Hi Christian, a pathological case. In DEV300 m51 Ivo moved and reformatted all sdf files. That's 40% of our code base. 145 MB compressed. A Gig or so uncompressed. Worst case ever. If you pull the changesets of a CWS that is rebased over the m51 bound you'll get pathological downloads. This will be in no way representative for the typical use later.
It remains that 'hg incoming' has to do download the changesets to be able to look into them - even if it discards them later. So maybe a Web based solution is better in your use case. Heiner Christian Lohmaier wrote: > Hi Heiner, *, > > On Mon, Aug 24, 2009 at 5:22 PM, Jens-Heiner > Rechtien<[email protected]> wrote: >>> [...] >> Oh, you can customize the output of "incoming" with the hg template >> engine. Incredible flexible. Just have a look here: >> http://hgbook.red-bean.com/read/customizing-the-output-of-mercurial.html > > Thanks for that - I gave it a shot and boy, was I disappointed. > > for this: > $ hg incoming --bundle ause103.bundle > http://hg.services.openoffice.org/hg/cws/ause103 > comparing with http://hg.services.openoffice.org/hg/cws/ause103 > searching for changes > changeset: 262967:349a62468998 > parent: 262762:f816fc756802 > user: hjs > date: Fri Jul 03 18:06:42 2009 +0200 > summary: # i103316# don't require minor extension; rename source root > > changeset: 262968:c25f516dabc1 > parent: 262967:349a62468998 > parent: 262807:d0e90b1453b9 > user: hjs > date: Fri Jul 03 18:51:57 2009 +0200 > summary: update to DEV300_m51 > > changeset: 262969:d72d38e8756b > parent: 262968:c25f516dabc1 > parent: 262929:33bae5801120 > user: hjs > date: Tue Aug 04 18:00:56 2009 +0200 > summary: merge to m54 > > changeset: 262970:9b7521a28e6e > user: hjs > date: Fri Aug 07 14:42:01 2009 +0200 > summary: #i103316#, #i103805# - avoid absolute path in tg_srs.mk; > rework localize.sdf detection > > changeset: 262971:d5d1973d3e8a > tag: tip > user: hjs > date: Fri Aug 07 14:48:36 2009 +0200 > summary: #i103316# - rework localize.sdf detection > > it did download full 900MB! (and of course took ages) - that is not > promising really. > Thankfully pulling the changes from the bundle resulted in an increase > of about 145MB "only", but still a little too much for my taste (and > for a net result of "added 5 changesets with 17 changes to 12 files > (+1 heads)" > > I'll try with some other cws - hoping that those won't cause such big > amount of wasted bandwidth.. > > ciao > Chrisitan > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
