wants to be taken serious as a platform for serious developers, admins and professionals then they should act in the same manner.

Yes but you like others are missing the point and the big picture. Developers develop. The follow certain project guidelines. It is not their job to stick to a RELEASE plan.

Developers are engineers and good engineers never stop engineering. For all I care, a good developer would never draw the line for a RELEASE. It is their job to improve software.

Project managers on the other hand have the responsibility to draw the lines and set the priorities.

Sales and Marketing people are the ones who are responsible to promote the end product and draw funds.

In other words it is the foundation's job to handle all those issues.
 
graudeejs said:
DutchDaemon, how about you close this thread for good.
Everything is said over and over and over already.

That would be the easy but not the correct approach. Think about it...
 
Status
Not open for further replies.
Back
Top