[tei-council] 2.2.0 / 2.1.0

[Martin Holmes]

On 12-10-26 03:11 PM, Sebastian Rahtz wrote:
> I just pulled the release from SF and did some searches; my view is that only the files in xml/tei/custom/schema
> are affected, but that they are (potentially) badly affected - they are the versions from June.

You're right that those are the ones affected as far as I can see.

> I feel like crying, to be honest :-{

I hope you're not serious. A few slightly borked schemas available for 
48 hours is not a disaster. Most people are still using P4 anyway. :-)

> We need a rapid decision - is this a rapid replacement reissue
> or version 2.3.0?  and who will do it.

I think we should rapidly replace the existing 2.2 schemas with the ones 
from the Oxford Jinks. Only James, Piotr and I think you can do that; 
the rest of us no longer have active logins for tei-c.org due to the new 
security thing.

Then I think we should do 2.2.1 within a few days. Kevin might try being 
release tech, if he's got time, and if not, I'm happy to do it. I 
wouldn't mind going through TCW22 again one more time while the recent 
experience is fresh in my mind.

> HOWEVER, I at least need some more time to check things are actually working
> properly now, I don't trust my patches and its too late at night. So someone
> make some decisions tomorrow am.

James's decision, ultimately, I think.

> PS if its not obvious, can I stress that this unfortunate event is no way
> the result of anything Piotr, James and Martin did during the release?

It's still not clear (to me anyway) whether it was something you did, or 
my decision back in June to comment out that one line in the P5 
Makefile. It might perfectly well be my fault.

As it stands, your Jinks (whose workspaces you wiped out) is building 
with the correct date, but mine (which I left intact) is not; it still 
has the old dates. That means that we haven't actually fixed the problem 
yet; if we don't get my Jinks to build correctly, then the next time we 
come to do this, we'll have the same issue (and lastSuccessfulArtifact 
schemas in between releases will be borked).

One thing is clear to me: I had intended to gain a better understanding 
of how the Makefiles work, but I've failed to do that. I understand 
individual blocks, but I don't have a clear idea of how all the 
different targets fit together. I think it might be a good idea for me 
to try making some flowcharts that map out how the builds proceed, so I 
have a clearer sense of where to look when something goes wrong.

Cheers,
Martin

> --
> Sebastian Rahtz
> Director (Research Support) of Academic IT Services
> University of Oxford IT Services
> 13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
>

-- 
Martin Holmes
mholmes at uvic.ca
UVic Humanities Computing and Media Centre