[Y-devel] Quick question... Is this project dead?!

[Phil]

Kimmo Sundqvist wrote:
> On Monday 17 January 2005 20:57, Jonathan Harper wrote:
> 
> 
>>The options are therefore 1) fork or 2) rewrite. Sitting around isn't going
>>to get us anywhere, as Phil and Andrew have repeatedly said, because it's
>>more of a private project than anything.

I sense that the myth of hidden code somewhere persists in your mind. 
Out! Out You Demons of Stupidity!!!

Mark has said he doesn't have a secret stash.  Andrew has said he 
doesn't have a secret stash.  Dave has said he doesn't have a secret 
stash.  I have said I don't have a secret stash.

Either there is no private code, or none of the "cabal" know of its 
existence.

> Did the "cabal" have some good ideas that are not said here yet, that they
> were planning to include in their rewrite?

Not that I am aware of.  I would like to see libsigc++ and whatever 
other C++ code is in there replaced with something in C; this is because 
in code which is this low-level using any language which does things 
"for" you usually ends up more of a liability than an asset.

> Btw, what would be the difference between a fork and a rewrite, now that 
> Mark&co aren't going to do it?

A fork uses some of the existing code, and a rewrite does not.  The 
architecture of GNU Arch (henceforth TLA) requires that all development 
be done in forks which are optionally later merged back to a central 
repository.  Thus, forking Y is not a big deal -- it's the expected 
development method.

Anyway, I don't think Mark, Andrew, Dave or I have ever said we "aren't 
going to do it".  The problem is that we can't predict when we will have 
sufficient time, and even if we could we would often be unable to tell you.

I'm not making excuses here, but I think it's important that I explain 
how being a professional (as in for money) software engineer interacts 
with working on Y.  I imagine that Mark, Andrew and Dave face similar 
restrictions.  There are a few important things you may not have realised:

I am not allowed to work on (or even look at) Y (or any other open 
project) without approval from my management chain and the company's IP 
& Law department.  This takes some time to come through, and must be 
re-obtained when certain circumstances change.  I did not reacquire this 
permission until last Thursday (and I only had permission for a 
fortnight the first time around).  The purpose of this is to ensure that 
I will not inadvertently reveal trade secrets or unwittingly incorporate 
GPL code into a commercial product.

If I am busy at work and asked to work overtime, I will be too coded-out 
to work on open-source projects.  I cannot tell you guys when this 
happens, or if it's happening when it is likely to end.  This is because 
my company's competitors can easily find out what product I work on and 
could possibly deduce when the next version is to be released from my 
work habits.

Of course, even without overtime a 40-hour week of developing reduces 
the amount of development drive I have left for the weekends.

I don't think any of this is unreasonable, I would not have agreed to 
the contract if I did.

Finally, of course, there are other demands on my time. In addition to 
the usual (family obligations etc.), I play Go competitively and need 
time to study to improve my game (and consequently my EGF ranking). 
Attending tournaments also takes time.

I hope this helps to explain why my development input thus far has been 
nil.  I still have a few weeks of paid vacation to book for this year, 
so unless something better comes up I will probably take some of those 
off to get some coding in.  Until then I would expect my useful input to 
be patchy (no pun intended) at best.  If anyone else posts a patch, I 
will be quite happy to review and test-build it on my x86-64 machine. 
I'm more likely to have the time and inclination for that as it's a bit 
less taxing than writing code.

Anyone still reading who cherishes hope or a rapid evolution from v0.2 
through v1.0 should probably give up that dream.  Think more along the 
lines of a HURD-scale development time-line.

Henceforth anyone asking "is this project dead?", "why haven't you made 
any changes?", "where is everybody?" or similar can consider themselves 
added to my revenge-list (which currently contains the designers of the 
Sun C compiler asm statement and the IE development team).

HTH,

Phil