Mail Archives: geda-user/2017/03/29/07:29:33
Hi --
I use gedasymbols on occasion, and I know if gedasymbols simply goes
away, I will at some point try to use it and be flummoxed when it
isn't there.
I vote for option 2, or a variant: Redirect gedasymbols to point at
EDAkrill so that when folks visit gedasymbols, they are redirected.
Thanks,
Stuart
On Wed, 29 Mar 2017, gedau AT igor2 DOT repo DOT hu wrote:
> Hi gEDA users,
>
> gedasymbols.org has ~3k footprints and ~2.5k symbols.
>
> Progress on gedasymbols-to-edakrill import: 876 items (footprints and symbols)
> got imported; edakrill now has more than 1k items available. The bare minimal
> import process should be finished within a week or two.
> (Help wanted with the tagging.)
>
> Meanwhile, we have still only one vote on what should happen to gedasymbols
> after the import. It's probably worth voting now, before I finish the import:
>
> On Sat, 25 Mar 2017, gedau AT igor2 DOT repo DOT hu wrote:
>
>> Hi all,
>>
>> I plan to import the gedasymbols.org database to EDAkrill (please see my
>> other mail about what EDAkrill is). Any file with a distribution license
>> permits this is a potential subject. After the import, we'll have multiple
>> choices about how to run our services.
>>
>> After consulting DJ, the maintainer of gedasymbols, I see three possible
>> ways to proceed:
>>
>> 1. keep them separate services, EDAkrill imports data from gedasymbols.org
>> (and gedasymbols.org imports data from EDAkrill if someone programs it)
>>
>> 2. merge gedasymbols into EDAkrill, keep gedasymbols as is, but read-only
>>
>> 3. migrate gedasymbols to EDAkrill, shut down the CVS based service
>>
>> I think the community should decide: please vote for on of the above
>> options, especially if you are a contributor or user of gedasymbols.org. At
>> the end I'd like DJ to make the final word based on the community
>> feedback/votes.
>>
>> For me any of these are fine. The import will happen in any case,
>> so if there's no decision we will probably end up with 1 implicitly.
>>
>>
>> Some technical details that may help the decision:
>>
>> - gedasymbols.org uses CVS; EDAkrill uses svn. If we go for 2. or 3.,
>> CVS contributors will need to switch to svn or web uploads. Downloads with
>> svn should be similar to CVS, just slightly different command line
>> (and less commands)
>>
>> - please do NOT propose switching EDAkrill to git/hg/your-favorite-VCS. It
>> won't happen, and I won't enter flame wars about this; if svn is a blocker
>> for you, then don't use the service, vote accordingly, set up your own
>> service based on different tools, etc.
>>
>> - other than that I am open to any discussion and questions about EDAkrill
>> or its relation to gedasymbols
>>
>> - solution 1. is the least intrusive change, but it would be sort of
>> effort duplication (which is not necessarily bad)
>>
>> - in case of 2., we would keep gedasymbols.org for compatibility but we
>> would recommend using EDAkrill - there wouldn't be new uploads to
>> gedasymbols.org, but for example tools and users who got used to make a CVS
>> checkout could go on doing that (but they won't see the new files uploaded
>> to EDAKrill)
>>
>> - in case of 3., we would probably redirect gedasymbols.org to EDAkrill, but
>> we wouldn't try to keep a compatible web page or CVS.
>>
>> - how integration is affected: as far as I know only pcb-rnd has gedasymbols
>> integration, and pcb-rnd will have EDAkrill integration too, so there won't
>> be a big change there.
>>
>> - stability of the service: repo.hu is hosted at a professional server hotel
>> in an air conditioned server room with redundant power (mains, UPS and AFAIK
>> even diesel). Network-wise it is very close to the Hungarian Internet
>> backbone. The international bandwidth is somewhat limited (for financial
>> reasons), but stability is excellent.
>>
>>
>> Regards,
>>
>> Igor2
>>
>
- Raw text -