Contents
- 22nd February 2000
- 15th February 2000
- 11th February 2000
- 8th February 2000
- 7th February 2000
- 6th February 2000
- 3rd February 2000
- 1st February 2000
Tuesday, 22nd February 2000
Well, it's been a hectic week all round.
The review daemon has gone - it kept on sending out details of old
changes, driving people off the mailing list. A few won't be
missed, to be honest, but most of the people lost have played a big role
in influencing my work (even if they don't realise it themselves).
The first part of the new Generic Setup code has been
completed. In GNQS_Devel, you can now use './configure' just as if
we are using GNU autoconf. Unfortunately, making GNQS_Devel
compile once again is a large task - one that might take a month or two
to complete.
The old SETUP software has also gone from GNQS_Devel.
We've had independent confirmation from a number of sites that the
'job stuck in arriving state' bug is definitely fixed, as are the weird
signal problems that plagued DEC Alpha users. I'm currently
putting together a list of other things to do before v3.50.9 goes out
the door.
I have two job interviews next week (at last!), and I'm also talking
to a few other people about other options.
Tuesday, 15th February 2000
Yesterday, I set up the Perforce Review daemon. It sends emails
to NQS-Developers whenever a change is made to either source tree.
It spent most of yesterday catching up by mailing out the 125 changes
that have already gone into the two trees. This unfortunately led
to a few people unsubscribing from NQS-Developers, including a couple of
people who's input has been very important in the past :( I hope
they come back, but the Review daemon is a necessary evil. Next
time - I'll warn people first!
The changes that I made late Thursday night and on Friday morning so
far appear to have finally nailed the 'jobs stuck in arriving state'
bug. However, v3.50.8 has only just gone out the door - another
release so soon is going to be too soon. Besides, I want to look
at optimising the main loop in GNQS, and tidying up some of the smaller
problems.
Rest of this week is going to be spent mainly working on a different
project.
Friday, 11th February 2000
I will be having visitors. No chance of GNQS at all over this
weekend.
Thursday, 10th February 2000
Spent the morning deleting all the dead email addresses from the
NQS mailing lists. Ho hum, what fun. It still amazes me
how much of a problem this is.
The afternoon was spent house-hunting, and then T'ai Chi. I
haven't played any T'ai Chi this week (looks guilty) so tonight is
going to be even harder work than normal!
Wednesday, 9th February 2000
Released GNQS v3.50.8. Posted announcements to Linux Weekly
News and Freshmeat.
Tuesday, 8th February 2000
Plans for today are to finish installing my Linux development
environment onto the new box, talk to a few more agencies, and then
put together a patch for the security hole. If the patch works,
I might even release v3.50.8 tonight.
--
Okay, the patch is done, and it works. I'm just going to hold
off releasing v3.50.8 for 24 hours in case Michael or Roman have work
in progress that I'll be interrupting.
Monday, 7th February 2000
A day dedicated to job hunting. Spent the morning in town
talking to recruitment agents. It's quite nice to be told that
I've priced myself too cheaply in the market ;-) A shame that
they dismiss the GNQS work as unimportant; and the UK wonders why it's
no longer a major player in computing.
Found lots more agents this afternoon - guess what I'm doing
tomorrow morning :(
Managed to get distracted. I've been building up a new
desktop machine to become my main Linux development box. Working
30 hours plus every week on the laptop has become quite
uncomfortable. Salvaged a motherboard from one friend's dead
machine, salvaged a supposedly dodgy disk from another friend's
machine, wired them up properly ;-), patched the Linux kernel, and hey
presto.
It's quite ironic really. The new machine is made up of parts
from the latest incarnation of myrddraal.gnqs.org.
Myrddraal was the first Linux box we built for GNQS work, back in
Spring 1995, and until this evening it still had one original part
left - the ISA network card. Depending on how things go over the
coming months, myrddraal may well be the last machine I do GNQS
work on.
I'd better explain.
I'm currently looking for work. That work will either be in a
UK firm, or (just hopefully) for a Linux company. The UK firm
will probably force me to sign a contract that prevents me legally
working on projects like GNQS anymore (an unfortunately common state
of affairs over here these days). The Linux company will
dominate my time completely (I'll be managerial as well as technical),
leaving little time for GNQS. Still, I'm sure it'll be possible
to work something out. We're just starting to get things rolling
forward with GNQS; I'd hate to have to leave the project.
Sunday, 6th February 2000
This should have been the day I did the great source tree
shuffle. Unfortunately, I spent the morning in the garden
loading up a skip, and the afternoon looking into a serious security
hole with Generic NQS.
Thursday, 3rd February 2000
I will be having visitors. No chance of any work on GNQS
today, tomorrow, or Saturday.
Tuesday, 1st February 2000
Can't believe it's February already. Where did January go?
Checked in the configuration stuff, even though it doesn't work
yet. That's not strictly true. What's there, works.
What isn't there will prevent GNQS from compiling.
Spent the rest of the day working on a document for a commercial
project I'm involved in. I get to make a phone call to someone
in the States next week, and that might determine whether I can get
the commercial project off the ground or not. I'll be kept very
busy with this project over the next few days.
I also found out today that, at some point this week, I'll have two
telephone interviews with Motorola. One is definitely on Friday,
the other I don't yet know about.
Spent the evening setting up Bugzilla. Urgh. I wonder
if anyone will pay me to write a nice replacement for it?
Anyway, it's up and available for testing at http://bugzilla.gnqs.org/.
|